This Article Abstract Full Text (PDF) Data Supplement All Versions of this Article: 94/10/1325    most recent 01.RES.0000127621.54132.AEv1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Barbuti, A. Articles by DiFrancesco, D. Search for Related Content PubMed PubMed Citation Articles by Barbuti, A. Articles by DiFrancesco, D. Related Collections Cell biology/structural biology Cell signalling/signal transduction Ion channels/membrane transport

(Circulation Research. 2004;94:1325.)
© 2004 American Heart Association, Inc.

Cellular Biology

Localization of Pacemaker Channels in Lipid Rafts Regulates Channel Kinetics

Andrea Barbuti, Biagio Gravante, Monica Riolfo, Raffaella Milanesi, Benedetta Terragni, Dario DiFrancesco

From the Department of Biomolecular Sciences and Biotechnology (A.B., B.G., M.R., R.M., B.T., D.D.), Laboratory of Molecular Physiology and Neurobiology, University of Milano; and INFM-Milano U. Unit (D.D.), Milano, Italy.

Correspondence to Dario DiFrancesco, Department of Biomolecular Sciences and Biotechnology, Laboratory of Molecular Physiology and Neurobiology, via Celoria 26, 20133 Milano, Italy. E-mail dario.difrancesco{at}unimi.it

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Lipid rafts are discrete membrane subdomains rich in sphingolipids and cholesterol. In ventricular myocytes a function of caveolae, a type of lipid rafts, is to concentrate in close proximity several proteins of the ß-adrenergic transduction pathway. We have investigated the subcellular localization of HCN4 channels expressed in HEK cells and studied the effects of such localization on the properties of pacemaker channels in HEK and rabbit sinoatrial (SAN) cells. We used a discontinuous sucrose gradient and Western blot analysis to detect HCN4 proteins in HEK and in SAN cells, and found that HCN4 proteins localize to low-density membrane fractions together with flotillin (HEK) or caveolin-3 (SAN), structural proteins of caveolae. Lipid raft disruption by cell incubation with methyl-ß-cyclodextrin (MßCD) impaired specific HCN4 localization. It also shifted the midpoint of activation of the HCN4 current in HEK cells and of I_f in SAN cells to the positive direction by 11.9 and 10.4 mV, respectively. These latter effects were not due to elevation of basal cyclic nucleotide levels because the cholesterol-depletion treatment did not alter the current response to cyclic nucleotides. In accordance with an increased I_f, MßCD-treated SAN cells showed large increases of diastolic depolarization slope (87%) and rate (58%). We also found that the kinetics of HCN4- and native f-channel deactivation were slower after lipid raft disorganization. In conclusion, our work indicates that pacemaker channels localize to lipid rafts and that disruption of lipid rafts causes channels to redistribute within the membrane and modifies their kinetic properties.

Key Words: HCN channels • pacemaker current • lipid rafts • caveolin • sinoatrial node

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
The sinoatrial node (SAN) is the region of the heart from which spontaneous action potentials originate and propagate to determine cardiac rhythm. This particular anatomical district is composed by specialized myocytes (pacemaker cells) whose activity is characterized by a slow diastolic depolarization phase at the end of the action potential. The pacemaker (I_f) current plays a key role in the generation of diastolic depolarization. f-Channels open toward the end of the action potential repolarization process and carry an inward current that depolarizes the membrane and drives the membrane potential up to threshold for initiating a new action potential. Although spontaneous activity is an intrinsic property of the heart, independent of innervation, fine modulation of heart rate is achieved through the release of the neurotransmitters norepinephrine (NE) and acetylcholine (ACh) by sympathetic and parasympathetic branches of the autonomic nervous system. It is well established that in cardiac cells NE and ACh, through specific ß-adrenergic (ß-AR) and muscarinic (mAChR) receptors, modulate intracellular level of cAMP. Direct binding of cAMP to f-channels shifts their activation curve toward more depolarized potentials, thus increasing the steepness of the diastolic depolarization.¹ cAMP is a widespread second messenger, which affects a number of intracellular processes. For this reason, a broad, generalized increase or decrease in cytoplasmic cAMP concentration would be wasteful for the cell. One way to avoid an indiscriminate increase in cAMP concentration is through localization of the factors involved in its synthesis and degradation in discrete membrane subdomains such as the lipid rafts. The lipidic composition of these structures is different from the rest of the membrane: lipid rafts² are rich in sphingolipids and cholesterol packed together, which results in a less permeable and less fluid environment.³ Cholesterol-binding agents like cyclodextrins can deplete cells of membrane cholesterol and thus disrupt lipid rafts and reorganize the spatial distribution of associated proteins.^4–6

Caveolae represent a morphologically identifiable type of lipid rafts and are described as stable flask-like membrane invaginations.² It has been recently shown that several elements of the ß-AR pathway are localized to caveolae in cardiac myocytes, cardiac fibroblasts, and HEK293 cells.⁵ Because SAN myocytes are rich in caveolae,^7,8 and because f-channels, and their molecular correlates HCN channels, are finely regulated by ß-AR and mAChR activation, we asked whether these channels, too, are localized to discrete membrane microenvironments such as the lipid rafts. To this aim, we investigated the subcellular localization of HCN channels heterologously expressed in HEK293 cells and studied the functional effects of disrupting lipid rafts on the properties of HCN channels in HEK293 cells and of native f-channels in isolated SAN myocytes.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Isolation of Rabbit SAN Cells
All the procedures adopted in this work conformed to guidelines for the care and use of laboratory animals as established by State (D.L. 116/1992) and European directives (86/609/CEE). We used a standard protocol to isolate rabbit SAN cells.⁹ Briefly, animals weighing 0.8 to 1 kg were anesthetized by intramuscular injection of xylazine 4.6 mg/kg and ketamine 60 mg/kg and euthanized by cervical dislocation and exsanguination. Hearts were removed and kept at 37°C in Tyrode solution (in mmol/L: 140 NaCl, 5.4 KCl, 1.8 CaCl₂, 5.5 D-glucose, and 5 HEPES-NaOH; pH 7.4). After dissection, myocytes were digested and isolated by enzymatic and mechanical trituration.

