Deleting the accessory subunit KChIP2 results in loss of I_to,f and increased I_K,slow that maintains normal action potential configuration

Abstract

Background

Four voltage-gated potassium currents: I_to,f (K_V4.2), I_to,s (K_V1.4), I_K,slow (K_V1.5+K_V2.1), and I_SS (TASK1) govern murine ventricular repolarization. Although the accessory subunit, KChIP2, influences I_to,f expression, in preliminary experiments we found that action potential duration (APD) is maintained in KChIP2 knockout mice.

Objective

We tested the role of KChIP2 in regulating APD and studied the underlying ionic currents.

Methods

We used microelectrode techniques, whole-cell patch clamp studies, and real-time PCR amplification to characterize ventricular repolarization and its determinants in WT and KChIP2^-/- mice.

Results

Despite comparable baseline action potentials, APD was more markedly prolonged by 4-aminopyridine (4-AP) in KChIP2^-/- preparations. Peak K⁺ current densities were similar in WT and KChIP2^-/- cells (mean±sem I_P: 28.3±2 (n=27) vs. 29.2±2 pA/pF (n=24), respectively; P>0.05). Heteropodatoxin-2 (HpTx-2, 1 μM) had no effect on current amplitude in KChIP2^-/- myocytes. The current fractions sensitive to 4-AP (50 μM and 1 mM) were larger in KChIP2^-/- than WT (P<0.05). Real-time PCR demonstrated absence of KChIP2 and increased K_V1.5 expression in KChIP2^-/- ventricular myocardium.

Conclusion

KChIP2 deficiency eliminated HpTx-2-sensitive I_to,f, but had little impact on total APD, secondary to upregulation of 4-AP-sensitive I_K,slow in association with increased K_V1.5 expression. There is increased sensitivity to 4-AP-mediated APD prolongation in KChIP2^-/-. Thus, KChIP2 appears important for murine repolarization in circumstances of reduced repolarization reserve.

Introduction

Voltage-gated potassium (K_V) currents are important determinants of mammalian myocardial repolarization that are differently regulated in a variety of cardiovascular pathologies.^1-4 In mice, at least 4 pharmacologically, molecularly and kinetically distinct K_V currents have been identified^{5, 6}: I_to,f (encoded by K_V4.2 and K_V4.3⁷), I_to,s (K_V1.4⁷), I_K,slow (K_V1.5⁸ and K_V2.1⁹), and I_SS (possibly involving the K_2P channel TASK1¹⁰). In contrast to larger animals, K_V4.3 is not required for functional I_to,f in the mouse.¹¹

The K_V gene family encodes pore-forming α subunits, four of which comprise the functional channel. A number of accessory subunits contribute to the physiologically native current.¹² Of the K_V channel-interacting proteins (KChIPs) which primarily bind to the cytoplasmic N-terminal domain of the K_V4 α subunits^{13, 14}, only KChIP2 is expressed in heart.^15-17 In mammalian cell lines, KChIP2 co-expression increases K_V4-encoded current density, slows inactivation, and accelerates recovery from inactivation,^{15, 16} and also modulates K_V1.5 expression.¹⁸

In human subjects, heart failure is associated with decreased I_to¹⁹, explained by decreased K_V4.3 and KChIP2 mRNA and protein levels.^{20, 21} A mutation in the pore region²² or deletion of the pore-forming segment²³ of K_V4.2 produce dominant-negative subunits that reduce I_to,f and prolong action potential duration (APD) of murine myocytes. The pore deletion is associated with ventricular hypertrophy that progresses to heart failure, whereas the pore mutation does not lead to functional or structural dysfunction.^{22, 23} Deleting the entire K_V4.2 gene completely abolishes I_to,f; however neither APD prolongation nor hypertrophy is observed.²⁴

Although KChIP2 knockout reportedly leads to loss of I_to in murine cardiomyocytes²⁵, our preliminary action potential recordings from intact ventricles were comparable to those in wild-type (WT) mice. Therefore, we designed the present studies to explore the K_V-subtype specific consequences of KChIP2 deficiency. Our results demonstrate that while I_to,f is eliminated, total K_V-current density remains unaltered. This is secondary to an upregulated 4-aminopyridine (4-AP) sensitive current, which translates to increased sensitivity to 4-AP-mediated APD prolongation.

