Caudal hindbrain “sensing” of glucoprivation activates central neural mechanisms that enhance systemic glucose availability, but the critical molecular variable(s) linked to detection of local metabolic insufficiency remains unclear. Central neurons and glia are metabolically coupled via intercellular trafficking of the glycolytic product lactate as a substrate for neuronal oxidative respiration. Using complementary in vivo models for experimental manipulation of lactate availability within the caudal hindbrain, we investigated the hypothesis that lactate insufficiency may be monitored by local metabolically “sensitive” neurons as an indicator of central nervous system energy imbalance. The data show that caudal fourth ventricular (CV4) administration of the monocarboxylate transporter inhibitor α-cyano-4-hydroxycinnamate (4CIN) resulted in dose-dependent increases in blood glucose in euglycemic animals, whereas the degree and duration of hypoglycemia elicited by insulin administration were exacerbated by exogenous l-lactate delivery to the CV4. Immunocytochemical processing of the hindbrain for the inducible c-fos gene product Fos revealed that 4CIN enhanced Fos immunoreactivity in the dorsal vagal complex (DVC), e.g., the nucleus of the solitary tract and dorsal vagal motor nucleus, and adjacent area postrema, sites where cells characterized by unique sensitivity to diminished glucose and/or glycolytic intermediate/end product levels reside, and in the medial vestibular nucleus (MV), and that CV4 l-lactate infusion increased Fos labeling within the DVC and MV after insulin-induced hypoglycemia. Together, these results support the view that lactate is a critical monitored metabolic variable in caudal hindbrain detection of energy imbalance resulting from glucoprivation and that diminished uptake and/or oxidative catabolism of this fuel activates neural mechanisms that increase systemic glucose availability.
- nucleus of the solitary tract
- central nervous system
metabolically “sensitive” neurons within characterized central nervous system (CNS) sites, including the caudal hindbrain dorsal vagal complex (DVC) and adjacent area postrema (AP), function to monitor cellular energy stasis (17, 18). Compensatory physiological and behavioral responses to glucoprivation are initiated, in part, by neural stimuli that originate in this area of the brain. Although the molecular mechanisms by which cellular metabolic insufficiency is detected in this and other CNS sites remain unclear, evidence for induction of hyperphagia and/or hyperglycemia (9, 24, 32) by fourth ventricular delivery of glucose antimetabolites or uptake inhibitors is consistent with the view that reductions in glucose uptake and/or catabolism within the periventricular hindbrain activate CNS-mediated compensatory mechanisms that function to restore glucostasis.
The high-energy demands of CNS neurons are supported primarily by mitochondrial aerobic metabolism (3). Because glucose occurs at high concentrations in the circulation and is rapidly transferred into the brain via astrocytic GLUT-1 transport activity (31), the conventional view holds that this molecule is the principal, if not sole substrate for neuronal oxidative respiration. However, emerging studies support the concept of cell-type compartmentation of glucose metabolism in the brain, involving the transport of metabolites between astrocytes and neurons (12). Although lactate uptake from the blood is insignificant, this oxidizable fuel is produced within astrocytes as an end product of glycolysis and is released by those cells into the extracellular compartment of the brain (8). Evidence for CNS neuronal reliance on intercellular lactate trafficking includes findings that lactate supports synaptic activity in the absence of glucose (11, 27), serves as the primary oxidative substrate when both fuels are present (5), and supports neuronal function during states of energy crisis (27, 28). Reports that exogenous lactate reverses glucoprivic stupor/coma and associated reductions in CNS levels of tricarboxylic acid cycle intermediates (31) imply a causal relationship between hypoglycemic brain dysfunction and lactate deficits due to precursor shortages. The present studies investigated the hypothesis that this substrate fuel may be a critical indicator of cellular energy homeostasis within the caudal hindbrain. To examine whether central neural mechanisms governing systemic glucose availability are activated in response to underutilization of lactate in this area of the brain, two complementary in vivo stereotaxic-based intracranial delivery models were used to evaluate the effects of pharmacological manipulation of local lactate uptake on the transcriptional status of caudal hindbrain metabolic “sensing” neurons and circulating glucose levels.
