- Email: [email protected]
Effects of lactate on glucose-sensing neurons in the solitary tract nucleus
Physiology & Behavior 74 (2001) 391 – 397
Effects of lactate on glucose-sensing neurons in the solitary tract nucleus Thami Himmi1, Jean Perrin, Michel Dallaporta, Jean-Claude Orsini* CNRS-LNB 7 and FSS, Universite´ de la Me´diterrane´e, 31 chemin Joseph Aiguier, 13402 Marseille ce´dex 20, France Received 23 February 2001; received in revised form 9 April 2001; accepted 19 June 2001
Abstract For nervous tissue, lactate is a valuable energy substrate that can be extracted from glucose by astrocytes and released for neuronal use. Therefore, we hypothesized that the glucose-sensing neurons that signal the glycemic changes involved in the control of body energy homeostasis may be responsive to extracellular lactate as well. To test this hypothesis, neuronal activity was recorded extracellularly in the solitary tract nucleus of anesthetized rats in order to compare the effects of microelectrophoretic applications of glucose and lactate and of moderate hyperglycemia and to assess the possible effects of lactate on the response to glucose. About 90% of the investigated neurons behaved in a similar manner after local ejections of glucose and lactate. Among them, most neurons activated by glucose were also activated by lactate and all neurons depressed by glucose were also depressed by lactate. This result suggests that the response to these two compounds is mediated by a common mechanism related to their utilization as oxidizible substrates. In half of the tested neurons, the response to glucose was eliminated or significantly reduced after repeated lactate ejections. This inhibitory effect is a likely result of a modification in glucose metabolism induced by a high extracellular lactate level. Most glycemia-sensitive neurons responded similarly to moderate hyperglycemia and to local lactate ejection, suggesting that high brain lactate levels might interfere with the brain mechanisms that mediate glucoprivic eating. D 2001 Elsevier Science Inc. All rights reserved. Keywords: Glucose; Lactate; Energy metabolism; Eating; Medulla oblongata; Electrophysiology; Iontophoresis; Rat
1. Introduction Behavioral and metabolic mechanisms that ensure the whole body’s energy homeostasis are largely controlled by the glucose supply to the brain, which is detected by specific central glucose-sensing neurons, as demonstrated by the effects of brain glucoprivation induced by intracerebroventricular 2-deoxyglucose injection [13]. More precisely, feeding behavior as well as various related endocrine secretions can be triggered by central glucoprivation limited to the hindbrain, since feeding is elicited by 5-thioglucose after injection into the fourth ventricle when the cerebral aqueduct was obstructed. Injection into the lateral ventricle was without effect in this situation [18,21]. The hindbrain glucose sensors mediating these responses
* Corresponding author. Tel.: +33-4-91-16-45-81; fax: +33-4-91-2208-75. E-mail address: [email protected] (J.-C. Orsini). 1 Invited professor at the Universite´ de la Me´diterrane´e, on leave from the Universite´ Cadi Ayyad, Beni Mellal, Morocco.
may be located in the caudal part of the nucleus tractus solitarii (NTS), where cFos-like immunoreactivity has been detected following intracerebroventricular 2-deoxyglucose injection [2]. In our previous electrophysiological in vivo studies [30,31], glycemia-sensitive neurons in this area were found to respond to moderate induced glycemic changes similar to those occurring under physiological conditions, i.e., preprandial hypoglycemia and postprandial hyperglycemia [10]. The sensitivity of NTS neurons to glycemic changes is most likely mediated by a local glucose-sensing mechanism, since they display similar responses to systemic and microelectrophoretic glucose applications [30]. For a majority of these glucose-sensing neurons, modulation of K-ATP channels is probably involved in the response to glucose [3], which thus reflects their cellular metabolic rate. Therefore, they may sense any other energy substrate available in their environment—that is able to be uptaken and metabolized by them. This is true of lactate, which, in the nervous tissue of vertebrates, has long been considered as a useless end-product of anaerobic glycolysis. There is now growing evidence from in vitro studies that it may be an alternative substrate for neuronal energy metabolism and
0031-9384/01/$ – see front matter D 2001 Elsevier Science Inc. All rights reserved. PII: S 0 0 3 1 - 9 3 8 4 ( 0 1 ) 0 0 5 7 3 - X
392
T. Himmi et al. / Physiology & Behavior 74 (2001) 391–397
can even be preferred over glucose. Thus, when both glucose and lactate are supplied at a concentration of 5 mM to excised sympathetic ganglia, lactate utilization is three times greater than glucose utilization [8]. In the retina, lactate extracted from glucose and released by Mu¨ller glial cells is metabolized by photoreceptors [19]. Normal synaptic function in hippocampal slices can be restored by exogenous lactate after glucose deprivation [25], and its recovery after hypoxia, in the presence of glucose, requires the release of lactate by glial cells [24]. In the same preparation, glial lactate also enables neurons to endure activation after excitation by glutamate [23]. In vivo measures in the rat hippocampus using a microsensor [4] and in the human visual cortex using magnetic resonance spectroscopy [20] have shown that the endogenous production of lactate in the CNS is increased in response to local neuronal activation. A model of brain energy supply has been recently proposed in which astrocytes take up glucose and export lactate, which is utilized by neurons as an aerobic substrate; this process is activity dependent [17]. Insofar as this model can be applied to NTS glucose-sensing neurons, their possible sensitivity to the local lactate level should allow them to signal the energy available in the brain tissue more efficiently. The present study was undertaken in an attempt to find support for this assumption. We identified NTS glucosesensing neurons responding to local glucose ejection, and among them, glycemia-sensitive neurons responding to moderate changes in glycemia, and we tested the firing rate of these cells to see whether it was also affected by local lactate ejection. Considering that the lactate level can dosedependently modify glucose utilization by nervous tissue [9], we also assessed the possible impact of lactate administration on the response to glucose.
