Diabetes mellitus (DM) is associated with increased plasma levels of arginine-vasopressin (AVP), which may aggravate hyperglycemia and nephropathy. However, the mechanisms by which DM may cause the increased AVP levels are not known. Electrophysiological recordings in supraoptic nucleus (SON) slices from streptozotocin (STZ)-induced DM rats and vehicle-treated control rats revealed that γ-aminobutyric acid (GABA) functions generally as an excitatory neurotransmitter in the AVP neurons of STZ rats, whereas it usually evokes inhibitory responses in the cells of control animals. Furthermore, Western blotting analyses of Cl− transporters in the SON tissues indicated that Na+-K+-2Cl– cotransporter isotype 1 (a Cl− importer) was upregulated and K+-Cl– cotransporter isotype 2 (KCC2; a Cl− extruder) was downregulated in STZ rats. Treatment with CLP290 (a KCC2 activator) significantly lowered blood AVP and glucose levels in STZ rats. Last, investigation that used rats expressing an AVP-enhanced green fluorescent protein fusion gene showed that AVP synthesis in AVP neurons was much more intense in STZ rats than in control rats. We conclude that altered Cl− homeostasis that makes GABA excitatory and enhanced AVP synthesis are important changes in AVP neurons that would increase AVP secretion in DM. Our data suggest that Cl− transporters in AVP neurons are potential targets of antidiabetes treatments.
Introduction
Diabetes mellitus (DM) is a group of metabolic diseases associated with serious complications such as nephropathy, retinopathy, and peripheral neuropathy. Characteristic features of DM include hyperglycemia and a high blood level of arginine-vasopressin (AVP), among others (1–5). Functional consequences of a high AVP level in the blood are unclear. Other than its well-known antidiuretic and vasoconstrictive actions, AVP has been shown to stimulate liver cells to increase their glucose output (6,7). AVP has also been shown to increase hepatic glucose output by stimulating the secretion of glucagon from pancreatic α-cells (8) and ACTH from the anterior pituitary (9,10). In addition, some evidence indicates that abundant circulating AVP contributes to increased glomerular filtration, albuminuria, and renal hypertrophy in DM (11) through its action at V2 receptors in the kidney (12,13). Furthermore, in rats with partial nephrectomy, the reduction of AVP secretion induced by increased water intake retarded the progression of renal failure (14). Thus elevated AVP in the blood in DM may contribute to disease progression (8,15) and directly drive serious DM complications such as renal failure (11).
It is unknown why AVP blood level is markedly elevated in DM, although it is thought that hyperglycemia, in the absence of insulin, greatly increases the hormonal output of AVP neurons by stimulating osmoreceptors (5,16). In this study we used the streptozotocin (STZ)-induced DM rat model to measure physiological changes in AVP neurons that could underlie the increased hormonal secretion. Here we present evidence that the Cl− importer Na+-K+-2Cl– cotransporter isotype 1 (NKCC1) is upregulated and the Cl− extruder K+-Cl– cotransporter isotype 2 (KCC2) is downregulated in AVP neurons, thus making γ-aminobutyric acid (GABA) excitatory in these cells, and that AVP synthesis is enhanced in magnocellular neurons in the paraventricular nucleus (PVN) and supraoptic nucleus (SON). Together, these changes would promote the secretion of AVP neurons and hence the elevation of AVP blood level found in DM. Furthermore, we show that the KCC2 activator CLP290 is an effective means of reducing blood levels of AVP and glucose in DM, thus raising the possibility that Cl− transporters in AVP neurons could be targets of antidiabetes treatment.
Research Design and Methods
Animals
Male Sprague-Dawley rats (6 weeks of age) from Orient Bio (Seongnam, Korea) and double-transgenic Wistar rats (6 weeks of age), which express both the AVP-enhanced green fluorescent protein (AVP-eGFP) fusion gene and the OXT-monomeric red fluorescent protein 1 (OXT-mRFP1) fusion gene (17), were used in this study. They were housed in a room with a 12-h light/12-h dark cycle for >1 week before being used for experiments and handled as previously described (18).
