Skip to main content
NCBI home page
The Journal of Physiology logoLink to The Journal of Physiology
. 2003 Jun 18;550(Pt 3):707–717. doi: 10.1113/jphysiol.2003.041962

G Protein-Independent Inhibition of GIRK Current by Adenosine in Rat Atrial Myocytes Overexpressing A1 Receptors after Adenovirus-Mediated Gene Transfer

Leif I Bösche 1, Marie-Cécile Wellner-Kienitz 1, Kirsten Bender 1, Lutz Pott 1
PMCID: PMC2343091  PMID: 12815176

Abstract

G protein-activated inwardly rectifying K+ (GIRK) channels, important regulators of membrane excitability in the heart and central nervous system, are activated by interaction with βγ subunits from heterotrimeric G proteins upon receptor stimulation. In atrial myocytes various endogenous receptors couple to GIRK channels, including the canonical muscarinic M2 receptor (M2AChR) and the A1 adenosine receptor (A1AdoR). Saturating stimulation of A1AdoR in atrial myocytes activates only a fraction of the GIRK current that is activated via M2AChR, which reflects a lower density of A1AdoR. In the present study A1AdoR were overexpressed by means of adenovirus-mediated gene transfer using green fluorescent protein (GFP) as the reporter. Confirmatory to a previous study, this resulted in an increased sensitivity of macroscopic GIRK current (ACh-activated K+ current (IK(ACh))) to stimulation by Ado. However, in the majority of GFP-positive myocytes, exposure to Ado at concentrations ≥ 1 μm resulted in activation of IK(ACh) followed by a rapid inhibition. In those cells a rebound activation of current was recorded upon washout of Ado. The inhibitory component could be recorded in isolation when IK(ACh) was activated by M2AChR-stimulation and brief pulses of Ado were superimposed. In myocytes loaded with GTP-γ-S, IK(ACh), irreversibly activated by brief exposure to agonist, was still reversibly inhibited by Ado, suggesting that inhibition is independent of G protein cycling. In myocytes co-transfected with adenoviral vectors encoding A1AdoR and GIRK4 subunit, no inhibition of GIRK current by Ado was observed. As acute desensitization of atrial GIRK current, which is reminiscent of the inhibition described here, has been shown to be absent in myocytes overexpressing GIRK4, this suggests that acute desensitization and the novel inhibition might share a common pathway whose target is the GIRK channel complex or its GIRK1 subunit.

G protein-activated inwardly rectifying K+ (GIRK1) channels are expressed in the heart, preferentially in the pacemaking and atrial myocytes and in various types of neurons and endocrine cells (Lesage et al. 1995; Yamada et al. 1998; Wickman et al. 2000; Milligan & White, 2001; Dascal, 2001). They represent important mediators of vagally induced bradycardia and synaptic inhibition in the central nervous system (Nicoll, 1988; Sodickson & Bean, 1996; Takigawa & Alzheimer, 1999; Yamada, 2002). These channels are assumed to be tetrameric complexes formed by assembly of GIRK subunits of which four mammalian types (GIRK1-4) have been identified. They are activated by direct interaction of their subunits with βγ subunits released from heterotrimeric G proteins upon agonist stimulation of appropriate 7-helix receptors. GIRK channels were the first targets of Gβγ to be identified. It is generally believed that these channels are preferentially activated by βγ released from pertussis toxin-sensitive Gi/o (Leaney et al. 2000).

In atrial myocytes whole-cell GIRK channel current can be activated, apart from the paradigmatic muscarinic M2 acetylcholine receptor (M2AChR), by purinergic A1 adenosine receptors (A1AdoR) and a sphingolipid receptor with high affinity for sphingosine-1-phosphate and sphingosyl-phosphorylcholine, presumably EDG3 (Bünemann et al. 1996b; Luscher et al. 1997; Himmel et al. 2000).

Stimulation of A1AdoR by saturating concentrations of Ado in atrial myocytes results in activation of only a fraction of the maximum current that is recorded upon exposure to ACh, which in atrial myocytes from rat amounts to ≈30 % (Kurachi et al. 1986; Bünemann & Pott, 1995; Takano & Noma, 1997; Wellner-Kienitz et al. 2000). It has been demonstrated that the expression level of A1AdoR limits responsiveness to Ado (Wellner-Kienitz et al. 2000); in this study the expression level of the receptor had been manipulated by transient transfection of atrial myocytes using Lipofectamine transfection of a conventional pcDNA3 construct encoding the receptor protein. In cultured atrial myocytes this yields transfection rates in terms of reporter (green fluorescent protein, GFP)-positive cells of ≤ 5 %.