tions of NaCl with the same osmolarity were systematically performed. 2.2. Microelectrophoresis Microelectrophoretic applications with calibrated currents (Bionic instrument SI 2101A and SI 2001A units) were performed using seven-barrel glass capillary pipettes (Clark, GC 150F) pulled in two stages using a Narishige PE2 vertical puller and broken to a tip diameter of 12– 14 mm. Two barrels were filled with 0.15 M NaCl for current balancing and iontophoretic ejection of sodium ions, respectively. Two other barrels were filled with glucose 0.5 M and L-lactic acid sodium salt 0.3 M (Sigma). Lactate ion was ejected by iontophoresis, and glucose by what is referred to as electroosmosis [15,27]. Retaining currents were routinely used. Control ejections of Na + and Cl were always carried out to confirm the specificity of the responses to glucose and lactate, respectively. 2.3. Single-unit recording The skull of the rat was fixed in a stereotaxic headholder according to the atlas of Paxinos and Watson [16]. A limited area of the mediodorsal part of the interparietal bone over the cerebellum was exposed and the multibarreled pipette was lowered to reach the subpostremal NTS. The dorsal and ventral limits of the NTS were located by monitoring the neuronal activity in the cerebellar cortex and dorsal vagal nucleus or hypoglossal nucleus. Extracellular action potentials were recorded either through the central barrel of the seven-barrel capillary pipette (4 – 14 MV) or through a tungsten electrode (4 – 5 MV) glued to the multibarrel
2. Methods 2.1. Animals and operative procedures Experiments were conducted on 20 adult male Wistar rats weighing 238 ± 48 g (mean ± S.D.). Anesthesia was induced by intramuscular injection of a mixture of 10 mg/ 100 g of ketamine hydrochloride (Imalge`ne 1000, RhoˆneMe´rieux) and 1.5 mg/100 g of xylazine (Rompun, Bayer). It was maintained by continuous infusion of ketamine hydrochloride into a femoral vein (10 mg/ml; flowrate: 6.7 –13.3 ml/min). Ketamine was chosen because it has been reported to have only minor effects on glycemia [1]. Rectal temperature was monitored by a thermistor and maintained between 36.5C and 37.5C by means of an electrically heated blanket. A catheter was inserted into the left jugular vein for a 4 min intravenous infusion of 20 mg glucose dissolved in 0.2 ml of saline. This infusion has been previously found to induce blood glucose level similar to postprandial hyperglycemia [31]. Control injec-
Fig. 1. Coronal section through the subpostremal NTS showing the location of the 41 recorded neurons. AP, area postrema; c, central canal; Cu, nucleus cuneatus; Gr, nucleus gracilis; ts, tractus solitarii; 10, motor nucleus of the vagus nerve; 12, nucleus of the hypoglossal nerve.
T. Himmi et al. / Physiology & Behavior 74 (2001) 391–397 Table 1 Number of NTS neurons (out of 41) displaying each type of response to local applications of glucose and lactate Response to glucose Response to lactate
Depressed
Activated
No response
Total
Depressed Activated No response Total
3 0 0 3
1 18 2 21
0 1 16 17
4 19 18 41
pipette. The number of spikes per 5 s was counted by a timer on an interface adapter board (National Instruments) in a computer (Macintosh, Apple), and the time course of the firing frequency was plotted on the screen.
393
2.4. Histological control procedure To confirm the location of the studied cells, the last recording site for the tungsten electrode was marked at the end of the experiment by passing an 8 –9 mA negative current through for 8 –10 s; for the glass electrode, this was achieved by passing a 50 mA negative current through for 10 –30 min (after replacing the barrel solution with concentrated HCl) [11]. Both procedures resulted in a small lesion around the tip of the electrode. The animal was intracardially perfused with 10% formalin and the brainstem was removed and stored for at least 24 h in the same fixative solution. Serial coronal sections (50 mm thick) were cut using a vibratome and stained with cresyl violet.
Fig. 2. Frequency – time histograms of a NTS neuron activated by local ejections of both glucose and lactate. The horizontal bars above the histograms represent the ejection duration and the numbers give the current intensity.
394
T. Himmi et al. / Physiology & Behavior 74 (2001) 391–397
2.5. Data analysis 2.5.1. Responses to local applications When computer graphs indicated modification of neuronal activity during and after the ejection and when this modification lasted more than 2 min, the mean discharge frequencies during this period of modified activity and the last 5 min before ejection were compared using Student’s t test. Changes in neuronal firing frequencies were considered to be responses if they were significant (significance level P < .05) and if they exceeded 30%. 2.5.2. Responses to intravenous glucose administration When computer graphs indicated modification of neuronal activity within 8 min after the onset of glucose
infusion and when this modification lasted more than 9 min, the mean discharge frequencies during the period of modified activity and a similar period of time before the onset of glucose infusion were compared using Student’s t test (significance level P < .05) to determine whether the response was significant.
3. Results The electrical activity of 41 units was recorded in the study. The histological control (Fig. 1) confirmed that all of them were located in the caudal part of the NTS, at the level of the area postrema (subpostremal NTS).
Fig. 3. Example of a NTS glucose-sensing and glycemia-sensitive neuron responding to lactate. The firing rate was increased by local ejections of both glucose (A) and lactate (B) and was also activated after moderate hyperglycemia induced by intravenous glucose injection (C).