Methods

Hearts were excised from anesthetized (0.1 mg/g ketamine, 0.01 mg/g xylazine, IP), adult (10-12 weeks), male C57BL6 WT control and KChIP2^-/- mice. Genotyping was performed using DNA isolated from tail samples (Allele Biotech, CA) followed by PCR amplification of KChIP2 and neomycin. All experiments were approved by the Institutional Animal Care and Use Committee and conformed to the Guide for the Care and Use of Laboratory Animals (US National Institutes of Health).

Transmembrane action-potential recordings

Multicellular left- (LV) or right- (RV) ventricular free wall preparations were superfused with oxygenated Tyrode's solution containing (mM): NaCl 131; NaHCO₃ 18; KCl 4; CaCl₂ 1.8; MgCl₂ 0.5; NaH₂PO₄ 1.8; dextrose 5.5 (pH 7.4; T 37°C). High resistance (>30 MΩ) microelectrodes filled with 3M KCl were used to record (sub)epicardial action potentials from preparations paced at 500 ms cycle length via bipolar surface electrodes. After 2h equilibration, we recorded action potentials from 9-12 sites per preparation. Recordings were repeated 30 min after adding 1 mM 4-AP to the superfusate.

Patch-clamp recordings

Myocytes were disaggregated from the entire LV free wall by enzymatic dissociation and mechanical dispersion. Cells were maintained in enzyme-free solution at room temperature and used within 6h of isolation. Recording pipettes (1-2 MΩ) contained (mM): KCl 135; EGTA 10; HEPES 10; and glucose 5 (pH 7.2). The superfusate contained (mM): NaCl 136; KCl 4; MgCl₂ 2; CaCl₂ 1; CoCl₂ 5; tetrodotoxin (TTX; Sigma, St. Louis, MO) 0.01, HEPES 10; glucose 10 (pH 7.4; T 24°C). Cell capacitances were 184±10 pF (WT) versus 172±9 pF (KChIP2^-/-; P>0.05). Series resistance was compensated electronically by 80-90%; time-constants of the compensated capacitance decay were comparable in WT and KChIP2^-/- myocytes (393±26 versus 412±22 μs, respectively; P>0.05). Leak currents were <100 pA and were not corrected. 4-AP (50 μM and 1 mM⁵; Sigma) and heteropoda toxin 2 (HpTx-2; 1 μM^{5, 26}; Sigma) were dissolved in water and added to the superfusate immediately before use.

Quantitative real-time PCR analysis

First strand cDNA was synthesized from 1 μg total mRNA isolated from WT and KChIP2^-/- hearts. K_V1.4, K_V1.5, K_V2.1, K_V4.2, K_V4.3, KChIP2, and Cyclophilin A expression was determined by quantitative real-time PCR (Roche, Basel, Switzerland). Amplicons were controlled by gel electrophoresis and sequencing. cDNA synthesis without reverse transcriptase served as negative control and showed no amplification. Results are presented as KChIP2^-/- normalized to WT levels. Samples were tested in duplicate.

Data analysis

Peak outward currents (I_P) were defined as maximal outward K⁺ current from 0 pA; I_K1 amplitudes were measured at peak inward current and at the end of the 400 ms voltage step. The decay phase of the total K_V current elicited by 4.5-s depolarization in LV free wall myocytes can be described by the sum of 2 exponential components, where the fast and slow τ are attributed to I_to,f and I_K,slow, respectively.^{6, 24} I_to,s is found exclusively in cells from the interventricular septum. The steady-state current, I_SS, does not inactivate and is described by a time-independent, constant current.⁷ Only fits with a correlation coefficient ≥98% were used. For each cell, current amplitudes are normalized to the whole-cell membrane capacitance, and current densities reported.

Data are presented as mean±SEM. Student's t test (when comparing 2 data sets) or 2-way ANOVA, followed by Bonferroni's test, when appropriate, were used to test differences statistically. P<0.05 was considered significant.

Results

Exaggerated APD prolongation by 4-AP in KChIP2^-/- ventricles

Maximum diastolic potential, action potential amplitude and V_max were comparable between WT and KChIP2^-/- groups (Table). WT and KChIP2^-/- APD were similar at control and prolonged significantly after exposure to 1 mM 4-AP (Table, Figure 1). However, 4-AP induced more marked APD prolongation in KChIP2^-/- LV and RV preparations. In RV, 4-AP effects were greater in KChIP2^-/- than WT (P<0.05; Table). Thus, the altered repolarizing currents in KChIP2^-/- mice manifest as APD changes only in the presence of 4-AP; baseline APDs are equivalent to WT. This prompted us to characterize the voltage-gated potassium currents in disaggregated myocytes from WT and KChIP2^-/- mice.