MATERIALS AND METHODS
Adult female Sprague-Dawley rats (220–250 g body wt) were maintained under a 14:10-h light-dark schedule (lights on at 0500) and allowed free access to standard laboratory rat chow and water. Fourteen days before experimentation, the animals were anesthetized with ketamine-xylazine (0.1 ml/100 g body wt ip, 100 mg ketamine-10 mg xylazine/ml; Henry Schein, Melville, NY) and implanted with midline 26-gauge stainless steel external cannula guides (Plastics One, Roanoke, VA) aimed above the caudal fourth ventricle, (CV4) utilizing three-dimensional coordinates from a rat brain atlas (30a). After surgery, the animals were transferred to individual cages. The patency of internal injection cannulas was verified at the conclusion of infusions, and the accuracy of cannula placements was determined by histological examination of sections through the hindbrain.
Experiment 1: Effects of CV4 infusion of graded doses of monocarboxylate transporter inhibitor α-cyano-4-hydroxycinnamate on blood glucose levels and hindbrain neuronal Fos expression.
Groups of rats (n = 5/group) were infused with graded doses of the monocarboxylate transporter (MCT) inhibitor α-cyano-4-hydroxycinnamate (4CIN; 10.0, 25.0, or 50.0 μg) or the vehicle propylene glycol (PG) in a 2.0-μl volume over a 30-min period beginning at ∼1300, via 33-gauge internal injection cannulas projecting 0.5 mm beyond the cannula guide into the CV4. The animals were tested in the fed state. Two hours after initiation of infusions, the animals were killed by transcardial perfusion under pentobarbital anesthesia with 2.0% sodium nitrite in 0.9% saline, followed by 0.1 M phosphate buffer, pH 7.6, containing 4% paraformaldehyde and 0.2% picric acid. Blood samples (0.4 ml) were collected before, during, and after infusion treatments in separate vehicle- or drug-treated groups of animals (n = 5/group) implanted with indwelling Silastic intravenous catheters 2 days before the study. Additional groups of animals were bled before and after CV4 infusion of the alternative MCT inhibitor phloretin (100 μg/2.0 μl; n = 4) or vehicle alone (n = 4). All blood samples were immediately centrifuged to obtain plasma for glucose assays; blood cells were resuspended in lactated Ringer solution and returned via cannula.
Experiment 2: Effects of graded CV4 lactate infusion on systemic insulin-induced hypoglycemia and hindbrain neuronal Fos expression.
At −10 min, groups of full-fed animals (n = 5/group) were pretreated by initiation of continuous CV4 infusion of graded concentrations of l-lactate (10, 25, or 50 μM at a rate of 4.0 μl/h) in artificial cerebrospinal fluid (aCSF) or aCSF alone. Actual administered doses of l-lactate were thus 7.2, 18.0, or 36.0 μg/2 h, respectively. Each rat was injected with regular insulin (Humulin R; 10 U/kg body wt sc; Henry Schein) at time 0 and killed by transcardial perfusion at +120 min. Separate groups of vehicle- or lactate-infused animals (n = 5/group) were implanted with indwelling intravenous catheters 2 days before these treatments and bled before (time 0) and at +30, 60, 90, and 120 min after insulin injections. Blood samples collected at each time point were centrifuged to collect plasma for glucose analyses, and blood cells were resuspended in lactated Ringer solution and returned via cannula. The metabolic specificity of l-lactate's effects on insulin-induced hypoglycemia (IIH) was determined by CV4 infusion of the metabolically active isomer l-lactate (50 μM/h; n = 5) vs. the nonactive isomer d-lactate (50 μM/h; n = 5) during IIH. The involvement of MCT-dependent mechanisms in exacerbation of IIH by l-lactate was addressed by 4CIN pretreatment before initiation of lactate infusion in insulin-injected animals (n = 4/group). Animals were treated with administration of 4CIN (50 μg) or the vehicle, aCSF, into the CV4 at −15 min, followed by initiation of lactate infusion at −10 min and injection of insulin at time 0; controls were treated by CV4 infusion of aCSF alone before and after insulin injection.
Plasma Glucose Measurements
Glucose concentrations were measured using Amplex red glucose spectrophotometric assay kit reagents (catalog no. A-22189; Molecular Probes, Eugene, OR). Plasma samples were diluted with reaction buffer and aliquoted into individual wells of 96-well plates. After incubation with Amplex red reagent-horseradish peroxidase-glucose oxidase working solution for 30 min, absorbencies were read at 570 nm in a plate reader.