In the present study the expression level of A1AdoR in rat atrial myocytes was increased using a simplified system for generating recombinant adenovirus vectors (pAdeasy, He et al. 1998). Following incubation of myocyte cultures with approximately 104 infectious particles ml−1, GFP-positive cells in the order of 80 % were identified 3 days later. Though not measured in quantitative terms, GFP-fluorescence of individual cells was markedly higher in the pAdeasy-transfected cells as compared with those transfected using conventional vectors.

Apart from an increase in Ado-induced current and a leftward shift of the concentration-response curve, in the majority of A1AdoR-overexpressing cells adenosine had dual actions, namely an activation of GIRK current, superimposed by a rapid reversible inhibition, resulting in a marked rebound activation upon washout of Ado. The inhibitory component could be recorded in isolation when the current was activated via M2AChR and Ado was applied in the presence of ACh. In myocytes loaded via the patch pipette with GTP-γ-S to irreversibly activate heterotrimeric G proteins, the inhibitory action of A1AdoR-stimulation persisted and remained fully reversible. This suggests that GTP/GDP-cycling is not required for GIRK current inhibition by overexpressed A1AdoR.

It is concluded that A1AdoR, and possibly other Gi/o-coupled receptors in cardiac cells, apart from initiating signals via heterotrimeric G protein α and βγ subunits confer an inhibitory signal to GIRK channels that is G protein-independent.

METHODS

Isolation and culture of atrial myocytes

Rats were killed following protocols approved by the Ruhr-University Bochum and the local authorities for the regulation of animal welfare (Regierungspärsident). Wistar Kyoto rats of either sex (around 200 g) were anaesthetized by I.V. injection of urethane (1 g kg−1). The chest was opened and the heart was removed and mounted on the cannula of a sterile Langendorff apparatus for coronary perfusion at constant flow. The method of enzymatic isolation of atrial myocytes and culture conditions have been described in detail elsewhere (e.g. Bünemann et al. 1996a). Myocytes were used experimentally from day 0 until day 5 after isolation.

Solutions and chemicals

For the patch-clamp (whole-cell) measurements of atrial GIRK current an extracellular solution of the following composition was used (mm): NaCl 120; KCl 20; CaCl2 0.5; MgCl2 1.0; Hepes 10.0, adjusted to pH 7.4 with NaOH. The solution for filling the patch-clamp pipettes for whole-cell measurements of G protein-activated K+ currents contained (mm): potassium aspartate 110; KCl 20; NaCl 5.0, MgCl2 1.0; Na2ATP 2.0; EGTA 2.0; GTP 0.01; Hepes 10.0, adjusted to pH 7.4 with KOH. Standard chemicals were from Merck (Darmstadt, Germany). EGTA, Hepes, MgATP, ACh, adenosine, GTP, and GTP-γ-S were from Sigma (Deisenhofen, Germany).

Current measurement

Membrane currents were measured using whole-cell patch-clamp. Pipettes were fabricated from borosilicate glass and were filled with the solution listed above (DC resistance 4–6 MΩ). Currents were measured by means of a patch-clamp amplifier (WPC-100, ESF, Göttingen, Germany). Signals were analog filtered (corner frequency of 1–3 KHz), digitally sampled at 5 KHz and stored on a computer, equipped with a hardware/software package (ISO2 by MFK, Frankfurt/Main, Germany) for voltage control and data acquisition. Experiments were performed at ambient temperature (22–24 °C). If not otherwise stated, cells were voltage-clamped at −90 mV, i.e. negative to the equilibrium potential for potassium (EK), resulting in inward K+ currents. Current-voltage relations were determined by means of voltage ramps between −120 and +60 mV. Rapid superfusion of the cells for application and withdrawal of agonist-containing solutions was performed by means of a solenoid-operated flow system that permitted switching between up to six different solutions with a half-time of exchange ≤ 100 ms.

Adenovirus constructs

The pAd-Easy1 virus and shuttle plasmid, pAd-TrackCMV, were obtained from Dr B. Vogelstein (Johns Hopkins University, Baltimore, MD, USA). Production and purification of the recombinant virus were performed as described in detail in by He et al. (1998). Briefly, the cDNA of a rat A1 receptor was subcloned into pAd-Track-CMV using Kpn I and Hind III to yield pAdTrack-A1AdoR. Recombinant adenovirus (pAd-A1AdoR) was generated by homologous recombination between pAdTrack-A1AdoR and pAd-Easy-1 in Escherichia coli. Constructs encoding for a muscarinic M2 receptor, GIRK1 and GIRK4 were similarly created (pAd-M2AChR; pAd-GIRK1; pAd-GIRK4).