T. Himmi et al. / Physiology & Behavior 74 (2001) 391–397
3.1. Comparative effects of glucose and lactate applied locally The 41 neurons were tested with electrophoretic glucose and lactate applications. Of them, 37 reacted similarly to both applications: 18 were activated by glucose and lactate (Fig. 2), 3 were depressed by both products and 16 responded to neither of them. Four neurons reacted in an unexpected way: one was activated by glucose and depressed by lactate, two were activated by glucose only and one by lactate only (Table 1). 3.2. Response to lactate of some neurons identified as glycemia sensitive Thirteen neurons activated by local glucose applications were tested at the end of the experiment with weak hyperglycemia induced by an intravenous glucose injection. Whereas five cells failed to respond to the intravenous glucose injection, the other eight were found to be glycemia
395
sensitive and were activated by both local and intravenous glucose administration. After the intravenous glucose injection, their firing frequency increased by 148 ± 29% (mean ± S.D.). The response had a latency of 290 ± 25 s and a duration of 979 ± 137 s. Of the eight neurons, local lactate activated six (Fig. 3), depressed one and had no effect on the last one. 3.3. Alteration of the response to glucose by local lactate applications In 20 neurons, the response to local glucose was tested before and after repeated lactate applications. In 10 cells (Fig. 4), the response was eliminated (five cells) or significantly decreased (five cells), whereas it displayed no significant change in seven neurons and significantly increased in three others. Four of the five neurons in which the response to glucose was significantly reduced by lactate were subsequently tested with a glucose ejection during the continu-
Fig. 4. Frequency – time histograms of a NTS neuron responding to local lactate and glucose ejections. The response to a 200 nA glucose ejection was abolished after repeated lactate applications.
396
T. Himmi et al. / Physiology & Behavior 74 (2001) 391–397
ous infusion of a subthreshold dose of lactate (i.e., an iontophoretic current insufficient to alter significantly neuronal activity). In all cases, an additional reduction in the glucose response was observed.
4. Discussion More than 90% of the recorded neurons reacted in the same way to repeated local applications of glucose and lactate. In particular, of the 24 neurons responsive to glucose, the firing rate of 21 was modified in the same direction in response to both substances. Likewise, it was recently observed by others [29] that the ventromedial hypothalamic neurons activated by glucose were also activated by lactate. The similarity of the responses to glucose and lactate can be explained by the fact that both responses reflect the level of substrate uptake and oxidation and the intracellular ratio of ATP/ADP. A coupling between glucose-sensing and cellular energy metabolism does exist and is mediated by ATP-sensitive potassium channels, both in peripheral tissues such as pancreatic b-cells [22] and in various central nervous structures (reviewed in Miller [12]). Its role in the activation of NTS neurons by glucose is supported by our recent pharmacological data [3]. As neural tissue can be fueled by lactate as well as glucose, it is conceivable that ATP generated by both substrates modulates the electrical activity of a given NTS neuron in a similar way. This process, resulting in the same neuronal response to both metabolites, can be achieved at the end of two possible metabolic pathways: (1) the neuron itself oxidizes glucose and lactate; (2) the neuron is mainly fueled by lactate, the ejected glucose being processed by the neighboring astrocytes where it is converted into lactate, which is released and then taken up and oxidized by the neuron. The latter possibility is in line with the Magistretti model of brain energy supply [17]. It does not completely exclude the former pathway: both effects of glucose on the neuron, exerted directly and through glial lactate production, can contribute to neuronal responses to this carbohydrate. However, when large amounts of lactate are exported by glia, lactate becomes the main neuronal energy substrate [23] and may play the main role in the neuronal response to glucose ejection. In 10 of the 20 neurons in which the effect of glucose was tested after repeated lactate ejections, the response to glucose was partially or totally eliminated. This inhibitory effect of lactate loading most likely results from a decrease in total glucose utilization in both neurons and astrocytes [28], with lactate becoming their main energy substrate [14]. It might also be due to a shift of glucose metabolism in the glial cells from glycolysis and lactate production to glycogen synthesis: the conversion of glucose into CO2 is inhibited in incubated nervous tissue
after lactate loading [6,9] and its incorporation into glycogen is stimulated [6]. The mechanisms by which, in the NTS, lactate is released by glia, is uptaken and metabolized by glucosesensing neurons and affects their activity need to be analyzed in future in vitro investigations, for example, by testing the effect of lactate export inhibition on the neuronal response to glucose. Similar responses to lactate and glucose were observed in most of the neurons that responded to moderate induced hyperglycemia. These glycemia-sensitive neurons may be involved in the feeding response to the preprandial hypoglycemia since their activity is modulated not only by hyperglycemia but also by hypoglycemia in the physiological range [30]. Therefore, lactate sensitivity might interfere with the brain mechanism that initiates eating, particularly when the lactate/glucose ratio is increased in the extracellular brain fluid. Food intake is inhibited after the postprandial rise in blood lactate, but the effect is mainly mediated by hepatic rather than central mechanisms [26], as hyperlactatemia per se does not evoke brain lactate uptake [5]. On the contrary, during intense physical exercise, blood lactate is also high, but its uptake by the brain is facilitated, probably due to an increase in cerebral metabolic demand [5]. Stimulation of brain glucose sensors by a high lactate level may then inhibit food intake and contribute to the ‘‘exercise-induced suppression of appetite’’ [7]. In summary, most glucose-sensing neurons recorded in the caudal NTS are similarly affected by local glucose and lactate: they can signal the brain concentrations of both substrates and hence the availability of energy in their environment. This finding and the observation that an increase in the extracellular lactate level often reduces neuronal sensitivity to glucose may reflect the relationship between the sensing mechanisms to both carbohydrates and cellular energy metabolism and the metabolic coupling between neurons and astrocytes. They also suggest the possible hypophagic effect of a high brain lactate level, for example, during the hyperlactatemia accompanying maximal muscular exercise.
