@misc{oai:ir.soken.ac.jp:00001059,
 author = {清水, 秀忠 and シミズ, ヒデタダ and SHIMIZU, Hidetada},
 month = {2016-02-17, 2016-02-17},
 note = {Sodium (Na) homeostasis is crucial for life and Na levels in body fluids are constantly monitored in the brain.  The subfornical organ (SFO) is the center of the sensing responsible for the control of Na-intake behavior, where Nax channels are expressed in specific glial cells as the Na-level sensor.  Nax channel is a concentration-sensitive Na channel with a threshold value of approximately 150 mM for the extracellular Na ion.  The Nax-positive glial cells are sensitive to an increase in the extracellular Na level in the physiological range, indicating that glial cells, not neurons, are the primary site of Na-level sensing.  However, the mechanism by which the Na signal sensed by “inexcitable” glial cells is transferred to neurons has remained to be elucidated.<br />  To gain insight into the cellular processes involving Nax in glial cells, in this doctor thesis, I started in my study with screening for molecules interacting with Nax using the yeast two-hybrid system with each of the cytoplasmic domains of mouse Nax as bait.  Among the positive clones isolated from a mouse DRG cDNA library by using the C-terminal region of Nax as bait, three clones coded for the α subunit of Na+/K+-ATPase.  A detailed analysis revealed that all these clones were identical and coded amino acid sequence of the region close to the cytoplasmic catalytic domain of the α1 subunit of Na+/K+-ATPase.  The direct interaction between Nax and the α1 subunit of Na+/K+-ATPase was verified by pull-down assays and the immunoprecipitation of the cell lysate.  <br />Coexpression of the α1 subunit of Na+/K+-ATPase and Nax channels was examined by double-fluorescent immunostaining using sections of the SFO and dissociated cells from the SFO.  The α1 subunit was broadly distributed throughout the SFO, overlapping with the expression of Nax channels.  The confocal microscopic analyses with isolated cells from the SFO showed that both molecules were colocalized in the plasma membrane.  Nax channels were expressed in large round cells, but not in small cells with neurite-like processes, indicating that Nax channels are expressed in glial cells including ependymal cells.<br />It is known that the α2 and α3 isoforms of Na+/K+-ATPase are also expressed in the brain.  Experiments using the yeast two-hybrid system showed that the cytoplasmic fragment of α2 corresponding to the region of the α1 isoform isolated also interacted with the C-terminal region of Nax, but that of α3 did not.  Thus, Nax has specific interaction with α1 and α2 isoforms of Na+/K+-ATPase.  By in situ hybridization, I verified that the mRNAs encoding the α1 and α2 isoforms of Na+/K+-ATPase were expressed in the SFO with a similar pattern to the Nax channels.  However, signals for the α3 isoform were not detected in the SFO. <br />I speculated that Nax may functionally regulate Na+/K+-ATPase through the close interaction between the two.  To examine this idea in vitro, I established cell lines using C6 glioma cells in which the expression of Nax channel is inducible under a tetracycline-based system.  I preferentially used one of the cell lines thus prepared, C6M16, for the following experiments.  C6M16 cells showed significant Na+ influx in response to an increase in the extracellular Na level within the physiological range (from 145 mM to 170 mM) specifically under the conditions where Nax channels are expressed.  <br />If the molecular properties of the Na+/K+-ATPase is changed by Nax, cellular metabolism should be affected accordingly, because cells in the CNS use ~50% of their energy resources to drive the Na+/K+-ATPase activity.  I then compared the cellular uptakes of a fluorescent glucose derivative (2-NBDG) in isotonic (145 mM) and hypertonic (170 mM) Na solutions.  The C6M16 cells with Nax expression showed approximately 1.6-fold greater activity for 2-NBDG uptake in the 170 mM Na solution as compared with in the 145 mM solution, while the uptake by the C6M16 cells without Nax expression was not increased in the 170 mM Na solution.  The increase in the uptake of 2-NBDG in Nax-expressing cells in the 170 mM Na solution was completely inhibited by ouabain, a specific inhibitor of Na+/K+-ATPase.<br />Next, I tested the effect of overexpression of the Nax-binding fragments of α1 and α2 subunits of Na+/K+-ATPase in C6M16 cells, because these fragments are expected to work as a competitor of Na+/K+-ATPase for binding to Nax channels.  