For infection, cells were incubated for 3 h, starting about 24 h after plating, with 1 ml culture medium containing 104 infectious particles. Electrophysiological recordings were made on days 3 and 4 after transfection. Transfected cells (≥ 80 %) were identified by epifluorescence of GFP (excitation wavelength 470 nm). In each set of experiments GFP-positive myocytes in time-matched cultures infected with the empty adenoviral vector served as controls. No significant differences in the properties of GIRK current (density and activation rates at various concentrations of ACh and Ado respectively) in these control cells and non-infected (native) cells were found in a separate series of experiments.

Real-time PCR

RNA was isolated from about 2 × 105 rat atrial myocytes on the day after plating. Integrity of isolated total RNA was verified by optical density (OD260 nm/280 nm absorption ratio). The target (A1AdoR) and reference gene (18S ribosomal subunit) were amplified using the isolated RNA as template. The reverse transcription and real-time PCR were performed by means of a one-step RT-PCR Kit according to the manufacturer's instructions (Qiagen GmbH, Hilden, Germany).

Target gene A1AdoR (sense primer: CTACTTCCACACACCTGCCTCA; antisense primer: GAGAATCCAGCAGCCAGCTA) and reference gene 18 s ribosomal subunit were amplified in a PCR cycler (GenAmp 5700) using the following conditions: one initial cycle for reverse transcription at 50 °C for 30 min, followed by an inactivation step (95 °C, 15 min), and a final three-step cycling, 94 °C for 15 s, 55 °C for 30 s, 72 °C for 30 s repeated 40 times. Amounts of PCR products were detected using the fluorescent reporter dye SYBR Green (Molecular Devices). RT-PCR amplification products were verified on a 2 % high ethidium bromide-stained agarose gel. Relative transcription levels were quantified using a formalism introduced by Pfaffl et al. (2002).

Statistical analysis

Wherever possible data are presented as means ± s.e.m. and were analysed using Student's t test for unpaired samples. A value of P < 0.05 was considered to be significant.

RESULTS

In line with previous studies, the current that can be activated by Ado in native rat atrial myocytes is significantly smaller than the total current activated upon stimulation of M2AChR if saturating concentrations of ACh (≥ 10 μm) and Ado (≥ 10 μm) are compared. This is illustrated by the representative current recording in Fig. 1A. Whereas ACh-evoked current is characterized by rapid activation (half-life 370 ms), a peak current density in the order of magnitude of 70 pA pF−1 and a rapid component of desensitization in the presence of agonist, activation of Ado-induced current is slow, reaching a plateau with a half-life of 1800 ms at the highest concentration tested (100 μm). The current amplitude corresponded to 42 % of peak Ach-activated K+ current (IK(ACh)) in the experiment shown. On average the amplitude of adenosine-activated K+ current (IK(Ado)) was 31.7 ± 6.8 % of IK(ACh) (n = 60). There was no difference in current, either regarding the amplitude or the activation kinetics between 10 and 100 μm Ado, demonstrating that 10 μm Ado does represent a saturating concentration of this agonist for activation of GIRK current at a physiological level of A1AdoR expression. Desensitization of the current, which is present in the transients evoked by ACh as a distinct kinetic component, appears to be absent when Ado is the activating agonist. As shown previously, however, desensitization does take place under this condition during slow activation (Wellner-Kienitz et al. 2000, see also Fig. 4).

Figure 1. Differential activation of atrial G protein-activated inwardly rectifying K+ (GIRK) current by acetylcholine (ACh) and adenosine (Ado).

Figure 1

Open in a new tab

A, inward current activated by rapid exposure to ACh (10−5 M) and two concentrations of Ado (10−5 M and 10−4 M). The rapid deflections in this and subsequent figures represent changes in membrane current due to ramp-shaped changes in membrane potential from −120 to + 60 mV within 500 ms, which were routinely applied once every 10 s to control stability of the recording conditions. B, response to 10−5 M ACh and onset of response to 10−5 M Ado on an expanded time scale. The response to 10−5 M Ado was scaled up to match the amplitude of the response to 10−5 M ACh. The arrow indicates the beginning of agonist exposure.

Figure 4. Differential effect of Ado on ACh-induced current in myocytes with and without rebound activation of Ado-induced current.