References [1] Aynsley-Green A, Biebuyck JF, Alberti KGMM. Anesthesia and insulin secretion: the effects of diethyl ether, halothane, pentobarbitone sodium and ketamine hydrochloride on intravenous glucose tolerance and insulin secretion in the rat. Diabetologia 1973;9: 274 – 81. [2] Briski KP, Marshall ES. Caudal brainstem Fos expression is restricted to periventricular catecholamine neuron-containing loci following intraventricular administration of 2-deoxy-D-glucose. Exp Brain Res 2000;133:547 – 51. [3] Dallaporta M, Perrin J, Orsini JC. Involvement of adenosine triphosphate-sensitive K+ channels in glucose-sensing in the rat solitary tract nucleus. Neurosci Lett 2000;278:77 – 80. [4] Hu Y, Wilson GS. A temporary local energy pool coupled to neuronal activity: fluctuations of extracellular lactate levels in rat brain moni-
T. Himmi et al. / Physiology & Behavior 74 (2001) 391–397
[5]
[6]
[7]
[8] [9]
[10] [11] [12] [13]
[14]
[15] [16] [17]
[18]
tored with rapid-response enzyme-based sensor. J Neurochem 1997;69:1484 – 90. Ide K, Schmalbruch IK, Quistorff B, Horn A, Secher NH. Lactate, glucose and O-2 uptake in human brain during recovery from maximal exercise. J Physiol 2000;522:159 – 64 (London). Ide T, Steinke J, Cahill GF. Metabolic interactions of glucose, lactate, and beta-hydroxybutyrate in brain slices. Am J Physiol 1969;217: 784 – 92. King NA, Tremblay A, Blundell JE. Effects of exercise on appetite control: implications for energy balance. Med Sci Sports Exercise 1997;29:1076 – 89. Larrabee MG. Lactate metabolism and its effects on glucose metabolism in an excised neural tissue. J Neurochem 1995;64:1734 – 41. Larrabee MG. Partitioning of CO2 production between glucose and lactate in excised sympathetic ganglia, with implications for brain. J Neurochem 1996;67:1726 – 34. Louis-Sylvestre J, Le Magnen J. A fall in blood glucose level precedes meal onset in free feeding rats. Neurosci Biobehav Rev 1980;4:13 – 5. McCance I, Phillis JW. The location of electrode tips in nervous tissue. Experientia 1965;21:108 – 9. Miller RJ. Glucose-regulated potassium channels are sweet news for neurobiologists. Trends Neurosci 1990;13:197 – 9. Muller EE, Panerai A, Cocchi D, Frohman LA, Antegaza P. Central glucoprivation: some physiological effects induced by the intraventricular administration of 2-deoxy-D-glucose. Experientia 1973;29: 874 – 5. Murata T, Waki A, Omata N, Fujibayashi Y, Sadato N, Yano R, Yoshimoto M, Isaki K, Yonekura Y. Dynamic changes in glucose metabolism by lactate loading as revealed by a positron autoradiography technique using rat living brain slices. Neurosci Lett 1998;249: 155 – 8. Oomura Y, Ono T, Ooyama H, Wayner MJ. Glucose and osmosensitive neurons of the rat hypothalamus. Nature 1969;222:282 – 4. Paxinos G, Watson C. The rat brain in stereotaxic coordinates. 2nd ed. Sydney: Academic Press, 1986. Pellerin L, Pellegri G, Bittar PG, Charnay Y, Bouras C, Martin JL, Stella N, Magistretti PJ. Evidence supporting the existence of an activity-dependent astrocyte-neuron lactate shuttle. Dev Neurosci 1998;20:291 – 9. Penicaud L, Pajot MT, Thompson DA. Evidence that receptors con-
[19]
[20]
[21]
[22] [23]
[24]
[25] [26]
[27] [28]
[29]
[30]
[31]
397
trolling growth hormone and hyperglycemic responses to glucoprivation are located in the hindbrain. Endocr Res 1990;16:461 – 75. Poitry-Yamate CL, Poitry S, Tsacopoulos M. Lactate released by Muller glial cells is metabolized by photoreceptors from mammalian retina. J Neurosci 1995;15:5179 – 91. Prichard J, Rothman D, Novotny E, Petroff O, Kuwabara T, Avison M, Howseman A, Hanstock C, Shulman R. Lactate rise detected by 1H NMR in human visual cortex during physiologic stimulation. Proc Natl Acad Sci USA 1991;88:5829 – 31. Ritter RC, Slusser PG, Stone S. Glucoreceptors controlling feeding and blood glucose: location in the hindbrain. Science 1981;213: 451 – 3. Rorsman P. The pancreatic beta-cell as a fuel sensor: an electrophysiologist’s viewpoint. Diabetologia 1997;40:487 – 95. Schurr A, Miller JJ, Payne RS, Rigor BM. An increase in lactate output by brain tissue serves to meet the energy needs of glutamateactivated neurons. J Neurosci 1999;19:34 – 9. Schurr A, Payne RS, Miller JJ, Rigor BM. Glia are the main source of lactate utilized by neurons for recovery of function posthypoxia. Brain Res 1997;774:221 – 4. Schurr A, West CA, Rigor BM. Lactate-supported synaptic function in the rat hippocampal slice preparation. Science 1988;240:1326 – 8. Silberbauer CJ, Surina-Baumgartner DM, Arnold M, Langhans W. Prandial lactate infusion inhibits spontaneous feeding in rats. Am J Physiol 2000;278:R646 – 53. Stone TW. Microiontophoresis and pressure ejection. New York: Wiley, 1985. Tabernero A, Vicario C, Medina JM. Lactate spares glucose as a metabolic fuel in neurons and astrocytes from primary culture. Neurosci Res 1996;26:369 – 76. Yang XJ, Kow LM, Funabashi T, Mobbs CV. Hypothalamic glucose sensor — similarities to and differences from pancreatic beta-cell mechanisms. Diabetes 1999;48:1763 – 72. Yettefti K, Orsini JC, Perrin J. Characteristics of glycemia-sensitive neurons in the nucleus tractus solitarii. Possible involvement in nutritional regulation. Physiol Behav 1997;61:93 – 100. Yettefti K, Orsini JC, El Ouazzani T, Himmi T, Boyer A, Perrin J. Sensitivity of nucleus tractus solitarius neurons to induced moderate hyperglycemia, with special reference to catecholaminergic regions. J Auton Nerv Syst 1995;51:191 – 7.