Table.

Action potential characteristics from intact LV and RV preparations.

LV	n	APA, mV	MDP, mV	Vmax, V/s	APD50, ms		APD90, ms
WT	4	73±3	-69±2	110±27	11±1		75±10
WT, 4-AP	4	76±3	-69±2	107±15	20±1*	77%	84±3*	13%
KChIP2-/-	3	70±3	-67±2	87±11	10±1		70±4
KChIP2-/-, 4-AP	3	74±6	-67±1	93±5	21±1*	115%	84±2*	21%
RV	n	APA, mV	MDP, mV	Vmax, V/s	APD50, ms		APD90, ms
WT	8	88±2	-77±1	215±12	4.1±0.3		31±3
WT, 4-AP	8	87±3	-73±1	196±15	8.9±0.8*	113%	41±4*	28%
KChIP2-/-	7	87±2	-77±1	192±20	3.8±0.3		39±4
KChIP2-/-, 4-AP	7	93±2	-74±2	160±13	13.6±0.9*†	262%	53±4*†	41%

APA, action potential amplitude; MDP, maximal diastolic potential; V_max, maximal upstroke velocity; APD₅₀ and APD₉₀, action potential duration to 50 and 90% repolarization. Relative increases in APD induced by 4-AP are indicated in percent. Pacing-cycle length: 500 ms.

^*, P<0.05 versus baseline.

^†, P<0.05 versus WT, 4-AP.

Figure 1.

APD prolongation by 4-AP is enhanced in KChIP2^-/- ventricles. Representative action potentials recorded from intact WT and KChIP2^-/- LV and RV in control and after addition of 1 mM 4-AP.

Comparable peak K_V currents in WT and KChIP2^-/- ventricular myocytes

Figure 2A shows representative K⁺ currents recorded from a WT and a KChIP2^-/- myocyte. Mean peak current (I_P) densities at +40 mV are 28.3±2.0 pA/pF in WT cells (n=27) versus 29.2±1.9 pA/pF in KChIP2^-/- cells (n=24; P>0.05). No differences in I_P density were seen using 4.5-s depolarizing steps between -40 and +50 mV (V_H=-70 mV; Figure 2B). Activation kinetics, measured as the time from depolarization to peak current, were significantly slower in KChIP2^-/- compared to WT, particularly at steps to positive voltages (Figure 2C).

Figure 2.

Comparison of K_V currents in WT and KChIP2^-/- myocytes. A, Representative current traces normalized to cell capacitance recorded from WT and KChIP2^-/- cells using 4.5-s depolarizing steps from V_H=-70 mV to V_step=-40 to +50 mV (Inset). Each step was preceded by a 25 ms depolarization to -20 mV to eliminate contamination by inward Na⁺ current not completely blocked by TTX. Arrowheads indicate 0 pA. B, I_P densities plotted as a function of V_test in WT (n=27) and KChIP2^-/- (n=24) myocytes. C, Time interval from cell depolarization to peak current as a function of V_test. *, P<0.05 versus WT. D, Representative current traces normalized to cell capacitance recorded from a WT and a KChIP2^-/- cell using 400-ms depolarizing steps. Although peak current amplitude is comparable, the WT cell displays faster current inactivation within the first ∼100 ms.

Significant differences were found in the double-exponential fit to the 4.5-s decay phase of the K_V currents. At V_step=+40 mV, the fast τ were 198±11 versus 226±19 ms in WT and KChIP2^-/- myocytes, respectively (P<0.05). The slow τ were 1486±57 versus 1248±69 ms in WT and KChIP2^-/- myocytes, respectively (P<0.05). The peak amplitudes of the fast component (9±2 versus 12±2 pA/pF, P>0.05), slow component (12±1 versus 12±1 pA/pF, P>0.05), and the constant component (6±1 versus 7±1 pA/pF, P>0.05), were all comparable between WT and KChIP2^-/- myocytes, respectively. When 400-ms depolarizations were used, a fast inactivating component was clearly visible in WT cells but absent in KChIP2^-/- myocytes (Figure 2D). This fast component had a time constant of inactivation of 38±2 ms, which is comparable to the previously reported τ for I_to,f.⁶

We measured peak and steady-state I_K1 to address whether other K⁺ currents were affected by KChIP2^-/- (Figure 3A). No measurable differences in the inwardly rectifying K⁺ current were observed in KChIP2^-/- (n=7-10; Figure 3B).