Cell Culture and Transfection
Modified HEK293 cells (Phoenix cells, referred to as HEK cells) were kept in a culture medium (D-MEM, Gibco) containing FBS 10%, (Gibco) and PenStrep solution 1% (Sigma) at 37°C in a 5% CO₂ incubator. Cells were transiently transfected according to a standard calcium phosphate protocol.¹⁰ For biochemistry, 24 µg of rabbit HCN4 (rbHCN4¹¹) cDNA was used for each 100-mm Petri dish. For electrophysiology, 4 µg of rbHCN4 cDNA and 2 µg of CD8 cDNA were cotransfected in 35-mm Petri dishes. Two to 3 days after transfection cells were selected by immunoreactivity with CD8 antibody-covered beads (Dynabeads).

Cholesterol Depletion
Membrane cholesterol depletion was achieved by methyl-ß-cyclodextrin (MßCD, Sigma) treatment (see the expanded Materials and Methods section in the online data supplement at http://circres.ahajournals.org).

Lipid Raft Isolation and Western Blot
Isolation of lipid raft–enriched membranes from either HEK cells or SAN tissue was performed according to a detergent-free method, as described previously.⁵ Samples were solubilized in 1x SDS-PAGE buffer, separated by 10% acrylamide SDS-PAGE (12% for SAN tissue), and transferred overnight to nitrocellulose. Nitrocellulose was cut transversely for incubation with anti-HCN4 primary antibody (1:200, Alomone) and anti–caveolin-3 primary antibody for SAN cells (1:1000, Transduction Laboratories) or anti-flotillin primary antibody for caveolin-free⁵ HEK cells (1:250, Transduction Laboratories), rinsed, and incubated with peroxidase-conjugated secondary antibodies. Signals were revealed by ECL (enhanced chemiluminescence) (Amersham Biosciences).

For Western blot analysis of native tissue, 12 to 15 rabbit SAN preparations were frozen in liquid nitrogen and crushed in a mortar. The sample thus obtained was resuspended in sodium carbonate (pH 11), homogenized with a loose-fitting tissue homogenizer (10 strokes) and a tip sonicator (three 10-second bursts), and centrifuged for 5 minutes at 400g. The supernatant was collected and total protein content measured (Biorad DC protein essay) before sucrose gradient loading.

Electrophysiology and Data Analysis
Methods for electrophysiology and data analysis are described as in the expanded Materials and Methods. Statistical analysis was performed by t test comparison of paired or unpaired data as appropriate. Significance level was set to P=0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
As a first approach to investigate whether hyperpolarization-activated channels localize to specific subcellular compartments, we expressed rbHCN4 (from here on referred to simply as HCN4) in HEK cells. HCN4 was chosen because the mRNA of this isoform has been reported to be the most abundantly expressed in native cardiac pacemaker tissue.¹² We then used a discontinuous sucrose gradient protocol in order to determine in which membrane density fraction HCN4 proteins localize.

Figure 1 shows data from a Western blot analysis of proteins precipitated from various fractions of the gradient. Fractions were grouped according to their density (see Figure 1 legend) and loaded onto different lanes. The first two lanes represent low-density membranes, and the third lane, high-density membranes. Lipid rafts have a high content in sphingolipids and cholesterol and tend to stratify in low-density fractions.⁵ Antibodies against flotillin, a protein found in lipid rafts, were used to identify fractions containing lipid raft–associated proteins.

View larger version (59K): [in this window] [in a new window]   Figure 1. HCN4 proteins localize to lipid raft–enriched fractions in HEK cells. HCN4-transfected HEK cells were lysed, membranes were separated by discontinuous sucrose gradient, and 13 fractions were collected and divided as follows: first and second lane, lipid raft (LR)–enriched fractions 4 to 5 and 6 to 7, respectively; third lane, nonlipid raft (NLR) fractions 8 to 13. A, Control cells: both HCN4 and flotillin were detected in LR membranes only. B, Cells incubated in 1% MßCD: this time HCN4 and flotillin were detected in NLR as well as in LR fractions.

Data from control HEK cells expressing HCN4 are shown in Figure 1A. Bands corresponding to the molecular weight of HCN4 subunit and flotillin were detected only in the first two lanes, suggesting that HCN4 channels are associated to lipid rafts. In each blot, the total protein content was checked by red-ponceau staining (not shown).

In Figure 1B, HCN4-expressing HEK cells were treated with 1% MßCD before running the sucrose gradient protocol. By its cholesterol-depleting action, MßCD disorganizes lipid rafts and associated proteins.⁶ As a confirmation that this procedure effectively led to disruption of lipid rafts, the flotillin signal in Figure 1B was detected in all the membrane fractions. Under these conditions, the HCN4 signal, too, was present in the high- as well as in the low-density fractions, confirming the hypothesis that HCN4 proteins are compartmentalized in lipid rafts.

We next asked if the specific localization of HCN4 channels could affect their properties. To address this question, we recorded the current induced by HCN4 expression in HEK cells before and after treatment with 1% MßCD. The mean density of HCN4 current expression did not change significantly: at –135 mV, it was 44.5±14.5 and 37.1±6.7 pA/pF in control (n=6) and MßCD-treated cells (n=4), respectively (not significantly different). In Figure 2A, mean activation curves for HCN4 current were measured (see online data supplement) and plotted for untreated (squares, n=7) and for treated cells (circles, n=4).

