• acetate;
  • astrocyte activation;
  • deoxyglucose;
  • β-hydroxybutyrate;
  • neuron–astrocyte interactions


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

To evaluate the response of astrocytes in the auditory pathway to increased neuronal signaling elicited by acoustic stimulation, conscious rats were presented with a unilateral broadband click stimulus and functional activation was assessed by quantitative autoradiography using three tracers to pulse label different metabolic pools in brain: [2-14C]acetate labels the ‘small’ (astrocytic) glutamate pool, [1-14C]hydroxybutyrate labels the ‘large’ glutamate pool, and [14C]deoxyglucose, reflects overall glucose utilization (CMRglc) in all brain cells. CMRglc rose during brain activation, and increased activity of the oxidative pathway in working astrocytes during acoustic stimulation was registered with [2-14C]acetate. In contrast, the stimulation-induced increase in metabolic activity was not reflected by greater trapping of products of [1-14C]hydroxybutyrate. The [2-14C]acetate uptake coefficient in the inferior colliculus and lateral lemniscus during acoustic stimulation was 15% and 18% (p < 0.01) higher in the activated compared to contralateral hemisphere, whereas CMRglc in these structures rose by 66% (p < 0.01) and 42% (p < 0.05), respectively. Calculated rates of brain utilization of blood-borne acetate (CMRacetate) are about 15–25% of total CMRglc in non-stimulated tissue and 10–20% of CMRglc in acoustically activated structures; they range from 28 to 115% of estimated rates of glucose oxidation in astrocytes. The rise in acetate utilization during acoustic stimulation is modest compared to total CMRglc, but astrocytic oxidative metabolism of ‘minor’ substrates present in blood can make a significant contribution to the overall energetics of astrocytes and astrocyte–neuron interactions in working brain.

Abbreviations used

blood–brain barrier


rate of acetate utilization


rate of glucose utilization


rate of oxygen utilization




integrated specific activity


monocarboxylic acid transporter


relative specific activity


tricarboxylic acid cycle

Compartmentation of glutamate metabolism in brain provides an important tool for in vivo studies of neuron–astrocyte interactions because different labeled compounds can be used alone or in combination to preferentially label metabolic pathways in astrocytes, neurons, or both. For example, after in vivo injections of tracer amounts of labeled glucose, pyruvate, lactate, acetoacetate, and β-hydroxybutyrate the ‘large’ glutamate pool is preferentially labeled, yielding a glutamine/glutamate specific activity ratio of less one (∼0.3–0.6) (O'Neal and Koeppe 1966; O'Neal et al. 1966; Cremer 1971; Cremer et al. 1978). In contrast, the relative specific activity (RSA) of glutamine is greater than one (ranging up to ∼5) after in vivo injections of labeled acetate, butyrate, propionate, leucine, or octanoate, revealing the existence of a ‘small’ glutamate pool that is the precursor for glutamine (O'Neal and Koeppe 1966; O'Neal et al. 1966; Berl and Frigyesi 1968; Patel and Balazs 1970; Cremer and Heath 1974; Cremer et al. 1977; Cremer et al. 1978). Assignment of the large and small glutamate pools to neurons and astrocytes, respectively, is based on different lines of evidence, including localization of glutamine synthetase in astrocytes (Martinez-Hernandez et al. 1977), preferential transport of acetate into cultured astrocytes compared to cultured neurons or synaptosomes (Waniewski and Martin 1998), greater in vivo autographic labeling of astrocytes with [3H]- and [14C]acetate compared to neuronal cell bodies and axons (Minchin and Beart 1975; Muir et al. 1986), higher glutamate levels in neurons and their processes compared to astrocytes and higher glutamine levels in glial cells in rat cerebellum (Ottersen et al. 1992), and 13C NMR studies confirming labeling the small pool by [13C]acetate in animals (e.g. Cerdan et al. 1990; Hassel et al. 1997) and humans (Lebon et al. 2002) and labeling the large pool by [2,4-13C2]β-hydroxybutyrate in humans (Pan et al. 2002). However, 2 h infusion of anesthetized rats with [U-13C4]hydroxybutyrate produces labeling patterns ascribed to the glial compartment (Künnecke et al. 1993). An intriguing finding noted by Cremer (1971) is that compounds transported across the blood–brain barrier (BBB) by the monocarboxylate transporter (MCT) can have different metabolic fates, for example ketone bodies, pyruvate, and lactate yield glutamine RSAs < 1, whereas acetate, butyrate, and propionate yield glutamine RSAs > 1. The basis for these differences is not known but might arise, in part, from differences in substrate preference, kinetics, and cellular distributions of MCT isoforms. Glucose, lactate, pyruvate, and β-hydroxybutyrate are readily oxidized in vitro by cultured astrocytes and neurons (e.g. Hertz 1982; Edmond et al. 1987; Edmond 1992), and glucose is the major fuel for all brain cells in vivo. However, differences between intact brain tissue and cultured cells, including maturation in vivo, transport barriers, and enzyme levels, would influence flow of substrates into different pathways. The glutamine RSA obtained after labeling with [U-14C]leucine rises postnatally, from 0.3 in newborn rats to 2.4 at 35 days (Patel and Balazs 1970), whereas β-hydroxybutyrate yields a glutamine RSA < 1 in newborn, developing, and adult brain in spite of the large shifts in levels (high during suckling; low thereafter) of its transport and metabolism proteins (Cremer 1971; Cremer and Heath 1974; DeVivo et al. 1975; Van den Berg and Ronda 1976b).

The present study used metabolic labeling in conscious rats to preferentially label the large and small glutamate pool in vivo and evaluate metabolic responses of astrocytes and neurons in brain structures that process acoustic information compared to other regions that would not be affected by stimulation of the auditory pathway. Neuronal synaptic activity and the rate of glucose utilization in auditory structures are enhanced when sound intensity increases (Sharp et al. 1981; Webster et al. 1985; Nudo and Masterton 1986), but the metabolic responses of astrocytes to sensory stimulation are not known. The objectives of the present study were: (i) to evaluate whether local metabolic activation of the oxidative pathway in astrocytes can be readily detected by quantitative autoradiography using [2-14C]acetate; (ii) to compare the calculated rate of acetate utilization during activation to glucose utilization (CMRglc) as a measure of the overall metabolic activity of all brain cells; (iii) to assess [1-14C]hydroxybutyrate metabolism during acoustic stimulation.

[2-14C]Acetate, 2-deoxy-d-[1-14C]glucose (DG) (specific activity 54–56 mCi/mmol), l-[14C(U)]glutamate (282 mCi/mmol), and l-[14C(U)]glutamine (273 mCi/mmol) were purchased from PerkinElmer Life Sciences (Boston, MA, USA). β-Hydroxy[1-14C]butyrate (55 mCi/mmol) was obtained from American Radiolabeled Chemicals (St. Louis, MO, USA).

On the day of the experiment, non-fasted male Wistar Hanover rats (Taconic Farms, Germantown, NY, USA) were anesthetized with halothane, catheters were inserted into a femoral artery and vein, and one tympanic membrane was punctured and the external auditory meatus plugged with bone wax. Monaural deprivation facilitates unilateral acoustic activation and determination of left–right differences in each subject. Rats were restrained via a hind limb plaster cast, and allowed to recover for at least 3 h; rectal temperature was monitored and maintained at 37°C. Immediately prior to the experiment, mean arterial blood pressure was measured with a calibrated Micro-Medical Analyzer (Louisville, KY, USA), and arterial blood was drawn for assay of PCO2, PO2, and pH (Model 248 pH/Blood Gas Analyzer, CIBA-Corning, Medfield, MA, USA), hematocrit, and plasma glucose and lactate levels (Model 2700 Dual Channel Analyzer, Yellow Springs Instruments, Yellow Springs, OH, USA).

