Modulation of the Cardiac Sodium Channel NaV1.5 by Fyn, a Src Family Tyrosine Kinase
http://www.100md.com
循环研究杂志 2005年第5期
The Department of Physiology, Institute of Hyperexcitability, Jefferson Medical College, Philadelphia, Pa.
Abstract
Dynamic modulation of ion channels can produce dramatic alterations of electrical excitability in cardiac myocytes. This study addresses the effects of the Src family tyrosine kinase Fyn on NaV1.5 cardiac sodium channels. Sodium currents were acquired by whole cell recording on HEK-293 cells transiently expressing NaV1.5. Acute treatment of cells with insulin caused a depolarizing shift in steady-state inactivation, an effect eliminated by the Src-specific tyrosine kinase inhibitor PP2. Sodium channels were coexpressed with either constitutively active (FynCA) or catalytically inactive (FynKD) variants of Fyn. FynCA caused a 10-mV depolarizing shift of steady-state inactivation compared with FynKD without altering the activation conductance-voltage relationship. Comparable effects of these Fyn variants were obtained with whole-cell and perforated-patch recording. Tyrosine phosphorylation of immunoprecipitated NaV1.5 was increased in cells expressing FynCA compared with FynKD. We show that Fyn is present in rat cardiac myocytes, and that NaV1.5 channels from these myocytes are tyrosine-phosphorylated. In HEK-293 cells the effect of FynCA on NaV1.5 inactivation is abolished by the single point mutation Y1495F, a residue located within the cytoplasmic linker between the third and fourth homologous domains of the sodium channel. We provide evidence that this linker is a substrate for Fyn in vitro, and that Y1495 is a preferred phosphorylation site. These results suggest that cardiac sodium channels are physiologically relevant targets of Src family tyrosine kinases.
Key Words: cardiac sodium channel tyrosine kinase Src Fyn phosphorylation
Introduction
Voltage-gated sodium channels play a pivotal role in cardiac excitability by controlling the upstroke of the cardiac action potential. These ion channels display exquisite sensitivity to changes in membrane potential and, as a consequence, minor alterations in their biophysical properties can have dramatic effects on cardiac function. Multiple isoforms of sodium channels are expressed in the heart, including the neuronal isoforms NaV1.1, 1.3, and 1.6, which have been shown to reside within T-tubules and participate in excitation-contraction coupling.1 However, in terms of conducting the cardiac action potential, the tetrodotoxin-resistant isoform, NaV1.5, is dominant in both its crude expression level and its contribution to the myocytes’ sodium current.1 For this reason, identification of regulatory factors that modify the behavior of NaV1.5 is crucial in understanding cardiac excitability.
Both Src family and receptor tyrosine kinases are known to be potent modulators of ion channels (see reviews2,3) Moreover, tyrosine kinase activity is enlisted in numerous signal transduction pathways. For example, multiple ligands such as insulin result in elevated tyrosine kinase activity in the heart.4,5 Relatively little is known, however, about the modulation of voltage-gated sodium channels by tyrosine kinases. In pheochromocytoma cells, acute application of receptor tyrosine kinase agonists results in decreased sodium current, caused by a hyperpolarizing shift in the voltage dependence of steady-state inactivation.6 Moreover, sodium channels from brain microsomes form a complex with the receptor protein tyrosine phosphatase (RPTP), and activation of RPTP shifts the voltage dependence of inactivation to depolarized potentials.7 Although different mechanisms may be at play, in each case, tyrosine phosphorylation of these sodium channels is associated with a hyperpolarizing shift in steady-state inactivation, resulting in fewer available channels for generating an action potential. Limited evidence suggests that cardiac sodium currents behave oppositely from their neuronal counterparts. Specifically, the application of tyrosine kinase inhibitors to cardiac myocytes causes a hyperpolarizing shift in the inactivation-voltage relationship, suggesting that the phosphorylated form of the cardiac channel displays enhanced excitability.8 We describe in this study the effects of the Src family kinase Fyn on the cardiac sodium channel NaV1.5, and Fyn’s potential target residue in the channel’s cytoplasmic inactivation gate.
Materials and Methods
DNA Clones, Transfection, and Cell Culture
NaV1.5 and the NaV1.5 Y1494F, NaV1.5 Y1495F point mutants were generated as described previously.9 A C-terminal 14-amino-acid V5 tag was introduced using standard cloning procedures.10 V5 had no effect on the fast inactivation of the sodium currents (data not shown). The pHOOK, pCS2 c-FynCA, and c-FynKD clones were gifts from Dr T. Holmes and Alberto Llamas (New York University, New York, NY).11 Calcium phosphate was used to transiently transfect HEK-293 cells. For electrophysiology, cells cotransfected with pHOOK (Invitrogen) were identified by their binding to pHOX (4-ethoxymethylene-2-phenyl-2-oxazolin-5-one)eCcoated 2.0 to 2.5 e magnetic beads (Spherotech), which were synthesized using previously described methods.12
Tissue Lysis, Immunoprecipitation, and Western Blots
HEK-293 cells were lysed 48 hours after transfection in ice-cold lysis buffer containing (in mmol/L) 25 TRIS, 150 NaCl, 100 CsF, 100 NaF, 5 EDTA, 1 Na3VO4, and 1% Triton X-100 (pH 7.5), supplemented with a protease inhibitor cocktail (Roche), and lysates clarified by centrifugation. For antiphosphotyrosine assays, cells were supplemented with 100 eol/L peroxide-treated Na3VO4 5 minutes before lysis. Minced adult rat hearts were homogenized 3x10 seconds at 10K with a Polytron homogenizer in lysis buffer without Triton X-100 and centrifuged to remove nuclei, then membranes pelleted by centrifugation at 50 000g. The membrane pellet was resuspended in lysis buffer containing 1% Triton X-100, then clarified by centrifugation. The resulting supernatant was assayed for protein using the Bradford assay.
Transfected sodium channels were immunoprecipitated with anti-V5 antibody (Invitrogen), whereas tyrosine-phosphorylated protein from rat heart was immunoprecipitated with the 4G10 antibody (Upstate) from 500 e lysate with protein G agarose (Sigma Chemical). Immunoprecipitated proteins were separated by SDS-PAGE and transferred to PDVF membranes (Millipore). Blots were probed with antibodies against V5, Fyn (Transduction Laboratories), or the nonconserved I-II loop of NaV1.5 (Chemicon), and detection was by enhanced chemiluminescence.
The top panel of Figure 1A shows a representative V5 Western blot of HEK-293 lysates from cells nontransfected (NT), expressing NaV1.5 alone, or coexpressed with either FynCA (Constitutively Active) or FynKD (Kinase Dead). NaV1.5 is robustly expressed in HEK-293 cells and cotransfection with either Fyn mutant has no effect on this expression. In the bottom panel of Figure 1 is a Fyn Western blot of the indicated cell types showing that transfected FynCA and FynKD are comparably expressed, ruling out differences in expression levels of these exogenous kinases as a possible explanation for their differing modulatory effects (see Results). It is worth noting the expression of an endogenous Fyn, albeit at reduced levels, in the nontransfected (NT) lane. The last lane of this panel demonstrates the presence of Fyn in a rat cardiac myocyte preparation. We next tested whether the engineered Fyn mutations had their expected effects when expressed in HEK-293 cells. Figure 1B shows cell lysates from HEK cells nontransfected or expressing either FynCA or FynKD. Phosphotyrosine levels in these cell types were assayed by 4G10. NT cells have a basal level of tyrosine phosphorylation that is markedly increased with the expression of FynCA and mildly suppressed by FynKD, consistent with the constitutively active and dominant-negative effects predicted for these mutant kinases.
Fusion Proteins and Autoradiogram Procedures
The sodium channel III-IV loop amino acids Gly1481 to Asp1523 were cloned into the pQE-30 Xa vector (Qiagen) resulting in a fusion protein containing a N-terminal 6-histidine repeat followed by the sequence: GSGSGSGIEGRPFNGTGSLGGQDIFMTEEQKKYYNAMKKLGSKKPQKPIPRPLNKFQGFIFDKLN. The Tyr-to-Phe mutations at Y1494 and Y1495 were produced sequentially with Quickchange site-directed mutagenesis (Stratagene). All clones contained an additional mutation at Y1517F. Fusion proteins (2 e), isolated on Ni-NTA columns (Qiagen), were incubated with 5 ng recombinant Fyn (Invitrogen) in a buffer containing (in mmol/L) 25 Tris-HCl, 30 MgCl2, 5 MnCl2, 0.5 EGTA, 0.05 Na3VO4, 0.5 DTT, pH 7.2, and 2.5 e藽i 32P-ATP. Reactions were incubated for 1 hour at 30°C, separated on 16.5% Tris-Tricene gels (BioRad), transferred to PVDF membranes and quantified with a phosphorimager.
