Selective Down-regulation of KV2.1 Function Contributes to Enhanced Arterial Tone during Diabetes*

Background: KV channels in vascular smooth muscle cells (VSMCs) regulate arterial tone. Results: KV2.1 in VSMCs is down-regulated via AKAP150-CaN-dependent NFATc3 signaling during diabetes. Conclusion: Transcriptional suppression of KV2.1 contributes to enhanced arterial tone in diabetes. Significance: AKAP150-CaN-dependent activation of NFATc3 may be a general mechanism for transcriptional regulation of K+ channels, and a valuable target to prevent and treat diabetic vascular complications. Enhanced arterial tone is a leading cause of vascular complications during diabetes. Voltage-gated K+ (KV) channels are key regulators of vascular smooth muscle cells (VSMCs) contractility and arterial tone. Whether impaired KV channel function contributes to enhance arterial tone during diabetes is unclear. Here, we demonstrate a reduction in KV-mediated currents (IKv) in VSMCs from a high fat diet (HFD) mouse model of type 2 diabetes. In particular, IKv sensitive to stromatoxin (ScTx), a potent KV2 blocker, were selectively reduced in diabetic VSMCs. This was associated with decreased KV2-mediated regulation of arterial tone and suppression of the KV2.1 subunit mRNA and protein in VSMCs/arteries isolated from HFD mice. We identified protein kinase A anchoring protein 150 (AKAP150), via targeting of the phosphatase calcineurin (CaN), and the transcription factor nuclear factor of activated T-cells c3 (NFATc3) as required determinants of KV2.1 suppression during diabetes. Interestingly, substantial reduction in transcript levels for KV2.1 preceded down-regulation of large conductance Ca2+-activated K+ (BKCa) channel β1 subunits, which are ultimately suppressed in chronic hyperglycemia to a similar extent. Together, our study supports the concept that transcriptional suppression of KV2.1 by activation of the AKAP150-CaN/NFATc3 signaling axis contributes to enhanced arterial tone during diabetes.

Non-insulin-dependent type 2 diabetes is a devastating disease affecting millions worldwide due to an aging population, sedentary lifestyle, and overnutrition. Vascular dysfunction, including enhanced arterial tone, is a leading cause of cardiovascular complications contributing to morbidity and mortality in the diabetic population (1). Although endothelial dysfunction has long been recognized as a major factor contributing to vascular dysfunction and enhanced arterial tone during diabetes (2)(3)(4)(5), abnormal VSMC function may also play a critical role, although the mechanisms for this remain poorly understood. Thus, advances in this area could prove valuable for the development of rational therapies to treat and prevent diabetic vascular complications.
The level of arterial tone is largely determined by vascular smooth muscle cell (VSMC) 2 membrane potential (V M ) and Ca 2ϩ entry via voltage-gated, L-type Ca 2ϩ channels (LTCCs) (6). A major regulator of V M is the activity of K V and BK Ca channels (7)(8)(9). Physiological activation of these channels hyperpolarizes VSMCs, thereby decreasing LTCC open probability and Ca 2ϩ influx leading to vasodilation, whereas their inhibition promotes vasoconstriction (10). Previous reports have shown that in vitro, short term exposure of coronary, cerebral, and mesenteric arteries to extracellular glucose concentrations that resemble diabetic hyperglycemia (e.g. 20 mM) inhibit K V and BK Ca channel activity in VSMCs (11)(12)(13)(14)(15). Thus, inhibition of these channels may contribute to enhanced arterial tone and vascular complications during diabetes. Consistent with this, our group and others have found that the activity of BK Ca channels is suppressed in VSMCs of several mouse models of diabetes (14 -17). However, whether impaired K V channel activity in VSMCs contributes to enhanced arterial tone during diabetes is currently unclear.
We recently demonstrated that suppression of BK Ca channel activity in a high fat diet (HFD) mouse model of type 2 diabetes proceeds through activation of the prohypertensive CaN/ NFATc3 signaling pathway (14). Activation of NFATc3 required anchoring of CaN by the scaffolding protein AKAP150 (murine ortholog of human AKAP79) in diabetic cells (14). Considering that this transcription factor is also known to regulate the expression of K V channels in VSMCs (18), we postulated that activation of this pathway may also modulate K V function during diabetes.
In the present study, we tested the hypothesis that K V channel function in VSMCs is impaired and contributes to increased arterial tone in a HFD mouse model of type 2 diabetes, and that the mechanism involves activation of the AKAP150-CaN/ NFATc3 signaling pathway. Consistent with this hypothesis, we show that K V channel function is decreased in VSMCs from wild type (WT) HFD mice due to a reduction in the expression of ScTx-sensitive K V 2.1, but not psora-4-sensitive K V 1.2 or K V 1.5 subunits. Genetic ablation of NFATc3, AKAP150, or disruption of the AKAP150-CaN interaction prevented down-regulation of the K V 2.1 subunit, suppression of ScTx-sensitive I Kv , and enhanced arterial tone in HFD mice. Furthermore, our data indicate that down-regulation of K V 2.1 occurs at an earlier time point (i.e. 1 h) compared with BK Ca ␤1 subunits under hyperglycemic conditions. These findings illuminate a critical role for AKAP150-anchored CaN, NFATc3, and K V 2.1 function in the regulation of arterial tone during diabetes.