An acoustic stimulus was presented to conscious rats in a sound-insulated box using a S10CTCM Click-Tone Module (Grass-Telefactor, W. Warwick, RI, USA) and two Grass Model 10H2S Audiometric Headphones. The stimulus was a broadband click (40 Hz−8 kHz tone presented as 100 µs duration square wave at a 99 Hz stimulus rate) at an intensity setting of 103 dB that corresponded to a measured level (Digital Sound Level Meter, Extech Instruments, Waltham, MA, USA) of ∼88 dB at the position of the rat; the ambient sound level in the box was ∼55 dB. The acoustic stimulus was given for 10 min before and maintained throughout the metabolic labeling period that was initiated with an intravenous pulse (100 µCi/kg) of [2-14C]acetate, [1-14C]DG, or [1-14C]hydroxybutyrate. Timed samples of arterial blood were drawn for assay of plasma glucose and lactate levels and total 14C contents in plasma (Packard Model 2550 liquid scintillation counter, external standardization). In a separate group of rats, larger samples of plasma were collected and stored at −80°C until used for purification of [14C]acetate and identification of labeled metabolites (see below). At the end of the experiment (5 min for 14C-labeled acetate and β-hydroxybutyrate or 45 min for [14C]DG), the rats were killed (pentobarbital, 200 mg/kg i.v.) and the brains were rapidly removed, frozen in isopentane, and stored at −80°C. The 45 min experimental period is necessary for fully quantitative assays of CMRglc with [14C]DG, but most of the labeling by [14C]DG occurs within the first 15 min after pulse labeling and activation is readily detected with [14C]DG after only 5 min of labeling in vivo (Adachi et al. 1995; Sokoloff et al. 1977; Sokoloff 1986). Also, preliminary studies in our laboratory show a similar (but only 20%) increase in labeling by [6-14C]glucose with a 5 min assay starting at either 10 or 50 min after the onset of acoustic stimulation. All animal use procedures were in strict accordance with the National Institutes of Health Guide for Care and Use of Laboratory Animals, and were reviewed and approved by the local animal care and use committee.

Local 14C concentrations in brain regions of interest were determined by quantitative autoradiography and used to calculate net uptake coefficients and metabolic rates. Serial coronal brain sections (20 µm-thick) were cut at about −25°C, dried onto glass coverslips at 60°C, and exposed to X-ray film along with calibrated 14C methacrylate microscales (Amersham, Piscataway, NJ, USA); 14C levels in regions of interest were determined by computer-assisted densitometry (Imaging Research, Inc., St. Catharines, Ontario, Canada). Local rates of glucose utilization (CMRglc) were determined according to the routine [14C]DG method (Sokoloff et al. 1977). Net uptake coefficients [i.e. plasma clearance rates, with units of mL/(g min)] for [2-14C]acetate and [1-14C]hydroxybutyrate were calculated by dividing the total 14C concentration (nCi/g) in the brain regions of interest by the integrated activity of total 14C in arterial plasma [(nCi/mL)min]. With an experimental period of 5 min, these net uptake coefficients mainly reflect metabolism because 90–95% of the 14C in brain at 5 min after a pulse of [14C]acetate is recovered in labeled metabolites (Dienel et al. 2001a,b) and half of the label recovered from rats given β-hydroxy[3-14C]butyrate is in amino acids at 3 min (Cremer 1971). Due to restricted transfer across the BBB and competition of [14C]acetate and [14C]hydroxybutyrate with monocarboxylic acids that share the same transport system, the calculated plasma time–activity integral overestimates the true precursor pool in brain and yields a minimal value for the net uptake coefficient and substrate utilization rate.

CMRacetate was calculated for representative auditory and non-auditory structures during rest and activation from the net uptake coefficients. Because the uptake values are based on time–activity integrals from measurements of total 14C in plasma, it was necessary to convert them to integrated specific activities (ISAs) and make corrections for the contribution of labeled metabolites in plasma and for overestimation of the brain time–activity integral by use of that in plasma; the overall correction was to multiply the uptake coefficient by 2.22. First, the uptake coefficient was multiplied by the mean plasma acetate concentration (0.94 µmol/mL), which was stable throughout the experimental interval, to convert the units from (nCi/mL)min to (nCi/µmol)min. This value was corrected for the contribution of labeled metabolites in arterial plasma by dividing by 0.66. To determine this correction factor, the time courses of total 14C and purified [14C]acetate were measured in each of five animals in a separate group of rats and the respective time–activity integrals calculated, i.e. 607 ± 39 and 402 ± 40 (nCi/mL)min (mean ± SD, p < 0.001, t-test); the mean of the ratios of the integral for purified [14C]acetate to total 14C determined in each animal was 0.66 ± 0.03 (n = 5). The resulting value was then divided by 0.64, the ratio of time–activity integral in brain compared to that in plasma. The ratio brain/plasma time–activity integrals was estimated from the time course of purified plasma acetate and calculations using rate constants (Lear and Ackermann 1990) for acetate transport (K1 = 0.13, k2 = 0.15) and metabolism in brain (k3 = 0.14); the estimates of acetate ISAs for brain and plasma were 274 and 428 (nCi/µmol)min, respectively. Values of rate constants used to estimate the brain ISA from measured values in plasma in different subjects influence the accuracy of the calculated rates and rate constants have a greater impact in short (5–10 min) compared to longer (30–45 min) labeling periods (Sokoloff et al. 1977; Sokoloff 1986; Adachi et al. 1995), so the calculated tissue time–activity integrals and metabolic rates might not be as accurate as desired. When calculated as described above, CMRacetate is a minimum value because it is not adjusted for competition of [14C]acetate in plasma with unlabeled lactate for transport across the BBB by the MCT. Plasma levels of both substrates are about 1 mm (see Results section), and the brain uptake indexes for acetate and lactate are similar (Oldendorf 1973). The influence of competitive inhibition of acetate transport by lactate was estimated using Michaelis–Menten kinetics and assuming that Vmax and Km are the same for both substrates so that Km = Ki; transport rates were calculated using a range of values for Km; reported values for lactate Km range from 1 mm (Cremer et al. 1979) to 6–14 mm(LaManna et al. 1993) in adult rat brain and 3–8 mm for MCT1 in other systems (Enerson and Drewes 2003). If the assumptions are valid, adjustment for acetate transport inhibition would raise the minimum CMRacetate by a factor of 1.5, 1.14, or 1.07 if the Km were 1, 6, or 14 mm, respectively; tabulated values are based on a mid-range value, Km = 6.

The specific activities of glutamine and glutamate were determined in a separate group of conscious rats to verify that autoradiographic assay of [2-14C]acetate uptake into brain mainly reflects metabolism via the small glutamate compartment. After 5 min pulse labeling with [2-14C]acetate during acoustic stimulation as described above, the rats were anesthetized (thiopental, 15 mg/kg, i.v.), their brains frozen in situ, and stored at −80°C; inferior colliculus and cerebral cortex were dissected out of each hemisphere of each brain in a cryobox at −25°C, powdered under liquid nitrogen, and weighed at −25°C. Metabolites were extracted with perchloric acid (Adachi et al. 1995), amino acids were separated from acidic and neutral compounds by Dowex-50-H+ column chromatography (Dienel et al. 2002), and amino acid levels were determined by HPLC with fluorometric detection (Shank et al. 1993). Because the high sample loads required for adequate 14C levels for counting interfered with amino acid analysis, the amino acids in the Dowex 50-H+ column eluates were separated by cellulose TLC with t-butanol : methylethylketone : formic acid : water (40 : 30 : 15 : 15) as the solvent system (Fink et al. 1963), eluted, and counted. This system resolves GABA, glutamate, aspartate, and glutamine (Rf values = 57, 45, 37, 31, respectively), but separation of glutamine and aspartate was not complete. The eluted glutamine fractions were therefore next adjusted to pH 7.0 and applied to a Dowex-1-acetate column to remove aspartate prior to 14C counting. Initial experiments in which the tissue extracts were incubated with and without glutaminase to convert glutamine to glutamate prior to TLC showed that labeling of aspartate was small compared to glutamine and no other labeled compounds remained in the glutamine fraction after its removal by glutaminase treatment. Completeness of the glutaminase reaction, and the Rf values and quantitative recoveries of glutamate and glutamine from the analytical procedure were first established using authentic 14C-labeled glutamate and glutamine.

Acetate in timed samples of arterial plasma was purified for determination of acetate concentration, the [14C]acetate time–activity integral, and 14C-labeled metabolites. Organic acids and anions were separated using a Dionex (Sunnyvale, CA, USA) DX-500 anion exchange chromatography system and an IonPac AS11-HC analytical column. The eluent flows sequentially through the column, suppressor (where the background is lowered by acid-base neutralization of the eluant), and conductivity detector to waste, where samples were collected for determination of 14C contents of timed fractions, with correction for the lag arising from tubing length. Total recovery of 14C in the detector effluent was determined by dividing 14C in all fractions by that applied to the column by the autosampler. We previously found that some sample components can cross the selectively permeable membranes of the suppressor, thereby causing label loss due to transfer into the suppressor waste fraction that does not flow to the conductivity detector where timed samples are collected for counting (Wang and Dienel 1994). Recoveries of 14C-labeled acetate, butyrate, and lactate standards in the detector effluent exceed 97%, and 0.2–1.5% was recovered in the suppressor waste line (n = 6 for each test compound). However, recovery of [14C]bicarbonate from the detector output ranged from 33 to 48%, and that from the suppressor waste line was 7–11%; total recoveries of 14C from injected [14C]bicarbonate were only 40–60% (n = 6), indicating some loss of 14CO2, even though all fractions were collected into vials containing NaOH to trap 14CO2 as carbonate.