Electrophysiology
In most experiments, standard whole-cell recording methods13 were used to record ionic currents. The patch pipette contained (in mmol/L) 105 CsF, 35 NaCl, 10 EGTA, and 10 Hepes (pH 7.4). The bath contained (in mmol/L) 150 NaCl, 2 KCl, 1.5 CaCl2, 1 MgCl2, and 10 Hepes (pH 7.4). Liquid junction potentials between the bath and the pipette solution were corrected. Electrode resistance was in the range of 1 to 1.5 M. Data were collected 10 to 15 minutes after the establishment of the whole cell configuration. In one set of experiments (Figure 5), perforated patch recording was used with the antibiotic amphotericin B, according to previously described methods.14 The pipette solution contained (in mmol/L) 75 Cs2SO4, 35 NaCl, 20 CsCl, 8 MgCl2, and 10 Hepes (pH 7.4) (junction potential=eC7.9 mV). The access resistance at the time of recording ranged from 2.0 to 8.7 M. The voltage errors because of series resistance in both standard whole cell and perforated patch recordings were always <3 mV after compensation. All experiments were performed at room temperature.
Data Analysis
Data were analyzed using pCLAMP (Axon Instruments) and ORIGIN 7.0 (OriginLab). Throughout the article, error bars represent the standard error of the mean. G-V relations were fitted to the Boltzmann equation:
where RT/F=25 mV at room temperature.
Results
Acute Modulation of NaV1.5 by Insulin and Reversal of This Effect by PP2
We investigated whether tyrosine kinases modulate the cardiac sodium channel by transiently expressing the NaV1.5 isoform in HEK-293 cells and assaying the expressed channels’ response to 1 e/mL insulin, a concentration 2-fold higher than the effective dose in vascular endothelium.15 The insulin receptor endogenous to HEK-293 cells is a receptor tyrosine kinase that has been shown to activate Src family tyrosine kinases.16 Bisindolylmaleimide (100 nmol/L) was present in both pretreatment and bath solutions to ablate any competing effects of conventional protein kinase C, whose activation has been implicated in the insulin response pathway of HEK-293 cells.17 The effects of insulin incubation on expressed cardiac sodium channels were ascertained by recording sodium currents in the whole-cell patch clamp configuration.
Figure 2A shows the effects of insulin and of the specific Src family kinase inhibitor PP2 on families of sodium currents resulting from depolarizing steps from a holding potential of eC140 mV. Steady-state inactivation curves are shown in Figure 2B. Insulin pretreatment caused a small but significant depolarizing shift of the inactivation midpoint (Vmid) from eC95.0±2.4 to eC89.2±1.9 mV for 7 control and 10 insulin-treated cells, respectively (P=0.039). This result is consistent with hyperpolarizing shifts caused by specific inhibitors of Src family kinases on cardiac myocytes.8 It is possible that this modulation is because of the tyrosine kinase activity of the insulin receptor itself or one of its downstream effectors. If the modulation is because of the activation of an endogenous Src family tyrosine kinase, the effect should be blocked by PP2. Consistent with this, 1 eol/L PP2 completely reversed the insulin-induced depolarizing shift back to control levels, Vmid=eC96.2±1.6 mV (n=6), even though insulin was present in pretreatment and bath solution (P<0.01 for PP2+insulin versus insulin alone; Figure 2A and 2B). Figure 2C displays the peak activation conductance-voltage (G-V) relationships for the 3 classes of cell treatments. In contrast to the modulation of sodium channel inactivation properties, no evidence of either the insulin or PP2 treatment was seen in the G-V curves. The rightward shift of inactivation alone by insulin should produce an enhanced window current, ie, an increased overlap of steady-state activation and inactivation curves, which would result in enhanced excitability in the region of the resting potential. These results suggest that the NaV1.5 sodium channel is a target of a Src family tyrosine kinase.
Modulation of NaV1.5 by Fyn Is Manifested in Inactivation Properties
To test whether NaV1.5 can be modulated by a Src family kinase, we coexpressed NaV1.5 in HEK-293 cells with either of 2 mutants of the kinase Fyn. The constitutively active mutant, FynCA, lacks its inhibitory carboxyl terminus, whereas the "kinase-dead" mutant, FynKD, has a single point mutation at K299M rendering it catalytically inactive.11 Figure 3A shows families of sodium currents from cells expressing NaV1.5 alone or with either of the 2 Fyn mutants. Consistent with our experiments with insulin, channel coexpression with FynCA causes an 11-mV depolarizing shift in the steady-state inactivation curve (Figure 3B) from an average Vmid of eC100.4±2.0 mV for 9 FynKD cells to eC89.4±0.9 mV for 12 FynCA cells (P<0.001). The steady-state inactivation curve for NaV1.5 alone (9 cells) is shown as a line without symbols for clarity. This line is a Boltzmann fit of the NaV1.5 data and is bounded on the right and left by the mutant kinases, suggesting that a baseline tyrosine phosphorylation of the channel exists when it is expressed alone in HEK-293 cells. This result also suggests that FynKD is capable of displacing endogenous Fyn from binding sites in the vicinity of its phosphorylation targets in the sodium channel. There were no significant differences in the G-V relationships for the 3 cell types shown (Figure 3C).
Although coexpression of FynCA clearly shifts the steady-state inactivation curve, it has a smaller effect on the rate of fast inactivation. Figure 3D shows that FynCA causes a slight slowing of fast inactivation. This difference is statistically insignificant (P>0.05). Consistent with the depolarizing shift of steady-state inactivation, however, the rate of recovery from inactivation was increased in NaV1.5 channels coexpressed with FynCA (Figure 4). Moreover, channels in the presence of FynCA had a decreased entry rate into a slow-inactivated state (Figure 3E). Here, significance (P<0.01) is shown as an asterisk at the relevant time point. These results, like those with insulin, provide evidence for modulation of NaV1.5 by a tyrosine kinase.
Our whole-cell recordings use a pipette solution containing 105 mmol/L CsF. To test whether our conclusions might be confounded by fluoride, a known inhibitor of serine-threonine phosphatases, we examined the effects of these 2 Fyn variants on NaV1.5 currents in the absence of fluoride using perforated patch recording, a technique that causes less perturbation of the intracellular milieu than whole-cell recording.14 Figure 5 shows that, in agreement with our standard whole-cell recording, FynCA produces a 10-mV depolarizing shift of steady-state inactivation without affecting the activation G-V relationships.
Tyrosine kinase modulation of an ion channel may be mediated either by phosphorylation of a tyrosine residue or by the allosteric consequences of the kinase binding to the channel.18 In our case, the latter of these 2 possibilities is unlikely, given that both of our kinase mutants contain unperturbed SH3 binding domains and differ only in their ability to phosphorylate tyrosine residues.11 Furthermore, biophysical consequences of the expressed FynCA on sodium channel inactivation were similar to those observed for an acute insulin treatment, suggesting that these effects are because of direct tyrosine phosphorylation of the channel rather than alterations of the channel protein during biogenesis and trafficking.
Biochemical Evidence for Tyrosine Phosphorylation of NaV1.5
Figure 6A shows the results of immunoprecipitation of V5-tagged NaV1.5 channels from cells also expressing FynCA or FynKD, an experiment designed to test whether kinase affects the levels of tyrosine phosphorylation of the coexpressed sodium channel. The V5 Western blots show that sodium channels were efficiently isolated in equal proportions from both cell types. Antiphosphotyrosine blots of the same samples demonstrate that the channels’ phosphotyrosine levels are markedly increased in cells coexpressing FynCA compared with those coexpressing FynKD.