MATERIALS AND METHODS
Animals-WT (C57Bl/6J), AKAP150 Ϫ/Ϫ , NFATc3 Ϫ/Ϫ , and knock-in mice expressing AKAP150 lacking its CaN binding site (⌬PIX) (19) were euthanized with a lethal dose of sodium pentobarbital (250 mg/kg; intraperitoneally), as approved by the University of California, Davis Institutional Animal Care and Use Committee. Mice were sustained on either a low fat (10% kcal; ct) or high fat (60% kcal) diet (Research Diets, New Brunswick, NJ) starting at 5 weeks of age for 24 -26 weeks. The composition of these diets and the propensity of mice maintained on this HFD to develop type 2 diabetes and induce vascular dysfunction of small resistance arteries have been well documented in previous studies (14,20,21). Cerebral arteries were used for functional experiments (i.e. electrophysiology, arterial diameter, and immunofluorescence), whereas 3rd and 4th order mesenteric arteries were used for electrophysiology experiments in VSMCs from WT LFD and HFD mice and for molecular biology experiments requiring larger tissue samples (i.e. Western blots).
Isolation of VSMCs-VSMCs were dissociated from arteries using enzymatic digestion techniques as described previously (18,22,23). Middle and posterior cerebral arteries, as well as third and fourth order mesenteric arteries were dissected in ice-cold dissection buffer composed of (in mM): 140 NaCl, 5 KCl, 2 MgCl 2 , 10 D-glucose, 10 HEPES, pH 7.4, with NaOH. Cerebral arteries were digested in a dissection solution supplemented with papain (1 mg/ml) and dithiothreitol (1 mg/ml) at 37°C for 7 min, then incubated in dissection buffer supplemented with collagenase type H (0.3 mg/ml) and collagenase type F (0.7 mg/ml) at 37°C for 7 min. Cells were then washed in ice-cold dissection buffer. Glass pipettes of decreasing diameters were used to gently triturate arteries and obtain single VSMCs. Isolated cells were maintained in ice-cold dissection buffer until use.
Electrophysiology-I Kv from freshly dissociated VSMCs were measured using the conventional whole cell patch-clamp technique with an Axopatch 200B amplifier (Molecular Devices, Sunnyvale, CA). Currents were evoked by 0.6-s depolarizing pulses from a holding potential of Ϫ70 mV to ϩ80 (for psora-4-sensitive currents) or to ϩ40 mV (for ScTx-sensitive currents) in increments of ϩ10 mV. A voltage error of 15 mV attributable to the liquid junction potential of the recording solutions used was corrected offline. I Kv were recorded in the continuous presence of the BK Ca channel blocker iberiotoxin (100 nM) to eliminate BK Ca channel activity. Bath solution consisted of the following components (in mM): 120 NaCl, 3 NaHCO 3 , 4.2 KCl, 1.2 KH 2 PO 4 , 2 MgCl 2 , 0.1 CaCl 2 , 10 D-glucose, and 10 HEPES (pH 7.4). Patch pipette solution was composed of (in mM): 110 K-gluconate, 30 KCl, 0.5 MgCl 2 , 10 EGTA, 5 HEPES, 5 NaATP, and 1 GTP (pH 7.2 with KOH). Experiments were carried out at room temperature (22-25°C). Currents were sampled at 20 kHz and low-pass filtered at 5 kHz. Electrophysiology recordings were analyzed using pCLAMP 10.
Quantitative Polymerase Chain Reaction-K V transcript expression was analyzed in single isolated cells using the Power SYBR Green Cells-to-CT kit (Life Technologies). For these experiments, single VSMCs were collected after enzymatic isolation using a glass micropipette. RT product was used for quantitative PCR. Specific primers to detect K V 2.1 (NM_008420; reference number QT00285971), K V 1.2 (NM_008417; reference number QT00128100), and K V 1.5 (NM_145983; reference number QT00268387) were acquired from Qiagen (Valencia, CA). ␤-Actin was used as an internal control (GenBank TM accession number V01217; sense nucleotide 2384 -2404 and antisense nucleotide 3071-3091). Amplification was performed using a Power SYBR Green PCR cocktail (Life Technologies) and an Applied Biosystems real-time PCR instrument. Expression for each gene was normalized to ␤-actin and expressed as a percentage of LFD.