DM Model
This model was produced by injecting once the antineoplastic drug STZ to the rats (65 mg/kg body weight, i.p.). Fresh STZ solutions were prepared before injections by dissolving the drug (65 mg) in 10 mL 50 mmol/L sodium citric acid (pH 4.5). Control rats received the vehicle. Rats were studied 3 weeks after STZ or vehicle injection. To confirm that STZ injection produced DM, we measured body weight change, blood and cerebrospinal fluid (CSF) glucose levels (blood glucose meter; Johnson & Johnson, Milpitas, CA), plasma and CSF Na+ concentrations ([Na+]) (i-Smart 30 VET; i-Sense, Seoul, Korea) and osmolalities (Fiske Micro-Osmometer model 210; Fiske Associates, Norwood, MA), and plasma AVP level (AVP ELISA kit; Enzo Life Sciences, Farmingdale, NY).
Hypothalamic Slice, Electrophysiological Recording, and Western Blotting
Electrophysiological recordings of SON in hypothalamic slices (350–400 μm thick) were obtained extracellularly or intracellularly using a gramicidin-perforated technique. Western blotting was performed for the SON tissues excised from hypothalamic slices. Detailed methods were previously described for these experiments (18,19).
Timing of Experiments
All experiments were performed during the light or projected light phase.
Drugs
We purchased all drugs and chemicals used in this study from Sigma-Aldrich, except for muscimol (Ascent Scientific, Cambridge, MA). VU0463271, CLP257, and CLP290 were gifts (see Acknowledgments).
Results
STZ treatment is commonly used to produce a DM model. To confirm the effectiveness of our treatment protocol, we compared the STZ-treated Sprague-Dawley rats (STZ rats) with vehicle-treated ones (control rats) with respect to various parameters related to DM (Table 1). The STZ rats gained significantly less body weight than the control rats in the 3-week interval between treatment and experiment. The glucose concentrations in the blood and CSF and the [Na+] in the CSF were much higher in the STZ rats than in the control rats, whereas plasma [Na+] was slightly but significantly lower in STZ rats. Consistent with these observations, the plasma and CSF osmolalities were also significantly higher in STZ rats. Last, the plasma AVP concentrations of STZ rats were ∼12 times higher than those of the control rats. Taken together, these results indicate that the STZ treatment produced DM in Sprague-Dawley rats.
Contents . | Control rats . | STZ rats . | ||
---|---|---|---|---|
Mean ± SD . | Rats (n) . | Mean ± SD . | Rats (n) . | |
Δ Body weight (g) | 138.6 ± 26.6 | 15 | 38.9 ± 39.5** | 15 |
Glucose (mg/dL) | ||||
Blood | 143.1 ± 42.4 | 15 | 512.7 ± 89.2** | 15 |
CSF | 103.7 ± 11.1 | 15 | 264.6 ± 68.8** | 15 |
[Na+] (mmol/L) | ||||
Plasma | 137.1 ± 2.5 | 15 | 132.2 ± 4.5** | 15 |
CSF | 148.2 ± 1.2 | 15 | 154.3 ± 5.4** | 15 |
Osmolality (mOsm/KgH2O) | ||||
Plasma | 322.3 ± 5.3 | 15 | 344.5 ± 13.3** | 15 |
CSF | 324.5 ± 4.5 | 15 | 340.7 ± 10.9** | 15 |
Plasma AVP (pg/mL) | 5.8 ± 5.6 | 9 | 68.6 ± 72.1* | 12 |
Contents . | Control rats . | STZ rats . | ||
---|---|---|---|---|
Mean ± SD . | Rats (n) . | Mean ± SD . | Rats (n) . | |
Δ Body weight (g) | 138.6 ± 26.6 | 15 | 38.9 ± 39.5** | 15 |
Glucose (mg/dL) | ||||
Blood | 143.1 ± 42.4 | 15 | 512.7 ± 89.2** | 15 |
CSF | 103.7 ± 11.1 | 15 | 264.6 ± 68.8** | 15 |
[Na+] (mmol/L) | ||||
Plasma | 137.1 ± 2.5 | 15 | 132.2 ± 4.5** | 15 |
CSF | 148.2 ± 1.2 | 15 | 154.3 ± 5.4** | 15 |
Osmolality (mOsm/KgH2O) | ||||
Plasma | 322.3 ± 5.3 | 15 | 344.5 ± 13.3** | 15 |
CSF | 324.5 ± 4.5 | 15 | 340.7 ± 10.9** | 15 |
Plasma AVP (pg/mL) | 5.8 ± 5.6 | 9 | 68.6 ± 72.1* | 12 |
Blood samples for glucose and [Na+] measurements were obtained from the tail vein without anesthesia, whereas the blood samples for plasma osmolality and AVP measurements were obtained by decapitation under urethane anesthesia (1.25 g/kg body weight, i.p.). To prevent clotting and protease action, these samples were collected in 3-mL Vacutainer tubes containing K2-EDTA (BD Biosciences, Franklin Lakes, NJ) and aprotinin (0.6 TIU/mL blood). The plasma was obtained by centrifuging the blood at 1,600g at 4°C for 15 min and was used immediately to measure osmolality or was stored in a deep freezer (−80°C) until used for AVP enzyme-linked immunosorbant assay. CSF samples were obtained through the atlanto-occipital membrane just before the blood sampling by decapitation.
