Redox- and Calmodulin-dependent S-Nitrosylation of the KCNQ1 Channel*

Nitric oxide (NO) is a gaseous signal mediator showing numerous important biological effects. NO has been shown in many instances to exhibit its action via the protein S-nitrosylation mechanism, in which binding of NO to Cys residues regulate protein function independently of activation of soluble guanylate cyclase. The direct link between protein S-nitrosylation and functional modulation, however, has been demonstrated only in limited examples. Furthermore, although most proteins have more than one Cys residue, the mechanism by which a certain Cys becomes a specific target residue of S-nitrosylation is poorly understood. We have previously reported that NO regulates currents through the cardiac slowly activating delayed rectifier potassium channel (IKs) irrespective of soluble guanylate cyclase activation. Here we demonstrate using a biotin-switch assay that NO induced S-nitrosylation of the α-subunit of the IKs channel, KCNQ1, at Cys445 in the C terminus. A redox motif flanking Cys445 and the interaction of KCNQ1 with calmodulin are required for preferential S-nitrosylation of Cys445. A patch clamp experiment shows that S-nitrosylation of Cys445 modulates the KCNQ1/KCNE1 channel function. Our data provide a molecular basis of NO-mediated regulation of the IKs channel. This novel regulatory mechanism of the IKs channel may play a role in previously demonstrated NO-mediated phenomenon in cardiac electrophysiology, including shortening in action potential duration in response to intracellular Ca2+ or sex hormones.

S-Nitrosylation is a nitric oxide (NO) 2 -induced post-translational modification in which a cysteinyl thiol (R-SH) is con-verted to a nitrosothiol (1)(2)(3) and acts as a regulatory mechanism of various classes of proteins, including ion channels, such as the skeletal muscle type ryanodine receptor (ryanodine receptor type 1) channel (4,5), the N-methyl-D-aspartate receptor channel (6,7), the cardiac L-type Ca 2ϩ channel (8), and the cardiac Na ϩ channel (9). We have previously reported that NO derived from endothelial NO synthase activates ion currents through the cardiac slowly activating delayed rectifier potassium channel (I Ks ) composed of the pore-forming ␣-subunit KCNQ1 and the auxiliary ␤-subunit KCNE1. The NO-dependent regulation of the I Ks channel plays a pivotal role in regulation of cardiac membrane potential by intracellular Ca 2ϩ (10) and by sex hormones (11)(12)(13). Because the NO-dependent I Ks activation was inhibited by an inhibitor of soluble guanylate cyclase, 1H-(1,2,4)oxadiazolo(4,3-a)quinoxlin-1-1 (ODQ), with only a limited magnitude but was robustly inhibited by a thiol-alkylating reagent, N-ethylmaleimide, and reversed by a reducing reagent, dithiothreitol, soluble guanylate cyclase-independent, the protein S-nitrosylation mechanism is posited to be mainly involved (14). However, the following issues remain to be addressed: (i) Is the I Ks channel S-nitrosylated? (ii) If so, then what is the target of S-nitrosylation between the ␣-subunit KCNQ1 and the ␤-subunit KCNE1? (iii) Among multiple Cys residues, which Cys is a target of S-nitrosylation? and (iv) How does NO specifically recognize the target Cys? In the present study, we used the biotin-switch assay and functional patch clamp experiment to answer these questions. Our data show that KCNQ1 is a target of S-nitrosylation, and the presence of a redox motif contributes to making the Cys at 445 in the C terminus of KCNQ1 a preferential target of S-nitrosylation.
To introduce a V5 epitope (-GLPIPNPLLGLDST-), the C-terminal KCNQ1 fragment with V5-epitope tag sequence was made by PCR using a full-length human KCNQ1 as a template, digested with MroI and BamHI, and subcloned into MroI-BamHI-digested KCNQ1/pcDNA3.1(Ϫ) (KCNQ1-V5). The pair of primers used was as follows: a sense strand, 5Ј-CAT-CGCCTCCTGCTTCTC; and an antisense strand, 5Ј-ggattctcaCGTAGAATCGAGACCGAGGAGAGGGTTAG-GGATAGGCTTACCGGACCCCTCATCGGGGCCCCT. Lowercase characters in the antisense primer indicate the BamHI recognition site, and those in italics indicate the stop codon. Underlined uppercase characters indicate the sequence encoding the V5 epitope. To facilitate identification of cells in which CaM had been transfected in patch clamp experiments, we subcloned either CaM or CaM 1234 with EGFP in the pIRES vector (Clontech) (EGFP-CaM/pIRES, EGFP-CaM 1234 /pIRES).