Western Blot Analysis-Cerebral and mesenteric arteries were homogenized in a Triton lysis buffer solution containing (in mM) 150 NaCl, 10 Na 2 HPO 4 , 1 EDTA with 1% deoxycholic acid, 0.1% sodium dodecyl sulfate and protease inhibitors (Complete Mini protease inhibitor cocktail Roche Applied Science), followed by sonication (20 min at 4°C). Tissue debris and nuclear fragments were removed by centrifugation at 10,000 ϫ g (10 min, 4°C) and whole cell lysates were obtained as the supernatant. An equal amount of protein was loaded for each tissue lysate. Proteins were separated under reducing conditions on a 10% polyacrylamide gel (Bio-Rad) by electrophoresis at 100 V for 1.5 h. Proteins were then electrophoretically transferred to a polyvinylidene difluoride membrane at 20 V (overnight, 4°C). Membranes were washed in Tris-buffered saline with 0.1% Tween 20 (TBS-T) and blocked with 10% nonfat milk in TBS-T (1 h, room temperature). Membranes were then incubated for 2 h at room temperature with subunit-specific antibodies from NeuroMab (University of California, Davis, Davis, CA). Antibodies and dilutions were as follows: mouse anti-K V 2.1 (73-014 clone K89/34; 1:2), anti-K V 1.2 (clone 14/16; 1:2), and K V 1.5 (clone K7/45; 1:2). Monoclonal antibody against ␤-actin (MA5-15739; 1:5,000) was from Pierce. Antibodies were prepared in TBS-T with 1% bovine serum albumin and 0.01% sodium azide. Membranes were then incubated (1 h, room temperature) with horseradish peroxidase-labeled goat anti-mouse (sc-2005; 1:5,000; Santa Cruz) in TBS-T containing 5% nonfat dried milk. Bands were identified by enhanced chemiluminescence and exposure to x-ray film. Densitometry for immunoreactive bands was performed with ImageJ software (National Institutes of Health). ␤-Actin was used as input control for normalization. Density was expressed as a percentage of control (LFD).
Immunofluorescence-Immunofluorescent labeling of freshly isolated VSMCs was performed as described previously (24) using a monoclonal antibody specific for K V 2.1 subunits (Neu-roMab; 75-159; clone K39/25, University of California, Davis). The secondary antibody was an Alexa Fluor 488-conjugated donkey anti-mouse (5 mg/ml) from Molecular Probes. Cells were imaged (512 ϫ 512 pixel images) using an Olympus FV1000 confocal microscope coupled with an Olympus ϫ60 oil immersion lens (NA ϭ 1.4) and a zoom of 3.5 (pixel size ϭ 0.1 m). Images were collected at multiple optical planes (z axis step size ϭ 0.5 m). The specificity of the primary antibody was tested in negative control experiments in which the primary antibody was substituted with PBS. K V 2.1-associated fluorescence was undetected under this experimental condition. Cells for each group were imaged with the same laser power, gain settings, and pinhole for all the treatments.
Arterial Diameter Measurements-Freshly isolated posterior cerebral arteries were cannulated on glass micropipettes mounted in a 5-ml myograph chamber (Living Systems Instrumentation, St. Albans, VT) as described previously (14). Arteries were pressurized to 20 mm Hg and allowed to equilibrate while continuously superfused (37°C, 30 min, 3-5 ml/min) with physiological saline solution consisting of (in mM): 119 NaCl, 4.7 KCl, 2 CaCl 2 , 24 NaHCO 3 , 1.2 KH 2 PO 4 , 1.2 MgSO 4 , 0.023 EDTA, and 10 D-glucose aerated with 5% CO 2 , 95% O 2 . Bath pH was closely monitored and maintained at 7.35-7.40. Following the equilibration period, intravascular pressure was increased to 60 mm Hg and arteries were allowed to develop myogenic tone. Arteries not exhibiting stable myogenic tone after ϳ1 h were discarded. To assess the contribution of K V 1.X and K V 2.1 channel function to regulation of arterial tone, the K V 1.X inhibitor psora-4 (500 nM) and K V 2.1 inhibitor ScTx (50 nM), respectively, were added to the superfusate. Arterial tone data are presented as a percent decrease in diameter relative to the maximum passive diameter at 60 mm Hg obtained at the end of each experiment using Ca 2ϩ -free saline solution containing nifedipine (1 M).
Chemicals and Statistical Analyses-All chemical reagents were from Sigma unless otherwise stated. Iberiotoxin was from Peptides International and ScTx-1 was from Alomone Labs. Data are expressed as mean Ϯ S.E. Data obtained using multiple vessels from the same animal were pooled for statistical analyses. Data were analyzed using GraphPad Prism software. Statistical significance was determined by Student's t test or one-way analysis of variance followed by Tukey multiple comparison test for comparison of multiple groups. p Ͻ 0.05 was considered statistically significant (denoted by * in figures).