Effects of lactate on glucose-sensing neurons in the solitary tract nucleus Thami Himmi1, Jean Perrin, Michel Dallaporta, Jean-Claude Orsini* CNRS-LNB 7 and FSS, Universite´ de la Me´diterrane´e, 31 chemin Joseph Aiguier, 13402 Marseille ce´dex 20, France Received 23 February 2001; received in revised form 9 April 2001; accepted 19 June 2001
Abstract For nervous tissue, lactate is a valuable energy substrate that can be extracted from glucose by astrocytes and released for neuronal use. Therefore, we hypothesized that the glucose-sensing neurons that signal the glycemic changes involved in the control of body energy homeostasis may be responsive to extracellular lactate as well. To test this hypothesis, neuronal activity was recorded extracellularly in the solitary tract nucleus of anesthetized rats in order to compare the effects of microelectrophoretic applications of glucose and lactate and of moderate hyperglycemia and to assess the possible effects of lactate on the response to glucose. About 90% of the investigated neurons behaved in a similar manner after local ejections of glucose and lactate. Among them, most neurons activated by glucose were also activated by lactate and all neurons depressed by glucose were also depressed by lactate. This result suggests that the response to these two compounds is mediated by a common mechanism related to their utilization as oxidizible substrates. In half of the tested neurons, the response to glucose was eliminated or significantly reduced after repeated lactate ejections. This inhibitory effect is a likely result of a modification in glucose metabolism induced by a high extracellular lactate level. Most glycemia-sensitive neurons responded similarly to moderate hyperglycemia and to local lactate ejection, suggesting that high brain lactate levels might interfere with the brain mechanisms that mediate glucoprivic eating. D 2001 Elsevier Science Inc. All rights reserved. Keywords: Glucose; Lactate; Energy metabolism; Eating; Medulla oblongata; Electrophysiology; Iontophoresis; Rat
1. Introduction Behavioral and metabolic mechanisms that ensure the whole body’s energy homeostasis are largely controlled by the glucose supply to the brain, which is detected by specific central glucose-sensing neurons, as demonstrated by the effects of brain glucoprivation induced by intracerebroventricular 2-deoxyglucose injection [13]. More precisely, feeding behavior as well as various related endocrine secretions can be triggered by central glucoprivation limited to the hindbrain, since feeding is elicited by 5-thioglucose after injection into the fourth ventricle when the cerebral aqueduct was obstructed. Injection into the lateral ventricle was without effect in this situation [18,21]. The hindbrain glucose sensors mediating these responses
* Corresponding author. Tel.: +33-4-91-16-45-81; fax: +33-4-91-2208-75. E-mail address: [email protected] (J.-C. Orsini). 1 Invited professor at the Universite´ de la Me´diterrane´e, on leave from the Universite´ Cadi Ayyad, Beni Mellal, Morocco.
may be located in the caudal part of the nucleus tractus solitarii (NTS), where cFos-like immunoreactivity has been detected following intracerebroventricular 2-deoxyglucose injection [2]. In our previous electrophysiological in vivo studies [30,31], glycemia-sensitive neurons in this area were found to respond to moderate induced glycemic changes similar to those occurring under physiological conditions, i.e., preprandial hypoglycemia and postprandial hyperglycemia [10]. The sensitivity of NTS neurons to glycemic changes is most likely mediated by a local glucose-sensing mechanism, since they display similar responses to systemic and microelectrophoretic glucose applications [30]. For a majority of these glucose-sensing neurons, modulation of K-ATP channels is probably involved in the response to glucose [3], which thus reflects their cellular metabolic rate. Therefore, they may sense any other energy substrate available in their environment—that is able to be uptaken and metabolized by them. This is true of lactate, which, in the nervous tissue of vertebrates, has long been considered as a useless end-product of anaerobic glycolysis. There is now growing evidence from in vitro studies that it may be an alternative substrate for neuronal energy metabolism and
0031-9384/01/$ – see front matter D 2001 Elsevier Science Inc. All rights reserved. PII: S 0 0 3 1 - 9 3 8 4 ( 0 1 ) 0 0 5 7 3 - X
392
T. Himmi et al. / Physiology & Behavior 74 (2001) 391–397
can even be preferred over glucose. Thus, when both glucose and lactate are supplied at a concentration of 5 mM to excised sympathetic ganglia, lactate utilization is three times greater than glucose utilization [8]. In the retina, lactate extracted from glucose and released by Mu¨ller glial cells is metabolized by photoreceptors [19]. Normal synaptic function in hippocampal slices can be restored by exogenous lactate after glucose deprivation [25], and its recovery after hypoxia, in the presence of glucose, requires the release of lactate by glial cells [24]. In the same preparation, glial lactate also enables neurons to endure activation after excitation by glutamate [23]. In vivo measures in the rat hippocampus using a microsensor [4] and in the human visual cortex using magnetic resonance spectroscopy [20] have shown that the endogenous production of lactate in the CNS is increased in response to local neuronal activation. A model of brain energy supply has been recently proposed in which astrocytes take up glucose and export lactate, which is utilized by neurons as an aerobic substrate; this process is activity dependent [17]. Insofar as this model can be applied to NTS glucose-sensing neurons, their possible sensitivity to the local lactate level should allow them to signal the energy available in the brain tissue more efficiently. The present study was undertaken in an attempt to find support for this assumption. We identified NTS glucosesensing neurons responding to local glucose ejection, and among them, glycemia-sensitive neurons responding to moderate changes in glycemia, and we tested the firing rate of these cells to see whether it was also affected by local lactate ejection. Considering that the lactate level can dosedependently modify glucose utilization by nervous tissue [9], we also assessed the possible impact of lactate administration on the response to glucose.