View larger version (24K): [in this window] [in a new window]   Figure 2. Effect of cholesterol depletion on the properties of HCN4 channels expressed in HEK cells. A, left, Sample current traces recorded from a holding potential of –35 to –95/–135 mV from an untreated (left bar) and an MßCD-treated cell (right bar), scaled to maximal amplitude. A, right, Current activation curves of cholesterol-depleted (circles) and control cells (squares). Individual curves were fitted to the Boltzmann equation and averaged (values in text). Shift thus obtained was 11.9 mV. B, left, Sample current traces recorded during steps to –105 and to –135/–55 mV in untreated (top) and MßCD-treated cells (bottom). B, right, Plot of the time constants of activation (filled symbols) and deactivation (open symbols) in untreated (squares) and MßCD-treated cells (circles). Deactivation was strongly slowed by MßCD treatment. Statistical analysis of individual cell data yielded significant differences for V1/2 values in A and for deactivation time constants in B (see text for details).

Treatment with the cholesterol-depleting agent shifted the current activation curve to more positive voltages. Mean values of half-activation voltage and inverse slope factor were V_1/2=–93.1±2.5 and –81.2±3.6 mV (significantly different), and s=9.9±0.7 and 12.0±0.4 mV for untreated (n=7) and MßCD-treated cells (n=4), respectively (not significantly different). We also found that the MßCD-treatment altered channel kinetics because current deactivation was strongly slowed, as shown in Figure 2B (open symbols; see also deactivation records at –55 mV in the left panels in Figure 2B). At voltages in the range –75/–35 mV, the deactivation time constants of MßCD-treated cells increased 1.9- to 2.6-fold relative to untreated cells (significantly different). Time constants of activation were slightly modified at negative potentials (Figure 2B, filled symbols), but changes were not statistically significant.

These results suggest that heterologously expressed channels are located in lipid rafts and that this location affects their kinetic properties. This prompted us to investigate whether a similar mechanism could also be operating for native f-channels in SAN cells.

We first applied the discontinuous sucrose gradient protocol to SAN membranes to verify if pacemaker channels and lipid rafts colocalize. SAN myocytes are rich in caveolae,^7,8 a type of lipid raft containing caveolin and characterized by flask-shaped membrane invaginations. We therefore choose caveolin-3 to identify lipid raft–enriched membranes.⁵ In Figure 3, Western blots of SAN proteins precipitated from the various fractions of the gradient show that both HCN4 and caveolin-3 localize to the lipid raft–enriched membrane fractions 4-5 and 6-7.

View larger version (57K): [in this window] [in a new window]   Figure 3. Localization of HCN4 to caveolin-enriched fractions in SAN cell membranes. SAN membranes were separated with the same procedure as for HEK cell membranes (see Materials and Methods). HCN4 (top) and caveolin-3 (bottom) were both found in fractions 4-5 and 6-7 only.

We then measured the I_f current from isolated SAN myocytes before and after treatment with MßCD (2%). MßCD treatment did not modify the I_f current density, which was –23.0±1.4 and –26.1±3.8 pA/pF at –115 mV in untreated (n=14) and MßCD-treated cells (n=13), respectively (not significantly different).

In Figure 4, mean activation curves of I_f current measured in untreated (squares) and treated SAN cells (circles) are shown. In cholesterol-depleted cells, the activation curve was shifted toward more depolarized voltages relative to untreated cells, as found in HEK cells (Figure 2A). The mean half-activation voltage (V_1/2) was –70.1±1.0 mV (n=24) in untreated cells and –59.7±1.7 mV (n=21) in MßCD-treated cells (statistically significant); the inverse slope factor(s) did not differ in the two groups (untreated, 12.1±0.5 mV; treated, 12.2±0.5 mV).

View larger version (18K): [in this window] [in a new window]   Figure 4. Mean activation curves of f-channels before and after disruption of lipid rafts. Activation curves were measured in SAN myocytes as described in Materials and Methods in control and after treatment with MßCD 2%. Boltzmann fitting (values in text) yielded a V1/2 shift of 10.4 mV. Top, Sample traces recorded from –20 to –75/–125 mV in an untreated (left bar) and an MßCD-treated myocyte (right bar), scaled to maximal amplitude.

Disruption of caveolae is known to alter the activity of adenylate-cyclase and guanylate-cyclase.^5,13 In endothelial cells, for example, interaction between caveolin-1 and eNOS (endothelial nitric oxide synthase) is inhibitory on the activity of the enzyme and disruption of caveolae has been reported to increase basal levels of NO, which in turn activates guanylate cyclase.¹³ The subsequent increase in cytosolic cGMP concentration could therefore modulate f-channel by binding to its cyclic nucleotide-binding domain¹ and/or indirectly through inhibition of phosphodiesterases.¹⁴ Disorganization of caveolae by 2-hydroxipropyl-ß-cyclodextrin has also been reported to lead to an increase in isoproterenol-, zinterol-, and forskolin-induced cAMP elevation in cardiac myocytes.⁵

We therefore checked whether the positive shift of the activation curve could not result from an increased concentration of intracellular nucleotides.^1,15 To this aim, we recorded I_f from SAN cells and analyzed the action on I_f of forskolin and of SNP, which are known to stimulate adenylate cyclase and guanylate cyclase and increase I_f in SAN cells by a positive shift of activation curve due to elevation of intracellular cAMP and cGMP, respectively.^16,17

In Figure 5A, hyperpolarizing steps were applied to near the midactivation voltage and I_f recorded before and during perfusion with either forskolin (50 µmol/L, left) or SNP (100 µmol/L, right) in untreated (top) and 2% MßCD-treated cells (bottom). Shifts of the I_f activation curve causing the current increase were measured as previously reported^18,19 (see Materials and Methods), averaged, and plotted in the bar graphs of Figure 5B. In neither case did the cholesterol-depleting treatment cause a significant change in responsiveness to the activating agents. Mean shifts caused by forskolin were 6.4±0.7 (n=6) and 6.1±0.1 mV (n=3), and those caused by SNP were 6±1.3 (n=4) and 6.2±0.9 mV (n=6) in untreated and MßCD-treated cells, respectively. These data rule against the view that the cholesterol-depleting maneuver induces a positive shift of the I_f activation curve by increasing basal levels of cyclic nucleotides.