RESULTS
Freshly isolated, small diameter (ϳ75-150 m) cerebral and mesenteric arteries and VSMCs from age-matched mice fed ad libitum with either a low fat (LFD, 10% kcal) or high fat (HFD, 60% kcal) diet were used for this study (14,20). This HFD model of type 2 diabetes was employed because it does not depend on genetic manipulations or chemical destruction of pancreatic ␤-cells. In addition, it recapitulates clinical features observed in patients with type 2 diabetes (20, 25), including enhanced arterial tone and increased blood pressure (14). WT mice on a HFD had significantly higher average body weight (42 Ϯ 1.0 g) than WT LFD (30 Ϯ 0.6 g; Table 1). Furthermore, non-fasting blood glucose levels were significantly increased in WT HFD (301 Ϯ 12 mg/dl) compared with WT LFD mice (145 Ϯ 4 mg/dl; Table 1). Fig. 1 shows I Kv evoked in cerebral VSMCs freshly dissociated from WT LFD and HFD mice in the presence of 100 nM iberiotoxin to inhibit BK Ca channels. The average capacitance of cerebral WT LFD and HFD cells was 19.4 Ϯ 1.2 and 19.0 Ϯ 1.2 pF, respectively (p Ͼ 0.05). I Kv were initially elicited by 600-ms depolarizing pulses from a holding potential of Ϫ85 mV to voltages ranging from Ϫ85 to ϩ65 mV in 10-mV increments. We found that the amplitude of control I Kv from WT HFD were significantly lower than that from WT LFD (Fig. 1Ai). Indeed, the current-voltage relationship of control I Kv revealed that current densities were smaller in WT HFD than in WT LFD at most voltages examined ( Impaired Stromatoxin Sensitivity in VSMCs and Arteries from WT Diabetic Mice-Previous studies indicate that K V 1.2, K V 1.5, and K V 2.1 subunits make predominant contributions to I Kv in VSMCs from cerebral and mesenteric arteries (8, 9, 26 -28). Thus, we examined whether impaired function of K V 1.X-and/or K V 2.1-containing channels contribute to suppressed I Kv in VSMCs during diabetes. This was achieved by selectively inhibiting either K V 1 channels with psora-4 (500 nM) (29) or K V 2 channels with ScTx (100 nM) (8,9). These agents have been used to evaluate the relative contribution of K V 1.X and K V 2 channels in the same VSMCs employed in the present study (8,9,30,31). I Kv were recorded from cerebral VSMCs as described above. The addition of psora-4 decreased I Kv at ϩ65 mV by ϳ30% in both LFD and HFD VSMCs (Fig. 1A, ii). However, the amplitude and current-voltage relationship of the psora-4-sensitive I Kv component was similar between these cells (Fig. 1, A and B, panels iii). Inhibition of K V 2 channels with ScTx becomes partial at positive potentials greater than ϩ30 mV (32). Thus, K V currents were evoked in the presence of ScTx by voltage steps from Ϫ85 to ϩ25 mV. Application of ScTx reduced I Kv at ϩ25 mV by ϳ70% in LFD and ϳ50% in HFD cells (Fig. 1, C and D, panels ii). Unlike psora-4-senstive I Kv , the amplitude of ScTx-sensitive I Kv was significantly smaller across a range of voltages in WT HFD compared with those in WT LFD (5.8 Ϯ 0.9 versus 14.2 Ϯ 2.0 pA/pF at ϩ25 mV, respectively; Fig. 1, C and D, panels iii; p Ͻ 0.05). Similar results were observed in mesenteric VSMCs (Fig. 1, B and D, insets). These data suggest that a reduction in I Kv observed in cerebral and mesenteric VSMCs from HFD mice is due to reduced ScTx-sensitive I Kv .

I Kv Function Is Suppressed in VSMCs from WT Diabetic Mice-
The functional significance of HFD-associated alterations in I Kv function was assessed by measuring psora-4 and ScTx sensitivity of arterial tone in cerebral arteries from WT LFD and HFD mice. At 60 mm Hg intravascular pressure, WT HFD arteries were significantly more constricted than WT LFD (Fig.  2, A and B). Whereas application of psora-4 (500 nM) caused a significant constriction in both WT LFD (13 Ϯ 2%) and HFD (17 Ϯ 4%) arteries (Fig. 2, A and C), ScTx (50 nM) induced constriction only in WT LFD arteries (Fig. 2, B and D). Indeed, ScTx had little effect on arteries from WT HFD (3 Ϯ 1%) as compared with LFD (11 Ϯ 2%; p Ͻ 0.05) mice. Arteries from all groups responded with robust constriction to 60 mM extracellular K ϩ concentration (Fig. 2E) or phenylephrine (14), suggest-ing that altered ScTx response was not due to differences in the magnitude of tone development. These results are consistent with impaired ScTx-sensitive I Kv contributing to enhanced arterial tone during diabetes.