Preparation of GST Fusion Protein and Biotin-Switch Assay-GST fusion proteins were purified as described previously (15). Briefly, expression of GST fusion proteins in Escherichia coli strain BL21(DE3) (Takara) was induced with 0.1 mM isopropyl 1-thio-␤-D-galactopyranoside overnight at 18 -20°C. Cells were harvested by centrifugation and were re-suspended in lysis buffer (50 mM Tris, pH 7.5, 100 mM NaCl, 5 mM EDTA). Cell lysis was achieved by treating the cells with lysozyme (1 mg/ml) for 1 h on ice, sonicating them with 1.5% Sarkosyl, and them subjecting them to a freeze-thaw cycle. Soluble fractions from a 10-min centrifugation at 24,000 ϫ g (4°C) were rotated for 1 h in the presence of GST-Sepharose beads (Amersham Biosciences) at 4°C. Beads containing bound fusion proteins in HEN buffer (100 mM HEPES, pH 7.7, 400 M EDTA, 40 M neocuproine) were incubated with 1 mM S-nitroso-N-acetyl penicillamine (SNAP, DOJINDO) for 1 h at a room temperature, and SNAP was removed by washing with HEN buffer five times.
The biotin-switch assay was performed as described previously by Jaffrey et al. (16) (Fig. 1). Briefly, after GST fusion proteins immobilized to GST-Sepharose beads were washed with HEN buffer containing 0.05% Triton X-100, and free thiol (R-SH) residues were blocked (R-S-S-CH 3 ) with 20 mM methylmethanethiol sulfonate (MMTS, Sigma-Aldrich) in HEN buffer containing 2.5% SDS for 30 min at 53°C. After a 1-min centrifugation at 2500 rpm (at room temperature), the supernatant was incubated for 30 min at Ϫ20°C with two volumes of pre-chilled acetone to remove MMTS. Only S-nitrosylated thiol (R-S-NO) residues were then reduced to free thiol (R-SH) with 1 mM sodium ascorbate in HEN buffer and were biotinylated (R-S-S-biotin) with 4 mM biotin-HPDP (Pierce). Biotinylated samples were electrophoresed in SDS-polyacrylamide gel, transferred to a polyvinylidene difluoride membrane (Amersham Biosciences), and immunoblotted with a 1:4000-diluted monoclonal anti-biotin antibody (BN-34, Sigma-Aldrich). The proteins were detected with an advanced enhanced chemiluminescence system (Amersham Biosciences) using a lumino-image analyzer, LAS-3000mini (Fujifilm). Densitometric analysis was performed with a software MultiGauge (Ver3.0, Fujifilm). For S-nitrosylation assay in cardiac myocytes, guinea pig left ventricle was washed with HEN buffer, minced, and homogenized in HEN buffer. After cell debris was removed by centrifugation, cell lysates were treated with 0.5 mM SNAP, 3-(2-hydroxy-1-methyl-2-nitrosohydrazino)-N-methyl-1-propanamide, or S-nitrosoglutathine for 30 min at room temperature, and NO donors were removed using a microcon YM column (Millipore Corp.).

Biotin-Switch Assay for Culture Cell Lysates and Cardiomyocyte
After treatment with each NO donor, lysates from culture cells and from guinea pig ventricles were incubated with protein G-Sepharose beads (Amersham Biosciences) and either polyclonal anti-KCNQ1 (C-20, Santa Cruz Biotechnology) or anti-KCNE1 (N-16, Santa Cruz Biotechnology) antibody overnight at 4°C. The biotin-switch assay for immune complexes was performed as described above for GST fusion proteins with minor modifications. Samples were blocked with 20 mM MMTS and biotinylated with 1 mM sodium ascorbate and 4 mM biotin-HPDP. After precipitation with pre-chilled acetone, streptavidin-agarose beads (Pierce) were added, and samples FIGURE 1. Schematic diagram of the biotin-switch assay. In the first step, free thiol (R-SH) residues were blocked with MMTS (R-S-S-CH 3 ). S-Nitrosylated thiol (R-S-NO) residues were reduced to free thiol (R-SH) residues with a weak reducing reagent, sodium ascorbate in the second step, and were biotinylated (R-S-S-biotin) with biotin-HPDP.
were incubated for 1 h at a room temperature. The beads were washed with HEN buffer containing 0.5% Triton X-100, and proteins were eluted by incubation with HEN buffer containing 100 mM ␤-mercaptoethanol at 37°C for 20 min. Eluted samples were electrophoresed in SDS-polyacrylamide gel, transferred to a polyvinylidene difluoride membrane, and immunoblotted with a 1:4000-diluted polyclonal anti-KCNQ1 antibody (C-20) or a 1:2000-diluted polyclonal anti-KCNE1 antibody (N-16).