*P < 0.05,**P < 0.001, Student t test or rank sum test.
GABA Is Excitatory in Most AVP Neurons of STZ Rats
AVP neurons in the PVN and SON are heavily innervated by GABAergic afferents (20), which originate from the perinuclear zones in the hypothalamus (20–22). Through the GABAA receptor, GABA exerts an excitatory, rather than inhibitory, effect in most AVP neurons in chronically salt-loaded, lactating, or hypertensive rats (18,19,23,24) in which AVP secretion is enhanced. Thus, in this study, we examined the possibility that GABA functions as an excitatory neurotransmitter in the AVP neurons of STZ rats as well, which would promote the secretory activities of these cells. To test this hypothesis, we first examined the effects of bath-applied bicuculline (30 μmol/L; a GABAA receptor antagonist) on the single-unit activities of the magnocellular neurons recorded extracellularly in the SON slices of control and STZ rats. In the cells (n = 16) of control rats (n = 3), bicuculline increased (n = 10) or decreased (n = 6) the unit activity, whereas in the STZ rats (n = 3) it decreased the unit activity in all cells (n = 16) examined (Fig. 1A and B). These results support the notion that, in general, GABA is excitatory in AVP neurons of STZ rats, whereas it is inhibitory in the cells of control rats. To obtain further evidence for this hypothesis, we next examined GABAA receptor–mediated postsynaptic potentials (PSP) recorded in the AVP neurons sampled in the SON slices of control and STZ rats. For the recording of the GABAergic PSP in isolation from glutamatergic excitatory PSP (EPSP), we used the gramicidin-perforated recording technique, which preserves the intracellular Cl− concentration ([Cl−]i) of the recorded cell (26), and included the N-methyl-D-aspartate (NMDA) receptor blocker DL-2-amino-5-phosphonopentanoic acid (50 μmol/L) and the non-NMDA receptor blocker 6,7-dinitroquinoxaline-2,3-dione (20 μmol/L) in the slice perfusion medium. AVP neurons were identified on the basis of their electrical properties, that is, little or no inward rectification at membrane potentials between −50 and −170 mV (27,28) and/or phasic patterns of firing (29) (Supplementary Fig. 1). It was confirmed that the recorded PSPs from AVP neurons were mediated by the GABAA receptor based on the inhibitory effects of bicuculline or gabazine on these synaptic potentials (Supplementary Fig. 2). Figure 1C and D illustrate that GABAergic PSP occurring in most of the AVP neurons of STZ rats were excitatory (i.e., EPSP), whereas those in the cells of control rats were mostly inhibitory (i.e., IPSP). This difference in GABAergic transmission was apparently due to the difference in the reversal potential of GABAA receptor–mediated responses (EGABA) (Fig. 1E and F); in the cells of STZ rats, the EGABA was positive by ∼13 mV, on average, to the action potential threshold, which was around −45 mV, whereas in the cells of control animals, it was negative to the threshold by ∼6 mV on average. Collectively, these results indicate that, owing to the depolarizing EGABA shift, GABA functions as an excitatory, instead of inhibitory, neurotransmitter in most AVP neurons of STZ rats.