Down-regulation of K V 2.1 Subunits in VSMCs during Diabetes-Impaired sensitivity to ScTx, but not psora-4, in diabetic cells and arteries could reflect selective down-regulation of K V 2.1 subunit expression. To test this, we evaluated K V 1.2, K V 1.5, and K V 2.1 transcript levels using quantitative PCR on single isolated cerebral and mesenteric VSMCs from WT LFD and HFD mice. We found that K V 2.1 transcript expression was ϳ65% lower in WT HFD than in WT LFD cells, with no detectable change in the expression of K V 1.2 and K V 1.5 transcripts (Fig. 3A). Consistent with this, K V 2.1, but not K V 1.2 or K V 1.5, protein levels were significantly reduced (ϳ75%) in WT HFD arterial lysates (Fig. 3B).
Immunofluorescence was also used to determine changes in K V 2.1 subunit expression in isolated cerebral VSMCs from WT LFD and HFD with a K V 2.1-specific antibody. As illustrated in Fig. 3C, we observed a strong intensity of the K V 2.1-associated FIGURE 2. Impaired ScTx-induced constriction in arteries from WT HFD mice. Representative diameter recordings from pressurized (60 mm Hg) WT cerebral arteries from LFD and HFD mice before and after (A) psora-4 (500 nM) or (B) ScTx (50 nM). C, bar plot summarizing arterial tone in WT LFD (n ϭ 5 arteries from 5 mice) and HFD (n ϭ 5 arteries from 5 mice) arteries in the absence (Ϫ) and presence (ϩ) of psora-4 (C). D, arterial tone in the absence (Ϫ) or presence (ϩ) of ScTx in WT LFD (n ϭ 7 arteries from 5 mice) and HFD (n ϭ 7 arteries from 6 mice) arteries. E, amalgamated data of arterial tone from WT LFD (n ϭ 8 arteries from 5 mice) and HFD (n ϭ 10 arteries from 5 mice) at 60 mm Hg in response to 60 mM extracellular K ϩ . *, p Ͻ 0.05.
fluorescence along the sarcolemma of this WT LFD cell. No K V 2.1-associated fluorescence was observed when the primary or secondary antibodies were excluded from the preparation. Conversely, the intensity of the K V 2.1-associated fluorescence along the sarcolemma, under the same experimental conditions, was markedly lower in cells from WT HFD mice (Fig. 3C). Altogether, these data indicate that suppression of ScTx-sensitive I Kv in VSMCs and impaired ScTx-induced vasoconstriction during diabetes results from down-regulation of K V 2.1 subunit expression.

AKAP150-anchored CaN and NFATc3 Activity Are Necessary for Impaired ScTx-sensitive I Kv Function and Down-regulation of K V 2.1 Subunit Expression during Diabetes-CaN and
NFATc3 are known to regulate transcription of K V 2.1 in VSMCs (18) and their activity is enhanced in diabetic arteries (14). Activation of CaN and CaN-mediated activation of NFATc3 in VSMCs of diabetic animals is dependent upon sarcolemmal phosphatase targeting by AKAP150 (18,33). Thus, to test the involvement of this pathway in the suppression of K V 2.1 in HFD cells, we first examined the role of AKAP150. To do so, we fed AKAP150-null (AKAP150 Ϫ/Ϫ ) mice either a LFD or HFD. Non-fasting blood glucose levels and body weight in AKAP150 Ϫ/Ϫ LFD and HFD mice were comparable with corresponding age-matched WT mice, and were elevated in HFD (Table 1).
We measured I Kv in VSMCs from AKAP150 Ϫ/Ϫ LFD and HFD mice before and after ScTx using the voltage protocol described above. VSMCs from AKAP150 Ϫ/Ϫ LFD mice produced robust I Kv (Fig. 4A) that were comparable with those observed in WT LFD cells, suggesting that AKAP150 does not influence basal I Kv in VSMCs of control mice. The average capacitance of AKAP150 Ϫ/Ϫ LFD and HFD cells was 17.7 Ϯ 1.2 and 16.4 Ϯ 0.6 pF, respectively, which is similar to WT cells (p Ͼ 0.05). In contrast to diabetic WT cells, the amplitude of control and ScTx-sensitive I Kv at ϩ25 mV in AKAP150 Ϫ/Ϫ HFD cells (13.0 Ϯ 2.0 pA/pF) was similar to the corresponding AKAP150 Ϫ/Ϫ LFD cells (14.0 Ϯ 2.0 pA/pF; Fig. 4, A and B). Likewise, no differences in ScTx-insensitive components were observed. Consistent with the electrophysiological data, K V 2.1 transcript and protein levels (Fig. 4, D and E) were similar in cells and lysates, respectively, from AKAP150 Ϫ/Ϫ LFD and HFD mice. Ablation of AKAP150 did not alter basal expression of K V 2.1 subunits (Fig. 4E). These data indicate that AKAP150 is required for suppression of K V 2.1 subunit expression and function in diabetic VSMCs.
