Reductions in the Cardiac Transient Outward K+ Current Ito Caused by Chronic β-Adrenergic Receptor Stimulation Are Partly Rescued by Inhibition of Nuclear Factor κB.

The fast transient outward potassium current ( I to,f ) plays a critical role in the electrical and contractile properties of the myocardium. I to,f channels are formed by the co-assembly of the pore-forming (cid:5) -subunits, Kv4.2 and Kv4.3, together with theaccessory (cid:3) -subunitKChIP2.Reductionsof I to,f arecommon in the diseased heart, which is also associated with enhanced stimulation of (cid:3) -adrenergic receptors ( (cid:3) -ARs). We used cultured neonatal rat ventricular myocytes to examine

Fast cardiac transient outward potassium currents are generated by channels comprised of voltage-gated ␣-pore-forming subunits (which in humans/canines is predominantly Kv4.3, and in rodents is Kv4.2 and Kv4.3) and the accessory ␤-subunit KChIP2 (1). These currents play a critical role in early cardiac repolarization (1), excitation contraction-coupling (2,3), and arrhythmias (4). I to,f and its molecular constituents are invariably reduced in cardiac hypertrophy and disease (1), and heart disease is also characterized by elevations in catecholamines and consequently enhanced activation of ␤-adrenergic receptors (␤-AR) 6 (5). Although chronic ␤-AR stimulation has been shown to decrease I to,f (6), the molecular mechanisms underlying regulation of I to,f by ␤-ARs remain unclear.
I to,f is regulated by several pathways activated in the diseased myocardium, such as calcineurin, nuclear factor-activated T-cells (NFAT) (6,7) and mitogen activated protein kinases (8). Although calcineurin/NFAT signaling regulates I to,f in a manner that is model dependent (6,9,10), this pathway does not mediate ␣-AR-induced changes in I to,f or its molecular subunits (9,10). We previously showed that ␣-AR stimulation activates the transcription factor nuclear factor B (NF-B), which mediates reductions in I to,f via strong repression of KChIP2 expression (9,10). The results of several studies also suggest that Ca 2ϩ /calmodulin-dependent protein kinase (CaMKII) may also regulate I to,f . For example, mice with transgenic overexpression of CaMKII␦c have reduced levels of I to,f as well as Kv4.2 and KChIP2 subunits (11), although these changes may be related to concurrent heart disease. In addition, CaMKII␦c activity is increased in human (12) and animal models (13) of heart disease and CaMKII has been shown to mediate NF-B activation (14). Another cardiac isoform of CaMKII, CaMKII␦b, has also been implicated in disease signaling path- . The authors declare that they have no conflicts of interest with the contents of this article. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. 1 Both authors contributed equally to the manuscript. 2 Supported by fellowships from H&SF of Canada and the H&SF/Richard Lewar Centre of Excellence. 3 Supported by NIH Grant T32GM007752. 4 Holds a Canada Research Chair in Molecular Cardiology. 5  ways (15,16), although its potential regulation of I to,f has not been investigated.
Although the regulation of I to,f has been studied in a number of experimental and disease models, the direct link between ␤-AR stimulation and I to,f has not been investigated. Therefore, we used cultured neonatal rat ventricular myocytes (NRVM) to examine the effects of the ␤-AR agonist, isoproterenol (ISO), on I to,f . Our data show that chronic ␤-AR stimulation reduces I to,f as well as mRNA levels for all three of the molecular subunits underlying I to,f channels. The reductions in I to,f and KChIP2 mRNA expression were rescued by NF-B inhibition. Stock concentrations of isoproterenol (ISO) (50 mol/liter) (Sigma) were dissolved in Milli-Q water and stored at Ϫ20°C. Multiple freeze/thawing was avoided by using small aliquots, and new stocks were prepared regularly.