To ensure that the observation of GABAergic transmission in the AVP neurons of the STZ rats described above was not strain specific, that the identification of AVP neurons was valid based on their electrical properties, and that the change in GABAergic transmission detected was selective for AVP over oxytocin (OXT) neurons, we next repeated the gramicidin-perforated recordings of GABAergic transmission in the AVP and OXT neurons of double-transgenic Wistar rats, which express both the AVP-eGFP and OXT-mRFP1 fusion genes. These rats were treated, as were Sprague-Dawley rats, with STZ or vehicle 3 weeks before SON slice preparation. AVP and OXT cells in the slice were identified first visually by their green and red fluorescence, respectively (Fig. 2A and B); then the visual identification was confirmed based on the electrophysiological properties of the recorded cell, that is, for AVP neurons, the presence of phasic firing and/or little or no inward rectification at membrane potentials between −50 and −170 mV, and for OXT cells, significant inward rectification at membrane potentials between −50 and −170 mV without phasic firing. As in the AVP neurons of Sprague-Dawley rats treated with STZ, the depolarizing shift of EGABA beyond the action potential threshold and the resultant emergence of GABAergic excitation occurred in most of the AVP neurons of the STZ-treated transgenic rats (Fig. 2A and C). It is interesting, however, that similar GABAergic transmission–related events did not occur in the OXT neurons of STZ-treated transgenic rats (Fig. 2B and D). These results indicate that the changes in GABAergic transmission following STZ treatment occur selectively in AVP neurons and are not strain dependent. Furthermore, they indicate that our identification of AVP neurons was valid based on their electrical properties.
Upregulation of NKCC1 and Downregulation of KCC2 Depolarize EGABA to Produce GABAergic Excitation in AVP Neurons of STZ Rats
In mammalian central nervous system neurons, the Cl− importer NKCC1 and Cl− extruder KCC2 play primary roles in regulating [Cl−]i, the major determinant of EGABA and the polarity of GABAA receptor–mediated PSP (30). To gain insight into the molecular mechanisms underlying the GABAergic excitation occurring in the AVP neurons of STZ rats, we investigated the expression levels of NKCC1 and KCC2 in the SON using Western blotting. The level of NKCC1 was significantly higher and the level of KCC2 was significantly lower in the STZ rats than in the control rats (Fig. 3A and B). These results suggest that, in the STZ rats, the upregulation of NKCC1 and the downregulation of KCC2 in AVP neurons depolarize the EGABA by increasing [Cl−]i, causing GABA to function as an excitatory neurotransmitter in these cells. To test this hypothesis more directly, we next examined the effects of the NKCC inhibitor bumetanide (31) and the KCC2 inhibitor VU0463271 (32) on the EGABA and GABAergic PSP profile in the AVP neurons of control and STZ rats. In the neurons (n = 13) of control rats (n = 5), the bath application of bumetanide (10 μmol/L) did not significantly alter the EGABA (Fig. 3C, top left), whereas in the cells (n = 16) of STZ rats (n = 3), it significantly hyperpolarized the EGABA (Fig. 3C, top right). The effects of bumetanide in these neurons were partially reversible and accompanied by a significant change in the GABAergic PSP profile; that is, the ratio of neurons showing EPSP and IPSP changed from 15:1 to 5:11 after bumetanide application (P < 0.001, Fisher exact test). Meanwhile, bath-applied VU0463271 (5 μmol/L) caused a significant depolarization of EGABA in the neurons (n = 8) of control rats (n = 3; Fig. 3C, bottom left) and altered the ratio of neurons showing GABAergic EPSP and IPSP from 1:7 to 7:1 (P < 0.001, Fisher exact test). In the cells (n = 7) of STZ rats (n = 3), however, VU0463271 did not significantly affect the EGABA (Fig. 3C, bottom right) and GABAergic PSP profile. These data, together with the results from the Western blotting experiments, indicate that the upregulation of NKCC1 and the downregulation of KCC2 are the molecular mechanisms that underlie the depolarizing shift of EGABA and the consequent emergence of GABAergic excitation in the AVP neurons of STZ rats.