Statistics-All numerical values are presented as mean Ϯ S.E. Statistical significance was evaluated by an analysis of variance followed by a Bonferroni multiple comparison test. p Ͻ 0.05 was taken as the significance level.

RESULTS
S-Nitrosylation of KCNQ1 but Not KCNE1-We first examined if either or both of KCNQ1 or/and KCNE1 is S-nitrosylated in culture cells. Human V5-tagged KCNQ1 or KCNE1 was expressed in HEK293 cells, immunoprecipitated with an anti-V5 antibody or an anti-KCNE1 antibody, and treated with an NO donor SNAP at a concentration of 0.5 mM. In the pilot experiments, we confirmed that transfection of HEK293 cells with V5-tagged KCNQ1 and KCNE1 yielded similar currents to those obtained by transfection with KCNQ1 and KCNE1. An immunopositive band was detected after avidin purification for KCNQ1 (lane 3, Fig. 2A). For KCNE1, although an immunopositive band was detected without application of a thiol-reacting reagent MMTS to block free thiol (R-SH) (lane 1, Fig. 2B), no such band was detected after blocking free thiol with MMTS (lane 2, Fig. 2B). We also found that exogenously applied SNAP induced S-nitrosylation of KCNQ1 of guinea pig ventricular myocytes (supplemental Fig. S1), which indicates that KCNQ1 is the target of S-nitrosylation also under a physiological condition.
Specific S-Nitrosylation of Cys 445 of KCNQ1-We next attempted to pinpoint the target Cys residue for S-nitrosylation among nine Cys residues present in KCNQ1. Because Cys residues that are subject to S-nitrosylation have been shown in a number of studies to reside in a juxtamembrane zone (6, 17), we focused on six such Cys residues (Cys 34 , Cys 122 , Cys 180 , Cys 381 , Cys 445 , and Cys 642 ) (closed circles in Fig. 3A). We attempted to create GST fusion proteins containing each Cys residue. In E. coli, although the GST fusion protein containing Cys 180 (Thr 167 -Pro 197 ) or Cys 642 (His 620 -Ser 670 ) became soluble, the GST fusion protein containing Cys 34 , Cys 122 , Cys 381 , or Cys 445 alone was hardly soluble. Alternatively, we created and examined the GST fusion protein with Cys 34 and Cys 122 (Met 1 -Tyr 125 of KCNQ1), and that with Cys 381 and Cys 445 (Val 355 -Leu 619 of KCNQ1). Each of four GST fusion proteins was purified from E. coli, incubated with an NO donor, and subjected to the biotin-switch assay. S-Nitrosylation was detected only in the GST fusion protein with Cys 381 and Cys 445 (Val 355 -Leu 619 of KCNQ1) (lane 3, Fig. 3B).
To examine if Cys 381 or/and Cys 445 were S-nitrosylated by an NO donor, we replaced Cys 381 with Ala (C381A) or Cys 445 with Ala (C445A) and carried out the biotin-switch assay. S-Nitrosylation remained detected for C381A mutant, but was completely eliminated for the C445A mutant (Fig. 3C). To further corroborate that Cys 445 is the main target of S-nitrosylation, we expressed V5-tagged full-length KCNQ1 in which Cys 445 had been replaced with Ala leaving remaining eight cysteines intact (KCNQ1(C445A)) in HEK293 cells. Cell lysates were treated with 0.5 mM SNAP and subjected to the biotin-switch assay. Application of an NO donor barely induced S-nitrosylation of KCNQ1(C445A) (lane 3, Fig. 3D), which verifies that Cys 445 is the main target of S-nitrosylation.