The brains were postfixed overnight in fresh fixative, sunk in 25% sucrose, cut into serial 25-μm sections on a Reichert-Jung freezing microtome, and stored at −20°C in cryoprotectant. After thorough rinsing in 0.05 M Tris-buffered saline, 0.9% NaCl, pH 7.6 (TBS), sections collected at 150-μm intervals through the hindbrain were preincubated for 30 min in normal goat serum (rabbit ABC Elite kit, product no. PK-6101; Vector Laboratories, Burlingame, CA) and then incubated for 48 h at 4°C with a rabbit polyclonal antiserum against human Fos4–17 (Ab-5; Oncogene Research Products, Cambridge, MA), diluted 1:100,000 in TBS containing 0.05% Triton X-100. After being rinsed with TBS, the tissues were sequentially incubated with biotinylated goat anti-rabbit second antibody and avidin-biotin-peroxidase complex, for 1 h each, at room temperature. Fos antigenic sites were visualized by incubation with filtered Sigma Fast cobalt-intensified 3,3′-diaminobenzidine tetrahydrochloride tablet sets (Sigma Chemical, St. Louis, MO). The sections were rinsed in TBS, mounted on gelatin-covered slides in saline, air-dried, dehydrated in graded alcohols, coverslipped with Permount, and examined with a Nikon E400 bright-field microscope. Digital images were captured with a Photometrics CoolSNAP cf monochrome camera and processed with Metamorph imaging software (Universal Imaging, Downingtown, PA). Bilateral counts of Fos-immunopositive neurons in individual neural loci were obtained using Metamorph-operated integrated morphometric analysis, with thresholds for intensity and hue set at 142 and 160, respectively.
Mean bilateral cell counts for individual neural structures were compared between treatment groups by one-way ANOVA followed by Tukey's honestly significant difference (HSD) test. Plasma glucose data were evaluated by two-way ANOVA followed by Tukey's HSD test.
Fos immunolabeling occurred in the hindbrain of animals treated within the MCT inhibitor 4CIN, namely, in the DVC, AP, and medial vestibular nucleus (MV), but was absent from other structures adjacent to the fourth ventricle. As shown in Fig. 1, this drug treatment resulted in significant, dose-proportionate increases in numbers of Fos-immunoreactivity (ir)-positive neurons, relative to the PG-treated controls, in the DVC, AP, and MV. Minimal Fos staining occurred in all three brain sites after vehicle infusion. Figures 2, 3, and 4 depict the effects of graded doses of 4CIN on patterns of Fos-ir within the nucleus of the solitary tract (NTS), AP, and MV, respectively; in Figs. 2–4, representative photomicrographs depict Fos-ir-positive neurons in the NTS 2 h after initiation of infusions of vehicle alone (B) or 4CIN at a dose of 10 (C) or 50 μg (D). 4CIN-induced Fos-ir in the DVC was observed in the medial (NTSm) and gelatinous (NTSge) parts of the NTS and in the dorsal motor nucleus of the vagus (DMV). Fos-ir was detected in the NTS at each drug dose but was only consistently present in the DMV in animals treated with the 50-μg dose of 4CIN. Groups treated with 4CIN were characterized by commensurate dose-related elevations in blood glucose levels. The data in Fig. 5 show that animals infused with the lowest drug dose, e.g., 10 μg, exhibited significantly higher plasma glucose levels, relative to baseline values, at +30 and +60 min, whereas doses of 25 or 50 μg 4CIN per animal elevated circulating glucose between +30 and +120 min after initiation of treatment. Figure 6 demonstrates that CV4 infusion of a different MCT inhibitor, phloretin, also significantly increased blood glucose levels above baseline.