Next, we directly tested the role of AKAP150-CaN interactions in K V 2.1 suppression during diabetes. To do this, we used VSMCs from LFD and HFD knock-in mice expressing mutant AKAP150 that is unable to bind CaN (⌬PIX) (14,19). Nonfasting glucose levels and body weight were significantly elevated in ⌬PIX HFD compared with ⌬PIX LFD mice (Table 1). Selective disruption of the AKAP150-CaN interaction abolished suppression of ScTx-sensitive I Kv function and K V 2.1 subunit expression during diabetes (Fig. 4, C-E). These results indicate that subcellular targeting of CaN by AKAP150 is necessary for suppression of K V 2.1 expression and function in VSMCs during diabetes.
The above findings are consistent with the concept that AKAP150-targeted CaN signals K V 2.1 transcriptional suppression via downstream activation of NFATc3. Therefore, we investigated the role of NFATc3 in suppression of K V 2.1 expression and function during diabetes by using NFATc3 null (NFATc3 Ϫ/Ϫ ) mice in LFD or HFD (Table 1). Fig. 5A demonstrated that NFATc3 Ϫ/Ϫ LFD and HFD myocytes produced robust I Kv . Consistent with a role for NFATc3, ScTx-sensitive I Kv function and K V 2.1 expression (Fig. 5, A-D) were similar in VSM and arteries from NFATc3 Ϫ/Ϫ LFD and HFD mice. Together, our data indicate that anchoring of CaN by AKAP150 mediates impairment of K V 2.1 expression and function during diabetes via downstream activation of NFATc3.
Loss of AKAP150, AKAP150-targeted CaN, or NFATc3 Prevents Enhanced Arterial Tone and Restores ScTx-sensitive Constriction in Diabetic Mice-We examined the functional relevance of the AKAP150-anchored CaN and NFATc3 signaling module on arterial tone and ScTx sensitivity. Diameter measurements at 60 mm Hg intravascular pressure were performed in cerebral arteries from AKAP150 Ϫ/Ϫ , ⌬PIX, and NFATc3 Ϫ/Ϫ mice in LFD and HFD. Unlike ScTx observations in arteries from WT mice (see Fig. 2, B and D), arterial tone development and ScTx-induced constriction were similar in LFD and HFD vessels from AKAP150 Ϫ/Ϫ , ⌬PIX, and NFATc3 Ϫ/Ϫ mice (Fig.  6, A-C). Altogether, these data indicate that disruption of the  (5-10 mM) and HFD (20 mM) mice. As expected, 20 mM D-glucose caused Ͼ60% reduction in the K V 2.1 transcript that was not observed when the non-metabo-lized L-glucose was substituted for D-glucose or when the CaN inhibitory peptide or VIVIT were present in cultured medium (Fig. 7A). These results suggest that activation of the CaN/ NFATc3 axis in VSMCs is necessary for transcriptional suppression of K V 2.1 during diabetic hyperglycemia.
In a previous study, we found that the regulatory BK Ca ␤1 subunit in VSMCs is down-regulated via activation of AKAP150-CaN/NFATc3 signaling in hyperglycemic animals on HFD leading to BK Ca channel impairment (14). Considering a potential common mechanism of AKAP150-CaN/NFATc3induced suppression of K V and BK Ca channels, an important question to be addressed is whether down-regulation of K V 2.1 and BK Ca ␤1 occurs on a similar time scale in response to hyperglycemic stimuli. To test this possibility, we examined the time course of K V 2.1 versus BK Ca ␤1 transcript suppression in FIGURE 4. AKAP150-anchored CaN is necessary for impaired K V 2.1 expression and function in HFD cells. A, representative whole cell K V currents from VSMCs isolated from LFD and HFD AKAP150 Ϫ/Ϫ mice recorded before and after application of ScTx (100 nM), and the corresponding ScTx-sensitive component (n ϭ 11 cells from 5 LFD mice and 9 cells from 5 HFD mice). Current-voltage relationships of I Kv before and after ScTx and the corresponding ScTx-sensitive component in LFD and HFD (B) AKAP150 Ϫ/Ϫ and (C) ⌬PIX cells. D, bar plot of K V 2.1 transcript levels in AKAP150 Ϫ/Ϫ and ⌬PIX HFD cells relative to LFD normalized to ␤-actin (n Ϸ 40 -50 VSMCs from 3 mice per condition). E, Western blot of immunoreactive bands of the expected molecular masses for K V 2.1 (ϳ110 kDa) and ␤-actin (ϳ43 kDa) in control, LFD, and HFD arteries from AKAP150 Ϫ/Ϫ and ⌬PIX mice (left) and corresponding (right) densitometry summary data (n ϭ 4 lysates from 4 mice per condition). *, p Ͻ 0.05.