To identify labeled glucose in selected samples of arterial plasma, glucose was derivatized to glucose-6-phosphate by incubation of portions of plasma samples with hexokinase, ATP, and Mg2+ and analyzed by the organic acid separation procedure described above. Unlabeled glucose is detectable by means of pulsed amperometry, and the identity of glucose as a major metabolite of [2-14C]acetate was confirmed using a second procedure, i.e. a Dionex AAA-Direct amino acid analyzer with an AminoPac PA10 L column. The time course of plasma [14C]glucose formed from [2-14C]acetate was not measured, and ISAplasma glucose and ISAbrain glucose were estimated (see Results section). The percentage of recovered 14C present in the glucose HPLC fraction at 1, 2, and 5 min was first multiplied by 0.82 to account for incomplete recovery of 14CO2 in the HPLC fractions (glucose levels are percentage of recovered, with an 82% overall recovery due mainly to loss of 14CO2 during HPLC analysis). These values were multiplied by the total 14C level in plasma at the corresponding times (see Results section) and divided by the mean plasma glucose level (12 µmol/mL) to obtain plasma glucose specific activities. ISAs were calculated using our estimates of rate constants (Adachi et al. 1995) for glucose transport (K1 =0.126, k2 = 0.163) and metabolism (k3 = 0.156); the respective values for plasma and brain were 9.6 and 3.8 (nCi/µmol)min. To evaluate potential interference of [14C]glucose in plasma formed by peripheral metabolism of [2-14C]acetate, the quantity of 14C in brain contributed by metabolism of both [2-14C]acetate and [14C]glucose was estimated as the sum of the products of the brain ISA and calculated CMR for each substrate.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Plasma glucose levels and physiological variables determined in each animal just prior to initiation of the experiment were similar in all groups and within the normal range (Table 1). The plasma levels of lactate, which competes with acetate and hydroxybutyrate for transport into brain by the MCT (Oldendorf 1973; Cremer et al. 1976; Cremer et al. 1979; Sarna et al. 1979), were relatively low and similar in all groups.

Table 1.  Physiological variables in conscious rats used for comparative studies of brain activation by acoustic stimulation
Group (n)Rectal temp. (°C)Arterial plasmaArterial blood
Glucose (mm)Lactate (mm)Pressure (mmHg)Hematocrit (%)pH (mmHg)pO2 (mmHg)pCO2
  1. Physiological variables were determined immediately prior to injection of the metabolic tracer; values are means ± SD for the indicated number (n) of rats/group.

 Broadband stimulus (5)37.5 ± 0.711.2 ± 1.80.8 ± 0.2118 ± 347 ± 27.43 ± 0.0194.0 ± 5.441.0 ± 2.9
 Ambient sound (3)36.9 ± 0.213.2 ± 2.10.7 ± 0.2127 ± 545 ± 17.43 ± 0.0196.8 ± 3.540.5 ± 1.8
 Broadband stimulus (5)37.3 ± 0.711.8 ± 1.70.7 ± 0.2119 ± 745 ± 37.44 ± 0.0197.5 ± 12.540.6 ± 2.0
 Ambient sound (3)37.1 ± 0.413.1 ± 1.80.8 ± 0.2124 ± 544 ± 17.45 ± 0.0193.8 ± 3.740.6 ± 3.7
 Broadband stimulus (5)37.5 ± 0.510.4 ± 1.10.8 ± 0.2120 ± 1248 ± 37.41 ± 0.0390.0 ± 7.936.2 ± 0.9

Assessment of side-to-side differences in each brain elicited by unilateral acoustic stimulation increases the sensitivity to detect small differences in metabolic activity because each hemisphere is exposed to the same arterial plasma time–activity integral and animal-to-animal differences (e.g. arising from variations in plasma levels of unlabeled acetate and lactate) are minimized. Representative autoradiographs in Fig. 1 illustrate the predominantly unilateral activation of acetate and glucose utilization in the auditory pathway by a broadband click stimulus (Figs 1b and d) compared to assays at ambient sound level (Figs 1a and c). Presentation of the acoustic stimulus enhanced net uptake of [2-14C]acetate in the lateral lemniscus and inferior colliculus by 18 and 15% (p < 0.01), respectively, in the activated compared to contralateral hemisphere, whereas the percentage increases in utilization of glucose were about threefold higher i.e. 41 and 66%, respectively (Table 2). These are minimal percent increases because crossover of nerve fibers between hemispheres may increase functional metabolism in the contralateral hemisphere. Right–left differences for the other auditory structures for acetate and [14C]DG were not statistically significantly different (Table 2). The acoustic stimulus enhanced both acetate and glucose utilization in the superior olive, lateral lemniscus, and inferior colliculus compared to the corresponding (sham) hemisphere in rats assayed under ambient sound levels (p < 0.05, two-way analysis of variance and Tukey's test). Right/left ratios for the non-auditory structures in all groups were close to one (Table 2).


Figure 1. Metabolic activation of astrocytes in auditory structures by broadband acoustic stimulus. Representative autoradiographs illustrate regional differences in metabolic activity in brains of conscious rats with auditory input to one ear blocked during exposure to ambient sound (a, c) or an acoustic stimulus (b, d). Increased processing of acoustic information in two structures in the auditory pathway in the activated hemisphere, the lateral lemniscus and inferior colliculus, is reflected by higher energy demand and increased rates of utilization of [2-14C]acetate by astrocytes (b) and of glucose by all cell types (d). Glucose utilization is reflected by [14C]deoxyglucose (DG); note the different color-coded scales in [14C]DG autoradiographs.

Download figure to PowerPoint

Table 2.  Acetate uptake and glucose utilization in brain of conscious rats during acoustic stimulation
Region of interestMinimum net uptake coefficient for [2-14C]acetate [mL/(100 g min)]Glucose utilization [µmol/(100 g min)]
Broadband click stimulus (n = 5)Ambient sound (n = 3)Broadband click stimulus (n = 5)Ambient sound (n = 3)
Activated (right) hemisphereContralateral (left) hemisphereRight/left ratioSham (right) hemisphereContralateral hemisphereRight/left ratioActivated (right) hemisphereContralateral (left) hemisphereRight/left ratioSham (right) hemisphereContralateral hemisphereRight/left ratio
  1. Unilateral acoustic activation with a broadband click stimulus (∼88 dB) was initiated 10 min prior to intravenous [14C]tracer injection into conscious rats kept in a sound-insulated environment; non-stimulated rats were assayed at ambient sound (∼55 dB) levels (see Methods). The net acetate uptake coefficient was calculated by dividing the 14C level in the region of interest by the time–activity integral for total 14C in arterial plasma. Values are mean ± SD for the number (n) of rats/group; statistically significant right–left differences were identified by the paired t-test (ap < 0.05, bp < 0.01). The right/left ratios are means of values calculated for each structure in each rat, not ratios of the mean values for structures in each hemisphere. The auditory structures in the right hemisphere of rats given the broadband stimulus had statistically significantly higher acetate (p < 0.05) and glucose (p < 0.01) utilization compared to auditory structures in right (sham) hemisphere of rats assayed during ambient sound levels (two-way anova). Plasma acetate levels in a parallel group of rats averaged 0.9 µmol/mL, and net acetate uptake coefficients can be converted to µmol/(100 g min) by multiplying by 0.9 (see Table 5).