View larger version (26K): [in this window] [in a new window]   Figure 5. Lack of effect of cholesterol depletion on If response to forskolin and SNP. A, Sample If traces recorded from untreated and 2% MßCD-treated SAN cells, in a control solution and during perfusion with either forskolin (*50 µmol/L) or SNP (#100 µmol/L). Activating step was –75 mV (untreated) or –65 mV (MßCD-treated). B, Bar graphs of mean shifts of activation curve elicited by application of either forskolin (left) or SNP (right). Mean values after MßCD treatment were not significantly different from control values (P>0.05).

Increase of I_f, such as the one induced by ßAR-mediated depolarizing shift of the activation curve, leads to acceleration of spontaneous rate.^9,18,20 To see if the I_f changes caused by the cholesterol-depleting procedure were accompanied by significant changes in rate, we measured activity from free-beating SAN myocytes.

Figure 6A shows representative action potentials recorded from a control myocyte (top) and from a cell treated with MßCD (2%) (bottom). The rate was about 58% faster after exposure to MßCD (277±25 bpm, n=5) than in untreated cells (175±16 bpm, n=7, significantly different) (Figure 6C). The rate acceleration was due mostly to a steeper slope of the diastolic depolarization, although the action potential duration was also affected (Figure 6B): the diastolic slope was 168±21 mV/s in MßCD-treated cells (n=5) and 90±10 mV/s in untreated cells (n=7; significantly different), with an approximate increase of 87% (Figure 6C).

View larger version (29K): [in this window] [in a new window]   Figure 6. Cholesterol depletion speeds spontaneous activity of SAN cells. A, Sample recordings of spontaneous activity from an untreated (top) and from an MßCD-treated SAN cell (bottom). Temperature was 37°C. B, Superimposition of action potentials on an expanded time scale highlights the acceleration of diastolic depolarization. Note also that the action potential duration was shorter in the MßCD-treated cell. C, Depletion of membrane cholesterol significantly increased spontaneous rate (left) and the slope of diastolic depolarization (right; values in text).

We then investigated if cholesterol depletion could also affect I_f kinetics in SAN cells. In Figure 7, time constants of activation (filled symbols) and deactivation (open symbols) are shown for untreated (squares) and MßCD-treated cells (circles).

View larger version (15K): [in this window] [in a new window]   Figure 7. MßCD slows deactivation kinetics of If in SAN myocytes. Activation/deactivation traces were recorded at various potentials and fitted by a monoexponential function after a brief delay (see online data supplement) and time constants were averaged. Datapoints negative to –50 mV are activation time constants (filled symbols), and datapoints positive to –50 mV are deactivation time constants (open symbols). MßCD-treated cells (circles) showed significantly slower time constants of deactivation relative to untreated cells (squares). On the left, sample traces during steps from –20 to –85 and –125/–35 mV are plotted from an untreated (top) and an MßCD-treated myocyte; deactivation records at –35 mV are plotted on expanded scales (right bars).

Exposure to MßCD strongly slowed deactivation without affecting the activation process, similar to the results with HCN4 in HEK cells in Figure 2B (see representative activation/deactivation records in the left panels of Figure 7). In MßCD-treated cells the mean deactivation time constant was 1.9- to 4-fold larger than in control cells in the range –45 to 25 mV (significantly different at all voltages), whereas no significant changes were found for the activation time constant.

To ensure that the effects observed were due to cholesterol depletion and not to some other unspecific action of MßCD on pacemaker channels, we treated SAN cells with 3 mmol/L MßCD conjugated to 0.3 mmol/L cholesterol. Previous work has shown that treatment with this solution does not lead to cholesterol depletion (rather, membranes are found to be enriched in cholesterol),²¹ nor does it disrupt lipid rafts.⁶

In Figure 8, activation curves (A) and time constant curves (B) measured from cells treated with MßCD/cholesterol and day-matched control cells are shown. We did not find significant differences in the activation curves of I_f (see also sample current records in the left panels of Figure 8). Half-activation voltages (V_1/2) were –67.5±2.3 and –67.9±2.4 mV, and inverse slope factors (s) were 13.8±0.5 and 12.1±0.6 mV in control (n=6) and MßCD/cholesterol-treated cells (n=6), respectively. Time constants in Figure 8B were also not significantly different except at V=–45 mV, where the deactivation time constant decreased from 1776 ms in control cells to 1336 ms in treated cells. It should be noticed that this change is opposite to that caused by treatment with MßCD alone (see Figures 2B and 7). Treatment with MßCD/cholesterol did not affect spontaneous activity: spontaneous rate was 189±31 and 170±12 bpm, and diastolic depolarization rate was 110.0±13.7 and 87.6±13.4 V/s in day-matched control (n=3) and treated cells (n=3), respectively (not significantly different). These data agree with the hypothesis that treatment with MßCD causes membrane cholesterol depletion and consequent disassembly of lipid rafts, and that this effect is responsible for modification of I_f kinetics.