Figure 4

Open in a new tab

A, representative recording of current from a pAd-A1AdoR-infected cell showing low intensity of GFP fluorescence. B, representative recording from a myocyte with high GFP fluorescence. ACh (10 μm) and Ado (10 μm) were applied as indicated by the horizontal bars.

Overexpression of A1AdoR using conventional transfection vector and liposome methodology has been shown to cause changes in all these parameters, resulting in whole-cell GIRK currents activated by Ado, which are indistinguishable from ACh-induced currents in native cells (Wellner-Kienitz et al. 2000). Half-times of activation of currents induced by ACh (20 μm) were increased from 244 ± 48 ms (mock-infected cells, n = 22) to 741 ± 95 ms (pAd-A1AdoR-infected cells, n = 15, P < 0.001), whereas this parameter was reduced from 2156 ± 244 ms (n = 19) to 229 ± 26 ms by overexpression of A1AdoR on Ado-induced GIRK current. This demonstrates, in line with previous studies (Bünemann & Pott, 1995; Bünemann et al. 1996a; Wellner-Kienitz et al. 2000) that the differences in Ado- vs. ACh-induced currents do not represent a genuine difference in signalling between these two receptors, but that they can be completely accounted for by a lower density of A1AdoR as compared with M2AChR.

In Fig. 2 activation of IK(Ado) by two different concentrations of Ado (0.1 and 10 μm) was compared in a control myocyte (A, infected with the empty viral construct) and two different representative myocytes infected with pAd-A1AdoR (B and C). Using real-time PCR the number of transcripts was found to be higher by a factor of approximately 900 compared with time-matched control cultures. The concentrations of Ado were chosen since they represent threshold and saturation, respectively, in native or mock-transfected cells. In the control cell the change in current upon exposure to 0.1 μm Ado was virtually zero (on average it amounted to 6.8 ± 1.8 %, n = 22), whereas the current elicited by 10 μm showed the properties as already introduced in Fig. 1. In pAd-A1AdoR-infected cells, exposure to 0.1 μm Ado always resulted in sizeable currents with half-lives of activation < 1 s. At the higher concentration two types of responses were recorded in different cells. The sample trace in Fig. 2B shows a response that was similar to the current usually elicited by a saturating concentration of ACh, namely its half-life of activation was < 500 ms, and peak current was followed by acute desensitization to a plateau. Upon washout of agonist, the current showed a monotone deactivation. This behaviour is not different from Ado-induced current in myocytes transfected with a conventional A1AdoR-encoding vector (Wellner-Kienitz et al. 2000). In the majority of positive cells, the current elicited by the high concentration of Ado was characterized by a rapid decay to a steady-state level, reminiscent of fast desensitization. However, upon washout of Ado, the current did not deactivate but increased, resulting in a second peak several seconds after switching to Ado-free solution (Fig. 2C). The degree of this rebound appeared to be qualitatively correlated with the intensity of GFP fluorescence, i.e. the transfection efficiency of the cells. By selecting cells with high GFP expression, this phenomenon was found throughout. Moreover, in cells that showed this phenomenon, it was concentration dependent. This is demonstrated in Fig. 3, which shows currents induced by 0.1, 1.0 and 10 μm Ado. Inward currents at 0.1 and 1.0 μm had identical amplitudes but differed in deactivation properties, which were monophasic at 0.1 μm but showed a pronounced rebound at 1.0 μm. The amplitude of current evoked by 10 μm was smaller by 25 %. This reduction in peak current is likely to result from inhibition or desensitization in parallel with activation, resulting in a blunted peak. Correspondingly, rebound activation, which - as will be shown later in this paragraph - represents recovery from inhibition, became more pronounced. We next studied if GIRK current activated by stimulation of endogenous M2AChR is sensitive to stimulation of overexpressed A1AdoR. In native (non-transfected) cells, an occlusive, non-additive behaviour is observed when both receptor agonists are applied simultaneously (Wellner-Kienitz et al. 2000). Qualitatively, this was confirmed in those pAd-A1AdoR-infected myocytes which did not show a rebound of GIRK current upon washout of Ado. A representative example is illustrated in Fig. 4A. Following a test pulse of Ado, the cell was exposed to ACh, which caused an inward current of less than half the amplitude of IK(Ado), which is in line with previously published data demonstrating a negative interference of overexpressed A1AdoR on functional expression of M2AChR (Wellner-Kienitz et al. 2000). In the continuous presence of ACh, exposure to Ado caused additional activation of inwardly rectifying current. Confirming published data, the total current (IK(ACh) + IK(Ado)) was substantially smaller than the Ado-induced current in the absence of ACh, which is supposed to be due to acute desensitization being superimposed on slow activation of IK(ACh). Using this type of experimental protocol we found an activation of current by Ado in the presence of ACh in pAd-A1AdoR-infected cells that did not show rebound activation upon washout of Ado. In pAd-A1AdoR-infected myocytes selected for strong GFP-fluorescence, exposure to Ado in the presence of ACh caused a reversible inhibition of the current, as shown by a representative sample recording in Fig. 4B. This clearly supports the notion that the rebound activation of current represents recovery from an A1AdoR-induced inhibition. In experiments like those illustrated in Fig. 2C and Fig. 3, recovery from inhibition apparently is faster than deactivation on washout of agonist. Analogous to acute desensitization this inhibition is independent of the receptor by which the system was activated. This heterologous nature suggests that inhibition occurs downstream of the receptor.