Insulin-injected animals treated with CV4 infusion of the vehicle, aCSF, exhibited low levels of Fos immunostaining in the DVC, e.g., the NTSm, NTSge, DMV, and MV (Fig. 6). As shown in Fig. 7, continuous infusion of l-lactate, over a dose range of 10, 25, or 50 μM/h, administered over a 2-h period beginning before and maintained during IIH, resulted in significant dose-related increases in numbers of Fos-ir-positive neurons in both the DVC and MV relative to the vehicle-treated IIH group. The representative photomicrographs in Figs. 8 and 9 depict the effects of CV4 infusion of graded doses of l-lactate on Fos immunolabeling within the DVC and MV, respectively, of insulin-injected rats; Fos-ir-positive neurons are depicted 2 h after a subcutaneous bolus injection of 10 U Humulin R/kg body wt and initiation of continuous intraventricular infusion of the vehicle, aCSF (B), or l-lactate over a concentration range of 10 (C) or 50 μM/h (D). Whereas Fos labeling of the NTS occurred in a dose-proportionate manner, staining of the DMV was observed primarily in response to the highest dose of l-lactate. The data in Fig. 10 show that administration of l-lactate at a concentration of either 25 or 50 μM/h resulted in plasma glucose levels that were significantly decreased in l-lactate- vs. vehicle-infused IIH animals at +60, +90, and +120 min after subcutaneous insulin injection. As shown in Fig. 11, enhancement of insulin-induced decrements in plasma glucose levels by the metabolically active isomer l-lactate was not replicated by treatment with the inactive isomer d-lactate, results that emphasize the metabolic specificity of action of the former molecule within the caudal hindbrain. The data in Fig. 11 also show that measurements of plasma glucose immediately before and 10 min after initiation of CV4 l-lactate or vehicle infusion reveal a lack of effect of intracerebroventricular administration of l-lactate alone on circulating glucose. Figure 12 shows that serial CV4 administration of 4CIN and l-lactate significantly elevated plasma glucose levels, compared with groups treated with l-lactate or aCSF, at time 0 and at +30 and +60 min after insulin injection. Between +90 and +150 min after insulin treatment, glucose levels were significantly decreased in rats given either 4CIN plus l-lactate or l-lactate only relative to those given aCSF alone. At both +120 and +150 min, rats treated with the combination of 4CIN and l-lactate exhibited significantly higher plasma glucose compared with animals treated with l-lactate only.
There is compelling evidence for neuronal reliance on uptake of the glucose metabolite lactate for aerobic respiratory function. The present studies show that pharmacological inhibition of caudal hindbrain MCT function in situ increases circulating glucose levels in parallel with transcriptional activation of neurons in periventricular structures of characterized significance for glucose “sensing.” The data also demonstrate that exogenous l-lactate delivery into the CV4 delays recovery from hypoglycemia and modifies DVC neuronal genomic responses to IIH. Together, the corroborative physiological and neuroanatomical results reported provide unique in vivo evidence in support of caudal hindbrain lactate utilization as a critical component of local mechanisms that serve to monitor cellular metabolic stasis and activate central neural pathways that regulate systemic glucostasis.
Glucose metabolism in the brain is compartmentalized and involves trafficking of the glycolytic metabolite lactate from astrocytes to neurons, which use this substrate to generate energy via aerobic respiration. The concept that pathophysiological consequences of inadequate glucose provision to the brain reflect, in part, corresponding downstream reductions in lactate uptake and utilization as an energy substrate is backed by evidence that exogenous lactate relieves neurological symptoms of hypoglycemia and reverses glucoprivic suppression of tricarboxylic acid cycle intermediates and ATP in the brain (31), and that reversal of aglycemic perturbances in retinal tissue ATP, aspartate, glutamate, and glutamine levels by lactate is abolished by 4CIN (38). Our observations of dose-proportionate induction of hyperglycemia by CV4 infusion of the MCT inhibitor 4CIN suggest that decreased monocarboxylate uptake and/or subsequent catabolism in the caudal periventricular hindbrain constitute a physiological signal(s) for activation of neural pathways that enhance systemic glucose availability. Our parallel findings that delivery of l-lactate, but not d-lactate, into this part of the brain exacerbates insulin-induced hypoglycemia and that pharmacological blockade of MCT attenuates this effect of CV4 l-lactate support the premise that MCT-mediated transport of this metabolizable molecule is a critical component of local metabolic sensing mechanisms. Although it is highly plausible that CV4 l-lactate intensifies hypoglycemia through actions on adjacent hindbrain tissue that alter neural control of glucose counterregulation, the possibility that this treatment suppresses blood glucose by modifying peripheral insulin action and/or metabolism cannot be overlooked. Recent evidence that l-lactate perfusion of the ventromedial hypothalamus inhibits counterregulatory hormone secretion during hypoglycemia (4) bolsters the view that in multiple metabolic sensing sites in the brain, lactate insufficiency may serve as a stimulus for activation of pathways controlling physiological responses to systemic glucose deficits. l-Lactate correspondingly enhances and diminishes the firing frequency of glucose-“stimulated” (GS) and -“inhibited” (GI) neurons in the ventromedial hypothalamus (34, 35). The DVC and AP contain GS and GI neurons (1); the former cells respond to elevated extracellular lactate with increased action potential frequencies (10), but the impact of lactoprivation on the electrophysiological function of the latter population is not known. Thus one potential mechanism by which caudal hindbrain lactoprivation and exogenous lactate administration may each alter blood glucose levels may involve modulation of energy-surfeit signaling by GS neurons in these hindbrain structures.