View larger version (21K): [in this window] [in a new window]   Figure 8. Cholesterol-saturated MßCD does not modify If kinetics. SAN myocytes were incubated with a solution of MßCD saturated with cholesterol (3 mmol/L MßCD/0.3 mmol/L cholesterol). A, Mean activation curve in cells treated with MßCD/cholesterol (circles) was unchanged relative to untreated cells (squares). Statistical analysis did not yield significant differences (values in text). On the left, representative current records during steps from –20 to –75/–125 mV from untreated (left bar) and MßCD/cholesterol-treated cells (right bar). B, Time constants of activation (filled symbols) and deactivation (open symbols) were similar in control cells (squares) and after MßCD/cholesterol treatment (circles). Statistical analysis did not yield significant differences in time constant values, except for the points at –45 mV (see text for details). Representative records during steps from –20 to –85 and –125/–35 mV are plotted on the left for an untreated (top) and an MßCD/cholesterol-treated cell (bottom).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
In SAN myocytes, I_f activation during the last fraction of action potential repolarization is a key process in the generation of the diastolic depolarization and spontaneous activity. The role of I_f in pacemaking is well established.^22–24 f-Channels play an essential role not only in the generation of pacemaker activity, but also, importantly, in the control of pacemaker rate by autonomic neurotransmitters. This mechanism is based on the modulation of intracellular cAMP by the opposite action of ß-adrenergic and muscarinic cholinergic stimuli.^1,9,20,25

f-Channels are activated by direct binding of cAMP molecules to channels.¹ This property likely reflects a need for fast channel modulation. This is not surprising, because the ability to respond quickly to autonomic stimulation represents an aspect of cardiac pacemaking that is of basic physiological relevance.

Fast, efficient responses to neurotransmitter-induced modulation mediated by cAMP require not only that the second messenger and its final target interact directly, but also that all the elements contributing to the modulatory mechanism are confined to a restricted space and are sufficiently close to each other.

Indeed, the notion of restricted localization of components of the cAMP-dependent biochemical pathway has been known for some time,^26–28 and in working cardiac myocytes there is evidence for restricted pools of cAMP functional to ß-AR modulation of Ca²⁺ channels and involved in contraction.^27,29

It has been recently shown that molecules involved in ß-AR signaling cascade in cardiac myocytes localize to caveolar lipid rafts.⁵ SAN cells have a large density of caveolae on their plasma membranes, which increase the surface area by about 100%, whereas in ventricular myocytes, the increase in surface due to caveolae is estimated to be about 30%.⁸ As in ventricular myocytes,⁵ caveolar rafts in SAN myocytes may therefore have a role in recruiting sets of proteins involved in specific membrane processes.

In this study, we have shown that HCN4 channels expressed in HEK cells or in SAN cells are localized to lipid rafts and that disruption of these membrane subdomains by cholesterol depletion affects cellular compartmentation of channels and their kinetic properties. In SAN cells, whose membranes are characterized by a high density of caveolar rafts according to morphological data,⁸ our data point to a caveolar localization of native f-channels.

Western blot analysis showed that HCN4 proteins are normally present exclusively in the lighter, raft-enriched fraction (first two lanes in Figure 1A). Treatment of HEK cells with the cholesterol-binding agent MßCD caused flotillin to lose its specific localization to lipid raft–enriched membranes. Under these conditions, HCN4 proteins were also distributed nonspecifically throughout all density fractions (Figure 1B), indicating that HCN4 channels are associated to lipid rafts.

Disruption of lipid rafts also affected HCN4- and f-channels kinetics. After MßCD, I_f and HCN4-induced currents activated at more depolarized voltages than in untreated cells (Figures 2 and 4). We found that both forskolin and SNP are as effective on I_f in cells treated with MßCD as they are in control cells (Figure 5), indicating that the MßCD-induced shift of the current activation curve was not due to increased basal levels of either cAMP or cGMP.

Because a rightward shift of I_f activation is responsible for the positive chronotropic action of sympathetic innervation, cholesterol depletion should affect spontaneous beating rate of isolated SAN myocytes. Indeed, MßCD-treated cells showed a much faster diastolic depolarization and beating rates than control cells (Figure 6).

As well as shifting the current activation curve, the cholesterol-depleting procedure slowed deactivation kinetics of both HCN4 current and I_f, without modifying appreciably activation (Figures 2 and 7 ). We do not have a ready explanation for this effect. It should be noted however that unequal modifications of activation and deactivation kinetics are known to occur for I_f. For example, the cAMP-induced shift of the I_f activation curve occurs in association with a shifting action that is stronger for deactivation than for activation time constants, suggesting that a decreased rate of channel closing, more than an increased rate of channel opening, contributes to the cAMP action.³⁰ Furthermore, deactivation, but not activation of HCN2 channels, is slowed when external K⁺ is replaced by Na⁺, Li⁺, or NMG⁺.³¹

Because changes in the concentration of cyclic-nucleotides are not responsible for the modification of HCN4/I_f kinetics, our results raise the question whether other channel modulating mechanisms may exist whose action depends on the integrity of lipid rafts.

A most natural candidate could be the MiRP1 protein (KCNE2).^32,33 If MiRP1, acting as a ß-subunit, modulates channel activity, then disruption of the lipid raft environment might impair its action. However, MiRP1 has a strong, nearly 2-fold slowing effect on HCN4 activation,³³ and impairment of MiRP1 action should therefore lead to a nearly 2-fold acceleration of channel kinetics, which is not compatible with the results we have obtained with MßCD treatment (Figure 2); secondly, we did not observe significant changes in HCN4 current density after MßCD, whereas roughly a 50% reduction should be expected if MiRP1 is involved.³³ These results indicate that MiRP1 is unlikely to be responsible for the slowing of deactivation seen on disruption of lipid rafts.

Regardless of the detailed mode of action, our data indicate the presence of an interaction between channels and lipid raft integrity that affects channel kinetics; this interaction appears to function essentially by accelerating channel deactivation.