Figure 3.

Similar inward rectifying K⁺ currents in WT and KChIP2^-/- myocytes. A, Representative current traces recorded from a WT and KChIP2^-/- myocyte. Arrowheads indicate 0 pA. Inset, voltage protocol. B, Peak inward I_K1 (left) and steady state (right) I_K1 in WT (n=7) and KChIP2^-/- (n=10) myocytes.

No HpTx-2-sensitive current and larger 4-AP-sensitive currents in myocytes from KChIP2^-/- mice

To evaluate the contribution of KChIP2 we used HpTx-2 (1 μM), which selectively blocks I_to,f without affecting other murine outward K⁺ currents.²⁶ Figure 4A shows native K_V currents in representative WT and KChIP2^-/- myocytes. Addition of HpTx-2 abolished the fast-inactivating component of the K_V-current decay in the WT cell and reduced I_P (Figure 4B). HpTx-2 blocked 11±2% of the I_P of the native K_V current in WT myocytes (n=13; Figure 4C), but reduced the current by only 3±2% in KChIP2^-/- myocytes (n=9; P<0.05 versus WT), suggesting KChIP2^-/- mice lack the HpTx-2 sensitive I_to,f.

Figure 4.

No effect of HpTx-2 on K_V currents in KChIP2^-/- myocytes. A, Representative baseline K_V currents recorded using 400-ms pulses from V_H=-70 mV to V_step between -40 and +50 mV (inset) in myocytes from WT and KChIP2^-/- mice. Arrowheads indicate 0 pA. B, Currents recorded in the same myocytes after addition of 1 μM HpTx-2 to the superfusion solution. C, The HpTx-2-sensitive currents obtained by digital offline subtraction of the current traces shown in A and B.

Since total K_V currents are comparable in WT and KChIP2^-/- mice, but I_to,f is missing in KChIP2^-/- myocytes, it is possible that a non-K_V4 current is increased in KChIP2^-/- mice. Therefore, we used a pharmacological approach to dissect the remaining currents in WT and KChIP2^-/- myocytes. 4-AP ≥5 mM abolishes all outward K⁺ currents in mice. At lower concentrations (IC₅₀ ≈ 50 μM), 4-AP most potently blocks I_K,slow, whereas I_to,f is blocked at higher concentrations (IC₅₀ ≈ 500 μM). I_SS is blocked only in the mM range.⁷

4-AP concentration-dependently decreased peak-current densities in WT and KChIP2^-/- myocytes (Figure 5). Figure 5A illustrates total K_V current before and after adding 50 μM 4-AP to the superfusate in WT and KChIP2^-/- myocytes. The 4-AP-sensitive current was obtained by digitally subtracting this current from the original K_V current (Figure 5B). Peak 4-AP-sensitive current amplitude was significantly larger in KChIP2^-/- myocytes than in WT myocytes at both 4-AP concentrations (Figure 5C). 4-AP-insensitive currents showed clear time-dependent inactivation in all WT myocytes (Figure 5A, left), whereas the 4-AP-insensitive current in KChIP2^-/- myocytes showed very little inactivation over the depolarizing step (Figure 5A, right). When I_P of 4-AP-insensitive (1 mM) currents were measured within the first 100 ms of the voltage step and compared to the mean current amplitude in the last 100 ms of the voltage step, 82±4% (n=12) of peak current remained in KChIP2^-/- cells, compared to 58±2% in the WT myocytes (n=9; P<0.05). This suggests the absence of time-dependent current in KChIP2^-/- myocytes exposed to 1 mM 4-AP.

Figure 5.

Larger fraction of K_V current is sensitive to 4-AP in KChIP2^-/- myocytes. A, Current traces normalized to cell capacitance recorded using 400-ms pulses from V_H=-70 mV to V_step=+40 mV from a WT and a KChIP2^-/- cell, before and after addition of 50 μM 4-AP to the superfusate. Arrowheads indicate 0 pA. Note the prominent inactivation of the 4-AP resistant current in WT. B, The current fraction sensitive to 50 μM 4-AP obtained by digital offline subtraction of the current traces shown in A. C, 4-AP-sensitive current relative to the full I_P in WT and KChIP2^-/- cells. *, P<0.05.