The function of the DVC and AP to monitor and signal glucoprivic disruption of energy stasis is supported by evidence that local glucose antimetabolite administration elicits hyperglycemia and hyperphagia (26) and by electrophysiological and immunocytochemical data demonstrating neuronal sensitivity within the NTS, DMV, and AP to direct manipulation of glucose availability (1, 6, 16, 36, 37). Our data show that 4CIN-induced Fos immunostaining in the DVC and AP is dose proportionate, evidence that transcriptional activation of local neurons is correlated with the degree of inhibition of caudal hindbrain lactate transport. The near-complete absence of Fos-ir in neighboring neural structures lacking metabolism- sensitive neurons, at any dose of 4CIN, supports the premise that this nuclear label identifies, in part, sensing neurons that participate in the initiation or relay of signaling of lactate insufficiency or, alternatively, suppression of energy-surfeit signaling, and argues against the possibility of nonspecific drug actions on local neurons. This view is supported by our recent report that DVC and AP tyrosine hydroxylase-ir-positive neurons, a phenotype characterized by electrophysiological reactivity to energy fuel availability, exhibit dose-proportionate colabeling for Fos in response to CV4 4CIN (21). Although we presume that GI neurons in the DVC and AP are susceptible to transcriptional activation by pharmacological lactoprivation, the possibility that GS cells also may be genomically reactive to this manipulation cannot be disregarded. Fos immunoexpression is a reliable indicator of altered cellular functional state, but activation of this stimulus-transcription cascade is not a definitive or automatic sign of altered neuronal firing rate. Thus it is possible that 4CIN-induced hindbrain Fos-ir may also reflect cellular reactions to lactoprivation that occur in the absence of altered neurotransmission, namely, metabolic or energetic adaptation. We also note that 4CIN and/or lactate-associated Fos labeling in the DVC may be due, in part, to visceral afferent signaling of secondary peripheral metabolic effects of the intracerebroventricularly administered compounds. CNS astrocytes and neurons express cell-specific monocarboxylate transporter variants, e.g., MCT1 and MCT2, respectively (7). Although cellular mapping of MCT2 gene transcript or protein expression within the rat DVC or AP has not been performed to date, ongoing studies in our laboratory aim to examine whether local neurons that express Fos-ir in response to pharmacological inhibition of MCT function exhibit this neuron-specific transporter, information that could reveal whether these cells are directly or indirectly susceptible to lactate deficiency. Evidence for a lack of expression of this transporter would imply that genomic activation of such neurons is mediated via afferent input from lactate-sensitive neurons located within the DVC and/or AP or outside the confines of these structures.