tions of NaCl with the same osmolarity were systematically performed. 2.2. Microelectrophoresis Microelectrophoretic applications with calibrated currents (Bionic instrument SI 2101A and SI 2001A units) were performed using seven-barrel glass capillary pipettes (Clark, GC 150F) pulled in two stages using a Narishige PE2 vertical puller and broken to a tip diameter of 12– 14 mm. Two barrels were filled with 0.15 M NaCl for current balancing and iontophoretic ejection of sodium ions, respectively. Two other barrels were filled with glucose 0.5 M and L-lactic acid sodium salt 0.3 M (Sigma). Lactate ion was ejected by iontophoresis, and glucose by what is referred to as electroosmosis [15,27]. Retaining currents were routinely used. Control ejections of Na + and Cl were always carried out to confirm the specificity of the responses to glucose and lactate, respectively. 2.3. Single-unit recording The skull of the rat was fixed in a stereotaxic headholder according to the atlas of Paxinos and Watson [16]. A limited area of the mediodorsal part of the interparietal bone over the cerebellum was exposed and the multibarreled pipette was lowered to reach the subpostremal NTS. The dorsal and ventral limits of the NTS were located by monitoring the neuronal activity in the cerebellar cortex and dorsal vagal nucleus or hypoglossal nucleus. Extracellular action potentials were recorded either through the central barrel of the seven-barrel capillary pipette (4 – 14 MV) or through a tungsten electrode (4 – 5 MV) glued to the multibarrel
2. Methods 2.1. Animals and operative procedures Experiments were conducted on 20 adult male Wistar rats weighing 238 ± 48 g (mean ± S.D.). Anesthesia was induced by intramuscular injection of a mixture of 10 mg/ 100 g of ketamine hydrochloride (Imalge`ne 1000, RhoˆneMe´rieux) and 1.5 mg/100 g of xylazine (Rompun, Bayer). It was maintained by continuous infusion of ketamine hydrochloride into a femoral vein (10 mg/ml; flowrate: 6.7 –13.3 ml/min). Ketamine was chosen because it has been reported to have only minor effects on glycemia [1]. Rectal temperature was monitored by a thermistor and maintained between 36.5C and 37.5C by means of an electrically heated blanket. A catheter was inserted into the left jugular vein for a 4 min intravenous infusion of 20 mg glucose dissolved in 0.2 ml of saline. This infusion has been previously found to induce blood glucose level similar to postprandial hyperglycemia [31]. Control injec-
Fig. 1. Coronal section through the subpostremal NTS showing the location of the 41 recorded neurons. AP, area postrema; c, central canal; Cu, nucleus cuneatus; Gr, nucleus gracilis; ts, tractus solitarii; 10, motor nucleus of the vagus nerve; 12, nucleus of the hypoglossal nerve.
T. Himmi et al. / Physiology & Behavior 74 (2001) 391–397 Table 1 Number of NTS neurons (out of 41) displaying each type of response to local applications of glucose and lactate Response to glucose Response to lactate
Depressed
Activated
No response
Total
Depressed Activated No response Total
3 0 0 3
1 18 2 21
0 1 16 17
4 19 18 41
pipette. The number of spikes per 5 s was counted by a timer on an interface adapter board (National Instruments) in a computer (Macintosh, Apple), and the time course of the firing frequency was plotted on the screen.
393
2.4. Histological control procedure To confirm the location of the studied cells, the last recording site for the tungsten electrode was marked at the end of the experiment by passing an 8 –9 mA negative current through for 8 –10 s; for the glass electrode, this was achieved by passing a 50 mA negative current through for 10 –30 min (after replacing the barrel solution with concentrated HCl) [11]. Both procedures resulted in a small lesion around the tip of the electrode. The animal was intracardially perfused with 10% formalin and the brainstem was removed and stored for at least 24 h in the same fixative solution. Serial coronal sections (50 mm thick) were cut using a vibratome and stained with cresyl violet.
Fig. 2. Frequency – time histograms of a NTS neuron activated by local ejections of both glucose and lactate. The horizontal bars above the histograms represent the ejection duration and the numbers give the current intensity.
394
T. Himmi et al. / Physiology & Behavior 74 (2001) 391–397
2.5. Data analysis 2.5.1. Responses to local applications When computer graphs indicated modification of neuronal activity during and after the ejection and when this modification lasted more than 2 min, the mean discharge frequencies during this period of modified activity and the last 5 min before ejection were compared using Student’s t test. Changes in neuronal firing frequencies were considered to be responses if they were significant (significance level P < .05) and if they exceeded 30%. 2.5.2. Responses to intravenous glucose administration When computer graphs indicated modification of neuronal activity within 8 min after the onset of glucose
infusion and when this modification lasted more than 9 min, the mean discharge frequencies during the period of modified activity and a similar period of time before the onset of glucose infusion were compared using Student’s t test (significance level P < .05) to determine whether the response was significant.
3. Results The electrical activity of 41 units was recorded in the study. The histological control (Fig. 1) confirmed that all of them were located in the caudal part of the NTS, at the level of the area postrema (subpostremal NTS).
Fig. 3. Example of a NTS glucose-sensing and glycemia-sensitive neuron responding to lactate. The firing rate was increased by local ejections of both glucose (A) and lactate (B) and was also activated after moderate hyperglycemia induced by intravenous glucose injection (C).