Although our data indicate an important contribution of I_f to MßCD-dependent changes of the rate of spontaneous activity, modifications of other currents or processes (such as Ca²⁺ homeostasis) cannot be ruled out. For example, it is known that Kv1.5 and Kv2.1 channels normally reside in lipid rafts both in native tissues and in heterologous expression systems, and that disruption of such structures alter their properties.^4,34 Furthermore, it has been reported that membrane cholesterol depletion alters the properties of the Na⁺-K⁺ pump in erythrocytes.³⁵

Physiological ß-AR modulation of rate in the SAN is thought to be mediated mostly by ß₁-adrenergic receptors, particularly in view of the much larger density of ß₁ than of ß₂ receptors,³⁶ although ß₂ stimulation also contributes to rate modulation in several species.^37–39 In ventricular cells, the largest fraction of caveolar ß-ARs belongs to the ß₂ subtype,⁵ and if a similar situation applied to SAN cells, our data would appear to suggest a preferential coupling of pacemaker channels to ß₂ receptors. However, the density of ß₁,⁴⁰ as well as of ß₂ receptors,³⁶ is reported to be several-fold higher in nodal than in surrounding atrial tissue, implying that the relative distribution of the two subtypes may be tissue-specific. The exact distribution of ß-AR subtypes in SAN cell membranes requires further detailed investigation.

In conclusion, our data in HEK and SAN cells indicate that pacemaker channels are located in lipid rafts, and that disruption of the interaction between channels and lipid rafts alters the kinetic properties of channels. Finally, concentration within lipid rafts agrees with evidence collected by single-channel experiments, indicating that f-channels are densely packed in restricted areas (hot-spots), rather than being uniformly spread across the whole surface of SAN membranes.⁴¹

	Acknowledgments

 
This work was supported by FIRB (Fondo per gli Investimenti della Ricerca di Base) 2001, MIUR (Ministero dell’ Istruzione, dell’ Università e della Ricerca) Cofin 2002, and MIUR ST/97 grants to Dario DiFrancesco. We thank M. Baruscotti, A. Moroni, Richard B. Robinson, and Susan F. Steinberg for advice and discussion.