We have demonstrated that Fyn alters the biophysical properties of NaV1.5 and that this modulation may involve the phosphorylation of 1 or more tyrosine residues on the channel. We next asked whether cardiac sodium channels are tyrosine phosphorylated in vivo. To this end, we used a commercially available antibody specific for NaV1.5 to identify these sodium channels in a rat cardiac membrane preparation. Lane 1 of Figure 6B shows a Western blot of this membrane preparation and identifies the expected endogenous sodium channel. To determine whether these sodium channels were tyrosine phosphorylated, we immunoprecipitated tyrosine-phosphorylated protein from this cardiac preparation with a phosphotyrosine antibody, 4G10, and probed with the NaV1.5 antibody. The results of this immunoprecipitation are in the adjacent lanes in Figure 6B and demonstrate that NaV1.5 is indeed tyrosine phosphorylated in vivo. These biochemical results lead us to suggest that the modulation of the inactivation properties of cardiac sodium channels is because of direct phosphorylation of tyrosine residues on the channel.
Target Tyrosine Residue for Fyn Is in the Cytosolic III-IV Loop
The NaV1.5 isoform contains 22 cytosolic tyrosines. Given that FynCA modulates the inactivation properties of the channel, we narrowed our attention to tyrosine residues in the cytoplasmic linker between domains III and IV, the putative inactivation gate.19 Figure 7A shows a schematic rendering of the NaV1.5 sodium channel highlighting the 4 homologous domains and the cytosolic linkers that connect them. The polypeptide sequence of the III-IV linker is partially expanded and 2 potential target residues at positions 1494 and 1495 are shown in bold. Also shown in bold is the IFM (Ile, Phe, Met) inactivation "ball" that is directly upstream from these 2 tyrosines. The structure of this sequence has been solved using solution nuclear magnetic resonance (NMR)20 and is shown in Figure 7B with the relevant residues marked for clarity. Previous work from our laboratory showed that the double Y1494Y1495/QQ mutation has pronounced effects on the rates and voltage dependence of inactivation.9 To test the importance of these twin residues in the modulatory effects we observed, each was mutated individually to phenylalanine, and the single point-mutant sodium channels were coexpressed with FynCA or FynKD. These single point mutations themselves have little effect on channel function, consistent with previous results.9
The effect of FynCA on the first mutant, Y1494F, is shown in Figure 7C. Y1494F is indistinguishable from the wild-type channel and responds to the kinase with a 12-mV rightward shift in the steady-state inactivation (Vmid from eC97.3±2.1 to eC85.4±1.6 mV for 7 FynKD and 7 FynCA cells, respectively). This result suggests that phosphorylation at tyrosine 1494 is unlikely to mediate the effect of the tyrosine kinase. In striking contrast, the effects of both FynCA and FynKD are abolished in the channel carrying the Y1495F mutation (Figure 7D). Y1495F channels had average Vmid values for steady-state inactivation of eC93.9±1.2 mV for 8 FynCA and eC93.3±1.8 mV for 7 FynKD expressing cells, respectively, both of which are indistinguishable from the Y1495F mutant alone, Vmid=eC92.1±1.2. As in the case with the wild-type channel, Fyn modulation primarily affects the properties of steady-state inactivation, leaving the activation G-V relationships of both mutants unchanged (Figure 7C and 7D). Although these data strongly suggest Y1495 to be the site of phosphorylation, we cannot rule out the possibility that, in actuality, both residues (as well as other tyrosine residues in the channel) are phosphorylated yet only 1 site, Y1495, governs the kinase effect on inactivation. In support of multiple targets of Fyn, immunoprecipitation of either Y1494F or Y1495F channels from cells coexpressing FynCA had roughly equal phosphotyrosine levels (data not shown), consistent with work showing multiple targets of Src family kinases on other ion channels.21,22 Nonetheless, the ability of the single Y1495F mutation to abolish Fyn modulation strongly implicates this site in mediating the functional effects of the kinase. This result can be rationalized by the proximity of Y1495, but not Y1494, to M1487, a critical residue involved in inactivation,23 as shown in the NMR structure (Figure 7B) of this segment of the III-IV linker.20
Sodium Channel III-IV Loop Is a Substrate for Fyn
To ascertain whether the Y1495 residue is a bona fide target of Fyn, we generated a series of 6xHis-tagged fusion proteins that contained the III-IV loop of NaV1.5 for use as in vitro substrates for Fyn. We tested 4 fragments altogether: a wild-type fragment, 2 single tyrosine to phenylalanine mutations at the 1494 and 1495 positions, and a double mutant with both 1494 and 1495 sites changed to phenylalanine. Figure 8A shows a representative Coommassie blueeCstained gel containing these fragments. Each fragment shows the same pattern with a prominent band at roughly the predicted size of 8.5 kDa for the full-length fusion protein and 2 smaller bands, both more weakly expressed. To verify the authenticity of the fusion protein, we used an antibody directed against the N-terminal 6xHis tag, which showed reactivity exclusively with the highest molecular weight band (data not shown). Before loading the fragments in the gel shown in Figure 8A, each fragment was used as a substrate for Fyn-catalyzed 32P incorporation. An autoradiogram of the same gel from Figure 8A is shown in Figure 8B. The band at 55 kDa corresponds to the autophosphorylated form of the kinase and demonstrates that there was robust and similar levels of kinase activity for each reaction. The fusion protein near the bottom of the gel gives a signal that is most intense for the wild-type fragment and diminishes with each point mutation. Using a phosphorimager, we measured the signal from the 8.5 kDa band and quantified the results from 9 32P incorporation experiments with the averages from this analysis shown in Figure 8C. The single Y1494F and Y1495F mutants showed phosphorylation levels that were significantly different (P<0.001) from both wild-type and double-mutant substrates. It is worth noting that their phosphorylation levels were significantly different from each other as well (P<0.01). Given that the signal of the Y1494F fusion protein arises from phosphorylation at the Y1495 site, our data suggest that Y1495 is the preferred phosphorylation target for Fyn. Therefore, our results support the idea that the cytoplasmic III-IV linker is a substrate for the Src family tyrosine kinase Fyn and that Y1495 is preferred to its neighboring tyrosine at position 1494.
Discussion
We show in this study that the voltage-gated cardiac sodium channel NaV1.5 is modulated by the Src family tyrosine kinase Fyn. The effect of the kinase on the channel is manifested in altered steady-state inactivation. Fyn most likely achieves this effect by increasing the rates of recovery from fast inactivated states. We further provide evidence that NaV1.5, either cotransfected with FynCA or isolated from cardiac myocytes, is tyrosine phosphorylated. Lastly, the biophysical effects of FynCA on NaV1.5 can be abolished with a single point mutation, Y1495F, in the cytosolic linker between domains III and IV.
Previous work on tyrosine kinase effects on voltage-gated sodium channels showed that modulation takes the form of altered inactivation.6eC8 The modulation of the steady-state inactivation of NaV1.5 by FynCA is in agreement with these previous observations. However, although the modulation we see is similar in magnitude, it is opposite in direction from that observed for other channel isoforms. Specifically, when neuronal sodium channel isoforms are modulated by tyrosine phosphorylation, whether through growth factors6 or via an association with protein tyrosine phosphatase ,7 the phosphorylated channel displays a more hyperpolarized steady-state inactivation relationship. However, the depolarizing shift we report in this study is what one would predict from the response of isolated cardiac myocytes to tyrosine kinase inhibitors.8 The apparent target tyrosine residue on the cardiac channel, Y1495, is highly conserved among sodium channels, suggesting that the phosphorylation of this residue could stabilize or destabilize fast inactivated states of the channel, depending on the sodium channel isoform. Alternatively, other tyrosine residues, or other proteins associated with sodium channels, may be responsible for these isoform differences. The effect of tyrosine kinases on the neuronal mutant homologous to Y1495F will prove useful in elucidating the mechanisms at play.