It is not clear whether decrements in plasma membrane transporter activity, conversion of internalized lactate to pyruvate, and/or subsequent oxidative catabolism of this product constitute stimuli that are transduced by specialized “monitoring” cells into neural regulatory signaling. The neuronal lactate dehydrogenase (LDH) enzyme variant, e.g., LDH-1, catalyzes the unidirectional conversion of lactate to pyruvate and is characteristic of tissues reliant on energy derived from oxidative respiration (14). In the rat brain, the LDH-1 gene is highly expressed in the NTS and other rare brain sites (13), data suggesting that neurons in these structures may function as lactate “sinks.” Further studies are needed to investigate whether this enzyme is expressed by NTS/AP neurons that are transcriptionally responsive to lactate deficits and, if so, whether level of enzyme activity is linked to signaling of lactoprivation. Although the primary route of glucose uptake into the brain occurs via the astrocytic cellular compartment (14), glucose is demonstrable within the CNS extracellular space (30). The ability of CNS neurons to directly internalize glucose via constitutive, e.g., GLUT-3 and -4, as well as energy-dependent, e.g., SGLT1, transporter function (22, 23) poses the pivotal question of whether metabolic signaling neurons in the DVC/AP deduce reductions in glucose availability via monitoring of lactate alone or, alternatively, uptake (or ratio of uptake) of both lactate and glucose. Additional research aimed at evaluating whether the same phenotypic neuron populations, or indeed individual neurons within this complex, exhibit comparable functional responses to systemic hypoglycemia vs. uptake of intercellular lactate is warranted. Forthcoming evidence that both manipulations elicit comparable responses from the same cells would imply that DVC/AP neuronal reactivity to systemic glucose deficits may be mediated primarily by downstream decrements in lactate trafficking or, alternatively, that a single population of “sensor” cells functions to monitor both lactate and glucose uptake. Alternatively, results demonstrating cellular sensitivity to hypoglycemia, but not decreased lactate availability, would support the existence of local neurons that function to monitor glucose uptake only. Recent studies comparing the electrophysiological responses of NTS neurons, under anesthesia, to microiontophoretically applied glucose or lactate reported that these treatments evoked similar responses from the majority of cells investigated, findings that support the likelihood that these substrate fuels dually influence the functional status of a subset of local neurons (10).
The current studies show that after CV4 infusion of vehicle during IIH, Fos immunostaining was minimal in the DVC and absent from the AP, results that concur with earlier studies investigating effects of a single bolus injection of regular insulin in nonfasted rats (19). The present disparity in magnitude of local Fos induction by pharmacological (4CIN) blockade of MCT vs. IIH-associated decreases in blood glucose may reflect discrepant reductions in hindbrain lactate utilization, i.e., decrements in lactate uptake may be larger due to the former versus the latter manipulation. This dissimilar Fos labeling also may be a sign of differential effects of these treatments on the ratio of neuronal uptake of lactate and glucose, given that only lactate transport is inhibited by 4CIN, whereas uptake of both lactate and glucose would be suppressed by IIH. CV4 lactate infusion initiated before IIH increased numbers of Fos-ir-positive neurons in the DVC relative to the vehicle-infused, hypoglycemic controls, despite the efficacy of this treatment in delaying restoration of euglycemia. Because both GS and GI neurons reside in the DVC, augmentation of hypoglycemia by exogenous l-lactate may reflect enhanced GS energy-surfeit signaling and/or diminished GI energy-deficit signaling. Given that DVC metabolic sensing neurons undergo minimal transcriptional activation during IIH alone, Fos-labeled neurons in the DVC of lactate-infused hypoglycemic animals may be GS that are reactive to elevations in extracellular lactate alone or relative to glucose. This possibility is supported by reports that elevated extracellular lactate levels increase neuronal firing frequencies in this structure (10). Patterns of Fos expression due to exogenous l-lactate may also reflect, in part, cellular biochemical adaptations by nonsensing neurons to the substrate fuel imbalance created under these particular circumstances. Although hypoglycemia, in the absence of lactate replacement, would be expected to diminish neuronal uptake of both substrate fuels, graded doses of lactate replacement during persistent neuroglucopenia could plausibly result in a dose-proportionate divergence between lactate and glucose availability. Our studies did not determine whether the infusion protocol utilized results in tissue lactate levels that mimic or exceed the normal range. If the latter outcome was achieved, the resultant dichotomy between these critical fuel molecules might stimulate genomic responses by neurons that signal metabolic imbalance as well as those that do not.