Auditory structures
Superior olive5.9 ± 1.55.7 ± 1.51.04 ± 0.054.6 ± 0.74.7 ± 0.70.99 ± 0.0187.3 ± 23.680.2 ± 25.41.11 ± 0.1756.1 ± 3.455.7 ± 2.61.01 ± 0.01
Lateral lemniscus6.0 ± 1.3b5.1 ± 1.41.18 ± 0.104.6 ± 0.84.4 ± 0.61.04 ± 0.0492.0 ± 29.4a63.6 ± 13.61.41 ± 0.2551.0 ± 4.546.9 ± 5.11.09 ± 0.06
Inferior colliculus6.1 ± 1.2b5.3 ± 1.01.15 ± 0.035.3 ± 1.0b5.0 ± 1.01.05 ± 0.01119.8 ± 36.9a70.8 ± 10.21.66 ± 0.3474.9 ± 8.166.4 ± 9.01.13 ± 0.09
Medial geniculate5.5 ± 1.9a5.3 ± 1.81.03 ± 0.014.5 ± 0.34.5 ± 0.41.00 ± 0.0263.9 ± 12.057.9 ± 9.51.11 ± 0.1361.1 ± 8.153.9 ± 6.11.13 ± 0.07
Auditory cortex5.6 ± 1.45.5 ± 1.51.02 ± 0.045.0 ± 0.64.8 ± 0.61.05 ± 0.0167.9 ± 5.065.0 ± 4.31.05 ± 0.0878.9 ± 8.5a70.0 ± 5.21.12 ± 0.04
Other structures
Visual cortex5.3 ± 2.15.4 ± 2.10.99 ± 0.023.9 ± 0.74.1 ± 0.80.97 ± 0.0351.5 ± 7.652.4 ± 7.50.98 ± 0.0441.8 ± 3.140.6 ± 2.31.03 ± 0.02
Sensory cortex5.4 ± 1.15.4 ± 1.21.01 ± 0.024.8 ± 0.54.8 ± 0.51.00 ± 0.0171.7 ± 2.071.6 ± 4.61.00 ± 0.0569.9 ± 6.667.7 ± 4.21.03 ± 0.08
Sensorimotor cortex5.6 ± 1.35.6 ± 1.41.00 ± 0.035.0 ± 0.55.0 ± 0.61.00 ± 0.0270.5 ± 5.271.7 ± 3.90.98 ± 0.0366.0 ± 7.765.1 ± 6.11.01 ± 0.03
Thalamus4.9 ± 2.34.8 ± 2.31.03 ± 0.023.9 ± 0.53.9 ± 0.50.99 ± 0.0562.4 ± 6.163.1 ± 5.30.99 ± 0.0261.2 ± 6.860.9 ± 7.21.01 ± 0.01
Caudate4.9 ± 2.64.9 ± 2.70.99 ± 0.013.5 ± 0.33.6 ± 0.30.98 ± 0.0269.7 ± 5.371.9 ± 4.70.97 ± 0.0364.3 ± 10.664.6 ± 11.51.00 ± 0.01

Calculation of the rate of acetate utilization from product accumulation in brain requires determination of plasma acetate concentration and ISA, assessment of the stability of the levels of unlabeled acetate and lactate in plasma throughout the experimental period, and evaluation of the possibility that labeled metabolites appear in plasma and contribute to metabolic labeling. The procedure designed to separate organic acids and anions in plasma (Fig. 2a) was used to verify the radiochemical purity of the [2-14C]acetate injectant (Fig. 2b) and assess recovery of 14C in the plasma acetate fraction. During the first half minute after the pulse intravenous injection essentially all of the 14C in plasma was recovered in the acetate fraction (Fig. 2c), but by 4–5 min a number of labeled compounds were detectable; two fractions contained most of the 14C, one eluting at 2–4 min (Metabolite 1), the another (Metabolite 2) eluting at 24–27 min (Fig. 2d).


Figure 2. Separation of [2-14C]acetate from 14C-labeled metabolites in arterial plasma. Acetate and lactate are separated from other anion standards (a). Virtually all of the 14C in the [2-14C]acetate injectant is recovered in the acetate fraction (b) and in blood sampled within the first 0.5 min after the intravenous pulse injection of [2-14C]acetate essentially all of the 14C was recovered in the acetate fraction (c). However, at later times most of the 14C was recovered in two other compounds (d); the two major metabolite fractions are designated as Metabolite 1, eluting between 2 and 4 min, and Metabolite 2, eluting at 24–27 min.

Download figure to PowerPoint

To determine ISA of [2-14C]acetate, the time course of its clearance from plasma was determined in a separate group of five non-fasted, conscious rats. Labeled acetate was quickly eliminated from plasma, and appearance of labeled metabolites were detectable within 1–2 min (Fig. 3a). The proportion of total 14C recovered in the metabolite fraction increased with time (Fig. 3b) due, in part, to continuous clearance of labeled acetate and constancy of the level of labeled metabolites (Fig. 3a inset). By 5 min, only about 6% of the 14C in plasma remained in the acetate fraction, whereas Metabolites 1 and 2 accounted for 66% and 9% of the total 14C recovered from the HPLC effluent, respectively (Table 3). The concentrations of unlabeled acetate and unlabeled lactate in arterial plasma were similar and relatively constant throughout the 5-min pulse-labeling period, averaging about 0.9 µmol/mL (Fig. 3c), which is similar to the range of published values for acetate, i.e. 0.2–0.8 mm (Knowles et al. 1974; Buckley and Williamson 1977; Tyce et al. 1981). The lactate concentration in this group of rats is similar to those measured in five separate groups prior to the experimental procedure (Table 1), indicating that competition of [14C]acetate with unlabeled acetate and lactate in plasma for uptake into brain would be nearly constant throughout the experimental period.


Figure 3. Temporal profiles of clearance of [2-14C]acetate from arterial plasma, appearance of [14C]metabolites, and levels of unlabeled monocarboxylic acids in plasma. Acetate was purified by HPLC in timed samples of arterial plasma after a pulse intravenous injection of [2-14C]acetate into conscious rats. Clearance of [2-14C]acetate from arterial plasma is very rapid (a). The proportion of labeled metabolites in plasma progressively increases and by 5 min after the intravenous pulse labeled metabolites accounted for about 95% of the 14C in plasma (b). Note that the total recovery of 14C from the HPLC system fell after the pulse injection (b), presumably due to formation and release to blood of 14CO2 by whole body metabolism of acetate (see Methods, Table 3). The levels of both unlabeled acetate and unlabeled lactate are relatively constant throughout the experimental period (c).

Download figure to PowerPoint

Table 3.  Distribution of 14C in arterial plasma among acetate and its metabolites at intervals after a pulse injection of [2-14C]acetate
HPLC fractionaPercentage of 14C recoveredb
1 min2 min5 min
  • Values are means ± SD (n = 5).

  • a

    HPLC elution profiles of major metabolites in plasma are shown in Fig. 2.

  • b

    Timed fractions of the conductivity detector effluent were collected from plasma samples drawn at the indicated times after pulse labeling, assayed for their 14C contents, and expressed as percentage of total 14C recovered in the detector effluent (see Methods). At early time points pooled samples were collected and some fractions (denoted by dashes) were not individually assayed.

  • c

    Total 14C recovered in the effluent was divided by the 14C content of the sample applied to the column; the injected amount was determined by sampling directly from the autosampler. Recoveries of 14C from samples of the injectant [2-14C]acetate and other monocarboxylic acid standards (butyrate, lactate) in the detector effluent were > 97% (see Methods), whereas recovery of 14C from the detector for plasma samples was lower. The low recovery probably arose from formation of 14CO2 and H14CO3 by metabolism of acetate in vivo and loss of 14CO2 to the suppressor output and release to air (see Methods). In separate experiments plasma metabolite labeling at 5 min after a pulse of [1-14C]acetate (n = 3) was assessed, and of the 14C recovered, about 10% was in acetate, 8% in lactate, 42% in the glucose fraction, and 13% in the carbonate fraction; recovery from the detector effluent was 63%. Lower 14C recovery with [1-14C]acetate compared [2-14C]acetate is consistent with 14CO2 loss, since label is more rapidly lost by decarboxylation of [1-14C]acetate (Pouteau et al. 1998; Dienel et al. 2001a,b).

Acetate67 ± 430 ± 26 ± 1
Lactate2 ± 1
Metabolite 1 (glucose) (2–4 min fraction)14 ± 344 ± 866 ± 13
Metabolite 2 (presumably CO3−2) (24–27 min fraction)9 ± 6
Recovery of 14C in conductivity detector effluent (% of injected)c82 ± 9

Metabolite 1 was identified as [14C]glucose by HPLC analysis of plasma samples before and after treatment with hexokinase plus Mg-ATP to enzymatically convert glucose to glucose-6-phosphate. Because incorporation of 14C from acetate into glucose could complicate use of acetate as an astrocyte marker due to metabolism of glucose in a different (mainly neuronal) compartment (see Introduction), two separation and detection systems were used to establish the identity of glucose as the major labeled metabolite of [2-14C]acetate in rat plasma. Unlabeled glucose is not detected by conductivity in the organic acid procedure used to assay acetate, and this analysis relied on counting the timed eluant fractions (Figs 4a, i and ii); on the other hand, glucose and glucose-6-phosphate are detected by pulsed amperometry, and both labeled and unlabeled glucose could be assayed directly in this system (Figs 4b, i and ii). When assayed in the conductivity system, hexokinase treatment removed virtually all of the 14C from the fractions corresponding to the [14C]glucose standard (Fig. 4a, i) and from HPLC fractions corresponding to Metabolite 1 (Fig. 4a, ii). Similarly, when assayed by amperometry, virtually all the endogenous unlabeled glucose in plasma was converted to glucose-6-phosphate by hexokinase treatment (Fig. 4b, i) and 14C previously contained in the underivatized glucose fraction was recovered in the glucose-6-phosphate fraction after hexokinase incubation (Fig. 4b, ii).