What is the physiological relevance of cardiac sodium channel modulation by a Src family tyrosine kinase First, we note that NaV1.51 and Src24,25 are both concentrated at the adherens junctions responsible for the electrical coupling between cardiac myocytes. Thus, both are well placed to influence action potential propagation through the myocardium. Moreover, activation of Src family kinases reduces junctional coupling, in part by phosphorylating connexins.25
An increase in excitability because of tyrosine phosphorylation may play a compensatory role in physiological or pathological situations in which cardiac excitability has been compromised. Tyrosine kinase activity in the heart can be elevated during decidedly nonpathological stimulation by adrenergic ligands,4 angiotensin II,26 epidermal growth factor,27 or insulin.5 Tyrosine kinases may be activated also through the pathological states associated with cardiac ischemia and reperfusion injury,28,29 and postinfarct left ventricular remodeling.30 Furthermore, end-stage dilated cardiomyopathy is accompanied by dramatic alterations in insulin signaling31 and tyrosine phosphorylation.32 In both physiological and pathological cases, tyrosine kinase activity is coupled to other signaling pathways, including downstream and upstream kinases. Moreover, cardiac myocytes contain numerous other isoforms of ion channels subject to modulation by kinases and phosphatases, resulting in an inherently complex system. We give evidence here that Src family tyrosine kinases provide the heart with a potent tool for fine tuning cardiac electrical excitability, demonstrating yet another adaptive pathway for regulation of heart function.
Acknowledgments
This work was supported by NIH grant AR41691 (to R.H.). We thank Todd Holmes for the FynCA and FynKD constructs and for generous and voluminous advice on the many pitfalls of studying tyrosine phosphorylation, and Rocky Kass for his insightful comments on the manuscript.
References
Maier SK, Westenbroek RE, Schenkman KA, Feigl EO, Scheuer T, Catterall WA. An unexpected role for brain-type sodium channels in coupling of cell surface depolarization to contraction in the heart. Proc Natl Acad Sci U S A. 2002; 99: 4073eC4078.
Siegelbaum SA. Channel regulation. Ion channel control by tyrosine phosphorylation. Curr Biol. 1994; 4: 242eC245.
Levitan IB. Modulation of ion channels by protein phosphorylation and dephosphorylation. Ann Rev Physiol. 1994; 56: 193eC212.
Ma YC, Huang XY. Novel signaling pathway through the beta-adrenergic receptor. Trends Cardiovasc Med. 2002; 12: 46eC49.
Zhang YH, Hancox JC. A novel, voltage-dependent nonselective cation current activated by insulin in guinea pig isolated ventricular myocytes. Circ Res. 2003; 92: 765eC768.
Hilborn MD, Vaillancourt RR, Rane SG. Growth factor receptor tyrosine kinases acutely regulate neuronal sodium channels through the src signaling pathway. J Neurosci. 1998; 18: 590eC600.
Ratcliffe CF, Qu Y, McCormick KA, Tibbs VC, Dixon JE, Scheuer T, Catterall WA. A sodium channel signaling complex: modulation by associated receptor protein tyrosine phosphatase beta. Nat Neurosci. 2000; 3: 437eC444.
Wang YG, Wagner MB, Kumar R, Cheng J, Joyner RW. Inhibition of fast sodium current in rabbit ventricular myocytes by protein tyrosine kinase inhibitors. Pflugers Arch. 2003; 446: 485eC491.
O’Leary ME, Chen L-Q, Kallen RG, Horn R. A molecular link between activation and inactivation of sodium channels. J Gen Physiol. 1995; 106: 641eC658.
Sambrook, J, Fritsch, EF, Maniatis, T. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 1989.
Nitabach MN, Llamas DA, Araneda RC, Intile JL, Thompson IJ, Zhou YI, Holmes TC. A mechanism for combinatorial regulation of electrical activity: Potassium channel subunits capable of functioning as Src homology 3-dependent adaptors. Proc Natl Acad Sci U S A. 2001; 98: 705eC710.
Chesnut JD, Baytan AR, Russell M, Chang MP, Bernard A, Maxwell IH, Hoeffler JP. Selective isolation of transiently transfected cells from a mammalian cell population with vectors expressing a membrane anchored single-chain antibody. J Immunol Method. 1996; 193: 17eC27.
Ding S, Horn R. Effect of S6 tail mutations on charge movement in Shaker potassium channels. Biophys J. 2003; 84: 295eC305.
Horn R, Korn SJ. Prevention of rundown in electrophysiological recording. Methods Enzymol. 1992; 207: 149eC155.
Booth G, Stalker TJ, Lefer AM, Scalia R. Elevated ambient glucose induces acute inflammatory events in the microvasculature: effects of insulin. Am J Physiol Endocrinol Metab. 2001; 280: E848eCE856.
Dowler S, Montalvo L, Cantrell D, Morrice N, Alessi DR. Phosphoinositide 3-kinase-dependent phosphorylation of the dual adaptor for phosphotyrosine and 3-phosphoinositides by the Src family of tyrosine kinase. Biochem J. 2000; 349: 605eC610.
Voss M, Weernink PA, Haupenthal S, Moller U, Cool RH, Bauer B, Camonis JH, Jakobs KH, Schmidt M. Phospholipase D stimulation by receptor tyrosine kinases mediated by protein kinase C and a Ras/Ral signaling cascade. J Biol Chem. 1999; 274: 34691eC34698.
Nitabach MN, Llamas DA, Thompson IJ, Collins KA, Holmes TC. Phosphorylation-dependent and phosphorylation-independent modes of modulation of Shaker family voltage-gated potassium channels by Src family protein tyrosine kinases. J Neurosci. 2002; 22: 7913eC7922.
Patton DE, West JW, Catterall WA, Goldin AL. Amino acid residues required for fast Na+-channel inactivation: Charge neutralizations and deletions in the III-IV linker. Proc Natl Acad Sci U S A. 1992; 89: 10905eC10909.
Rohl CA, Boeckman FA, Baker C, Scheuer T, Catterall WA, Klevit RE. Solution structure of the sodium channel inactivation gate. Biochemistry. 1999; 38: 855eC861.
Holmes TC, Fadool DA, Levitan IB. Tyrosine phosphorylation of the Kv1.3 potassium channel. J Neurosci. 1996; 16: 1581eC1590.
Li Y, Langlais P, Gamper N, Liu F, Shapiro MS. Dual phosphorylations underlie modulation of unitary KCNQ K+ channels by Src tyrosine kinase. J Biol Chem. 2004; 279: 45399eC45407.
West JW, Patton DE, Scheuer T, Wang Y, Goldin AL, Catterall WA. A cluster of hydrophobic amino acid residues required for fast Na+-channel inactivation. Proc Natl Acad Sci U S A. 1992; 89: 10910eC10914.
Holmes TC, Fadool DA, Ren R, Levitan IB. Association of Src tyrosine kinase with a human potassium channel mediated by SH3 domain. Science. 1996; 274: 2089eC2091.
Toyofuku T, Akamatsu Y, Zhang H, Kuzuya T, Tada M, Hori M. c-Src regulates the interaction between connexin-43 and ZO-1 in cardiac myocytes. J Biol Chem. 2001; 276: 1780eC1788.
Sadoshima J, Qiu Z, Morgan JP, Izumo S. Angiotensin II and other hypertrophic stimuli mediated by G protein-coupled receptors activate tyrosine kinase, mitogen-activated protein kinase, and 90-kD S6 kinase in cardiac myocytes: the critical role of Ca2+-dependent signaling. Circ Res. 1995; 76: 1eC15.
Wu JY, Yu H, Cohen IS. Epidermal growth factor increases if in rabbit SA node cells by activating a tyrosine kinase. Biochimica Biophysica Acta. 2000; 1463: 15eC19.
Ping P, Zhang J, Zheng YT, Li RC, Dawn B, Tang XL, Takano H, Balafanova Z, Bolli R. Demonstration of selective protein kinase C-dependent activation of Src and Lck tyrosine kinases during ischemic preconditioning in conscious rabbits. Circ Res. 1999; 85: 542eC550.
Yamaura G, Turoczi T, Yamamoto F, Siddqui MA, Maulik N, Das DK. STAT signaling in ischemic heart: a role of STAT5A in ischemic preconditioning. Am J Physiol Heart Circ Physiol. 2003; 285: H476eCH482.
Melillo G, Lima JA, Judd RM, Goldschmidt-Clermont PJ, Silverman HS. Intrinsic myocyte dysfunction and tyrosine kinase pathway activation underlie the impaired wall thickening of adjacent regions during postinfarct left ventricular remodeling. Circulation. 1996; 93: 1447eC1458.
Shah A, Shannon RP. Insulin resistance in dilated cardiomyopathy. Rev Cardiovasc Med. 2003; 4: S50eCS57.