Fos immunocytochemical analysis of the hindbrain was performed to evaluate the responsiveness of characterized populations of metabolically sensitive neurons in this region of the brain to pharmacological lactoprivation. However, we have observed that CV4 4CIN elicits dose-dependent Fos-ir in several extrahindbrain loci that govern energy balance, including the lateral hypothalamic area, hypothalamic paraventricular, dorsomedial, and arcuate nuclei, and thalamic paraventricular nucleus, in patterns that coincide with those elicited by glucose antimetabolite treatment (unpublished observation). These findings support the view that the central autonomic circuitry controlling autonomic, neuroendocrine, and behavioral responses to metabolic imbalance is sensitive to hindbrain lactoprivation. This manipulation also elicited dose-proportionate Fos labeling of the MV. In the absence of proof that metabolically sensitive neurons exist in the MV, one interpretation of these data is that those MV neurons characterized by genomic responsiveness to 4CIN may reside downstream of caudal hindbrain structures, namely, the DVC and AP, that signal lactate deficiency. Because neither the NTS nor AP directly innervate the MV, indirect communication of lactate imbalance may occur, possibly involving relay by higher-order forebrain structures. Further work is needed to examine the relevance of lactoprivic genomic activation of MV neurons to central control of counterregulatory hormone secretion and blood glucose levels. An alternative possibility is that MV neuronal activation may be due to AP signaling of potential toxic effects of the administered compounds; however, the transient, dose-dependent increase in blood glucose and associated enhancement of neuronal Fos immunolabeling in the MV argue against nonspecific mechanisms of drug action within the caudal hindbrain. We also comment that because Fos-ir does not occur in the MV after intracerebroventricular glucose antimetabolite administration (6), the current data support the possibility that selective impairment of lactate utilization may activate neural circuitries that function to appraise the vestibulocerebellar system of potential metabolic imbalance.
Although C1/A1 and C3 neuron populations have been functionally implicated in metabolic monitoring (26) and express Fos-ir in response to systemic injection of the antimetabolic glucose analog 2-deoxy-d-glucose (2DG) (25), these neurons were not labeled for Fos in response to either CV4 administration of 4CIN or IIH in the presence or absence of lactate in the current studies. These disparate results most likely reflect differences in the 2DG model of pharmacological glucoprivation and the experimental paradigms we utilized. Although CV4 4CIN infusion inhibits lactate transport within a circumscribed area of the periventricular neuroaxis, peripheral injection of 2DG evokes both peripheral and central signaling of cellular energy and imbalance and impairs global utilization of both glucose and its metabolite, lactate, by CNS neurons. Alternatively, the present lack of Fos labeling of C1/A1 and C3 neurons in the 4CIN-treated animals may reflect their discriminative sensitivity to glucose but not lactate deficits. Furthermore, the systemic 2DG and IIH models diverge in that they elicit hyperglycemia vs. hypoglycemia, respectively, coincident with glucopenia. The present data concur with recent studies from our laboratory that report the absence of Fos labeling from C1/A1 and C3 neurons after injection of the intermediate-acting insulin, Humulin NPH (20). Together, these observations are consistent with a recent review of IIH-associated patterns of Fos-ir in the brain denoting the absence of labeling of A1 neurons in response to this manipulation (19).
In light of in vitro reports that 4CIN inhibits mitochondrial, as well as plasma membrane monocarboxylate transport (15), we cannot discount the possibility that the in vivo drug effects we observed may not solely reflect diminished yield of pyruvate from nonneuronal lactate but, rather, may be correlated to overall reduction of cellular oxidative catabolism and related energy production in the DVC and AP because of diminished utilization of pyruvate generated from either extracellular lactate or neuronal glycolysis. Further studies are needed to compare, in the in vivo context evaluated in this study, the relative magnitude of effects of 4CIN on intracellular vs. plasma membrane sites of action, e.g., on net mitochondrial uptake of pyruvate vs. plasma membrane transport inhibition.
In summary, the current studies show that CV4 infusion of the MCT inhibitor 4CIN in euglycemic animals elicits dose-dependent increases in blood glucose levels, whereas the degree and magnitude of hypoglycemia ensuing after insulin administration are augmented by exogenous lactate delivery to this brain site. The immunocytochemical data show that DVC and AP neurons express Fos-ir in response to 4CIN and that IIH-associated patterns of Fos labeling in that structure are modified by exogenous infusion of lactate. Together, these results provide compelling evidence that lactate is a critical monitored metabolic variable in caudal hindbrain detection of cellular energy imbalance resulting from systemic glucoprivation and support the novel view that diminished uptake and/or oxidative catabolism of this fuel activates neural mechanisms that increase systemic glucose availability.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 the American Physiological Society