Role of Redox Motif for Specific Cys 445 S-Nitrosylation-Several mechanisms have previously been implicated for the sitespecific S-nitrosylation in several proteins, which include the presence of Cys residue in the hydrophobic milieu (18), flanking of the Cys residue by a redox (acid-base) motif (19,20), allosteric effects by protein-protein interaction (4,5), and compartmentalization of NO sources and targets (21,22). The Kyte-Doolittle hydropathy profiling indicates that Cys 445 does not reside in the hydrophobic milieu. Cys 445 , but none of the other eight cysteines in KCNQ1, is immediately followed by an acidic amino acid. Such a sequence is partially matched to the proposed consensus sequence for S-nitrosylation, (Lys/Arg/His)-Cys-(Asp/Glu), in which the presence of an acidic amino acid immediately after Cys is proposed to be especially critical (19). To explore if an acidic amino acid, Asp 446 , has some integral roles for S-nitrosylation of Cys 445 , we made a mutant GST fusion protein of the KCNQ1 C terminus (Val 355 -Leu 619 ) with replacement of Asp 446 with Asn (D446N) leaving Cys 445 intact, and performed in vitro biotin-switch assay. An S-nitrosylated band was not detected (lane 3 in Fig. 4, A and B). When Asp 446 was replaced with another acidic amino acid Glu (D446E), S-nitrosylation was clearly detected (lane 5 in Fig. 4A and column 2 in Fig. 4B), suggesting that S-nitrosylation of Cys 445 requires a negative charge at KCNQ1 residue 446.
Because the basic amino acid closest to Cys 445 is His 442 , we replaced His 442 with Ala (H442A). We could still observe S-ni-trosylation in the H442A mutant (lane 2 in Fig. 4A and column 4 in Fig. 4B). Collectively, the acidic amino acid Asp 446 , but not the basic amino acid His 442 , appears to be required for sitespecific S-nitrosylation of Cys 445 .
CaM Is Required for S-Nitrosylation in Living Cells-To assess whether Cys 445 also is the S-nitrosylation target and whether redox-motif is critical for S-nitrosylation of Cys 445 also in living cells, biotin-switch assays were performed in lysates obtained after application of an NO donor to HEK293 cells   FEBRUARY 27, 2009 • VOLUME 284 • NUMBER 9 JOURNAL OF BIOLOGICAL CHEMISTRY 6017 expressing full-length KCNQ1WT, KCNQ1(C445A), or KCNQ1(D446N) (Fig. 5A). To our surprise, not only KCNQ1 mutants but also KCNQ1WT were not S-nitrosylated (lane 2, Fig. 5A), suggesting that an additional factor for S-nitrosylation of the KCNQ1 channel is required in living cells.

S-Nitrosylation of the KCNQ1 Channel
Skeletal muscle type ryanodine receptor type 1 requires interaction with CaM to be S-nitrosylated (4,5). KCNQ1 possesses two CaM binding domains in the C terminus (23), and its gating kinetics is regulated by CaM binding (24,25). Cys 445 , an S-nitrosylation target, locates between these two CaM binding sites (26). I Ks values are responsive to intracellular Ca 2ϩ (27), and we have previously demonstrated in separate experiments that I Ks regulation by Ca 2ϩ is mediated by NO (10) and that the presence of CaM is essential for Ca 2ϩ regulation (28). In the latter study, we suggest a role of an allosteric effect of CaM, but not activation of CaM-dependent protein kinase in I Ks regulation by intracellular Ca 2ϩ (28). Taken together, these pre-existing conditions led us to speculate that CaM might be required for S-nitrosylation of KCNQ1 and that the amount of CaM endogenously present in HEK293 cells is not sufficient for KCNQ1 S-nitrosylation. We co-transfected CaM with KCNQ1 into HEK293 cells and performed biotin-switch experiments.
Western blot analysis indicates that CaM level is ϳ8 times higher in CaM-transfected cells than in CaM non-transfected cells (supplemental Fig. S2). In the presence of CaM, KCNQ1WT was clearly S-nitrosylated (lane 2 in Fig. 5B), whereas KCNQ1(C445A) or KCNQ1(D446N) was not (lanes 3 and 4 in Fig. 5B). When we co-transfected CaM 1234 mutant with KCNQ1WT, in which four Ca 2ϩ binding motifs are disrupted (29), S-nitrosylation was not observed even in KCNQ1WT (lane 2 in Fig. 5C). Thus, interaction with CaM in the Ca 2ϩ -bound form is critical for S-nitrosylation of KCNQ1 in living cells. We found using co-immunoprecipitation experiments that the interaction with CaM was intact for KCNQ1(D446N) (supplemental Fig. S3), indicating that the failure of an NO donor to S-nitrosylate KCNQ1(D446N) is due to the disruption of a redox motif but not to the disturbed interaction with CaM.