Podewski EK, Hilfiker-Kleiner D, Hilfiker A, Morawietz H, Lichtenberg A, Wollert KC, Drexler H. Alterations in Janus kinase (JAK)-signal transducers and activators of transcription (STAT) signaling in patients with end-stage dilated cardiomyopathy. Circulation. 2003; 107: 798eC802.(Christopher A. Ahern, Ji-)
Abstract
Dynamic modulation of ion channels can produce dramatic alterations of electrical excitability in cardiac myocytes. This study addresses the effects of the Src family tyrosine kinase Fyn on NaV1.5 cardiac sodium channels. Sodium currents were acquired by whole cell recording on HEK-293 cells transiently expressing NaV1.5. Acute treatment of cells with insulin caused a depolarizing shift in steady-state inactivation, an effect eliminated by the Src-specific tyrosine kinase inhibitor PP2. Sodium channels were coexpressed with either constitutively active (FynCA) or catalytically inactive (FynKD) variants of Fyn. FynCA caused a 10-mV depolarizing shift of steady-state inactivation compared with FynKD without altering the activation conductance-voltage relationship. Comparable effects of these Fyn variants were obtained with whole-cell and perforated-patch recording. Tyrosine phosphorylation of immunoprecipitated NaV1.5 was increased in cells expressing FynCA compared with FynKD. We show that Fyn is present in rat cardiac myocytes, and that NaV1.5 channels from these myocytes are tyrosine-phosphorylated. In HEK-293 cells the effect of FynCA on NaV1.5 inactivation is abolished by the single point mutation Y1495F, a residue located within the cytoplasmic linker between the third and fourth homologous domains of the sodium channel. We provide evidence that this linker is a substrate for Fyn in vitro, and that Y1495 is a preferred phosphorylation site. These results suggest that cardiac sodium channels are physiologically relevant targets of Src family tyrosine kinases.
Key Words: cardiac sodium channel tyrosine kinase Src Fyn phosphorylation
Introduction
Voltage-gated sodium channels play a pivotal role in cardiac excitability by controlling the upstroke of the cardiac action potential. These ion channels display exquisite sensitivity to changes in membrane potential and, as a consequence, minor alterations in their biophysical properties can have dramatic effects on cardiac function. Multiple isoforms of sodium channels are expressed in the heart, including the neuronal isoforms NaV1.1, 1.3, and 1.6, which have been shown to reside within T-tubules and participate in excitation-contraction coupling.1 However, in terms of conducting the cardiac action potential, the tetrodotoxin-resistant isoform, NaV1.5, is dominant in both its crude expression level and its contribution to the myocytes’ sodium current.1 For this reason, identification of regulatory factors that modify the behavior of NaV1.5 is crucial in understanding cardiac excitability.
Both Src family and receptor tyrosine kinases are known to be potent modulators of ion channels (see reviews2,3) Moreover, tyrosine kinase activity is enlisted in numerous signal transduction pathways. For example, multiple ligands such as insulin result in elevated tyrosine kinase activity in the heart.4,5 Relatively little is known, however, about the modulation of voltage-gated sodium channels by tyrosine kinases. In pheochromocytoma cells, acute application of receptor tyrosine kinase agonists results in decreased sodium current, caused by a hyperpolarizing shift in the voltage dependence of steady-state inactivation.6 Moreover, sodium channels from brain microsomes form a complex with the receptor protein tyrosine phosphatase (RPTP), and activation of RPTP shifts the voltage dependence of inactivation to depolarized potentials.7 Although different mechanisms may be at play, in each case, tyrosine phosphorylation of these sodium channels is associated with a hyperpolarizing shift in steady-state inactivation, resulting in fewer available channels for generating an action potential. Limited evidence suggests that cardiac sodium currents behave oppositely from their neuronal counterparts. Specifically, the application of tyrosine kinase inhibitors to cardiac myocytes causes a hyperpolarizing shift in the inactivation-voltage relationship, suggesting that the phosphorylated form of the cardiac channel displays enhanced excitability.8 We describe in this study the effects of the Src family kinase Fyn on the cardiac sodium channel NaV1.5, and Fyn’s potential target residue in the channel’s cytoplasmic inactivation gate.
Materials and Methods
DNA Clones, Transfection, and Cell Culture
NaV1.5 and the NaV1.5 Y1494F, NaV1.5 Y1495F point mutants were generated as described previously.9 A C-terminal 14-amino-acid V5 tag was introduced using standard cloning procedures.10 V5 had no effect on the fast inactivation of the sodium currents (data not shown). The pHOOK, pCS2 c-FynCA, and c-FynKD clones were gifts from Dr T. Holmes and Alberto Llamas (New York University, New York, NY).11 Calcium phosphate was used to transiently transfect HEK-293 cells. For electrophysiology, cells cotransfected with pHOOK (Invitrogen) were identified by their binding to pHOX (4-ethoxymethylene-2-phenyl-2-oxazolin-5-one)eCcoated 2.0 to 2.5 e magnetic beads (Spherotech), which were synthesized using previously described methods.12
Tissue Lysis, Immunoprecipitation, and Western Blots
HEK-293 cells were lysed 48 hours after transfection in ice-cold lysis buffer containing (in mmol/L) 25 TRIS, 150 NaCl, 100 CsF, 100 NaF, 5 EDTA, 1 Na3VO4, and 1% Triton X-100 (pH 7.5), supplemented with a protease inhibitor cocktail (Roche), and lysates clarified by centrifugation. For antiphosphotyrosine assays, cells were supplemented with 100 eol/L peroxide-treated Na3VO4 5 minutes before lysis. Minced adult rat hearts were homogenized 3x10 seconds at 10K with a Polytron homogenizer in lysis buffer without Triton X-100 and centrifuged to remove nuclei, then membranes pelleted by centrifugation at 50 000g. The membrane pellet was resuspended in lysis buffer containing 1% Triton X-100, then clarified by centrifugation. The resulting supernatant was assayed for protein using the Bradford assay.
Transfected sodium channels were immunoprecipitated with anti-V5 antibody (Invitrogen), whereas tyrosine-phosphorylated protein from rat heart was immunoprecipitated with the 4G10 antibody (Upstate) from 500 e lysate with protein G agarose (Sigma Chemical). Immunoprecipitated proteins were separated by SDS-PAGE and transferred to PDVF membranes (Millipore). Blots were probed with antibodies against V5, Fyn (Transduction Laboratories), or the nonconserved I-II loop of NaV1.5 (Chemicon), and detection was by enhanced chemiluminescence.
The top panel of Figure 1A shows a representative V5 Western blot of HEK-293 lysates from cells nontransfected (NT), expressing NaV1.5 alone, or coexpressed with either FynCA (Constitutively Active) or FynKD (Kinase Dead). NaV1.5 is robustly expressed in HEK-293 cells and cotransfection with either Fyn mutant has no effect on this expression. In the bottom panel of Figure 1 is a Fyn Western blot of the indicated cell types showing that transfected FynCA and FynKD are comparably expressed, ruling out differences in expression levels of these exogenous kinases as a possible explanation for their differing modulatory effects (see Results). It is worth noting the expression of an endogenous Fyn, albeit at reduced levels, in the nontransfected (NT) lane. The last lane of this panel demonstrates the presence of Fyn in a rat cardiac myocyte preparation. We next tested whether the engineered Fyn mutations had their expected effects when expressed in HEK-293 cells. Figure 1B shows cell lysates from HEK cells nontransfected or expressing either FynCA or FynKD. Phosphotyrosine levels in these cell types were assayed by 4G10. NT cells have a basal level of tyrosine phosphorylation that is markedly increased with the expression of FynCA and mildly suppressed by FynKD, consistent with the constitutively active and dominant-negative effects predicted for these mutant kinases.
Fusion Proteins and Autoradiogram Procedures
The sodium channel III-IV loop amino acids Gly1481 to Asp1523 were cloned into the pQE-30 Xa vector (Qiagen) resulting in a fusion protein containing a N-terminal 6-histidine repeat followed by the sequence: GSGSGSGIEGRPFNGTGSLGGQDIFMTEEQKKYYNAMKKLGSKKPQKPIPRPLNKFQGFIFDKLN. The Tyr-to-Phe mutations at Y1494 and Y1495 were produced sequentially with Quickchange site-directed mutagenesis (Stratagene). All clones contained an additional mutation at Y1517F. Fusion proteins (2 e), isolated on Ni-NTA columns (Qiagen), were incubated with 5 ng recombinant Fyn (Invitrogen) in a buffer containing (in mmol/L) 25 Tris-HCl, 30 MgCl2, 5 MnCl2, 0.5 EGTA, 0.05 Na3VO4, 0.5 DTT, pH 7.2, and 2.5 e藽i 32P-ATP. Reactions were incubated for 1 hour at 30°C, separated on 16.5% Tris-Tricene gels (BioRad), transferred to PVDF membranes and quantified with a phosphorimager.