Figure 4. [14C]Glucose is the major metabolite of [2-14C]acetate in rat plasma in brief metabolic labeling experiments. The elution time of a [14C]glucose standard in the organic acid separation procedure (see Methods) was about 3 min, and it was shifted to about 25 min after incubation with hexokinase, ATP, and Mg2+ to convert [14C]glucose to [14C]glucose-6-phosphate (a,i). When a plasma sample drawn from rats at about 4–5 min after a pulse i.v. injection of [2-14C]acetate was similarly treated, the 14C eluting at 2–3 min (a,ii) and recovered in the Metabolite 1 fraction (Fig. 1d), was completely removed from the early eluting fraction by incubation with hexokinase and it was recovered at 25 min (a,ii). Because unlabeled glucose is detectable by pulsed amperometry, the Dionex amino acid analysis system provided a second procedure to confirm the identity of labeled glucose in plasma. Incubation of a plasma sample from rats labeled with [2-14C]acetate with hexokinase converted all of the unlabeled glucose (b,i) and 14C-labeled glucose derived from [2-14C]acetate (b,ii) to glucose-6-phosphate, which eluted at later times. Note that the plasma samples had lower counts than the standards and could not be diluted as much as the [14C]glucose standard used in panel a(i); the higher salt concentrations in the plasma samples incubated in hexokinase assay system altered the elution profiles (a,ii; b,ii) compared to the samples with standards (a,i) or unlabeled glucose in blood (b,i), which ranged from 11 to 13 mm (Table 1).

Download figure to PowerPoint

14CO2 is formed in vivo by oxidative metabolism of [2-14C]acetate in body tissues (Hetenyi et al. 1982; Pouteau et al. 1998) and the elution time of plasma Metabolite 2 corresponded to that of carbonate (Fig. 2d), which would be formed from labeled CO2 and HCO3 in plasma during HPLC analysis due to the high pH of the NaOH eluant. Because recovery of total 14C in the plasma samples from the detector was not complete (82%, Table 3), some 14CO2 was probably lost via the suppressor membrane or during fraction collection, as we observed for the H14CO3 standard in initial experiments (see Methods). Assuming that the 18–25% of the 14C in plasma not accounted for in the recovery analysis (Fig. 3b, Table 3) corresponded to loss of 14CO2 the fraction of 14CO2 would rise from 9 to 25–35%. The plasma lactate fraction was slightly labeled by [2-14C]acetate (Figs 2d and 3b); the identities of minor labeled compounds (Fig. 2d), which could include amino acids, urea, and carboxylic acids (Tyce et al. 1981; Hetenyi et al. 1982; Pouteau et al. 1998), were not determined.

The RSA of glutamine averaged 2.4 in the activated and contralateral inferior colliculus and 5.2 in entire cerebral cortex (Table 4), indicating that the autoradiographs reflect mainly metabolism of [14C]acetate via the small glutamate pool. The concentrations of tricarboxylic acid cycle-derived amino acids did not change during acoustic stimulation (Table 4).

Table 4.  Amino acid levels and glutamine/glutamate specific activity ratios during acoustic stimulation
 Inferior colliculusCerebral cortex
Right (activated)LeftRightLeft
  1. Conscious rats were given an intravenous pulse injection of [2-14C]acetate during unilateral broadband click acoustic stimulation, and their brains frozen in situ 5 min later; and amino acid levels and specific activities determined in dissected samples of inferior colliculus or entire cerebral cortex from the non-activated (left) and activated (right) hemispheres of each animal (see Methods); note that acetate metabolism was not altered in cerebral cortex by the acoustic stimulus (Table 2). Values are means ± SD of three samples for each structure, each sample consisted of brain powders pooled from 2 to 3 rats.

Amino acid level (µmol/g)
Glutamate10.0 ± 1.010.3 ± 1.413.0 ± 2.211.1 ± 1.1
Glutamine6.3 ± 0.26.1 ± 0.46.8 ± 0.86.3 ± 0.6
Aspartate3.8 ± 0.43.7 ± 0.64.0 ± 0.54.1 ± 0.9
GABA2.8 ± 0.42.6 ± 0.52.4 ± 0.22.4 ± 0.8
Specific activity ratio
Glutamine/Glutamate2.4 ± 0.42.4 ± 0.45.1 ± 1.65.3 ± 1.5

Calculated rates of utilization blood-borne acetate in representative non-stimulated and activated auditory structures and in non-auditory tissue yielded values ranging from about 11–16 µmol/(100 g min); these values are 10–25% of the rate of glucose utilization by all brain cells in the same structure (Table 5). CMRacetate is within the order of magnitude of the estimated rate of glucose oxidation by astrocytes (Vox-Glc-astro), and the lower the overall rate of glucose utilization the more closely CMRacetate approximated Vox-Glc-astro (e.g. in the non-stimulated medial geniculate and visual cortex, Table 5). Some of the [14C]glucose in plasma will enter the brain and contribute to the level of total 14C assayed by autoradiography. However, the estimated contribution of [2-14C]acetate to the total 14C in non-activated and acoustically stimulated structures was 89–94% (Table 5). Also, 72–84% of the net rise in total 14C in the activated compared to contralateral auditory structures was estimated to arise from acetate metabolism (Table 5 footnote e). Metabolism of [14C]glucose in brain will also label the large glutamate pool. The contribution of [14C]glucose to the total label in glutamate was estimated to be 6 and 10% in the non-stimulated and activated colliculus, respectively (Table 5 footnote e); this amount would be smaller in structures with lower rates of glucose utilization.

Table 5.  Calculated rates of utilization of blood-borne [2-14C]acetate in brain of conscious rats and contribution of [14C]acetate to the total 14C in brain measured by quantitative autoradiography
StructureCMRglcaVox-Glc astrocytebAcetate net uptake coefficientaCalculated CMRacetatec,d% of total 14C in brain from acetatee
Range (minimum – adjusted for BBB transport competition)c% CMRglcd%Vox-Glc-astrod
  • Units of CMRglc, Vox-Glc, and CMRacetate are µmol/(100 g min); acetate minimum net uptake coefficient units are ml/(100 g min).

  • a

    Data are from Table 2.

  • b

    Vox-Glc astrocyte, the rate of oxidative metabolism of glucose in astrocytes, is estimated to be 10–20% of total glucose oxidation in human brain (Gruetter 2004; Shulman et al. 2004). Under resting conditions almost all of the glucose is oxidized and the TCA cycle rate is twice that of glucose utilization; the Vox range = 0.1 and 0.2 times 2CMRglc. During stimulation the CMRO2/CMRglc ratio falls in many conditions (Dienel and Cruz 2004), and if this occurs during stimulation of auditory structures, total Vox-Glc would be overestimated.

  • c

    Values are a range representing the minimum CMRacetate calculated by multiplying the net uptake coefficients by 2.22 and a higher value obtained by multiplying the minimum CMRacetate by 1.14 to adjust for competition of [14C]acetate with unlabeled lactate across the blood brain barrier (BBB), assuming Km = Ki = 6 mm (see Methods).

  • d

    Minimum CMRacetate is expressed as percentage of CMRglc and percentage of the range of Vox-Glc in astrocytes.

  • e

    The fraction of the total 14C in brain contributed by metabolism of [2-14C]acetate was estimated after calculating the sum of the separate contributions of [14C]acetate and [14C]glucose; these values were estimated as the products of the respective integrated specific activities in brain and utilization rates; minimum CMRacetate was used for this calculation. Also, the percentage of the net increase (activation–non-activated) in total tissue 14C content during stimulation derived from [14C]acetate was estimated as 72% and 84% for inferior colliculus and medial geniculate, respectively. [14C]Glucose would preferentially label the large glutamate pool in brain, and the fraction of 14C in glutamate derived from [14C]glucose in the inferior colliculus was estimated as 6 and 10% in non-stimulated and activated tissue, respectively; this calculation was based on recovery of 31 and 36% of the 14C in brain metabolites labeled by [14C]glucose in glutamate (Dienel et al. 2002).

Inferior colliculus
Medial geniculate
Visual cortex52.410–205.412.0–13.72357–11594

In contrast to the robust activation of glucose utilization and modest rise in acetate metabolism (Fig. 1, Table 2), β-hydroxy[1-14C]butyrate autoradiographs did not show an increase in labeling of the activated inferior colliculus during acoustic activation (Fig. 5) and the net uptake coefficients for [1-14C]hydroxybutyrate were identical in the activated and contralateral inferior colliculus (Table 6). The values in various gray matter structures were similar to each other but were about twice those of white matter (Table 6); all values are comparable to those reported for fed rats (Hawkins et al. 1986). The regional uptake coefficient values for β-hydroxybutyrate were less than half of the corresponding coefficients for acetate (compare Tables 2 and 6).