S-Nitrosylation on KCNQ1 Channel Function-We employed a patch clamp experiment to investigate functional consequences of S-nitrosylation of Cys 445 of the KCNQ1 channel. HEK 293 cells expressing KCNQ1 and KCNE1 conduct slowly activating outward currents with very similar kinetics to those of I Ks in cardiac myocytes (Fig. 6A). Without co-transfection of CaM, the amplitude of outward currents increased monotonically during the depolarization pulse (upper panel in Fig. 6A), whereas, with co-transfection of CaM, outward currents appeared to saturate at the later phase of the depolarization pulse (lower panel in Fig. 6A), in agreement with the previous reports that CaM induces inactivation of the KCNQ1/KCNE1 current (24,25). With co-transfection of CaM, application of SNAP increased KCNQ1/KCNE1 current amplitudes in a dosedependent manner (red traces in Fig. 6, A and B, second bar in Fig. 6C), whereas SNAP at 100 nM did not significantly enhance KCNQ1(C445A)/KCNE1 currents (third bar in Fig. 6C). S-nitrosoglutathione, a different class of NO donor, also enhanced KCNQ1/KCNE1 currents (data not shown). Even without application of SNAP (0 nM), I K tail amplitude was slightly increased (first bar in Fig. 6B): this slight increase appeared to be due to the nonspecific effect of DMSO used as a vehicle, because in the time control experiment application of DMSO slightly enhanced KCNQ1/KCNE1 currents (data not shown).
Without co-transfection of CaM (fourth bar in Fig. 6C) or with co-transfection of CaM 1234 mutant (fifth bar in Fig. 6C), SNAP at 100 nM failed to enhance KCNQ1/KCNE1 currents. To examine the effects of intracellular Ca 2ϩ , we replaced 11 mM EGTA with equimolar BAPTA. In the presence of BAPTA, I Ks amplitude suffered from significant rundown, which agrees with the finding that injection of BAPTA into oocytes (estimated intracellular concentration of 5-10 mM) reduced KCNQ1/KCNE1 currents by Ͼ50% (25). Although the BAPTA-induced rundown made the effects of SNAP difficult to assess, we could at least conclude that SNAP did not enhance I Ks tail amplitude in the presence of BAPTA (sixth bar in Fig. 6C).
With co-transfection with wild-type CaM, SNAP slightly enhanced KCNQ1(C445A)/KCNE1 currents, although not significantly (third bar in Fig. 6C). We had previously reported that, although NO donor-induced enhancement of I Ks was mostly inhibited by an alkylating reagent, N-ethylmaleimide, and reversed by a reducing reagent, dithiothreitol, some frac- tion was inhibited by an inhibitor of soluble guanylate cyclase, ODQ (14). We tested the hypothesis if small SNAP-induced enhancement for KCNQ1(C445A)/KCNE1 currents is due to a minor contribution of soluble guanylate cyclase-dependent mechanism. In the presence of ODQ, SNAP failed to enhance KCNQ1(C445A)/KCNE1 currents (seventh bar in Fig. 6C), indicating that the SNAP-induced enhancement of KCNQ1(C445A)/KCNE1 currents is likely due to the cGC-dependent mechanism, but not due to the effects on redox-sensitive Cys other than Cys 445 .

DISCUSSION
Emerging evidence demonstrates that protein S-nitrosylation is an important NO-mediated regulatory mechanism of various classes of proteins (1-3), including ion channels (4 -9). However, the direct link between protein S-nitrosylation and functional relevance has been proven only for limited examples. Furthermore, the mechanism underlying preferential S-nitrosylation of the target Cys is not fully understood. In the present study, we demonstrate that an NO donor induces S-nitrosylation at Cys 445 in the C terminus of the pore-forming ␣-subunit KCNQ1. We provide convincing evidence to show that the redox motif flanking Cys 445 is required for the sitespecific S-nitrosylation of Cys 445 . S-Nitrosylation at Cys 445 of the KCNQ1 channel functionally regulates the KCNQ1/ KCNE1 complex channel.