Electrophysiology
In most experiments, standard whole-cell recording methods13 were used to record ionic currents. The patch pipette contained (in mmol/L) 105 CsF, 35 NaCl, 10 EGTA, and 10 Hepes (pH 7.4). The bath contained (in mmol/L) 150 NaCl, 2 KCl, 1.5 CaCl2, 1 MgCl2, and 10 Hepes (pH 7.4). Liquid junction potentials between the bath and the pipette solution were corrected. Electrode resistance was in the range of 1 to 1.5 M. Data were collected 10 to 15 minutes after the establishment of the whole cell configuration. In one set of experiments (Figure 5), perforated patch recording was used with the antibiotic amphotericin B, according to previously described methods.14 The pipette solution contained (in mmol/L) 75 Cs2SO4, 35 NaCl, 20 CsCl, 8 MgCl2, and 10 Hepes (pH 7.4) (junction potential=eC7.9 mV). The access resistance at the time of recording ranged from 2.0 to 8.7 M. The voltage errors because of series resistance in both standard whole cell and perforated patch recordings were always <3 mV after compensation. All experiments were performed at room temperature.
Data Analysis
Data were analyzed using pCLAMP (Axon Instruments) and ORIGIN 7.0 (OriginLab). Throughout the article, error bars represent the standard error of the mean. G-V relations were fitted to the Boltzmann equation:
where RT/F=25 mV at room temperature.
Results
Acute Modulation of NaV1.5 by Insulin and Reversal of This Effect by PP2
We investigated whether tyrosine kinases modulate the cardiac sodium channel by transiently expressing the NaV1.5 isoform in HEK-293 cells and assaying the expressed channels’ response to 1 e/mL insulin, a concentration 2-fold higher than the effective dose in vascular endothelium.15 The insulin receptor endogenous to HEK-293 cells is a receptor tyrosine kinase that has been shown to activate Src family tyrosine kinases.16 Bisindolylmaleimide (100 nmol/L) was present in both pretreatment and bath solutions to ablate any competing effects of conventional protein kinase C, whose activation has been implicated in the insulin response pathway of HEK-293 cells.17 The effects of insulin incubation on expressed cardiac sodium channels were ascertained by recording sodium currents in the whole-cell patch clamp configuration.
Figure 2A shows the effects of insulin and of the specific Src family kinase inhibitor PP2 on families of sodium currents resulting from depolarizing steps from a holding potential of eC140 mV. Steady-state inactivation curves are shown in Figure 2B. Insulin pretreatment caused a small but significant depolarizing shift of the inactivation midpoint (Vmid) from eC95.0±2.4 to eC89.2±1.9 mV for 7 control and 10 insulin-treated cells, respectively (P=0.039). This result is consistent with hyperpolarizing shifts caused by specific inhibitors of Src family kinases on cardiac myocytes.8 It is possible that this modulation is because of the tyrosine kinase activity of the insulin receptor itself or one of its downstream effectors. If the modulation is because of the activation of an endogenous Src family tyrosine kinase, the effect should be blocked by PP2. Consistent with this, 1 eol/L PP2 completely reversed the insulin-induced depolarizing shift back to control levels, Vmid=eC96.2±1.6 mV (n=6), even though insulin was present in pretreatment and bath solution (P<0.01 for PP2+insulin versus insulin alone; Figure 2A and 2B). Figure 2C displays the peak activation conductance-voltage (G-V) relationships for the 3 classes of cell treatments. In contrast to the modulation of sodium channel inactivation properties, no evidence of either the insulin or PP2 treatment was seen in the G-V curves. The rightward shift of inactivation alone by insulin should produce an enhanced window current, ie, an increased overlap of steady-state activation and inactivation curves, which would result in enhanced excitability in the region of the resting potential. These results suggest that the NaV1.5 sodium channel is a target of a Src family tyrosine kinase.
Modulation of NaV1.5 by Fyn Is Manifested in Inactivation Properties
To test whether NaV1.5 can be modulated by a Src family kinase, we coexpressed NaV1.5 in HEK-293 cells with either of 2 mutants of the kinase Fyn. The constitutively active mutant, FynCA, lacks its inhibitory carboxyl terminus, whereas the "kinase-dead" mutant, FynKD, has a single point mutation at K299M rendering it catalytically inactive.11 Figure 3A shows families of sodium currents from cells expressing NaV1.5 alone or with either of the 2 Fyn mutants. Consistent with our experiments with insulin, channel coexpression with FynCA causes an 11-mV depolarizing shift in the steady-state inactivation curve (Figure 3B) from an average Vmid of eC100.4±2.0 mV for 9 FynKD cells to eC89.4±0.9 mV for 12 FynCA cells (P<0.001). The steady-state inactivation curve for NaV1.5 alone (9 cells) is shown as a line without symbols for clarity. This line is a Boltzmann fit of the NaV1.5 data and is bounded on the right and left by the mutant kinases, suggesting that a baseline tyrosine phosphorylation of the channel exists when it is expressed alone in HEK-293 cells. This result also suggests that FynKD is capable of displacing endogenous Fyn from binding sites in the vicinity of its phosphorylation targets in the sodium channel. There were no significant differences in the G-V relationships for the 3 cell types shown (Figure 3C).
Although coexpression of FynCA clearly shifts the steady-state inactivation curve, it has a smaller effect on the rate of fast inactivation. Figure 3D shows that FynCA causes a slight slowing of fast inactivation. This difference is statistically insignificant (P>0.05). Consistent with the depolarizing shift of steady-state inactivation, however, the rate of recovery from inactivation was increased in NaV1.5 channels coexpressed with FynCA (Figure 4). Moreover, channels in the presence of FynCA had a decreased entry rate into a slow-inactivated state (Figure 3E). Here, significance (P<0.01) is shown as an asterisk at the relevant time point. These results, like those with insulin, provide evidence for modulation of NaV1.5 by a tyrosine kinase.
Our whole-cell recordings use a pipette solution containing 105 mmol/L CsF. To test whether our conclusions might be confounded by fluoride, a known inhibitor of serine-threonine phosphatases, we examined the effects of these 2 Fyn variants on NaV1.5 currents in the absence of fluoride using perforated patch recording, a technique that causes less perturbation of the intracellular milieu than whole-cell recording.14 Figure 5 shows that, in agreement with our standard whole-cell recording, FynCA produces a 10-mV depolarizing shift of steady-state inactivation without affecting the activation G-V relationships.
Tyrosine kinase modulation of an ion channel may be mediated either by phosphorylation of a tyrosine residue or by the allosteric consequences of the kinase binding to the channel.18 In our case, the latter of these 2 possibilities is unlikely, given that both of our kinase mutants contain unperturbed SH3 binding domains and differ only in their ability to phosphorylate tyrosine residues.11 Furthermore, biophysical consequences of the expressed FynCA on sodium channel inactivation were similar to those observed for an acute insulin treatment, suggesting that these effects are because of direct tyrosine phosphorylation of the channel rather than alterations of the channel protein during biogenesis and trafficking.
Biochemical Evidence for Tyrosine Phosphorylation of NaV1.5
Figure 6A shows the results of immunoprecipitation of V5-tagged NaV1.5 channels from cells also expressing FynCA or FynKD, an experiment designed to test whether kinase affects the levels of tyrosine phosphorylation of the coexpressed sodium channel. The V5 Western blots show that sodium channels were efficiently isolated in equal proportions from both cell types. Antiphosphotyrosine blots of the same samples demonstrate that the channels’ phosphotyrosine levels are markedly increased in cells coexpressing FynCA compared with those coexpressing FynKD.