Figure 5. Labeling of inferior colliculus with [1-14C]hydroxybutyrate during unilateral acoustic stimulation. The representative autoradiograph illustrates the lack of increased metabolic trapping of products of [1-14C]hydroxybutyrate in brain of conscious rats with auditory input to one ear blocked during exposure to an acoustic stimulus (see Methods and compare to Fig. 1).

Download figure to PowerPoint

Table 6.  [1-14C]Hydroxybutyrate net uptake coefficients in brain of conscious rats during acoustic stimulation
Region of interestActivated (right) hemisphereContralateral (left) hemisphere Right/left ratio
  1. Minimum net uptake coefficients [mL/(100 g min)] for [1-14C]hydroxybutyrate (means ± SD, n = 5) were determined autoradiographically at 5 min after an intravenous pulse injection of tracer (see Methods). Right/left ratios are means ± SD of the ratios calculated for each structure in each animal.

Auditory structures
Inferior colliculus1.6 ± 0.21.6 ± 0.21.00 ± 0.02
Medial geniculate1.6 ± 0.21.6 ± 0.30.96 ± 0.04
Auditory cortex1.7 ± 0.21.8 ± 0.20.95 ± 0.04
Other structures
Sensorimotor cortex1.7 ± 0.31.7 ± 0.31.00 ± 0.02
Sensory cortex1.7 ± 0.21.7 ± 0.20.98 ± 0.06
Thalamus1.4 ± 0.21.4 ± 0.20.98 ± 0.05
Caudate1.2 ± 0.21.2 ± 0.21.00 ± 0.02
Corpus callosum0.8 ± 0.20.8 ± 0.21.00 ± 0.01


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Acetate is a normal constituent of blood derived mainly from metabolic activity of bacteria in the digestive tract and, along with other short-chain fatty acids, acetate can provide a large fraction of the caloric requirements of various animal species, especially ruminants (Knowles et al. 1974; McNeil 1984; Bergman 1990). Acetate is readily taken up into brain and rapidly metabolized via the small glutamate compartment to label tricarboxylic acid (TCA) cycle-derived amino acids and yield a glutamine RSA > 1 (see Introduction). In our studies 90–98% of the 14C is in brain in metabolites, with 60–70% in amino acids and a glutamine RSA of 2–5 at 5 min after a pulse intravenous injection of [1-14C]- or 2-14C]acetate (Dienel et al. 2001a,b; Table 4). Glutamate is a neurotransmitter in the inferior colliculus (Faingold et al. 1989; Hu et al. 1994) and acoustic stimulation increases synaptic activity and CMRglc in this structure (Sharp et al. 1981; Webster et al. 1985; Nudo and Masterton 1986). The constancy of the glutamine RSA after acoustic activation of the inferior colliculus suggests that the rate of labeling of glutamine by [2-14C]acetate matches the (presumed) increase in conversion of [14C]glutamine to neurotransmitter [14C]glutamate via the glutamate–glutamine cycle during acoustic stimulation as well as the estimated 6–10% direct labeling of glutamate via the large glutamate pool by metabolism of [14C]glucose (Table 5 footnote e). There probably was some labeling of brain by [14C]glucose, but within the limits of analysis in the present study, even the large rise in CMRglc relative to CMRacetate during stimulation in the inferior colliculus [14C]glucose did not appear to significantly interfere with the autoradiographic assay or analysis of relative labeling of glutamine and glutamate by [2-14C]acetate. Similar conclusions were drawn for the first 15 min after an intraperitoneal injection of [1-14C]- or [2-14C]acetate in mice (Van den Berg et al. 1969; Van den Berg and Garfinkel 1971; Van den Berg and Ronda 1976a).

The present study extends previous autoradiographic assays to localize and assess acetate metabolism in brain in vivo (Muir et al. 1986; Lear and Ackermann 1990), and demonstrates that metabolic activation of astrocytes by sensory stimulation is detectable with [2-14C]acetate in vivo. Oxidative metabolism of blood-borne acetate may contribute significantly to energy production in working astrocytes even though the magnitude of the calculated acetate utilization rate during rest and the incremental increase in CMRacetate elicited by acoustic stimulation are much smaller than the corresponding values for CMRglc. The energy yield from acetate can be substantial and biologically significant for working astrocytes because calculated rates of acetate oxidation are within the range of estimates of oxidative metabolism of glucose in astrocytes in vivo (Table 5). From these data it can be inferred that assessment of the energetics of working astrocytes based only on glucose metabolism could substantially underestimate the ATP generated by the astrocytic TCA cycle.

The energy contribution of acetate oxidation relative to that of glucose would rise during activation if glycolysis were enhanced, as suggested by the fall in the CMRO2/CMRglc ratio during activation in many experimental conditions (Dienel and Hertz, 2001; Dienel and Cruz 2004). For example, the apparently discrepant labeling by DG and hydroxybutyrate (see below for more detailed discussion) could arise from a large increase in glycolysis during acoustic activation in neurons, astrocytes, or both cell types. This possibility is consistent with our on-going (unpublished) studies of acoustic activation in which labeling of the inferior colliculus by [6-14C]glucose increases by only 20%, compared to the 70% rise registered by [14C]DG (Table 2). Lack of label retention suggests product loss, and the lactate level in the extracellular fluid of inferior colliculus does rise during acoustic stimulation. Lactate loss might contribute to underestimation of the rise in CMRglc with [6-14C]glucose compared to [14C]DG. If glycolytic lactate production accounts for a substantial fraction of the additional glucose consumed in the inferior colliculus during acoustic stimulation over and above the ‘resting’ level, the yield of ATP from glycolysis would be much less than from acetate. For example, if half of the net rise in CMRglc is glycolytic [0.5 × 50 µmol/(100 g min), Table 2] the ATP yield would be 50 µmol/(100 g min), which is only about one-third of that generated by oxidation of acetate [i.e. 13.6 µmol acetate oxidized/(100 g min) × 10 ATP per mole acetate]. In addition to acetate, other substrates metabolized by astrocytes can contribute to their energy budget under different conditions, especially glycogen (Dienel and Cruz 2003, 2004). Other short- and medium-chain fatty acids present in small amounts in blood and tissue that are quickly oxidized via the small glutamate pool in vivo (e.g. butyrate, propionate, octanoate; Cremer et al. 1977; Ebert et al. 2003) might also be minor energy sources for astrocytes in vivo. Increased oxidative metabolism in activated astrocytes in vivo is consistent with results of studies of cultured astrocytes showing stimulation of their oxygen consumption by increased extracellular potassium (Hertz 1982; Hertz et al. 1986) and glutamate (Eriksson et al. 1995) and greater oxidative metabolism of glucose, glutamate, and other compounds under stimulatory conditions and glutamate loading (e.g. Yu et al. 1982; Hertz and Peng 1992; McKenna et al. 1996; Peng et al. 2001). Further investigation of astrocyte metabolism during brain activation in vivo is important because if astrocytic energy production is much higher than generally recognized it is likely that unidentified ATP-requiring processes participate in functional activation of astrocytes during brain activation, and these need to be included in working models of neuron–astrocyte interactions.

Ketone bodies, like glucose, preferentially label the large (neuronal) glutamate pool in vivo during short (3–20 min) labeling periods after an injection of tracer amounts of substrate (see Introduction). Overall CMRglc in the inferior colliculus rose markedly during acoustic stimulation, and lack of registration of increased metabolism by [1-14C]hydroxybutyrate in auditory structures was therefore unexpected. Failure to detect metabolic activation with [1-14C]hydroxybutyrate could arise for a number of reasons.