Experimental Procedures
Isolation, Culture, and Treatment of Neonatal Rat Ventricular Myocytes (NRVM)-All studies were in accordance with the guidelines set by the Canadian Council on Animal Care and the University of Toronto. Hearts from 1-2-day-old Sprague-Dawley rats were rapidly excised after cervical dislocation and immediately placed in chilled CBHFF buffer. The ventricles were isolated by removing the atria, and were subsequently washed five times in CBHFF. The ventricles were gradually minced and placed in CBHFF buffer containing 1 mg/ml of trypsin (Life Technologies) at 4°C for 3 h with gentle rocking. To dissociate myocytes, minced ventricles were washed and placed in 5 ml of pre-warmed buffer containing 1 mg/ml of trypsin plus 0.15 mg/ml collagenase type II (Worthington) and stirred for 10 min 3-5 times. NRVM were pelleted at 1000 rpm for 5 min, resuspended in Dulbecco's Modified Eagle Medium (DMEM) and Ham F12 (1:1 v/v) containing 5% fetal bovine serum (FBS) (Invitrogen), and pre-plated for 1 h at 37°C, and 4 ϫ 10 5 myocytes were plated on 35-mm dishes containing glass coverslips coated with 0.6% gelatin for electrophysiology studies. For mRNA experiments, 7.0 ϫ 10 5 myocytes were plated on 60-mm dishes. NRVMs were cultured in 5% FBS/ DMEM/F12 containing 0.1 mmol/liter bromodeoxyuridine (BrdU) (Sigma) and penicillin-streptomycin. The medium was changed to DMEM/F12 (without serum or BrdU), supplemented with 1ϫ insulin-transferrin-selenium-X (Life Technologies), 25 g/ml ascorbic acid and 1 nmol/liter LiCl after ϳ21 h and then treated (or not) with 50 nmol/liter ISO. The medium and treatments were replaced daily, and after 40 h in serum free supplemented media and treatment NRVMs were harvested for quantitative real-time PCR (qRT-PCR) or used for electrophysiology.
Electrophysiology-The whole-cell configuration (17) was used to record ionic currents from NRVM and all experiments were performed at room temperature. The electrophysiological setup was as described previously (9,18). Glass obtained from World Precision Instruments (TW150F-4) was used for patch pipettes to generate a resistance ranging from 2-5 M⍀. Patch pipettes were filled with K INT solution. Modified Tyrode solution (see Solutions) continuously superfused NRVMs and only beating cells were used for recordings. Whole-cell currents were digitized at 10 kHz and filtered at 2 kHz (Molecular Devices, 200B amplifier). Applied voltages were not corrected for a liquid junction potential of Ϫ5 mV. The following protocol was designed to exclusively evoke the fast transient outward potassium currents (I to,f ) since I to,f recovers more quickly from inactivation compared with I to,s (19). The protocol is as follows: a 150-ms step to Ϫ80 mV, from a holding potential of Ϫ20 mV, followed by a 50-ms pre-pulse to Ϫ40 mV (to discharge the Na ϩ current) and then a series of 500-ms depolarizing steps from ϩ60 mV to Ϫ40 mV. I to,s was not examined in this study. Subtracting the peak outward current from the current at the end of the depolarizing step was used to determine I to,f .
Quantitative RT-PCR-MIQE guidelines (20) were followed for sample preparation and qRT-PCR. Each sample group contained 3 replicates and the control samples were infected with GFP only without any drug treatment. Following the manufacturer's recommendations, TRIzol Reagent (Life Technologies) was used to isolate RNA from cultured NRVM. To assess the quality of the mRNA samples we used a NanoDrop spectrophotometer (Thermo Scientific) to measure absorption ratios of 260 nm and 280 nm (A 260 /A 280 Ͼ1.8) and we also examined the 28S/18S rRNA ratio on a 1.2% formaldehyde gel. The cDNA was synthesized from 1-2 g of RNA (Superscript TM III Reverse Transcriptase, Life Technologies) and stored at Ϫ80°C. SYBR green (Life Technologies) was used to quantify the quantities of Kv4.3 and Kv4.2 as well as GAPDH (reference gene), while a TaqMan master mix (Life Technologies) was used for KChIP2. The Viia 7 Real-Time thermal cycler PCR System (Life Technologies) was used to amplify the cDNA. The protocol consisted of the following four stages: 50°C for 2 min, 95°C for 10 min, 95°C for 15 s followed by 60°C for 1 min, repeated 40 times, and a final dissociation step at 95°C for 15 s followed by 60°C for 15 s. All "no template control" wells were negative in all PCR runs.
The following primer/probes were used: GAPDH: 5Ј- The Viia 7 Software version 1.2 (Life Technologies) was used to analyze qRT-PCR results. The ⌬⌬Ct method with GAPDH as the reference gene was used to quantify relative mRNA levels of the triplicate samples (21). Dissociation curves and agarose electrophoresis were used to confirm specificity. We also used 18S rRNA (Life Technologies) levels to normalize the amplicon levels. The results were identical when normalization was performed using GAPDH compared with 18S rRNA.