AVP Synthesis Is Enhanced in AVP Neurons of STZ Rats
The results described above demonstrate that GABAergic inhibition is converted to excitation in the AVP neurons of STZ rats, thus indicating that the process of AVP release is enhanced in these cells. To maintain increased hormone secretion in a given neuron, the synthesis of the hormone, as well as its release process, should be enhanced. To determine whether AVP synthesis was enhanced in the AVP neurons of STZ rats, we compared the STZ-treated Wistar rats expressing both AVP-eGFP and OXT-mRFP1 fusion genes with the vehicle-treated control animals with respect to the intensities of green and red fluorescence in the SON and the magnocellular division of the PVN. As expected, the green fluorescence intensity was significantly higher in both hypothalamic regions of the STZ rats than in the control rats (Fig. 4A and B). On the other hand, red fluorescence intensity in both the SON and PVN was not different between the two rat groups (Fig. 4A and C). These results indicate that the synthesis of AVP, but not OXT, is enhanced in hypothalamic magnocellular neurosecretory cells of the STZ rats.
Systemic Administration of the KCC2 Activator CLP290 in STZ Rats Lowers AVP and Glucose Levels in the Blood
The results presented above indicate that the inhibitory-to-excitatory switch in GABAergic transmission in the AVP neurons of STZ rats, an electrophysiological change that would promote the release of AVP from these cells, arises from the upregulation of NKCC1 and the downregulation of KCC2. Thus, we reasoned that the blockade of NKCC1 or the activation of KCC2 might decrease AVP release by preventing the GABAergic excitation of AVP neurons. Moreover, we thought that they might lower the blood glucose level as well, because AVP increases blood glucose by stimulating hepatic gluconeogenesis and glycogenolysis (6–8). To determine whether KCC2 activation in STZ rats can lower AVP and glucose levels in the blood, we examined the effects of CLP290, a prodrug of the KCC2 activator CLP257, on these variables (33). Before we tested the effects of CLP290 on the AVP and glucose levels in the blood, we first investigated whether KCC2 activation with CLP257 prevented GABAergic excitation by assessing the effects of this agent on the EGABA and GABAergic PSP profile in the AVP neurons recorded in the SON slices of STZ rats. The effects of CLP257 were evaluated by comparing the electrophysiological data from SON slices incubated in this agent (25 μmol/L) for 2 h with those from slices kept in the vehicle dimethyl sulfoxide (0.025%) for the same duration. The AVP neurons sampled in the slices treated with CLP257 had more hyperpolarized EGABA (Fig. 5A) and a significantly lower ratio of GABAergic EPSP to IPSP than the ones recorded in vehicle-treated slices (3:6 vs. 9:1; P = 0.02, Fisher exact test). On the other hand, the AVP neurons sampled in the SON slices from control rats treated with CLP257 were not different from the cells in the vehicle-treated slices from the same rat group with respect to EGABA (Fig. 5A) and GABAergic PSP profile (ratio of GABAergic EPSP to IPSP 1:8 to 0:11). These results indicate that KCC2 activation with CLP257 prevents GABAergic excitation in the AVP neurons of STZ rats. Therefore, we continued to examine the effects of CLP290 on AVP and glucose levels in the blood. CLP290 (100 mg/kg body weight) or vehicle (20% hydroxypropyl-β-cyclodextrin) was injected intraperitoneally 4 h before blood sampling to measure AVP and glucose concentrations. As illustrated in Fig. 5B and C, CLP290 treatment in STZ rats significantly lowered AVP and glucose levels, whereas in control rats it resulted in a significant increase in the plasma AVP level without affecting the glucose concentration.
Discussion
The elevated level of AVP in the blood in DM can aggravate hyperglycemia (6–9) and contribute to the generation of diabetic nephropathy (11). Despite such potential significance of high AVP blood level, the mechanisms underlying the elevation of AVP level have not been fully identified. In this study we focused on the possible changes occurring in AVP neurons that would promote their hormone secretion under disease conditions. Secretion of a neurohormone or neurotransmitter from a neuron involves two processes: the synthesis and the release of the secreted substance. In mature central nervous system neurons, the release process is governed by action potential firing, which is in turn regulated mainly by the balance of excitatory glutamatergic and inhibitory GABAergic transmission. Given that AVP neurons are densely innervated by GABAergic afferents (20) and that GABA functions as an excitatory, instead of inhibitory, neurotransmitter in the majority of AVP neurons of chronically salt-loaded, lactating, or hypertensive rats (18,19,23,24), in which AVP secretion is enhanced, we scrutinized GABAergic transmission in this study. We found that the EGABA depolarized significantly, and therefore GABA became excitatory in the majority of AVP neurons of our DM model rats. On the other hand, GABA was mostly inhibitory in the cells of normal (i.e., control) rats, as previously reported (19,23). Furthermore, we discovered that the synthesis of AVP is greatly increased in the SON and in the magnocellular region of the PVN of DM model rats. Thus, our results suggest that increased secretory activities of AVP neurons contribute to the increase in the blood level of AVP in DM as a result of the enhancement of AVP synthesis and the modulation of GABAergic transmission.