T. Himmi et al. / Physiology & Behavior 74 (2001) 391–397
3.1. Comparative effects of glucose and lactate applied locally The 41 neurons were tested with electrophoretic glucose and lactate applications. Of them, 37 reacted similarly to both applications: 18 were activated by glucose and lactate (Fig. 2), 3 were depressed by both products and 16 responded to neither of them. Four neurons reacted in an unexpected way: one was activated by glucose and depressed by lactate, two were activated by glucose only and one by lactate only (Table 1). 3.2. Response to lactate of some neurons identified as glycemia sensitive Thirteen neurons activated by local glucose applications were tested at the end of the experiment with weak hyperglycemia induced by an intravenous glucose injection. Whereas five cells failed to respond to the intravenous glucose injection, the other eight were found to be glycemia
395
sensitive and were activated by both local and intravenous glucose administration. After the intravenous glucose injection, their firing frequency increased by 148 ± 29% (mean ± S.D.). The response had a latency of 290 ± 25 s and a duration of 979 ± 137 s. Of the eight neurons, local lactate activated six (Fig. 3), depressed one and had no effect on the last one. 3.3. Alteration of the response to glucose by local lactate applications In 20 neurons, the response to local glucose was tested before and after repeated lactate applications. In 10 cells (Fig. 4), the response was eliminated (five cells) or significantly decreased (five cells), whereas it displayed no significant change in seven neurons and significantly increased in three others. Four of the five neurons in which the response to glucose was significantly reduced by lactate were subsequently tested with a glucose ejection during the continu-
Fig. 4. Frequency – time histograms of a NTS neuron responding to local lactate and glucose ejections. The response to a 200 nA glucose ejection was abolished after repeated lactate applications.
396
T. Himmi et al. / Physiology & Behavior 74 (2001) 391–397
ous infusion of a subthreshold dose of lactate (i.e., an iontophoretic current insufficient to alter significantly neuronal activity). In all cases, an additional reduction in the glucose response was observed.
4. Discussion More than 90% of the recorded neurons reacted in the same way to repeated local applications of glucose and lactate. In particular, of the 24 neurons responsive to glucose, the firing rate of 21 was modified in the same direction in response to both substances. Likewise, it was recently observed by others [29] that the ventromedial hypothalamic neurons activated by glucose were also activated by lactate. The similarity of the responses to glucose and lactate can be explained by the fact that both responses reflect the level of substrate uptake and oxidation and the intracellular ratio of ATP/ADP. A coupling between glucose-sensing and cellular energy metabolism does exist and is mediated by ATP-sensitive potassium channels, both in peripheral tissues such as pancreatic b-cells [22] and in various central nervous structures (reviewed in Miller [12]). Its role in the activation of NTS neurons by glucose is supported by our recent pharmacological data [3]. As neural tissue can be fueled by lactate as well as glucose, it is conceivable that ATP generated by both substrates modulates the electrical activity of a given NTS neuron in a similar way. This process, resulting in the same neuronal response to both metabolites, can be achieved at the end of two possible metabolic pathways: (1) the neuron itself oxidizes glucose and lactate; (2) the neuron is mainly fueled by lactate, the ejected glucose being processed by the neighboring astrocytes where it is converted into lactate, which is released and then taken up and oxidized by the neuron. The latter possibility is in line with the Magistretti model of brain energy supply [17]. It does not completely exclude the former pathway: both effects of glucose on the neuron, exerted directly and through glial lactate production, can contribute to neuronal responses to this carbohydrate. However, when large amounts of lactate are exported by glia, lactate becomes the main neuronal energy substrate [23] and may play the main role in the neuronal response to glucose ejection. In 10 of the 20 neurons in which the effect of glucose was tested after repeated lactate ejections, the response to glucose was partially or totally eliminated. This inhibitory effect of lactate loading most likely results from a decrease in total glucose utilization in both neurons and astrocytes [28], with lactate becoming their main energy substrate [14]. It might also be due to a shift of glucose metabolism in the glial cells from glycolysis and lactate production to glycogen synthesis: the conversion of glucose into CO2 is inhibited in incubated nervous tissue
after lactate loading [6,9] and its incorporation into glycogen is stimulated [6]. The mechanisms by which, in the NTS, lactate is released by glia, is uptaken and metabolized by glucosesensing neurons and affects their activity need to be analyzed in future in vitro investigations, for example, by testing the effect of lactate export inhibition on the neuronal response to glucose. Similar responses to lactate and glucose were observed in most of the neurons that responded to moderate induced hyperglycemia. These glycemia-sensitive neurons may be involved in the feeding response to the preprandial hypoglycemia since their activity is modulated not only by hyperglycemia but also by hypoglycemia in the physiological range [30]. Therefore, lactate sensitivity might interfere with the brain mechanism that initiates eating, particularly when the lactate/glucose ratio is increased in the extracellular brain fluid. Food intake is inhibited after the postprandial rise in blood lactate, but the effect is mainly mediated by hepatic rather than central mechanisms [26], as hyperlactatemia per se does not evoke brain lactate uptake [5]. On the contrary, during intense physical exercise, blood lactate is also high, but its uptake by the brain is facilitated, probably due to an increase in cerebral metabolic demand [5]. Stimulation of brain glucose sensors by a high lactate level may then inhibit food intake and contribute to the ‘‘exercise-induced suppression of appetite’’ [7]. In summary, most glucose-sensing neurons recorded in the caudal NTS are similarly affected by local glucose and lactate: they can signal the brain concentrations of both substrates and hence the availability of energy in their environment. This finding and the observation that an increase in the extracellular lactate level often reduces neuronal sensitivity to glucose may reflect the relationship between the sensing mechanisms to both carbohydrates and cellular energy metabolism and the metabolic coupling between neurons and astrocytes. They also suggest the possible hypophagic effect of a high brain lactate level, for example, during the hyperlactatemia accompanying maximal muscular exercise.