AKAP150 Suppresses Vascular K V 2.1 Expression in Diabetes
VSMCs from WT arteries incubated at different time points in medium containing 10 or 20 mM extracellular D-glucose. We found that increasing the extracellular D-glucose concentration to 20 mM decreased K V 2.1 expression by ϳ60% within the first hour of incubation, but had no effect on BK Ca ␤1 (Fig. 7B). Transcript levels for both genes were subsequently down-regulated to a similar extent after 6, 12, 24, and 48 h in 20 mM D-glucose (Fig. 7B). Consistent with a role for AKAP150-anchored CaN in K V 2.1 and BK Ca ␤1 suppression during diabetic hyperglycemia, 20 mM D-glucose had no effect on transcript levels for either gene in VSMCs from ⌬PIX mice (p Ͼ 0.05; Fig.  7C). Together, these data suggest a distinct temporal profile of early K V 2.1 suppression during diabetic hyperglycemia, with subsequent concomitant down-regulation of both genes to a similar extent that is mediated by direct activation of the AKAP150-anchored CaN/NFAT signaling pathway.

DISCUSSION
In this study, we provide evidence implicating suppression of K V channel function as a significant contributor of enhanced arterial tone during diabetes (Fig. 7D). We report the following novel findings. First, I Kv in cerebral and mesenteric VSMCs are significantly suppressed during diabetes. Second, impaired I Kv in these cells leading to enhanced arterial tone during diabetes occurs due to selective down-regulation of K V 2.1 expression. Third, AKAP150-anchored CaN and NFATc3 are obligatory components in the signaling pathway underlying suppression of I Kv and K V 2.1 expression during diabetes. Fourth, although K V 2.1 and BK Ca ␤1 transcripts are reduced to a similar extent during established diabetic hyperglycemia, our data suggest that K V 2.1 suppression may occur at an earlier stage. The implications of these findings are discussed below.
Considerable data attribute impairment of endothelium-dependent vasodilatory mechanisms as a significant contributor to vascular dysfunction and enhanced arterial tone during diabetes (2)(3)(4)(5). However, abnormal VSM function may also critically contribute to enhancement of arterial tone during this pathological condition. For instance, enhanced arterial tone could result from a reduction in outward K ϩ conductance in VSMCs (34). Consistent with this idea, our initial experiments demonstrate that K V channel function is suppressed in diabetic VSMCs (Fig. 1). This discovery is significant, as a reduction in K V channel function will result in membrane depolarization, and increased activity of LTCCs, ultimately leading to elevated global [Ca 2ϩ ] i , and VSM contraction (6,7). Although an increase in global cytosolic [Ca 2ϩ ] i could promote activation of BK Ca channels to compensate for the loss of K V channel function via negative feedback control of V M depolarization, BK Ca channel activity is also suppressed in WT HFD mice (14). Thus, synergistic impairment of K V and BK Ca activity in diabetic animals may substantially reduce feedback membrane potential hyperpolarization leading to VSM contraction and enhanced arterial tone during diabetes (Fig. 7D). Ultimately, this could significantly contribute to raise mean arterial blood pressure or limit blood flow in diabetic patients.
In this study, we found that psora-4, a selective K V 1.X inhibitor (29 -31, 35), induced robust constriction in pressurized arteries from LFD and HFD mice. Yet, ScTx, which is a selective K V 2 inhibitor in VSM (8,9), had a reduced effect on WT HFD vessels indicating impaired K V 2 channel function during diabetes. Interestingly, we observed a trend toward a greater psora-4-induced constriction in pressurized arteries from WT HFD as compared with WT LFD mice, perhaps reflecting compensatory activation of additional K V 1.X at more depolarized membrane potentials in pressurized arteries from diabetic mice. This difference, however, was not statistically significant. Note that vasoconstriction of cerebral vessels in response to K V 2 or K V 1 channel blockers is independent of functional endothe-lium (8,13). Considering this, it follows that reduced endothelial vasodilatory function, as observed in several experimental models of diabetes, would not significantly impact the contribution of K V 2 channels to the regulation of arterial tone. Thus, altered response to ScTx during diabetes observed in this study likely reflects reduced expression of functional K V 2 channels. Consistent with this, ScTx-sensitive and psora-4-insensitive (presumably produced by K V 2.1 channels) current densities were significantly suppressed at most membrane potentials in HFD as compared with LFD cells. Conversely, the amplitude of the ScTx-insensitive currents, which presumably are predominantly mediated by K V 1.X, and psora-sensitive currents were not different in VSM isolated from LFD and HFD mice. These data indicate that K V 2.1, but not K V 1.X channel function is decreased in VSMCs from HFD mice. This conclusion is further supported at the molecular level upon demonstration of reduced K V 2.1, but not K V 1.2 or K V 1.5, transcript and proteins in HFD cells/arteries when compared with LFD cells/arteries. Altogether, these results are the first indication that selective K V 2.1 down-regulation underlies reduced I Kv in VSM contributing to enhanced arterial tone during diabetes, and point to impairment of K V function as a mediator of vascular dysfunction.