AVP neurons are heavily innervated not only by GABAergic inputs but also by glutamatergic afferents originating from various brain regions, including the organum vasculosum of the lamina terminalis, the subfornical organ, and the median preoptic nucleus (34). In a previous study, Di and Tasker (35) showed that dehydration induced by chronic salt loading, which leads to enhanced AVP secretion, increased excitatory postsynaptic current amplitude and frequency in magnocellular neurons of the rat SON. Thus the increased secretory activities of AVP neurons in DM may depend on the modulation of glutamatergic (as well as GABAergic) transmission. The determination of the importance of glutamatergic transmission in enhancing AVP secretion in this model awaits future studies.
In this study we discovered that STZ treatment resulted in the depolarizing shift of EGABA and the consequent emergence of GABAergic excitation in AVP but not OXT neurons. This finding contrasts with the observations made in our previous study performed with rats in which EGABA depolarized and GABAergic excitation emerged in both AVP and OXT neurons after chronic salt loading (19). The absence of such changes in the OXT neurons of DM model rats makes sense because the emergence of GABAergic excitation in this neuronal population would enhance the secretion of OXT, a hormone known to stimulate the secretion of atrial natriuretic hormone from the atrium (36) and act directly in the kidney to promote natriuresis (37,38). The enhanced secretion of OXT would aggravate the hyponatremia present in these animals.
The strength and polarity of synaptic responses mediated by the GABAA receptor are determined primarily by [Cl−]i. The results obtained from our Western blotting and neurophysiological experiments indicate that the depolarizing shift of EGABA and the resultant emergence of GABAergic excitation detected in the AVP neurons of DM model rats—electrophysiological changes that would promote the secretory activities of these cells—were due to the increase in [Cl−]i that resulted from the upregulation of the Cl− importer NKCC1 and downregulation of the Cl− extruder KCC2. This conclusion is corroborated by the finding that the systemic administration of the KCC2 activator CLP290 resulted in a significant reduction of the blood level of AVP in the DM model rats. In this study we did not investigate the possible mechanisms that could lead to changes in NKCC1 and KCC2 expression. The mechanism that causes NKCC1 upregulation may involve the activation of autoreceptors by AVP, a paracrine released from the somata and dendrites of AVP neurons in response to hyperglycemia, and perhaps to the increase in osmolality of and [Na+] in the CSF. In accordance with this idea, Ludwig (39) demonstrated that hyperosmotic stress is a good stimulus for AVP release from the somata and dendrites of hypothalamic magnocellular neurons in the rat. In addition, we showed that, in rats subjected to chronic hyperosmotic stress, the cerebroventricular administration of the V1a AVP receptor antagonist d(CH2)5[Tyr(Me)2,Ala-NH29]AVP significantly lowers the proportion of magnocellular neurons in the SON that exhibit GABAergic excitation, a phenomenon resulting from the upregulation of NKCC1 (19). Furthermore, Wakamatsu and colleagues (40) showed that AVP increases the expression of NKCC1 in the outer medullary collecting duct of the rat nephron.