Elevated K_V1.5 transcription in KChIP2^-/- mice

The altered pharmacological responsiveness combined with an unaltered native I_P in KChIP2^-/- myocytes suggest that currents contributing to overall repolarization may change subsequent to KChIP2 knockout. Therefore, we studied the molecular remodeling in KChIP2^-/- LV, using quantitative real-time PCR with K_V-subunit specific primers to quantify transcriptional activity in WT (n=6) and KChIP2^-/- (n=6) tissue. We found significant increases in mRNA levels for K_V4.2, K_V4.3, and K_V1.5 in KChIP2^-/- ventricles versus WT (Figure 6). As expected, KChIP2 transcription was not detected in KChIP2^-/- LV. K_V2.1 and K_V1.4 mRNA were comparable in WT and KChIP2^-/- LV. Cyclophilin A levels (control) remained unaltered.

Figure 6.

Compensatory remodeling potentially includes elevated K_V1.5 transcription. Expression levels of the major ion channels underlying murine repolarization in KChIP2^-/- (n=6) relative to WT (n=6) ventricles. Grey horizontal line indicates equal expression levels in WT and KChIP2^-/- ventricles. KChIP2 was not detected in KChIP2^-/- myocytes. *, P<0.05.

Discussion

The significant findings of this study are that myocytes from KChIP2^-/- mice have no I_to,f, but show no change in I_P density, versus WT myocytes. An increased 4-AP-sensitive current is likely responsible for the maintained I_P in KChIP2^-/- mice. Transmembrane action potentials recorded from intact ventricular preparations under control conditions are comparable in KChIP2^-/- and WT; however, 4-AP-induced APD prolongation is intensified in KChIP2^-/- preparations. Furthermore, the transcript level of the K_V1.5 subunit is increased in KChIP2^-/- ventricles. In total, this is consistent with an augmented I_K,slow in KChIP2^-/- myocytes, maintaining normal repolarization reserve.

KChIP2 is responsible for maintenance of mouse ventricular I_to,f

Using 4.5s depolarizing steps, K_V current decay was described by 2 exponential components in KChIP2^-/- and WT myocytes, with only minor differences in time constants of inactivation. A very fast component of inactivation (τ<50 ms) was present in WT myocytes within the first 100 ms of depolarization, and by using 400-ms pulses, this rapid component was amplified relative to the total current decay (compare Figure 2A and 2D). This component was blocked by HpTx-2, which had no effect in KChIP2^-/- myocytes (Figure 4). Thus, we found no evidence that K_V4.2 or K_V4.3 produce functional I_to,f in the absence of KChIP2 in the mouse, despite increased mRNA levels of these α subunits in the LV (Figure 6).

The unaltered I_P and APD in the absence of K_V antagonists in KChIP2^-/- mice suggest upregulation of a secondary repolarizing K⁺ current that compensates for the absent I_to,f. Interestingly, 4-AP had a larger effect on K_V current amplitude in KChIP2^-/- myocytes, translating to exaggerated APD prolongation upon 4-AP-treatment. Previous studies suggest that KChIP2 associates with the K_V1.5 α subunit in myocytes and that KChIP2 co-expression decreases K_V1.5-encoded K⁺ current in cell lines, suggesting a multifunctional role of KChIP2 in regulating murine repolarizing currents.¹⁸ Thus, by relieving KChIP2-mediated suppression of K_V1.5 in KChIP2^-/- mice, an increased I_K,slow would be expected. We observe a 3-fold increase in K_V1.5 mRNA in KChIP2^-/- ventricles (Figure 6) and an augmented 4-AP-sensitive I_K,slow (Figure 5). The lack of an anti-K_V1.5 antibody that can detect endogenous K_V1.5 in cardiomyocytes reliably²⁷ precluded analysis of K_V1.5 protein expression. Furthermore, it is unclear whether direct relief of suppression contributes to the increased I_K,slow in KChIP2^-/- myocytes.

In vitro, KChIP2 increases K_V4-encoded current amplitude, slows inactivation, and accelerates recovery from inactivation.^{15, 16} Hence, if a K_V4-dependent current plays a considerable role in murine outward voltage-gated K⁺ currents, corollaries of KChIP2^-/- in vivo should include decreased I_P density, faster current decay, and slowed recovery of I_P from inactivation. Despite transcription of K_V4.2 and K_V4.3, we found no residual I_to,f in KChIP2^-/- mice. Thus, although a small, fast inactivating current is recorded in cell lines transfected with K_V4.2 or K_V4.3 in the absence of KChIP2, no such current is found in KChIP2^-/- mice, suggesting an essential role for KChIP2 in translating K_V4 protein into functional current in vivo.