Western Blot Analysis-Dishes containing NRVM were washed with PBS and cells were scraped and suspended in the same buffer. The cells were then centrifuged at 1000 ϫ g for 5 min, and the supernatant was discarded. Cells were lysed using buffer consisting of 50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate, 10% protease inhibitor mixture (P8340, Sigma) at 4°C for 1 h. After centrifugation for 15 min at 12,000 ϫ g, supernatants were extracted and protein concentration was determined using DC Protein Assay kit (#500-0112, Bio-Rad). Exactly 20 g of each sample was resolved by 10% SDS-PAGE and transferred to nitrocellulose membranes. Blots were incubated with primary antibodies: Kv4.2 (1:1000, ab123543, abcam), KChIP2 (1:1000, ab99041, abcam), GAPDH (1:2000, abs16, EMD Millipore) for 2 h at room temperature. For Kv4.2 blots, anti-rabbit secondary (1:2000, 7074S, Cell Signaling) was used, and blots were incubated at room temperature for 1 h. For KChIP2 detection, anti-mouse secondary antibody (1:2000, 7076S, Cell Signaling) was used. For GAPDH detection, blots for Kv4.2 and KChIP2 were first stripped using commercial stripping buffer (ST010, GeBa) according to manufacturer's instructions. After incubation with primary antibody for GAPDH, anti-rabbit secondary antibody was used and blots were incubated at room temperature for 1 h. Amersham Biosciences ECL Prime Western blotting Detection Reagent (RPN2232, GE Healthcare) was used for detection and analysis of bands was performed using densitometry plugin for ImageJ software. For statistical analysis, the ISO treated samples were compared with the control (normalized to 1.0) for each experiment. For data presentation, errors are shown for controls to relative inter-experimental variation.
Luciferase Assay-NRVMs were transfected with a KChIP2 promoter luciferase reporter plasmid containing binding elements for KChIP2 (KChIP2 luc) (9) with or without eukaryotic expression plasmids encoding wild type p65 NF-B or infected with adenovirus encoding a phosphorylation defective mutant of IB␣ designated (IB␣SA) as we reported (23). After 24 h transfection or infection, myocytes were treated with isoproterenol (ISO, 10 mol/liter) for 24 h. Cardiac cell lysates were harvested for luciferase assay. Luciferase activity was normalized to ␤-galactosidase activity to control for any differences in transfection efficiency among the groups.
Data Analysis and Statistics-Excel 2011 (Microsoft and Clampfit 10.2 (Molecular Devices) were used to analyze data, while Prism 5 (GraphPad) was used for statistical analysis. Statistical tests were considered significant at the p Ͻ 0.05 levels, and data were presented as means Ϯ S.E. of the mean. To test for differences between two sets of data, an unpaired Student's t test was used. For multiple comparisons, a one-way analysis of variance (ANOVA) test was performed and, if significant differences were observed, a Bonferroni post-test was used to determine which means differed.
A two-way ANOVA analysis was performed to test for an interaction between the drug (ISO) and a particular inhibitor (e.g. IB␣SA, CAIN, CAMKII␦b, or CAMKII␦c). The two-way ANOVA was set up as follows: GFP (control), GFPϩISO, ISOϩinhibitor and inhibitor alone, and a particular parameter (e.g. current or mRNA) was measured for each condition. A significant interaction (p Ͻ 0.05) suggested that the effects of ISO in combination with an inhibitor (i.e. ISOϩinhibitor) on a particular parameter were different than the effects GFPϩISO and the inhibitor alone on the same parameter.