An interesting observation in this study is that K V 2.1 and BK Ca ␤1 genes are differentially suppressed during diabetic hyperglycemia (Fig. 7B). Our data indicate that within the first hour of incubation in a hyperglycemic solution that resembles the levels of plasma glucose in HFD mice, K V 2.1, but not BK Ca ␤1 expression, is significantly down-regulated. Subsequently, concomitant suppression of both genes occurs to a similar extent within 48 h of hyperglycemic exposure. These results suggest that K V 2.1 and BK Ca ␤1 genes have different thresholds for hyperglycemia-induced transcriptional suppression. Mechanisms of early temporal differences in K V 2.1 and BK Ca ␤1 expression during hyperglycemic stimuli are unclear, but may involve dissimilar thresholds for NFATc3-dependent suppression or differential mRNA degradation independent of the rate or magnitude of transcriptional suppression (36). Although future studies should further investigate this issue, it is intriguing to speculate that early initiation of K V 2.1 suppression could represent a key feature of pre-diabetic hyperglycemia, which drives pathophysiological engagement of the AKAP150-CaN/ NFATc3 signaling pathway via membrane depolarization, LTCC activity, and Ca 2ϩ influx (37,38). Subsequently, further NFATc3 nuclear accumulation reaches levels sufficient for BK Ca ␤1 down-regulation. Data demonstrating that disruption of the AKAP150-CaN interaction or specific inhibition of CaN and NFAT in VSMCs prevents the down-regulation of K V 2.1 and BK Ca ␤1 expression (see also Ref. 14) and impaired arterial tone during diabetic hyperglycemia and in HFD gives credence to this hypothesis. However, future experiments should investigate the relationship between NFATc3 activity and K V 2.1 and BK Ca ␤1 expression in further detail.
Our data support a mechanistic model whereby targeting of CaN by AKAP150 drives NFATc3 activation during diabetes. Once activated, NFATc3 translocates into the nucleus of VSMCs where it reduces K V 2.1 and subsequently BK Ca ␤1 expression (Fig. 7D). Consistent with a critical role for AKAP150-anchored CaN, disruption of the interaction between these two proteins prevented NFATc3 dephosphorylation and nuclear accumulation during diabetes (14). This is correlated with restoration of K V 2.1 (Figs. 4, 6, and 7) and BK Ca ␤1 subunit expression and function, and attenuation of blood pressure in diabetic AKAP150 Ϫ/Ϫ and ⌬PIX mice (Fig. 7, B and C, and Ref. 14).
Although our data argue against a role for AKAP150 in basal regulation of K V or BK Ca channel function (see Figs. 1 and 4 and Ref. 14), this anchoring protein is known to interact with LTCCs (39). Hence, AKAP150 may also function to position CaN near LTCCs (23) to efficiently activate Ca 2ϩ -dependent CaN/NFATc3 signaling during diabetes. Indeed, we have previously found that NFATc3 is preferentially activated in VSMCs by LTCC-dependent Ca 2ϩ microdomains (i.e. Ca 2ϩ sparklets (38,40,41)), which are significantly elevated during diabetes (42), rather than by elevations in global [Ca 2ϩ ] i (40). Similar increases in LTCC-dependent Ca 2ϩ microdomains have been observed in angiotensin II-induced hypertensive VSMCs (24,40) and after activation of reactive oxygen species (43). Thus, increases in local LTCC activity leading to activation of the AKAP-CaN/NFATc3 signaling pathway may repre-sent a wide ranging mechanism for development of vascular dysfunction in many pathological conditions. In the present model, NFATc3 nuclear accumulation and subsequent K V 2.1 and BK Ca ␤1 down-regulation is a dynamic process highly dependent on NFAT nuclear import and export rate (44), as well as non-fasting plasma glucose levels in diabetic animals that could be sufficient for NFAT activation.
In addition to transcriptional suppression of K V 2.1 subunit expression, post-translational modification of this (and other K V ) subunit may also contribute to a reduction in I Kv during diabetes (11)(12)(13)(45)(46)(47). Hence, multiple mechanisms may potentially synergize and contribute to impaired K V function, and vascular complications during diabetes. Note, however, that activation of divergent mechanisms may vary between vessels and animal models of diabetes. Thus, the relative contribution of transcription-dependent and -independent pathways to altered K V expression and function during diabetes warrants further investigation.
To summarize, our data demonstrate that suppression of K V channel function via selective down-regulation of K V 2.1 contributes to enhanced arterial tone during diabetes. AKAP150-CaN/NFATc3 signaling is central to impaired K V 2.1 expression and function. Our findings also support the view that activation of this signaling pathway may be a general mechanism for transcriptional regulation of K ϩ channels and that with K V 2.1, may be novel therapeutic targets to prevent and/or treat vascular complications during diabetes, and perhaps other pathological conditions.