The mechanism underlying KCC2 downregulation in the AVP neurons of DM model rats may involve the action of brain-derived neurotrophic factor (BDNF), which is released locally from glia and/or neurons in the PVN and SON. It has been shown that hyperosmotic stimulus increases the local release of BDNF in the SON (41) and that the BDNF receptor tropomyosin receptor kinase B (TrkB) is present in the SON (23,41,42). Furthermore, it has been reported that chronic salt loading in the rat prevents the GABAergic inhibition of AVP neurons by KCC2 downregulation via the BDNF-dependent activation of TrkB (23). In this study we demonstrated that CLP257, a drug that activates KCC2 by apparently inhibiting BDNF action (33), hyperpolarized EGABA and hence converted GABAergic excitation back to inhibition in the AVP neurons of DM model rats. Last, it has been shown that, in a rat model of neuropathy produced by applying polyethylene cuffs to the sciatic nerve, BDNF from microglia downregulates KCC2 expression through TrkB, enabling GABAergic excitation to occur in neurons in the lamina I of the spinal dorsal horn. This paradoxical excitation contributes to the generation of neuropathic pain in this animal model (43). More recently, however, Morgardo and colleagues (44) reported that, in STZ-induced diabetic rats, BDNF level is not altered in the spinal cord and neuropathic pain is not much affected by the BDNF sequester TrkB/Fc, although spinal KCC2 downregulation, which causes the inhibitory-to-excitatory switch of GABAergic action to occur, underlies the generation of neuropathic pain (45,46). Thus they suggested that BDNF-induced KCC2 downregulation is less likely to be responsible for the generation of GABAergic excitation in the spinal neurons and of neuropathic pain in this animal model. Further investigation is required to determine whether the mechanisms underlying the changes in NKCC1 and KCC2 expression in the AVP neurons of our DM model rats involve the AVP activation of autoreceptors and the BDNF activation of TrkB.
In this study we found that intraperitoneal injection of the KCC2 activator CLP290 in STZ-induced DM rats lowered the blood level of AVP by ∼50% and of glucose by ∼15%. These results not only support the notion that the downregulation of KCC2 is partly responsible for the depolarizing shift of EGABA and the emergence of GABAergic excitation in the AVP neurons of DM model rats (see results), but they also suggest that this drug is an effective means of lowering the blood levels of AVP and glucose in DM. The smaller effect of CLP290 on blood glucose than on AVP may be related to the fact that the effects of AVP on glycogenolysis and gluconeogenesis (7,8) in the liver are reduced in DM because of the downregulation of hepatic V1a receptors (47,48) and because AVP is not the only factor contributing to the increase of blood glucose in DM.
The increase in the plasma AVP level of control rats after CLP290 treatment was an unexpected finding, raising the possibility that nonspecific effects on cells other than AVP neurons are involved in CLP290 administration in vivo. However, it seems unlikely that the increase in AVP level resulted from GABAergic excitation in AVP neurons in light of the observation that the KCC2 activator CLP257 did not depolarize the EGABA in these cells. Also, it is unlikely that the plasma AVP concentration was increased by osmotic stimulation because the intraperitoneally injected drug solution was rather hypo-osmotic (∼270 mOsm/kg H2O).
In summary, we provide evidence that altered Cl− homeostasis in AVP neurons and the enhanced AVP synthesis in these cells are important changes underlying the increase in plasma AVP level in DM. With growing evidence that elevated AVP in the blood in DM is a risk factor for metabolic disorders (8,15) and progression of serious diabetic complications such as nephropathy (11), future studies are warranted to evaluate the possible effects of drugs including CLP290 that target Cl− transporters against diabetic complications.
Y.-B.K. and W.B.K. contributed equally to this work.
Article Information
Acknowledgments. The authors are grateful to Dr. Y. Ueta (University of Occupational and Environmental Health, Kitakyushu, Japan), Dr. Y. De Koninck (Université Laval, Québec, Canada), and Dr. C. Lidsley (Vanderbilt University, Nashville, TN) for providing rats expressing AVP-eGFP and OXT-mRFP1 fusion genes, KCC2 activators (CLP257, CLP290), and the KCC2 inhibitor (VU0463271), respectively.
Funding. This work was supported by National Research Foundation of Korea (NRF) grants funded by the Korean government (Ministry of Science, Information and Communications Technology and Future Planning) (2016R1D1A1B03932771 to Y.-B.K. and 2014R1A2A1A11049900 and 2017R1A2B2002277 to Y.I.K.). Y.-B.K., W.B.K., Y.S.K., and Y.I.K. were supported by the Brain Korea 21 Project from 2009 to 2017. Y.S.K. and C.S.C. received support from the O’Keefe Foundation.
Duality of Interest. No potential conflicts of interest relevant to this article were reported.
Author Contributions. Y.-B.K., W.B.K., H.C.H., G.D.B., C.S.C., and Y.I.K. conceived this project. Y.-B.K., W.B.K., W.W.J., X.J., Y.S.K., and B.K. performed the experiments and analyzed the results. Y.-B.K., W.B.K., C.S.C., and Y.I.K. wrote the manuscript. Y.I.K. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.