Results
To understand how ␤-AR stimulation regulates I to,f , we treated cultured rat neonatal ventricular myocytes (NRVMs) with 50 nmol/liter isoproterenol (ISO) for 48 h. These NRVMs were infected with adenoviruses expressing either GFP or various constructs to interfere with selected cell signaling pathways (see below). To quantify I to,f , cells were held at a holding potential of Ϫ20 mV and then repolarized to Ϫ80 mV for 150 ms prior to the application of depolarizing voltage steps to activate I to,f . As explained further in the "Discussion," this protocol successfully permits nearly full inactivation recovery of I to,f (i.e. Ͼ99%) while minimizing the appearance of the overlapping currents generated by I to,s in these cells, thereby allowing accurate quantification of I to,f (19). As shown in Fig. 1, ISO decreased (p Ͻ 0.001) I to,f current densities by 62% in association with marked reductions in the mRNA levels of all three of its molecular constituents KChIP2 (p Ͻ 0.001), Kv4.2 (p Ͻ 0.01), and Kv4.3 (p Ͻ 0.001). These changes in I to,f and mRNA were associated with reductions of Kv4.2 (p Ͻ 0.05) and KChIP2 (p ϭ 0.062) protein (Fig. 2). Unfortunately, we could not reliably quantify Kv4.3 protein in NRVMs due to non-specificity of commercially available antibodies.
Previous studies have established that ␤-AR stimulation leads to CaMKII activation (16,24,25) which may underlie the changes in I to,f seen with CaMKII activation. We found that ␤-AR stimulation, over a time course of 30 and 60 min (standard for this assay), increases the CaMKII activity in NRVMs (Fig. 3), consistent with previous studies (25). Next, we determined whether CaMKII activation mediates the changes in I to,f induced by ␤-AR stimulation. This was initially examined by overexpressing the dominant-negative CaMKII␦c construct (dn-CaMKII␦c) to inhibit CaMKII␦c (26), a major cardiac isozyme of CaMKII (27). Overexpressing dn-CaMKII␦c caused increases in I to,f (p Ͻ 0.01) in the absence of ISO (Fig. 4), establishing that I to,f is regulated by CaMKII␦c under basal conditions. Interestingly, in the presence of ISO dn-CaMKII␦c caused even further reductions (p Ͻ 0.05) in I to,f (Fig. 4). These findings support the conclusion that CaMKII␦c does not mediate the ISO-induced I to,f changes, although CaMKII␦c does regulate I to,f . Indeed, the pattern of changes in the mRNA expression of the molecular constituents of I to,f , induced by CaMKII␦c inhibition, was similar in the presence and absence of ␤-AR stimulation (Fig. 4C). Although complex, the link between the changes in I to,f and its molecular constituents seen in these studies can be readily understood by considering the molecular requirements for channel production, as suggested previously (see "Discussion") (9,10).
To examine the role of the other major cardiac isoform of CaMKII, CaMKII␦b, we also performed studies using dn-CaMKII␦b to inhibit CaMKII␦b. In contrast to CaMKII␦c, CaMKII␦b inhibition decreased I to,f (Fig. 5) in the absence of ISO. Moreover, dn-CaMKII␦b also failed to reverse the effects of ISO on I to,f (p ϭ 0.82) (Fig. 5), suggesting that CaMKII␦b does not mediate the effects of ISO on I to,f . Similar to CaMKII␦c inhibition, the pattern of mRNA changes of either KChIP2 (which was elevated) or Kv4.2 (which was reduced), induced by dn-CaMKII␦b overexpression, did not depend on whether ␤-ARs were stimulated. Interestingly, CaMKII␦b did, however, prevent the Kv4.3 and KChIP2 reductions induced by ␤-AR stimulation when ISO was present, suggesting that CaMKII␦b participates in the ISO-mediated reductions of Kv4.3 and KChIP2. Consistent with this suggestion, a two-way ANOVA analysis revealed positive interactions (p Ͻ 0.05) between the actions of ISO and dn-CaMKII␦b for Kv4.3 and KChIP2 subunit expression. Taken together, these findings support the conclusion that the ISO-mediated reductions of Kv4.3 and KChIP2 mRNA involve CaMKII␦b, whose effects appear to impact minimally on the I to,f reductions induced by ␤-AR stimulation. Again, the complex dependence of I to,f channel production on its molecular constituents is discussed more fully in "Discussion" (9, 10).   Data were expressed as a percentage of the maximal CaMKII activity in order to normalize for differences in total CaMKII protein present in each NRVM sample. Activity of CaMKII is increased in response to ISO at two time-points, 30 and 60 min.*, p Ͻ 0.05 versus vehicle.