As was expected, the transfection of an expression vector carrying the fragments of the α1 and α2 subunits significantly suppressed the metabolic response in the C6M16 cells with Nax expression in the 170 mM Na solution.  In contrast, overexpression of the fragment of the α3 subunit, which was negative for interaction with Nax channel, did not affect the metabolic activation.  <br />The C-terminal fragment of Nax was also expected to work as a competitor for the binding of Nax channels to Na+/K+-ATPase.  Unexpectedly but intriguingly, overexpression of the C-terminal fragment of Nax further enhanced the 2-NBDG uptake, when it was coexpressed in Nax-positive cells.  This suggests that the C-terminal region of Nax is also able to support Na+/K+-ATPase, as well as the native Nax.  However, the expression of the C-terminal fragment of Nax by itself (without concomitant expression of the native Nax) exerted no effect on the 2-NBDG uptake.  This strongly suggests that a function of the native Nax channel (presumably Na+-influx activity) is also essential for the upregulation of the metabolic state, in addition to the function which is substitutable with the C-terminal region of Nax.<br />To estimate the contribution of Na+ influx itself to the metabolic activation, I tested the effect of the influx generated by a Na-specific ionophore, monensin, on the uptake of glucose.  At a concentration of 0.5 μM, monensin triggered a small Na+ influx into cells comparable to that of the C6M16 cells expressing Nax when stimulated in the 170 mM solution.  However, the application of 0.5 μM monensin to C6M16 cells without Nax-channel expression did not enhance the 2-NBDG uptake, and higher concentrations of monensin were not effective either.  In contrast, when Nax-expressing cells were treated with 0.5 μM and higher concentrations of monensin, the 2-NBDG uptake was markedly enhanced dose-dependently.  These results clearly indicate that the increase of the Na-ion concentration in the cell is not enough by itself to trigger the uptake of glucose (metabolic stimulation), and that the presence of Nax channel protein is required for the stimulation of glucose uptake by the cells.  Importantly, the C-terminal fragment of Nax induced markedly enhanced 2-NBDG uptake under the condition without Nax-channel expression with 0.5 μM monensin.  This indicates that the full-length Nax channel can be replaced by the C-terminal fragment under the condition where Na+ influx was secured by monensin.  Taken together, it is probable that both pre-stimulation of Na+/K+-ATPase (by interaction with Nax channels through the C-terminal region of Nax) and Na+ influx (through Nax channels or monensin) are essential for the activation of the Na+/K+-ATPase and the cellular metabolic stimulation.<br />To examine whether the Nax channel is indeed involved in the energy-control system in the Nax-positive glial cells in vivo, I performed an imaging analysis of the uptake of glucose in the SFO using 2-NBDG.  In the wild-type SFO, incubated with 2-NBDG in the 170 mM Na solution, an intensively labeled mesh-like structure became apparent, suggesting that fine glial processes in the SFO actively took up the fluorescent derivative of glucose.  These results clearly indicate that the SFO tissue has activity to take up glucose in response to a Na-level increase, and the Nax channel is an essential component for this mechanism.  <br />Next, I examined the dissociated cells from the SFO of wild-type and Nax-KO mice to confirm that the cells showing the enhancement of 2-NBDG uptake express Nax channels.  Only among the wild-type cells, cellular populations that intensively took up 2-NBDG in the 170 mM Na solution were present, and these cells were all positive for Nax and GFAP.  These results clearly indicate that the Nax channel is an essential component for the upregulation of energy demand in the SFO under the high Na condition, as observed in the C6M16 cells.  In support of this view, cells dissociated from the SFO of wild-type mice showed a markedly enhanced uptake of 2-NBDG in the presence of 0.5 μM monensin, while the same stimulation of the cells from Nax-KO mice induced little enhancement of the uptake.  <br />Increased demand for glucose by cells means that cellular glycolysis is enhanced to yield lactate (or pyruvate).  To confirm this idea, I next measured the amounts of lactate and pyruvate released from the SFO as another parameter of the metabolic activity.  The SFO tissues removed from mice of both genotypes were incubated in a modified Ringer solution (containing 145 mM or 160 mM Na) at 37°C for 30 hr.  