	Footnotes

 
Original received October 27, 2003; resubmission received March 1, 2004; revised resubmission received March 25, 2004; accepted March 29, 2004.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. DiFrancesco D, Tortora P. Direct activation of cardiac pacemaker channels by intracellular cyclic AMP. Nature. 1991; 351: 145–147.[CrossRef][Medline] [Order article via Infotrieve]
2. Lai EC. Lipid rafts make for slippery platforms. J Cell Biol. 2003; 162: 365–370.[Abstract/Free Full Text]
3. Harder T, Simons K. Caveolae, DIGs, and the dynamics of sphingolipid-cholesterol microdomains. Curr Opin Cell Biol. 1997; 9: 534–542.[CrossRef][Medline] [Order article via Infotrieve]
4. Martens JR, Navarro-Polanco R, Coppock EA, Nishiyama A, Parshley L, Grobaski TD, Tamkun MM. Differential targeting of Shaker-like potassium channels to lipid rafts. J Biol Chem. 2000; 275: 7443–7446.[Abstract/Free Full Text]
5. Rybin VO, Xu X, Lisanti MP, Steinberg SF. Differential targeting of ß-adrenergic receptor subtypes and adenylyl cyclase to cardiomyocyte caveolae: a mechanism to functionally regulate the cAMP signaling pathway. J Biol Chem. 2000; 275: 41447–41457.[Abstract/Free Full Text]
6. Oshikawa J, Toya Y, Fujita T, Egawa M, Kawabe J, Umemura S, Ishikawa Y. Nicotinic acetylcholine receptor ₇ regulates cAMP signal within lipid rafts. Am J Physiol Cell Physiol. 2003; 285: C567–C574.[Abstract/Free Full Text]
7. Masson-Pevet M, Bleeker WK, Gros D. The plasma membrane of leading pacemaker cells in the rabbit sinus node: a qualitative and quantitative ultrastructural analysis. Circ Res. 1979; 45: 621–629.[Abstract/Free Full Text]
8. Masson-Pevet M, Gros D, Besselsen E. The caveolae in rabbit sinus node and atrium. Cell Tissue Res. 1980; 208: 183–196.[Medline] [Order article via Infotrieve]
9. DiFrancesco D, Ferroni A, Mazzanti M, Tromba C. Properties of the hyperpolarizing-activated current (if) in cells isolated from the rabbit sino-atrial node. J Physiol. 1986; 377: 61–88.[Abstract/Free Full Text]
10. Vaccari T, Moroni A, Rocchi M, Gorza L, Bianchi ME, Beltrame M, DiFrancesco D. The human gene coding for HCN2, a pacemaker channel of the heart. Biochim Biophys Acta. 1999; 1446: 419–425.[Medline] [Order article via Infotrieve]
11. Viscomi C, Altomare C, Bucchi A, Camatini E, Baruscotti M, Moroni A, DiFrancesco D. C terminus-mediated control of voltage and cAMP gating of hyperpolarization-activated cyclic nucleotide-gated channels. J Biol Chem. 2001; 276: 29930–29934.[Abstract/Free Full Text]
12. Shi W, Wymore R, Yu H, Wu J, Wymore RT, Pan Z, Robinson RB, Dixon JE, McKinnon D, Cohen IS. Distribution and prevalence of hyperpolarization-activated cation channel (HCN) mRNA expression in cardiac tissues. Circ Res. 1999; 85: e1–e6.[Medline] [Order article via Infotrieve]
13. Goligorsky MS, Li H, Brodsky S, Chen J. Relationships between caveolae and eNOS: everything in proximity and the proximity of everything. Am J Physiol Renal Physiol. 2002; 283: F1–F10.[Abstract/Free Full Text]
14. Accili EA, Redaelli G, DiFrancesco D. Differential control of the hyperpolarization-activated current (If) by cAMP gating and phosphatase inhibition in rabbit sino-atrial node myocytes. J Physiol. 1997; 500: 643–651.[Medline] [Order article via Infotrieve]
15. DiFrancesco D, Mangoni M. Modulation of single hyperpolarization-activated channels (I_f) by cAMP in the rabbit sino-atrial node. J Physiol. 1994; 474: 473–482.[Abstract/Free Full Text]
16. DiFrancesco D, Tromba C. Muscarinic control of the hyperpolarization-activated current (I_f) in rabbit sino-atrial node myocytes. J Physiol. 1988; 405: 493–510.[Abstract/Free Full Text]
17. Yoo S, Lee SH, Choi BH, Yeom JB, Ho W, Earm YE. Dual effect of nitric oxide on the hyperpolarization-activated inward current (I_f) in sino-atrial node cells of the rabbit. J Mol Cell Cardiol. 1998; 30: 2729–2738.[CrossRef][Medline] [Order article via Infotrieve]
18. Accili EA, Redaelli G, DiFrancesco D. Activation of the hyperpolarization-activated current (I_f) in sino-atrial node myocytes of the rabbit by vasoactive intestinal peptide. Pflugers Arch. 1996; 431: 803–805.[Medline] [Order article via Infotrieve]
19. Bois P, Renaudon B, Baruscotti M, Lenfant J, DiFrancesco D. Activation of f-channels by cAMP analogues in macropatches from rabbit sino-atrial node myocytes. J Physiol. 1997; 501: 565–571.[CrossRef][Medline] [Order article via Infotrieve]
20. Brown HF, DiFrancesco D, Noble SJ. How does adrenaline accelerate the heart? Nature. 1979; 280: 235–236.[CrossRef][Medline] [Order article via Infotrieve]
21. Sooksawate T, Simmonds MA. Effects of membrane cholesterol on the sensitivity of the GABA_A receptor to GABA in acutely dissociated rat hippocampal neurones. Neuropharmacology. 2001; 40: 178–184.[CrossRef][Medline] [Order article via Infotrieve]
22. DiFrancesco D. Pacemaker mechanisms in cardiac tissue. Annu Rev Physiol. 1993; 55: 455–472.[CrossRef][Medline] [Order article via Infotrieve]
23. Robinson RB, Siegelbaum SA. Hyperpolarization-activated cation currents: from molecules to physiological function. Annu Rev Physiol. 2003; 65: 480.
24. Accili EA, Proenza C, Baruscotti M, DiFrancesco D. From funny current to HCN channels: 20 years of excitation. News Physiol Sci. 2002; 17: 32–37.[Abstract/Free Full Text]
25. DiFrancesco D, Tromba C. Acetylcholine inhibits activation of the cardiac hyperpolarizing-activated current, If. Pflugers Arch. 1987; 410: 139–142.[CrossRef][Medline] [Order article via Infotrieve]
26. Buxton IL, Brunton LL. Compartments of cyclic AMP and protein kinase in mammalian cardiomyocytes. J Biol Chem. 1983; 258: 10233–10239.[Abstract/Free Full Text]
27. Jurevicius J, Fischmeister R. cAMP compartmentation is responsible for a local activation of cardiac Ca²⁺ channels by ß-adrenergic agonists. Proc Natl Acad Sci U S A. 1996; 93: 295–299.[Abstract/Free Full Text]
28. Steinberg SF, Brunton LL. Compartmentation of G protein-coupled signaling pathways in cardiac myocytes. Annu Rev Pharmacol Toxicol. 2001; 41: 751–773.[CrossRef][Medline] [Order article via Infotrieve]
29. Zaccolo M, Pozzan T. Discrete microdomains with high concentration of cAMP in stimulated rat neonatal cardiac myocytes. Science. 2002; 295: 1711–1715.[Abstract/Free Full Text]
30. DiFrancesco D. Dual allosteric modulation of pacemaker (f) channels by cAMP and voltage in rabbit SA node. J Physiol. 1999; 515: 367–376.[Abstract/Free Full Text]
31. Azene E, Xue T, Li RA. Molecular basis of the effect of potassium on heterologously expressed pacemaker (HCN) channels. J Physiol. 2003; 547: 349–356.[Abstract/Free Full Text]
32. Yu H, Wu J, Potapova I, Wymore RT, Holmes B, Zuckerman J, Pan Z, Wang H, Shi W, Robinson RB, El Maghrabi MR, Benjamin W, Dixon J, McKinnon D, Cohen IS, Wymore R. MinK-related peptide 1: a ß subunit for the HCN ion channel subunit family enhances expression and speeds activation. Circ Res. 2001; 88: e84–e87.
33. Decher N, Bundis F, Vajna R, Steinmeyer K. KCNE2 modulates current amplitudes and activation kinetics of HCN4: influence of KCNE family members on HCN4 currents. Pflugers Arch. 2003; 466: 633–640.
34. Martens JR, Sakamoto N, Sullivan SA, Grobaski TD, Tamkun MM. Isoform-specific localization of voltage-gated K⁺ channels to distinct lipid raft populations: targeting of Kv1.5 to caveolae. J Biol Chem. 2001; 276: 8409–8414.[Abstract/Free Full Text]
35. Bastiaanse EM, Hold KM, Van der LA. The effect of membrane cholesterol content on ion transport processes in plasma membranes. Cardiovasc Res. 1997; 33: 272–283.[Abstract/Free Full Text]
36. Rodefeld MD, Beau SL, Schuessler RB, Boineau JP, Saffitz JE. ß -Adrenergic and muscarinic cholinergic receptor densities in the human sinoatrial node: identification of a high ß2-adrenergic receptor density. J Cardiovasc Electrophysiol. 1996; 7: 1039–1049.[Medline] [Order article via Infotrieve]
37. Hall JA, Petch MC, Brown MJ. Intracoronary injections of salbutamol demonstrate the presence of functional ß2-adrenoceptors in the human heart. Circ Res. 1989; 65: 546–553.[Abstract/Free Full Text]
38. Lemoine H, Kaumann AJ. Regional differences of ß1- and ß2-adrenoceptor-mediated functions in feline heart: a ß2-adrenoceptor-mediated positive inotropic effect possibly unrelated to cyclic AMP. Naunyn Schmiedebergs Arch Pharmacol. 1991; 344: 56–69.[Medline] [Order article via Infotrieve]
39. Lowe MD, Lynham JA, Grace AA, Kaumann AJ. Comparison of the affinity of ß-blockers for two states of the ß1-adrenoceptor in ferret ventricular myocardium. Br J Pharmacol. 2002; 135: 451–461.[CrossRef][Medline] [Order article via Infotrieve]
40. Beau SL, Hand DE, Schuessler RB, Bromberg BI, Kwon B, Boineau JP, Saffitz JE. Relative densities of muscarinic cholinergic and ß-adrenergic receptors in the canine sinoatrial node and their relation to sites of pacemaker activity. Circ Res. 1995; 77: 957–963.[Abstract/Free Full Text]
41. DiFrancesco D. Characterization of single pacemaker channels in cardiac sino-atrial node cells. Nature. 1986; 324: 470–473.[CrossRef][Medline] [Order article via Infotrieve]