Although our results with dn-CaMKII␦b and dn-CaMKII␦c support the conclusion that the reductions in I to,f induced by ISO are not mediated primarily by CaMKII, it is conceivable that these two CaMKII isozymes act together to regulate I to,f following ␤-AR stimulation. Therefore, we also co-transfected NRVMs with dn-CaMKII␦c and dn-CaMKII␦b. Unfortunately, dual-infections caused extensive cell death, thus preventing reliable investigations into the changes in I to,f and its molecular components induced by   ISO. Thus, we cannot draw any conclusions regarding the interactions between these two CaMKII isozymes.
To understand the molecular basis for the decreases in I to,f induced by ISO, we next considered calcineurin, a Ca 2ϩ -dependent phosphatase which is activated in response to ␤-AR stimulation. Consistent with previous studies (9, 10), we found that calcineurin inhibition with CAIN (an endogenous inhibitor), under baseline conditions (i.e. without ISO), caused I to,f reductions (p Ͻ 0.01) in association with reductions in both KChIP2 (p Ͻ 0.05) and Kv4.2 (p Ͻ 0.01) mRNA expression and slight elevations (p Ͻ 0.05) in Kv4.3 expression (Fig. 6). More importantly, when ISO was present, CAIN caused further reductions in I to,f (p Ͻ 0.05), as well as KChIP2 (p Ͻ 0.05) with no effect (p ϭ 0.53) on Kv4.3 mRNA levels, and Kv4.2 tended to decrease (p ϭ 0.07). These results demonstrate that the I to,f reductions seen with ␤-AR activation are not mediated by calcineurin, even though calcineurin, like CaMKII␦b and CaMKII␦c, can regulate the levels of I to,f and its molecular constituents in both the presence and absence of ␤-AR stimulation.
We previously demonstrated that NF-B mediates ␣-AR-induced reductions in I to,f via down-regulation KChIP2 (10). To explore the potential involvement of NF-B in ␤-AR regulation of I to,f , NF-B signaling was inhibited using the phosphorylation-deficient mutant of IB␣ (IB␣SA) (23). As shown previously (10), NF-B inhibition with IB␣SA had no effect on baseline I to,f but caused elevations (p Ͻ 0.001) in KChIP2 expression along with reductions (p Ͻ 0.001) in Kv4.2 expression and moderate increases (p Ͻ 0.01) in Kv4.3 (Fig. 7). In addition, IB␣SA diminished (p Ͻ 0.05) the extent of I to,f reductions induced by ISO in correlation with restoration of KChIP2 mRNA to basal levels (p Ͻ 0.05), without affecting either Kv4.2 or Kv4.3 expression (Fig. 7). These results support the hypothesis that NF-B partly mediates reductions in I to,f induced by ␤-AR stimulation. Interestingly, consistent with this conclusion, two-way ANOVA analysis revealed an interaction (p Ͻ 0.05) between ISO and IB␣SA for the regulation of KChIP2 mRNA expression, with this trend (p ϭ 0.063) towards an interaction for I to,f . Furthermore, the KChIP2 promoter activity in NRVMs, assessed via luciferase reporter measurements, was reduced by forced over-expression of the NF-B subunit p65, (p Ͻ 0.001) as well as by ISO treatment (p Ͻ 0.05). Conversely, inhibition of p65 NF-B with IB␣SA increased luciferase activity and abrogated the inhibitory effects of ISO (p Ͻ 0.05) (Fig. 8A). ISO treatment also increased the phosphorylation status of p65 subunit (Fig. 8B), which has been shown to increase the transcriptional activity of NF-B (28,29).

Discussion
To study the effects of ␤-AR stimulation on I to,f , it was essential to quantify and separate I to,f from other potentially overlapping currents in our patch-clamp studies. To accomplish this goal, we quantified I to,f using voltage protocols that exploited the vast differences in recovery from inactivation that exist between I to,f and I to,s , which have -values for recovery from inactivation at ϳ45 ms and ϳ3800 ms, respectively (19). Specifically, cardiomyocytes were held at Ϫ20 mV in order to ensure complete inactivation of both I to,f and I to,s after which a brief 150-ms pulse was applied to Ϫ80 mV, thereby assuring that Ͼ99% of I to,f has recovered from inactivation with minimal recovery (ϳ3.8%) of I to,s (19). Consequently, subsequent depolarization steps (i.e. Ϫ40 to ϩ60 mV) to activate I to,f ensured that the peak currents gave reliable estimations of I to,f . We con- ␤-Adrenergic Regulation of I to FEBRUARY 19, 2016 • VOLUME 291 • NUMBER 8

JOURNAL OF BIOLOGICAL CHEMISTRY 4161
tend that these I to,f measurements, along with Kv4.2/4.3 and KChIP2 mRNA levels, in our cultured NRVMs provide trustworthy data on the regulation of I to,f by ␤-AR stimulation.