Among nine Cys in KCNQ1, only Cys 445 is immediately followed by the acidic amino residue, and the in vitro biotinswitch assay indicates that the presence of acidic amino acid at the 446th residue is required for S-nitrosylation of Cys 445 . His 442 is the closest basic amino acid, but the replacement of His 442 to a neutral amino acid Ala does not disrupt S-nitrosylation of Cys 445 . We are not sure if the presence of basic amino acids is not required or if other distal basic amino acids possibly located close to Cys 445 in the three-dimensional structure are required. In living cells, an NO donor failed to S-nitrosylate WT KCNQ1 in the absence of CaM, and even in the presence of CaM, an NO donor failed to S-nitrosylate KCNQ1(D446N), implying that both of the redox motif flanking Cys 445 and the presence of CaM are required. We have previously demonstrated that endothelial NO synthase and KCNQ1 co-localize in the caveolae fraction of guinea pig hearts (12), indicating the close proximity of the NO donor and the target protein. Thus, multiple factors orchestrate to make Cys 445 of KCNQ1 a target of molecule-specific and site-specific S-nitrosylation. The presence of CaM is also a pre-requisite for S-nitrosylation of the skeletal muscle type ryanodine receptor type 1 (4,5). It is suggested that the binding of Ca 2ϩ -bound CaM, but not apo-CaM, unmasks the target Cys residue or provide hydrophobic milieu (4,5). CaM frequently acts at the level of electron transfer, such as that between the FAD reductase and heme domains to regulate NOS activity (30). Thus, it might be an alternative possibility that CaM acts as an electron transfer cofactor between an NO donor and Cys 445 surrounded by the redox motif. Co-transfection of CaM is required for S-nitrosylation of KCNQ1 in living cells, but not in cell lysates, for which we have currently no clear explanation. It might be a potential explanation that to induce S-nitrosylation in the reducing intracellular milieu, the presence of CaM would be required. However, more work will clearly be needed to clarify the role of CaM for S-nitrosylation of KCNQ1.
An exogenously applied NO donor increased the amplitude of KCNQ1/KCNE1 channel currents. An NO donor failed to enhance KCNQ1(C445A)/KCNE1 currents or KCNQ1/ KCNE1 currents without co-transfection of CaM or with cotransfection of CaM 1234 . Because each of the three experimental conditions dislocates S-nitrosylation of the KCNQ1 channel in the biotin-switch assay, these data serve as evidence for the direct link between S-nitrosylation of Cys 445 and the activation of the KCNQ1/KCNE1 channel. We used a high concentration of SNAP (0.5 mM) to assess S-nitrosylation of KCNQ1, because a similar concentration (1 mM) of SNAP was used in the original biotin-switch assay report (16). However, this concentration was much higher than that (1-100 nM) used for patch clamp experiments, which would certainly raise a question on the causative linkage between S-nitrosylation and KCNQ1 current enhancement. We, therefore, examined the concentra- tion dependence for KCNQ1 S-nitrosylation by SNAP: SNAP induced S-nitrosylation of KCNQ1 above 10 nM in a concentration-dependent manner (supplemental Fig. S3), suggesting that S-nitrosylation of the KCNQ1 channel and its activation occur within a similar dosage of SNAP.
The underlying mechanism for S-nitrosylation-induced activation of the KCNQ1 channel remains to be addressed. Biochemical, spectroscopic, and crystallographic analyses have recently provided the structure model for the C terminus of KCNQ1, which consists of four ␣-helices, helices A, B, C, and D, and an unstructured loop between helices A and B (31). Helices A and B provide binding sites for CaM, and the Cys 445 , the target of S-nitrosylation resides in the midst of the unstructured loop between helix A and B. Because CaM constitutively binds to both helix A and helix B (31), bridging of two helices by CaM might create a milieu for S-nitrosylation of Cys 445 in the midst of the loop. Although functional assembly of KCNQ1 channels has been shown to require its interaction with CaM (24,25), similar KCNQ1/KCNE1 channel currents were recorded in the absence and presence of CaM co-expression. Endogenous levels of CaM in HEK293 cells was about one-eighth of those in CaM-transfected cells (supplemental Fig. S2), which might be sufficient for functional assembly of KCNQ1 but not for KCNQ1 S-nitrosylation. Certainly, more work will be needed to decipher the role of CaM for S-nitrosylation and the mechanism linking S-nitrosylation of the KCNQ1 channel to its activation.
The S-nitrosylation of the KCNQ1 channel is the novel regulatory mechanism of this channel that may have a great impact on physiological regulation of cardiac electrical activity, because the I Ks channel composed of KCNQ1 and KCNE1 has a crucial regulatory role in cardiac electrophysiology, including response to autonomic nervous stimulation (32) and adaptation to heart rate changes (33). In fact, we have previously depicted the role of NO-dependent regulation of the I Ks channel in feedback regulation of Ca 2ϩ homeostasis (9) and the gender difference in the life-threatening cardiac arrhythmias (10 -13).