I_to remodeling in KChIP2^-/- mice reduces repolarization reserve

The KChIP2^-/- mouse adds to the animal models and patient data in which a normal QT interval is excessively prolonged by a challenge (e.g. I_Kr block), reflecting a reduced repolarization reserve.^{29, 30} Many patients with congenital or drug-induced long-QT syndromes have normal QT intervals in the absence of provocation, referred to as genotype-phenotype discordance.^{31, 32} Furthermore, downregulation of I_to is prominent in heart failure^{1, 4}, where I_Kr-blocking drugs can induce excess QT interval prolongation.³³ Additionally, the significant cardiac plasticity in cardiac memory in dogs³⁴, includes decreased I_to and downregulated K_V4.3 and KChIP2 transcription in myocardium near ventricular pacing sites.^{35, 36} Here, I_Kr blockade causes more marked QT prolongation than in control dogs.³⁷

In the present study, removal of I_to,f from the assembly of murine repolarizing currents caused no overt reduction in total current (Figure 2), due to I_K,slow upregulation. The phenotypic consequence of this remodeling became evident only when action potentials were recorded in the presence of 4-AP (Figure 1). Thus, a reduced repolarization reserve with respect to I_to increases susceptibility to pharmacological prolongation of repolarization. Finally, the differences in action potential morphology and 4-AP response between LV and RV (Figure 1) likely reflect regional differences in repolarizing currents reported in cells isolated from the ventricles.⁵

Relationship to other studies on mouse pathologies

Kuo et al.²⁵ reported decreased total I_to density and significantly prolonged APD in disaggregated RV myocytes from KChIP2^-/- mice. In their study, I_Na and I_Ca,L were allowed to contribute to the current recordings.^{25, 38} In contrast, our study evaluated the relative contributions of the different voltage-dependent K⁺ currents to repolarization in KChIP2^-/- mice. This intent prompted us to use solutions containing antagonists for I_Na (TTX) and I_Ca,L (Co²⁺).⁶ Using protocols and solutions identical to those of Kuo et al., we observed a smaller total outward current in KChIP2^-/- myocytes (data not shown). The discrepancy in APD likely reflects the difference between the uncoupled disaggregated myocytes in the previous study²⁵ and the multicellular preparations in the present study. However, our results complement the previous report of unaltered QT intervals or effective refractory periods in anesthetized, intact KChIP2^-/- mice.²⁵

In agreement with earlier reports on KChIP2^-/- mice, we observed no evidence of cardiac hypertrophy or impaired function.²⁵ This is paralleled by results from K_V4.2^-/- mice²⁴ and mice expressing a K_V4.2-pore mutant.²² In contrast, mice expressing a carboxy-terminal deletion of K_V4.2 display marked hypertrophy and ventricular dilatation progressing to congestive heart failure.²³ Whereas the two mutants display prolonged APD, K_V4.2^-/- mice have an upregulated I_to,s and increased transcription of K_V1.4.²⁴ In our experiments, we found neither I_to,s nor increased K_V1.4 mRNA in KChIP2^-/- myocytes. Thus, the compensatory mechanisms clearly differ between K_V4.2^-/- and KChIP2^-/- mice and the signal transduction causing this discrepant remodeling is of interest. Furthermore, the absence of functional remodeling in the two knockout mice argues against a direct link between deficient I_to,f and cardiac hypertrophy and heart failure.

Limitations

Our recordings of WT I_P are smaller and I_to,f are reduced versus other reports⁶, despite identical mouse strains, patch-clamp solutions, and voltage protocols. Different cell-disaggregation protocols or storage conditions between myocytes disaggregation and patch-clamp experiments (plating on laminin-coated coverslips, stored in 5% CO₂ incubator in up to 24 h)⁶ could contribute to the differences seen between laboratories. Nevertheless, we studied WT and KChIP2^-/- mice under identical conditions and the differences reported are consequent to KChIP2 deficiency.

Although increased mRNA expression may not be identical to increased protein translation and –trafficking, our functional data of upregulated I_K,slow density complements the increased K_V1.5 mRNA levels.

Conclusions

Loss of KChIP2 has little impact on total murine K_V current, and the genotype is reflected phenotypically only during pharmacological challenges to repolarization. I_to,f is absent in KChIP2^-/- mice, whereas 4-AP-sensitive I_K,slow is upregulated. This is associated with excess prolongation of APD in the presence of 4-AP and with increased expression of K_V1.5. Thus, KChIP2 and I_to,f appear critical to action potential morphology in the mouse only in circumstances of decreased repolarization reserve.