Previous studies suggest that the Ca 2ϩ /calmodulin-dependent protein kinase (CaMKII) may regulate I to,f (16,24,25). In addition, in heart disease that is characterized by enhanced activation of ␤-adrenergic receptors (␤-AR) (5) is associated with both CaMKII activation (12) and reduced I to,f (1). Consistent with these previous observations, we found that ␤-AR activation for 48 h in NRVMs caused profound I to,f reductions, along with decreased mRNA levels of its three molecular subunits (i.e. KChIP2, Kv4.2, and Kv4.3). We also observed ϳ50% reductions in both Kv4.2 and KChIP2 protein levels with ISO, although the changes in KChIP2 did not quite reach significance.
Although inhibition of either CaMKII or calcineurin could not rescue the ISO-dependent reductions in I to,f , despite a clear ability of these pathways to regulate the mRNA expression of the molecular components of I to,f , NF-B inhibition did partly reverse the I to,f reductions induced by ISO, in association with restoration of the KChIP2 subunit mRNA, but not the Kv4.2 or Kv4.3 subunits (summarized in Fig. 9). These results support the conclusion that NF-B activation contributes to the ISOmediated I to,f reductions by suppressing KChIP2 mRNA expression. In support of this, a two-way ANOVA analyses established an interaction between NF-B inhibition and ISO application in the regulation of KChIP2 expression, while a strong trend toward an interaction between NF-B inhibition and ISO was also observed for I to,f currents. Further evidence for our conclusion is provided by the observation that overexpression of the p65 NF-B subunit reduced KChIP2 promoter activity while ␤-AR stimulation increased phosphorylation of p65 NF-B and depressed KChIP2 promoter activity. Notably, the inhibitory effects of NF-B on KChIP2 transcription were abrogated by inhibition of NF-B with IBSA.
The involvement of NF-B in ␤-AR-mediated effects on I to,f is consistent with previous studies showing that KChIP2 facili-   tates insertion of I to,f channels into cell membranes (30) while also limiting channel degradation (31). Importantly, NF-B inhibition had no measureable effects on I to,f densities in the absence of ␤-AR stimulation, despite causing elevations of KChIP2 mRNA levels under these conditions. These differences in I to,f responses associated with the NF-B-mediated alterations in KChIP2 mRNA (observed with and without ␤-AR stimulation) can be explained by considering the connection between channel production and the absolute levels of the various subunits. Specifically, previous studies (10, 32) have concluded that Kv4.2 is the major limiting factor for channel production under baseline conditions while KChIP2 becomes far more influential as a limiting factor when Kv4.2 expression is low (32), as seen with ␤-AR stimulation. This interpretation is also consistent with our CaMKII␦b/c and calcineurin inhibition studies, as discussed below.
Although acute stimulation of ␤-ARs activated CaMKII (Fig.  3), we found that inhibition of neither CaMKII␦b nor CaMKII␦c was able to restore the I to,f reductions caused by chronic ␤-AR stimulation. Nevertheless, our results establish that CaMKII␦b and CaMKII␦c oppositely regulate I to,f as well as Kv4.2 and KChIP2 under baseline conditions (i.e. in the absence of ISO). Specifically, CaMKII␦b inhibition caused profound elevations of KChIP2 and reductions in Kv4.2 while CaMKII␦c elevated Kv4.2 and reduced KChIP2 (as well as Kv4.3). Again, largely the same patterns were seen with inhibition of either CaMKII␦b or CaMKII␦c when ␤-ARs were stimulated. We acknowledge that it is conceivable that CaMKII is not actually activated with chronic ␤-AR stimulation (48 h), despite being activated following acute ISO treatment (Ͻ1 h) (Fig. 3). However, our two-way ANOVA analysis establishes that interactions occur between dn-CaMKII␦b and ISO treatment in the regulation of both Kv4.3, and KChIP2 mRNA levels. Thus, our studies are consistent with the notion that CaMKII␦b participates in the ISO-induced reduction in both the KChIP2 and Kv4.3 subunits (summarized Fig. 9).