References [1] Aynsley-Green A, Biebuyck JF, Alberti KGMM. Anesthesia and insulin secretion: the effects of diethyl ether, halothane, pentobarbitone sodium and ketamine hydrochloride on intravenous glucose tolerance and insulin secretion in the rat. Diabetologia 1973;9: 274 – 81. [2] Briski KP, Marshall ES. Caudal brainstem Fos expression is restricted to periventricular catecholamine neuron-containing loci following intraventricular administration of 2-deoxy-D-glucose. Exp Brain Res 2000;133:547 – 51. [3] Dallaporta M, Perrin J, Orsini JC. Involvement of adenosine triphosphate-sensitive K+ channels in glucose-sensing in the rat solitary tract nucleus. Neurosci Lett 2000;278:77 – 80. [4] Hu Y, Wilson GS. A temporary local energy pool coupled to neuronal activity: fluctuations of extracellular lactate levels in rat brain moni-
T. Himmi et al. / Physiology & Behavior 74 (2001) 391–397
[5]
[6]
[7]
[8] [9]
[10] [11] [12] [13]
[14]
[15] [16] [17]
[18]
tored with rapid-response enzyme-based sensor. J Neurochem 1997;69:1484 – 90. Ide K, Schmalbruch IK, Quistorff B, Horn A, Secher NH. Lactate, glucose and O-2 uptake in human brain during recovery from maximal exercise. J Physiol 2000;522:159 – 64 (London). Ide T, Steinke J, Cahill GF. Metabolic interactions of glucose, lactate, and beta-hydroxybutyrate in brain slices. Am J Physiol 1969;217: 784 – 92. King NA, Tremblay A, Blundell JE. Effects of exercise on appetite control: implications for energy balance. Med Sci Sports Exercise 1997;29:1076 – 89. Larrabee MG. Lactate metabolism and its effects on glucose metabolism in an excised neural tissue. J Neurochem 1995;64:1734 – 41. Larrabee MG. Partitioning of CO2 production between glucose and lactate in excised sympathetic ganglia, with implications for brain. J Neurochem 1996;67:1726 – 34. Louis-Sylvestre J, Le Magnen J. A fall in blood glucose level precedes meal onset in free feeding rats. Neurosci Biobehav Rev 1980;4:13 – 5. McCance I, Phillis JW. The location of electrode tips in nervous tissue. Experientia 1965;21:108 – 9. Miller RJ. Glucose-regulated potassium channels are sweet news for neurobiologists. Trends Neurosci 1990;13:197 – 9. Muller EE, Panerai A, Cocchi D, Frohman LA, Antegaza P. Central glucoprivation: some physiological effects induced by the intraventricular administration of 2-deoxy-D-glucose. Experientia 1973;29: 874 – 5. Murata T, Waki A, Omata N, Fujibayashi Y, Sadato N, Yano R, Yoshimoto M, Isaki K, Yonekura Y. Dynamic changes in glucose metabolism by lactate loading as revealed by a positron autoradiography technique using rat living brain slices. Neurosci Lett 1998;249: 155 – 8. Oomura Y, Ono T, Ooyama H, Wayner MJ. Glucose and osmosensitive neurons of the rat hypothalamus. Nature 1969;222:282 – 4. Paxinos G, Watson C. The rat brain in stereotaxic coordinates. 2nd ed. Sydney: Academic Press, 1986. Pellerin L, Pellegri G, Bittar PG, Charnay Y, Bouras C, Martin JL, Stella N, Magistretti PJ. Evidence supporting the existence of an activity-dependent astrocyte-neuron lactate shuttle. Dev Neurosci 1998;20:291 – 9. Penicaud L, Pajot MT, Thompson DA. Evidence that receptors con-
[19]
[20]
[21]
[22] [23]
[24]
[25] [26]
[27] [28]
[29]
[30]
[31]
397
trolling growth hormone and hyperglycemic responses to glucoprivation are located in the hindbrain. Endocr Res 1990;16:461 – 75. Poitry-Yamate CL, Poitry S, Tsacopoulos M. Lactate released by Muller glial cells is metabolized by photoreceptors from mammalian retina. J Neurosci 1995;15:5179 – 91. Prichard J, Rothman D, Novotny E, Petroff O, Kuwabara T, Avison M, Howseman A, Hanstock C, Shulman R. Lactate rise detected by 1H NMR in human visual cortex during physiologic stimulation. Proc Natl Acad Sci USA 1991;88:5829 – 31. Ritter RC, Slusser PG, Stone S. Glucoreceptors controlling feeding and blood glucose: location in the hindbrain. Science 1981;213: 451 – 3. Rorsman P. The pancreatic beta-cell as a fuel sensor: an electrophysiologist’s viewpoint. Diabetologia 1997;40:487 – 95. Schurr A, Miller JJ, Payne RS, Rigor BM. An increase in lactate output by brain tissue serves to meet the energy needs of glutamateactivated neurons. J Neurosci 1999;19:34 – 9. Schurr A, Payne RS, Miller JJ, Rigor BM. Glia are the main source of lactate utilized by neurons for recovery of function posthypoxia. Brain Res 1997;774:221 – 4. Schurr A, West CA, Rigor BM. Lactate-supported synaptic function in the rat hippocampal slice preparation. Science 1988;240:1326 – 8. Silberbauer CJ, Surina-Baumgartner DM, Arnold M, Langhans W. Prandial lactate infusion inhibits spontaneous feeding in rats. Am J Physiol 2000;278:R646 – 53. Stone TW. Microiontophoresis and pressure ejection. New York: Wiley, 1985. Tabernero A, Vicario C, Medina JM. Lactate spares glucose as a metabolic fuel in neurons and astrocytes from primary culture. Neurosci Res 1996;26:369 – 76. Yang XJ, Kow LM, Funabashi T, Mobbs CV. Hypothalamic glucose sensor — similarities to and differences from pancreatic beta-cell mechanisms. Diabetes 1999;48:1763 – 72. Yettefti K, Orsini JC, Perrin J. Characteristics of glycemia-sensitive neurons in the nucleus tractus solitarii. Possible involvement in nutritional regulation. Physiol Behav 1997;61:93 – 100. Yettefti K, Orsini JC, El Ouazzani T, Himmi T, Boyer A, Perrin J. Sensitivity of nucleus tractus solitarius neurons to induced moderate hyperglycemia, with special reference to catecholaminergic regions. J Auton Nerv Syst 1995;51:191 – 7.