We have demonstrated that Fyn alters the biophysical properties of NaV1.5 and that this modulation may involve the phosphorylation of 1 or more tyrosine residues on the channel. We next asked whether cardiac sodium channels are tyrosine phosphorylated in vivo. To this end, we used a commercially available antibody specific for NaV1.5 to identify these sodium channels in a rat cardiac membrane preparation. Lane 1 of Figure 6B shows a Western blot of this membrane preparation and identifies the expected endogenous sodium channel. To determine whether these sodium channels were tyrosine phosphorylated, we immunoprecipitated tyrosine-phosphorylated protein from this cardiac preparation with a phosphotyrosine antibody, 4G10, and probed with the NaV1.5 antibody. The results of this immunoprecipitation are in the adjacent lanes in Figure 6B and demonstrate that NaV1.5 is indeed tyrosine phosphorylated in vivo. These biochemical results lead us to suggest that the modulation of the inactivation properties of cardiac sodium channels is because of direct phosphorylation of tyrosine residues on the channel.
Target Tyrosine Residue for Fyn Is in the Cytosolic III-IV Loop
The NaV1.5 isoform contains 22 cytosolic tyrosines. Given that FynCA modulates the inactivation properties of the channel, we narrowed our attention to tyrosine residues in the cytoplasmic linker between domains III and IV, the putative inactivation gate.19 Figure 7A shows a schematic rendering of the NaV1.5 sodium channel highlighting the 4 homologous domains and the cytosolic linkers that connect them. The polypeptide sequence of the III-IV linker is partially expanded and 2 potential target residues at positions 1494 and 1495 are shown in bold. Also shown in bold is the IFM (Ile, Phe, Met) inactivation "ball" that is directly upstream from these 2 tyrosines. The structure of this sequence has been solved using solution nuclear magnetic resonance (NMR)20 and is shown in Figure 7B with the relevant residues marked for clarity. Previous work from our laboratory showed that the double Y1494Y1495/QQ mutation has pronounced effects on the rates and voltage dependence of inactivation.9 To test the importance of these twin residues in the modulatory effects we observed, each was mutated individually to phenylalanine, and the single point-mutant sodium channels were coexpressed with FynCA or FynKD. These single point mutations themselves have little effect on channel function, consistent with previous results.9
The effect of FynCA on the first mutant, Y1494F, is shown in Figure 7C. Y1494F is indistinguishable from the wild-type channel and responds to the kinase with a 12-mV rightward shift in the steady-state inactivation (Vmid from eC97.3±2.1 to eC85.4±1.6 mV for 7 FynKD and 7 FynCA cells, respectively). This result suggests that phosphorylation at tyrosine 1494 is unlikely to mediate the effect of the tyrosine kinase. In striking contrast, the effects of both FynCA and FynKD are abolished in the channel carrying the Y1495F mutation (Figure 7D). Y1495F channels had average Vmid values for steady-state inactivation of eC93.9±1.2 mV for 8 FynCA and eC93.3±1.8 mV for 7 FynKD expressing cells, respectively, both of which are indistinguishable from the Y1495F mutant alone, Vmid=eC92.1±1.2. As in the case with the wild-type channel, Fyn modulation primarily affects the properties of steady-state inactivation, leaving the activation G-V relationships of both mutants unchanged (Figure 7C and 7D). Although these data strongly suggest Y1495 to be the site of phosphorylation, we cannot rule out the possibility that, in actuality, both residues (as well as other tyrosine residues in the channel) are phosphorylated yet only 1 site, Y1495, governs the kinase effect on inactivation. In support of multiple targets of Fyn, immunoprecipitation of either Y1494F or Y1495F channels from cells coexpressing FynCA had roughly equal phosphotyrosine levels (data not shown), consistent with work showing multiple targets of Src family kinases on other ion channels.21,22 Nonetheless, the ability of the single Y1495F mutation to abolish Fyn modulation strongly implicates this site in mediating the functional effects of the kinase. This result can be rationalized by the proximity of Y1495, but not Y1494, to M1487, a critical residue involved in inactivation,23 as shown in the NMR structure (Figure 7B) of this segment of the III-IV linker.20
Sodium Channel III-IV Loop Is a Substrate for Fyn
To ascertain whether the Y1495 residue is a bona fide target of Fyn, we generated a series of 6xHis-tagged fusion proteins that contained the III-IV loop of NaV1.5 for use as in vitro substrates for Fyn. We tested 4 fragments altogether: a wild-type fragment, 2 single tyrosine to phenylalanine mutations at the 1494 and 1495 positions, and a double mutant with both 1494 and 1495 sites changed to phenylalanine. Figure 8A shows a representative Coommassie blueeCstained gel containing these fragments. Each fragment shows the same pattern with a prominent band at roughly the predicted size of 8.5 kDa for the full-length fusion protein and 2 smaller bands, both more weakly expressed. To verify the authenticity of the fusion protein, we used an antibody directed against the N-terminal 6xHis tag, which showed reactivity exclusively with the highest molecular weight band (data not shown). Before loading the fragments in the gel shown in Figure 8A, each fragment was used as a substrate for Fyn-catalyzed 32P incorporation. An autoradiogram of the same gel from Figure 8A is shown in Figure 8B. The band at 55 kDa corresponds to the autophosphorylated form of the kinase and demonstrates that there was robust and similar levels of kinase activity for each reaction. The fusion protein near the bottom of the gel gives a signal that is most intense for the wild-type fragment and diminishes with each point mutation. Using a phosphorimager, we measured the signal from the 8.5 kDa band and quantified the results from 9 32P incorporation experiments with the averages from this analysis shown in Figure 8C. The single Y1494F and Y1495F mutants showed phosphorylation levels that were significantly different (P<0.001) from both wild-type and double-mutant substrates. It is worth noting that their phosphorylation levels were significantly different from each other as well (P<0.01). Given that the signal of the Y1494F fusion protein arises from phosphorylation at the Y1495 site, our data suggest that Y1495 is the preferred phosphorylation target for Fyn. Therefore, our results support the idea that the cytoplasmic III-IV linker is a substrate for the Src family tyrosine kinase Fyn and that Y1495 is preferred to its neighboring tyrosine at position 1494.
Discussion
We show in this study that the voltage-gated cardiac sodium channel NaV1.5 is modulated by the Src family tyrosine kinase Fyn. The effect of the kinase on the channel is manifested in altered steady-state inactivation. Fyn most likely achieves this effect by increasing the rates of recovery from fast inactivated states. We further provide evidence that NaV1.5, either cotransfected with FynCA or isolated from cardiac myocytes, is tyrosine phosphorylated. Lastly, the biophysical effects of FynCA on NaV1.5 can be abolished with a single point mutation, Y1495F, in the cytosolic linker between domains III and IV.
Previous work on tyrosine kinase effects on voltage-gated sodium channels showed that modulation takes the form of altered inactivation.6eC8 The modulation of the steady-state inactivation of NaV1.5 by FynCA is in agreement with these previous observations. However, although the modulation we see is similar in magnitude, it is opposite in direction from that observed for other channel isoforms. Specifically, when neuronal sodium channel isoforms are modulated by tyrosine phosphorylation, whether through growth factors6 or via an association with protein tyrosine phosphatase ,7 the phosphorylated channel displays a more hyperpolarized steady-state inactivation relationship. However, the depolarizing shift we report in this study is what one would predict from the response of isolated cardiac myocytes to tyrosine kinase inhibitors.8 The apparent target tyrosine residue on the cardiac channel, Y1495, is highly conserved among sodium channels, suggesting that the phosphorylation of this residue could stabilize or destabilize fast inactivated states of the channel, depending on the sodium channel isoform. Alternatively, other tyrosine residues, or other proteins associated with sodium channels, may be responsible for these isoform differences. The effect of tyrosine kinases on the neuronal mutant homologous to Y1495F will prove useful in elucidating the mechanisms at play.
What is the physiological relevance of cardiac sodium channel modulation by a Src family tyrosine kinase First, we note that NaV1.51 and Src24,25 are both concentrated at the adherens junctions responsible for the electrical coupling between cardiac myocytes. Thus, both are well placed to influence action potential propagation through the myocardium. Moreover, activation of Src family kinases reduces junctional coupling, in part by phosphorylating connexins.25
An increase in excitability because of tyrosine phosphorylation may play a compensatory role in physiological or pathological situations in which cardiac excitability has been compromised. Tyrosine kinase activity in the heart can be elevated during decidedly nonpathological stimulation by adrenergic ligands,4 angiotensin II,26 epidermal growth factor,27 or insulin.5 Tyrosine kinases may be activated also through the pathological states associated with cardiac ischemia and reperfusion injury,28,29 and postinfarct left ventricular remodeling.30 Furthermore, end-stage dilated cardiomyopathy is accompanied by dramatic alterations in insulin signaling31 and tyrosine phosphorylation.32 In both physiological and pathological cases, tyrosine kinase activity is coupled to other signaling pathways, including downstream and upstream kinases. Moreover, cardiac myocytes contain numerous other isoforms of ion channels subject to modulation by kinases and phosphatases, resulting in an inherently complex system. We give evidence here that Src family tyrosine kinases provide the heart with a potent tool for fine tuning cardiac electrical excitability, demonstrating yet another adaptive pathway for regulation of heart function.