The wild-type SFO showed an increase in lactate secretion by ~60% compared with the Nax-KO SFO, in 160 mM Na solution.  On the other hand, amounts of pyruvate released into the medium were 10-fold lower than those of lactate and did not differ under the two different Na conditions.  This indicates that anaerobic glycolysis was stimulated in Nax-positive glial cells of the SFO under the high Na condition.  <br />Neurons in the SFO of Nax-KO mice are hyperactivated under dehydrated conditions compared with wild-type mice.  In the SFO, GABAergic neurons are one of the major neuronal types surrounded by Nax-positive glial cells.  Then I examined the neuronal activity of GABAergic neurons in the SFO using patch-clamp techniques in the cell-attached mode.  I prepared acute slices containing the SFO from GAD-GFP mice and GAD-GFP/Nax-KO, in which the GABAergic neurons bear enhanced green fluorescent protein (eGFP) as marker, and selected the eGFP-positive cells under a fluorescence microscope.  The GABAergic neurons in the SFO of both wild-type and Nax-KO mice showed spontaneous firing at a similar frequency (~4 Hz) under the 145 mM Na condition.  After the extracellular Na-concentration was raised to 160 mM, the firing frequency of the GABAergic neurons in the SFO of wild-type mice gradually increased 2-fold, but that of Nax-KO mice did not show a significant change.<br />Because metabolic activation leads to the release of lactate from Nax-positive glial cells, I next checked the possibility that lactate mediates the signal from the glial cells to GABAergic neurons to control the SFO activity.  When lactate was added at 1 mM to the perfusate, the firing frequency of GABAergic neurons in the SFO of both wild-type and Nax-KO mice increased.  Furthermore, when 1 mM of lactate was added under the high Na condition, no additive effect on the neuronal activity was observed.  These results indicate that lactate and Na share a common pathway in the stimulation of GABAergic neurons in the SFO.  Lactate was most effective at ~1 mM in promoting the firing rate, and at higher concentrations, the firing was rather suppressed.<br />The neural activation induced by the Na-level increase was inhibited by α-Cyano-4-hydroxycinnamic acid, an inhibitor of monocarboxylate transporters (MCTs).  These results clearly indicate that the Na-dependent stimulation of GABAergic neurons is largely mediated by MCTs.  I also examined the effect of the other metabolic monocarboxylates, pyruvate and acetate, both known to be transported by MCTs.  When pyruvate was added at 1 mM to the perfusate, the firing frequency of GABAergic neurons in the SFO of both wild-type and Nax-KO mice similarly increased.  By contrast, when acetate was added at 1 mM to the perfusate, the firing frequency was not significantly changed in either genotype.<br />I further explored the activation mechanism underlying the increase in the firing rate of the GABAergic neurons.  The finding that lactate and pyruvate are equally effective suggests that the GABAergic neurons are energetically stimulated.  Moreover, I found that Na-dependent potentiation of the firing activity of the GABAergic neurons were reduced by diazoxide, an opener of the ATP-sensitive K channel (Kir6.2 / KATP channel):  The KATP channel closes in response to the increase of intracellular ATP level and depolarizes the cell.  So, I examined the membrane potential of the GABAergic neurons during the application of lactate or high Na solution.  The membrane potential was depolarized by both lactate and Na, and the depolarization effect was expectedly reduced by diazoxide.  These data thus support the view that lactate serves as an energy substrate to up-regulate the firing activity of the GABAergic neurons.<br />From these results, the following cellular mechanism for the signaling from glial cells to neurons became clear.  Na-level-dependent Na+ influx through Nax and direct interaction between Nax and Na+/K+-ATPase are the basis for activation of Na+/K+-ATPase in the glial cells.  Activation of Na+/K+-ATPase stimulates anaerobic metabolism of glucose by the glial cells, which produces lactate as the end product.  There exist GABAergic neurons spontaneously firing in the SFO.  Lactate released from the glial cells functions as the substance signaling to the neurons for activation.  To my knowledge, this study is the first to show that glial cells take the initiative in the regulation of neural activity using lactate as a signaling substance.<br /><br /><br />, application/pdf, 総研大甲第1101号},
 title = {Glial Nax channels control lactate signaling to neurons for brain [Na+] sensing},
 year = {}
}