This article has been cited by other articles:

				M. E. Mangoni and J. Nargeot Genesis and Regulation of the Heart Automaticity Physiol Rev, July 1, 2008; 88(3): 919 - 982. [Abstract] [Full Text] [PDF]

				A. Younes, A. E. Lyashkov, D. Graham, A. Sheydina, M. V. Volkova, M. Mitsak, T. M. Vinogradova, Y. O. Lukyanenko, Y. Li, A. M. Ruknudin, et al. Ca2+-stimulated Basal Adenylyl Cyclase Activity Localization in Membrane Lipid Microdomains of Cardiac Sinoatrial Nodal Pacemaker Cells J. Biol. Chem., May 23, 2008; 283(21): 14461 - 14468. [Abstract] [Full Text] [PDF]

				D. P. McEwen, Q. Li, S. Jackson, P. M. Jenkins, and J. R. Martens Caveolin Regulates Kv1.5 Trafficking to Cholesterol-Rich Membrane Microdomains Mol. Pharmacol., March 1, 2008; 73(3): 678 - 685. [Abstract] [Full Text] [PDF]

				S. Langlois, K. N. Cowan, Q. Shao, B. J. Cowan, and D. W. Laird Caveolin-1 and -2 Interact with Connexin43 and Regulate Gap Junctional Intercellular Communication in Keratinocytes Mol. Biol. Cell, March 1, 2008; 19(3): 912 - 928. [Abstract] [Full Text] [PDF]

				C.-H. Li, Q. Zhang, B. Teng, S. J. Mustafa, J.-Y. Huang, and H.-G. Yu Src tyrosine kinase alters gating of hyperpolarization-activated HCN4 pacemaker channel through Tyr531 Am J Physiol Cell Physiol, January 1, 2008; 294(1): C355 - C362. [Abstract] [Full Text] [PDF]

				H. Tsujikawa, Y. Song, M. Watanabe, H. Masumiya, S. A. Gupte, R. Ochi, and T. Okada Cholesterol depletion modulates basal L-type Ca2+ current and abolishes its -adrenergic enhancement in ventricular myocytes Am J Physiol Heart Circ Physiol, January 1, 2008; 294(1): H285 - H292. [Abstract] [Full Text] [PDF]

				P. J. Mohler and X. H. T. Wehrens Mechanisms of Human Arrhythmia Syndromes: Abnormal Cardiac Macromolecular Interactions Physiology, October 1, 2007; 22(5): 342 - 350. [Abstract] [Full Text] [PDF]

				J. Abi-Char, A. Maguy, A. Coulombe, E. Balse, P. Ratajczak, J.-L. Samuel, S. Nattel, and S. N. Hatem Membrane cholesterol modulates Kv1.5 potassium channel distribution and function in rat cardiomyocytes J. Physiol., August 1, 2007; 582(3): 1205 - 1217. [Abstract] [Full Text] [PDF]

				W. Lin, U. Laitko, P. F. Juranka, and C. E. Morris Dual Stretch Responses of mHCN2 Pacemaker Channels: Accelerated Activation, Accelerated Deactivation Biophys. J., March 1, 2007; 92(5): 1559 - 1572. [Abstract] [Full Text] [PDF]

				A. Knaus, X. Zong, N. Beetz, R. Jahns, M. J. Lohse, M. Biel, and L. Hein Direct Inhibition of Cardiac Hyperpolarization-Activated Cyclic Nucleotide-Gated Pacemaker Channels by Clonidine Circulation, February 20, 2007; 115(7): 872 - 880. [Abstract] [Full Text] [PDF]

				A. Hinzpeter, J. Fritsch, F. Borot, S. Trudel, D.-L. Vieu, F. Brouillard, M. Baudouin-Legros, J. Clain, A. Edelman, and M. Ollero Membrane Cholesterol Content Modulates ClC-2 Gating and Sensitivity to Oxidative Stress J. Biol. Chem., January 26, 2007; 282(4): 2423 - 2432. [Abstract] [Full Text] [PDF]

				M. Vatta, M. J. Ackerman, B. Ye, J. C. Makielski, E. E. Ughanze, E. W. Taylor, D. J. Tester, R. C. Balijepalli, J. D. Foell, Z. Li, et al. Mutant Caveolin-3 Induces Persistent Late Sodium Current and Is Associated With Long-QT Syndrome Circulation, November 14, 2006; 114(20): 2104 - 2112. [Abstract] [Full Text] [PDF]

				J. P. Dekker and G. Yellen Cooperative Gating between Single HCN Pacemaker Channels J. Gen. Physiol., November 1, 2006; 128(5): 561 - 567. [Abstract] [Full Text] [PDF]

R. C. Balijepalli, J. D. Foell, D. D. Hall