  • i. 
    Oxidative metabolism in the neurons that take up and oxidize hydroxybutyrate under the resting condition does not increase in response to the auditory stimulus. This is probably unlikely because labeling of glutamate in cerebral cortex by [6-14C]glucose rose during generalized sensory stimulation, consistent with higher flux of glucose into the oxidative pathway in activated neurons (Dienel et al. 2002). However, as discussed above, glycolysis might be the primary pathway activated under the conditions of the present study.
  • ii. 
    Metabolism of hydroxybutyrate might be slow compared to glucose and acetate so that an increase in labeling is not registered in brief activation experiments, but incorporation of 14C from hydroxybutyrate into amino acids appears to be comparable to that of glucose and acetate in adult rat brain. At 3 min after an intravenous injection of tracer amounts of [3-14C]hydroxybutyrate, 50% of the total 14C in brain was recovered in amino acids (Cremer 1971), and additional label would be in TCA cycle carboxylic acids. At 5 min after pulse labeling with [6-14C]glucose, 38–45% (the range represents values at rest and stimulation; Dienel et al. 2002) of the total 14C in brain was recovered in amino acids, and 60–70% is in amino acids at 5 min after labeling with [1-14C]- or [2-14C]acetate (Dienel et al. 2001a,b).
  • iii. 
    If metabolism of [1-14C]hydroxybutyrate rises during activation, 14CO2 loss via TCA cycle decarboxylation reactions might be fast enough so that product trapping is incomplete and an increased oxidative flux is not registered by dilution of label into the large unlabeled amino acid pools. Complete trapping in amino acid pools requires fast exchange of label from the TCA cycle intermediates to amino acids and rapid mitochondrial–cytoplasmic pool mixing reactions. Some label loss within 5 min is likely because [1-14C]hydroxybutyrate is metabolized to [1-14C]acetyl CoA, which is also formed from [2-14C]glucose. Label from [2-14C]glucose is lost from brain more rapidly than that from [1-14C]- or 6-14C]glucose in brief experiments (Hawkins et al. 1985; Lear and Ackermann 1988). These findings are also consistent with lower (by about 30%) retention of label from [1-14C]acetate compared to [2-14C]acetate in 5-min labeling experiments (Dienel et al. 2001a). Taken together, the above results support the notion that exchange and mixing reactions in both neurons and astrocytes may not be rapid enough to quantitatively trap labeled products of [1-14C]acetate and [1-14C]hydroxybutyrate. If exchange-mixing were very fast relative to the TCA cycle rate, nearly all of the label entering the TCA cycle should exchange on the first turn of the TCA cycle and product trapping would be expected to be similar for [1-14C]- or 6-14C]glucose compared to [2-14C]glucose and for [1-14C]acetate compared to [2-14C]acetate. This equivalence is not, however, observed.
  • iv. 
    Finally, endogenous substrates (e.g. pyruvate and lactate) derived from metabolism of blood glucose and brain glycogen might compete with the [14C]hydroxybutyrate for transport into brain cells or into mitochondria, thereby preventing increased labeling during activation. Experimental resolution of the above issues and further use of labeled compounds preferentially metabolized in astrocytes and neurons should help improve both the understanding and modeling of astrocyte–neuron interactions in vivo.


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

This work was supported, in part, by National Science Foundation grant IBN-9728171 and NIH grants NS36728 and NS38230 (GD) and HD16596 (HRZ).