The pattern of changes in I to,f and its molecular constituents associated with CaMKII inhibition warrant further brief discussion. As already mentioned with respects to our NF-B results, the precise connection between the changes in mRNA expression and I to,f densities induced by CaMKII␦b or CaMKII␦c inhibition are complex. Nevertheless, our findings are generally consistent with previous studies and show remarkable internal consistency. For example, under basal conditions, the increase in I to,f seen with CaMKII␦c inhibition correlated with increases in Kv4.2 and occurred despite reductions in KChIP2, which is expected if Kv4.2 subunits are the limiting factors under baseline conditions, as reported previously (9,32,33). On the other hand, the I to,f levels seen in the presence of ISO are actually reduced further by CaMKII␦c inhibition which can be readily explained if KChIP2 becomes the rate-limiting factor for channel production when KChIP2 levels fall excessively, an assumption consistent with previous studies in mice showing that I to,f is completely eliminated when KChIP2 is ablated (34). Similar patterns are seen with CaMKII␦b inhibition. Specifically, CaMKII␦b inhibition in the absence of ISO results in reduced I to,f , which is expected from the profound reduction in the mRNA levels of the Kv4.2 subunits (which are channel limiting), despite elevated KChIP2. Similarly, the elevations in KChIP2 induced by CaMKII␦b inhibition are unable to increase I to,f in the presence of ISO because of the low levels of Kv4.2 mRNA under these conditions.
We also found that calcineurin did not mediate the I to,f reductions induced by ␤-AR stimulation, despite the observation that calcineurin regulates I to,f and its molecular constituents in the presence and absence of ␤-AR stimulation. Our results confirm previous studies showing that calcineurin and NFAT transcriptional factors regulate I to,f and its molecular constituents. In particular, we found that calcineurin inhibition with CAIN reduced I to,f as well as KChIP2 and Kv4.2, but not Kv4.3, in both the presence and absence of ISO. The precise quantitative concordance between changes in I to,f and its molecular constituents with calcineurin inhibition was complex and, because they essentially followed the patterns discussed for CaMKII␦b/c and NF-B, are not discussed further. More importantly, the similar responses in the presence and absence of ␤-AR stimulation when calcineurin was inhibited supports the conclusion that calcineurin does not mediate the I to,f changes seen with chronic ␤-AR stimulation.
Despite our very interesting and compelling findings, we cautiously interpreted our data with the following caveats. First, we used NRVMs as a model instead of adult myocytes. While adult myocytes have advantages in terms of maturity, NRVMs were used in our study because they have been utilized previously in many studies for exploring multiple cardiac signaling pathways (8,9,18,30,33). In addition, NRVMs are more amenable to viral infection and maintenance over a several days without significant de-differentiation, which is the case with adult myocytes. Second, it is important to note that we did not measure Kv1.4 mRNA in these experiments, and Wagner et al. found that Kv1.4 subunits were up-regulated in the hearts of mice overexpressing CaMKII␦c (11). Indeed, Kv1.4 channels form the basis for I to,s , which recovers slowly from inactivation, compared with I to,f (Kv4.2/4.3 channels), and thus the modulation of I to,s by ␤-AR simulations and CaMKIIs warrants further investigation. Finally, despite the finding of clear correlations between subunit mRNA and protein levels, our major conclusions are based on mRNA measurements alone. This was our goal since the mediators we tested are transcriptional modulators and thus we expect alterations in mRNA of I to,f subunits. While it is conceivable that some post-translational changes occur with our molecular interventions, the principal points of the study are supported by both I to,f and subunit mRNA measurements. Additional studies will be needed to determine post-translation changes in I to,f channel proteins in the setting of CaMKII and NF-B inhibitions.
In conclusion, several new findings arise from our studies. NF-B contributes to the I to,f changes following ␤-AR stimulation in correlation with changes in KChIP2 mRNA expression. Additional studies will be needed to determine whether modulation of NF-B will be an effective treatment for prevention of deleterious reductions in I to,f channels caused by heart disease. We also found, for the first time, that the CaMKII isoforms oppositely regulate I to,f and its molecular constituents in the absence of ␤-AR stimulation, while CaMKII␦b contributes partly to changes in Kv4.3 and KChIP2 mRNA expression induced by with ISO.