Acknowledgments
This work was supported by NIH grant AR41691 (to R.H.). We thank Todd Holmes for the FynCA and FynKD constructs and for generous and voluminous advice on the many pitfalls of studying tyrosine phosphorylation, and Rocky Kass for his insightful comments on the manuscript.
References
Maier SK, Westenbroek RE, Schenkman KA, Feigl EO, Scheuer T, Catterall WA. An unexpected role for brain-type sodium channels in coupling of cell surface depolarization to contraction in the heart. Proc Natl Acad Sci U S A. 2002; 99: 4073eC4078.
Siegelbaum SA. Channel regulation. Ion channel control by tyrosine phosphorylation. Curr Biol. 1994; 4: 242eC245.
Levitan IB. Modulation of ion channels by protein phosphorylation and dephosphorylation. Ann Rev Physiol. 1994; 56: 193eC212.
Ma YC, Huang XY. Novel signaling pathway through the beta-adrenergic receptor. Trends Cardiovasc Med. 2002; 12: 46eC49.
Zhang YH, Hancox JC. A novel, voltage-dependent nonselective cation current activated by insulin in guinea pig isolated ventricular myocytes. Circ Res. 2003; 92: 765eC768.
Hilborn MD, Vaillancourt RR, Rane SG. Growth factor receptor tyrosine kinases acutely regulate neuronal sodium channels through the src signaling pathway. J Neurosci. 1998; 18: 590eC600.
Ratcliffe CF, Qu Y, McCormick KA, Tibbs VC, Dixon JE, Scheuer T, Catterall WA. A sodium channel signaling complex: modulation by associated receptor protein tyrosine phosphatase beta. Nat Neurosci. 2000; 3: 437eC444.
Wang YG, Wagner MB, Kumar R, Cheng J, Joyner RW. Inhibition of fast sodium current in rabbit ventricular myocytes by protein tyrosine kinase inhibitors. Pflugers Arch. 2003; 446: 485eC491.
O’Leary ME, Chen L-Q, Kallen RG, Horn R. A molecular link between activation and inactivation of sodium channels. J Gen Physiol. 1995; 106: 641eC658.
Sambrook, J, Fritsch, EF, Maniatis, T. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 1989.
Nitabach MN, Llamas DA, Araneda RC, Intile JL, Thompson IJ, Zhou YI, Holmes TC. A mechanism for combinatorial regulation of electrical activity: Potassium channel subunits capable of functioning as Src homology 3-dependent adaptors. Proc Natl Acad Sci U S A. 2001; 98: 705eC710.
Chesnut JD, Baytan AR, Russell M, Chang MP, Bernard A, Maxwell IH, Hoeffler JP. Selective isolation of transiently transfected cells from a mammalian cell population with vectors expressing a membrane anchored single-chain antibody. J Immunol Method. 1996; 193: 17eC27.
Ding S, Horn R. Effect of S6 tail mutations on charge movement in Shaker potassium channels. Biophys J. 2003; 84: 295eC305.
Horn R, Korn SJ. Prevention of rundown in electrophysiological recording. Methods Enzymol. 1992; 207: 149eC155.
Booth G, Stalker TJ, Lefer AM, Scalia R. Elevated ambient glucose induces acute inflammatory events in the microvasculature: effects of insulin. Am J Physiol Endocrinol Metab. 2001; 280: E848eCE856.
Dowler S, Montalvo L, Cantrell D, Morrice N, Alessi DR. Phosphoinositide 3-kinase-dependent phosphorylation of the dual adaptor for phosphotyrosine and 3-phosphoinositides by the Src family of tyrosine kinase. Biochem J. 2000; 349: 605eC610.
Voss M, Weernink PA, Haupenthal S, Moller U, Cool RH, Bauer B, Camonis JH, Jakobs KH, Schmidt M. Phospholipase D stimulation by receptor tyrosine kinases mediated by protein kinase C and a Ras/Ral signaling cascade. J Biol Chem. 1999; 274: 34691eC34698.
Nitabach MN, Llamas DA, Thompson IJ, Collins KA, Holmes TC. Phosphorylation-dependent and phosphorylation-independent modes of modulation of Shaker family voltage-gated potassium channels by Src family protein tyrosine kinases. J Neurosci. 2002; 22: 7913eC7922.
Patton DE, West JW, Catterall WA, Goldin AL. Amino acid residues required for fast Na+-channel inactivation: Charge neutralizations and deletions in the III-IV linker. Proc Natl Acad Sci U S A. 1992; 89: 10905eC10909.
Rohl CA, Boeckman FA, Baker C, Scheuer T, Catterall WA, Klevit RE. Solution structure of the sodium channel inactivation gate. Biochemistry. 1999; 38: 855eC861.
Holmes TC, Fadool DA, Levitan IB. Tyrosine phosphorylation of the Kv1.3 potassium channel. J Neurosci. 1996; 16: 1581eC1590.
Li Y, Langlais P, Gamper N, Liu F, Shapiro MS. Dual phosphorylations underlie modulation of unitary KCNQ K+ channels by Src tyrosine kinase. J Biol Chem. 2004; 279: 45399eC45407.
West JW, Patton DE, Scheuer T, Wang Y, Goldin AL, Catterall WA. A cluster of hydrophobic amino acid residues required for fast Na+-channel inactivation. Proc Natl Acad Sci U S A. 1992; 89: 10910eC10914.
Holmes TC, Fadool DA, Ren R, Levitan IB. Association of Src tyrosine kinase with a human potassium channel mediated by SH3 domain. Science. 1996; 274: 2089eC2091.
Toyofuku T, Akamatsu Y, Zhang H, Kuzuya T, Tada M, Hori M. c-Src regulates the interaction between connexin-43 and ZO-1 in cardiac myocytes. J Biol Chem. 2001; 276: 1780eC1788.
Sadoshima J, Qiu Z, Morgan JP, Izumo S. Angiotensin II and other hypertrophic stimuli mediated by G protein-coupled receptors activate tyrosine kinase, mitogen-activated protein kinase, and 90-kD S6 kinase in cardiac myocytes: the critical role of Ca2+-dependent signaling. Circ Res. 1995; 76: 1eC15.
Wu JY, Yu H, Cohen IS. Epidermal growth factor increases if in rabbit SA node cells by activating a tyrosine kinase. Biochimica Biophysica Acta. 2000; 1463: 15eC19.
Ping P, Zhang J, Zheng YT, Li RC, Dawn B, Tang XL, Takano H, Balafanova Z, Bolli R. Demonstration of selective protein kinase C-dependent activation of Src and Lck tyrosine kinases during ischemic preconditioning in conscious rabbits. Circ Res. 1999; 85: 542eC550.
Yamaura G, Turoczi T, Yamamoto F, Siddqui MA, Maulik N, Das DK. STAT signaling in ischemic heart: a role of STAT5A in ischemic preconditioning. Am J Physiol Heart Circ Physiol. 2003; 285: H476eCH482.
Melillo G, Lima JA, Judd RM, Goldschmidt-Clermont PJ, Silverman HS. Intrinsic myocyte dysfunction and tyrosine kinase pathway activation underlie the impaired wall thickening of adjacent regions during postinfarct left ventricular remodeling. Circulation. 1996; 93: 1447eC1458.
Shah A, Shannon RP. Insulin resistance in dilated cardiomyopathy. Rev Cardiovasc Med. 2003; 4: S50eCS57.
Podewski EK, Hilfiker-Kleiner D, Hilfiker A, Morawietz H, Lichtenberg A, Wollert KC, Drexler H. Alterations in Janus kinase (JAK)-signal transducers and activators of transcription (STAT) signaling in patients with end-stage dilated cardiomyopathy. Circulation. 2003; 107: 798eC802.(Christopher A. Ahern, Ji-)