  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  • Adachi K., Cruz N. F., Sokoloff L. and Dienel G. A. (1995) Labeling of metabolic pools by [6-14C]glucose during K+-induced stimulation of glucose utilization in rat brain. J. Cereb. Blood Flow Metab. 15, 97110.
  • Bergman E. N. (1990) Energy contributions of volatile fatty acids from the gastrointestinal tract in various species. Physiol. Rev. 70, 567590.
  • Berl S. and Frigyesi T. L. (1968) Metabolism of [14C]leucine and [14C]acetate in sensorimotor cortex, thalamus, caudate nucleus and cerebellum of the cat. J. Neurochem. 15, 965970.
  • Buckley B. M. and Williamson D. H. (1977) Origins of blood acetate in the rat. Biochem. J. 166, 539545.
  • Cerdan S., Künnecke B. and Seelig J. (1990) Cerebral metabolism of [1,2–13C2]acetate as detected by in vivo and in vitro13C NMR. J. Biol. Chem. 265, 12 91612 926.
  • Cremer J. E. (1971) Incorporation of label from d-β-hydroxyl[14C]butyrate and [3-14C]acetoacetate into amino acids in rat brain in vivo. Biochem. J. 122, 135138.
  • Cremer J. E. and Heath D. F. (1974) The estimation of rates of utilization of glucose and ketone bodies in the brain of the suckling rat using compartmental analysis of isotopic data. Biochem. J. 142, 527544.
  • Cremer J. E., Braun L. D. and Oldendorf W. H. (1976) Changes during development in transport processes of the blood–brain barrier. Biochim. Biophys. Acta 448, 633637.
  • Cremer J. E., Teal H. M., Heath D. F. and Cavanagh J. B. (1977) The influence of portocaval anastomosis on the metabolism of labeled octanoate, butyrate, and leucine in rat brain. J. Neurochem. 28, 215222.
  • Cremer J. E., Sarna C. S., Teal H. M. and Cunningham V. J. (1978) Amino acid precursors: Their transport into brain and initial metabolism, in Amino Acids as Chemical Transmitters. NATO Advanced Study Institutes series: Series A, Life Sciences (FonnumF., ed.), Vol. 16, pp. 669689. Plenum Press, New York.
  • Cremer J. E., Cunningham V. J., Pardridge W. M., Braun L. D. and Oldendorf W. H. (1979) Kinetics of blood–brain barrier transport of pyruvate, lactate and glucose in suckling, weanling and adult rats. J. Neurochem. 33, 439445.
  • DeVivo D. C., Leckie M. P. and Agrawal H. C. (1975) d-beta-Hydroxybutyrate: a major precursor of amino acids in developing rat brain. J. Neurochem. 25, 161170.
  • Dienel G. A. and Cruz N. F. (2003) Neighborly interactions of metabolically-activated astrocytes in vivo. Neurochem. Int. 43, 339354.
  • Dienel G. A. and Cruz N. F. (2004) Nutrition during brain activation: does cell-to-cell lactate shuttling contribute significantly to sweet and sour food for thought? Neurochem. Int. 45, 321351.
  • Dienel G. A. and Hertz L. (2001) Glucose and lactate metabolism during brain activation. J. Neurosci. Res. 66, 824838.
  • Dienel G. A., Liu K. and Cruz N. F. (2001a) Local uptake of 14C-labeled acetate and butyrate in rat brain in vivo during spreading cortical depression. J. Neurosci. Res. 66, 812820.
  • Dienel G. A., Popp D., Drew P. D., Ball K., Krisht A. and Cruz N. F. (2001b) Preferential labeling of glial and meningial brain tumors with [2–14C]acetate. J. Nucl. Med. 42, 12431250.
  • Dienel G. A., Wang R. Y. and Cruz N. F. (2002) Generalized sensory stimulation of conscious rats increases labeling of oxidative pathways of glucose metabolism when the brain glucose-oxygen uptake ratio rises. J. Cereb. Blood Flow Metab. 22, 14901502.
  • Ebert D., Haller R. G. and Walton M. E. (2003) Energy contribution of octanoate to intact rat brain metabolism measured by 13C nuclear magnetic resonance spectroscopy. J. Neurosci. 23, 59285935.
  • Edmond J. (1992) Energy metabolism in developing brain cells. Can. J. Physiol. Pharmacol. 70, S118S129.
  • Edmond J., Robbins R. A., Bergstrom J. D., Cole R. A. and De Vellis J. (1987) Capacity for substrate utilization in oxidative metabolism by neurons, astrocytes, and oligodendrocytes from developing brain in primary culture. J. Neurosci. Res. 18, 551561.
  • Enerson B. E. and Drewes L. R. (2003) Molecular features, regulation, and function of monocarboxylate transporters: implications for drug delivery. J. Pharm. Sci. 92, 15311544.
  • Eriksson G., Peterson A., Iverfeldt K. and Walum E. (1995) Sodium-dependent glutamate uptake as an activator of oxidative metabolism in primary astrocyte cultures from newborn rat. Glia 15, 152156.
  • Faingold C. L., Hoffmann W. E. and Caspary D. M. (1989) Effects of excitant amino acids on acoustic responses of inferior colliculus neurons. Hear. Res. 40, 127136.
  • Fink K., Cline R. E. and Fink R. M. (1963) Paper chromatography of several classes of compounds. Anal. Chem. 35, 389398.
  • Gruetter R. (2004) Principles of the measurement of neuro-glial metabolism using in vivo 13C NMR spectroscopy. Adv. Mol. Cell. Biol. 31, 409433.
  • Hassel B., Bachelard H., Jones P., Fonnum F. and Sonnewald U. (1997) Trafficking of amino acids between neurons and glia in vivo. Effects of inhibition of glial metabolism by fluoroacetate. J. Cereb. Blood Flow Metab. 17, 12301238.
  • Hawkins R. A., Mans A. M., Davis D. W., Vina J. R. and Hibbard L. S. (1985) Cerebral glucose use measured with [14C]glucose labeled in the 1, 2, or 6 position. Am. J. Physiol. 248, C170C176.
  • Hawkins R. A., Mans A. M. and Davis D. W. (1986) Regional ketone body utilization by rat brain in starvation and diabetes. Am. J. Physiol. 250, E169E178.
  • Hertz L. (1982) Astrocytes, in Handbook of Neurochemistry (LajthaA., ed.) 2nd edn, Vol. 1,Chemical and Cellular Architecture, pp. 319355. Plenum Press, New York.
  • Hertz L. and Peng L. (1992) Energy metabolism at the cellular level of the CNS. Can. J. Physiol. Pharmacol. 70, S145S157.
  • Hertz E., Shargool M. and Hertz L. (1986) Effects of barbiturates on energy metabolism by cultured astrocytes and neurons in the presence of normal and elevated concentrations of potassium. Neuropharmacology 25, 533539.
  • Hetenyi G. Jr, Lussier B., Ferrarotto C. and Radziuk J. (1982) Calculation of the rate of gluconeogenesis from the incorporation of 14C atoms from labelled bicarbonate or acetate. Can. J. Physiol. Pharmacol. 60, 16031609.
  • Hu B., Senatorov V. and Mooney D. (1994) Lemniscal and non-lemniscal synaptic transmission in rat auditory thalamus. J. Physiol. 479, 217231.
  • Knowles S. E., Jarrett I. G., Filsell O. H. and Ballard F. J. (1974) Production and utilization of acetate in mammals. Biochem. J. 142, 401411.
  • Künnecke B., Cerdan S. and Seelig J. (1993) Cerebral metabolism of [1,2-13C2]glucose and [U-13C4]3-hydroxybutyrate in rat brain as detected by 13C NMR spectroscopy. NMR Biomed. 6, 264277.
  • LaManna J. C., Harrington J. F., Vendell L. M., Abi-Saleh K., Lust W. D. and Harik S. I. (1993) Regional blood–brain lactate influx. Brain Res. 614, 164170.
  • Lear J. L. and Ackermann R. F. (1988) Comparison of cerebral glucose metabolic rates measured with fluorodeoxyglucose and glucose labeled in the 1, 2, 3–4, and 6 positions using double label quantitative digital autoradiography. J. Cereb. Blood Flow. Metab. 8, 575585.
  • Lear J. L. and Ackermann R. F. (1990) Evaluation of radiolabeled acetate and fluoroacetate as potential tracers of cerebral oxidative metabolism. Metab. Brain Dis 5, 4556.
  • Lebon V., Petersen K. F., Cline G. W., Shen J., Mason G. F., Dufour S., Behar K. L., Shulman G. I. and Rothman D. L. (2002) Astroglial contribution to brain energy metabolism in humans revealed by 13C nuclear magnetic resonance spectroscopy: elucidation of the dominant pathway for neurotransmitter glutamate repletion and measurement of astrocytic oxidative metabolism. J. Neurosci. 22, 15231531.
  • Martinez-Hernandez A., Bell K. and Norenberg M. (1977) Glutamine synthetase: glial localization in brain. Science 195, 13561358.
  • McKenna M. C., Sonnewald U., Huang X., Stevenson J. and Zielke H. R. (1996) Exogenous glutamate concentration regulates the metabolic fate of glutamate in astrocytes. J. Neurochem. 66, 386393.
  • McNeil N. I. (1984) The contribution of the large intestine to energy supplies in man. Am. J. Clin. Nutr. 39, 338342.
  • Minchin M. C. W. and Beart P. M. (1975) Compartmentation of amino acid metabolism in the rat dorsal root ganglion: a metabolic and autoradiographic study. Brain Res. 83, 437449.
  • Muir D., Berl S. and Clarke D. D. (1986) Acetate and fluoroacetate as possible markers for glial metabolism in vivo. Brain Res. 380, 336340.
  • Nudo R. J. and Masterton R. B. (1986) Stimulation-induced [14C]2-deoxyglucose labeling of synaptic activity in the central auditory system. J. Comp. Neurol. 245, 553565.
  • O'Neal R. M. and Koeppe R. E. (1966) Precursors in vivo of glutamate, aspartate and their derivatives of rat brain. J. Neurochem. 13, 835847.
  • O'Neal R. M., Koeppe R. E. and Williams E. L. (1966) Utilization in vivo of glucose and volatile fatty acids by sheep brain for the synthesis of acidic amino acids. Biochem. J. 101, 591597.
  • Oldendorf W. H. (1973) Carrier-mediated blood–brain barrier transport of short-chain monocarboxylic organic acids. Am. J. Physiol. 224, 14501453.
  • Ottersen O. P., Zhang N. and Walberg F. (1992) Metabolic compartmentation of glutamate and glutamine: morphological evidence obtained by quantitative immunocytochemistry in rat cerebellum. Neuroscience 46, 519534.
  • Pan J. W., De Graaf R. A., Petersen K. F., Shulman G. I., Hetherington H. P. and Rothman D. L. (2002) [2,4-13C2]-beta-Hydroxybutyrate metabolism in human brain. J. Cereb. Blood Flow Metab. 22, 890898.
  • Patel A. J. and Balazs R. (1970) Manifestation of metabolic compartmentation during the maturation of the rat brain. J. Neurochem. 17, 955971.
  • Peng L., Swanson R. A. and Hertz L. (2001) Effects of l-glutamate, d-aspartate, and monensin on glycolytic and oxidative glucose metabolism in mouse astrocyte cultures: further evidence that glutamate uptake is metabolically driven by oxidative metabolism. Neurochem. Int. 38, 437443.
  • Pouteau E., Maugere P., Darmaun D., Marchini J. S., Piloquet H., Dumon H., Nguyen P. and Krempf M. (1998) Role of glucose and glutamine synthesis in the differential recovery of 13CO2 from infused [2–13C]- versus [1–13C]acetate. Metabolism 47, 549554.
  • Sarna C. S., Bradbury M. W. B., Cremer J. E., Lai J. C. K. and Teal H. M. (1979) Brain metabolism and specific transport at the blood–brain barrier after portacaval anastomosis in the rat. Brain Res. 160, 6983.
  • Shank R. P., Leo G. C. and Zielke H. R. (1993) Cerebral metabolic compartmentation as revealed by nuclear magnetic resonance analysis of d-[1-13C]glucose metabolism. J. Neurochem. 61, 315323.
  • Sharp F. R., Ryan A. F., Goodwin P. and Woolf N. K. (1981) Increasing intensities of wide band noise increase [14C]2-deoxyglucose uptake in gerbil central auditory structures. Brain Res. 230, 8796.
  • Shulman R. G., Rothman D. L., Behar K. L. and Hyder F. (2004) Energetic basis of brain activity: implications for neuroimaging. Trends Neurosci. 27, 489495.
  • Sokoloff L. (1986) Cerebral circulation, energy metabolism, and protein synthesis. General characteristics and principles of measurement, in Positron Emission Tomography and Autoradiography: Principles and Applications for the Brain and Heart (PhelpsM., MazziottaJ. and SchelbertH., eds), pp. 171. Raven Press, New York.
  • Sokoloff L., Reivich M., Kennedy C., Des Rosiers M. H., Patlak C. S., Pettigrew K. D., Sakurada O. and Shinohara M. (1977) The [14C]deoxyglucose method for the measurement of local glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat. J. Neurochem. 28, 897916.
  • Tyce G. M., Ogg J. and Owen C. A. Jr (1981) Metabolism of acetate to amino acids in brains of rats after complete hepatectomy. J. Neurochem. 36, 640650.
  • Van den Berg C. J. and Garfinkel D. (1971) A simulation study of brain compartments. Metabolism of glutamate and related substances in mouse brain. Biochem. J. 123, 211218.
  • Van den Berg C. J. and Ronda G. (1976a) The incorporation of double-labelled acetate into glutamate and related amino acids from adult mouse brain: compartmentation of amino acid metabolism in brain. J. Neurochem. 27, 14431448.
  • Van den Berg C. J. and Ronda G. (1976b) Metabolism of glutamate and related amino acids in the 10-day-old mouse brain. experiments with labelled acetate and beta-hydroxybutyrate. J. Neurochem. 27, 14491453.
  • Van den Berg C. J., Krzalic L., Mela P. and Waelsch H. (1969) Compartmentation of glutamate metabolism in brain. Evidence for the existence of two different tricarboxylic acid cycles in brain. Biochem. J. 113, 281290.
  • Wang R. Y. and Dienel G. A. (1994) Ion chromatography with suppressed conductivity detection: recoveries of 14C-labeled metabolites. Biotechniques 17, 106113.
  • Waniewski R. A. and Martin D. L. (1998) Preferential utilization of acetate by astrocytes is attributable to transport. J. Neurosci. 18, 52255233.
  • Webster W. R., Serviere J., Martin R. and Brown M. (1985) Uncrossed and crossed inhibition in the inferior colliculus of the cat. a combined 2-deoxyglucose and electrophysiological study. J. Neurosci. 5, 18201832.
  • Yu A. C., Schousboe A. and Hertz L. (1982) Metabolic fate of 14C-labeled glutamate in astrocytes in primary cultures. J. Neurochem. 39, 954960.