Figure 2. Effect of A1AdoR overexpression on Ado-induced GIRK current.

Figure 2

Open in a new tab

Comparison of current changes induced by 0.1 and 10 μm Ado in a representative control myocyte infected with the empty vector (A) and two myocytes infected with pAd-A1AdoR (B and C). The arrows in C indicate the start of washout of Ado.

Figure 3. Current changes induced by three concentrations of Ado in a representative highly GFP-positive pAd-A1AdoR-infected myocyte.

Figure 3

Open in a new tab

The arrows mark the start of washout of Ado. The dotted line was drawn to demonstrate identity of peak currents at 0.1 and 1.0 μm Ado.

We next studied whether inhibition is mediated by a G protein-dependent pathway. In this case irreversible activation of heterotrimeric G proteins by GTP-γ-S, loaded into the cell via the patch-clamp pipette, should not only result in irreversible activation of the current, but should also render the inhibition irreversible, i.e. once fully activated by brief exposure(s) to agonist (Ado or ACh), subsequent challenges with Ado should not exert an inhibitory effect. This was tested in a series of experiments using a patch-clamp filling solution supplemented with GTP-γ-S. In native or mock-infected cells this protocol resulted in stable activation of GIRK current after a few brief exposures to Ado or ACh. As shown in Fig. 5A and B, once a stable level of activation was reached, the current became insensitive to subsequent exposures to agonist, reflecting irreversible activation of the entire population of relevant Gα subunits (e.g. Meyer et al. 2001). A typical result, representative of 26/26 pAd-A1AdoR-infected cells selected for strong GFP fluorescence, is illustrated in Fig. 5C. As in Fig. 5A and B, exposure to Ado resulted in a stepwise activation of GIRK current. The first exposure to Ado resulted in a response composed of activation, inhibition and rebound activation. Upon the next three pulses of Ado there was less activation and subsequent deactivation upon washout, unmasking the inhibitory action of Ado. After the fourth exposure to Ado the current remained stably activated. A subsequent exposure induced the inhibitory component only, which was still rapidly reversible. This inhibitory signal was not altered during long-term recordings of up to 22 min from GTP-γ-S-loaded cells (not shown), supporting the notion that it is G protein-independent.

Figure 5. Reversible inhibition of current in GTP-γ-S-loaded myocytes.

Figure 5

Open in a new tab

A, recording of membrane current from a myocyte infected with the empty (GFP-encoding) vector. The pipette filling solution was supplemented with the non-hydrolysable GTP analogue GTP-γ-S (500 μm). Ado was applied as indicated by the horizontal bars. After stable and irreversible activation of the current, further application of Ado had no effect. B, an analogous control experiment using ACh (20 μm) to activate the current. C, representative recording of membrane current from a pAd-A1AdoR-infected myocyte. Ado was applied as indicated by horizontal bars. D, background-subtracted current–voltage relation of Ado-activated current and fraction of current that was inhibited by Ado from trace C. Em, membrane potential. E, representative recording demonstrating concentration-dependent inhibition by Ado (10 μm) of current in a pAd-A1AdoR-infected myocyte loaded with GTP-γ-S.

The reduction in net inward current represents inhibition of GIRK current but not activation of a putative current that is outward at holding potential (−90 mV) as verified by comparing the different current-voltage relations of the total activated current and the inhibited fraction, which are perfectly superimposable (Fig. 5D). Inhibition of GIRK current, also under the condition of GTP-γ-S loading, is dependent on the concentration of Ado as shown by the representative experiment depicted in Fig. 5E. Note that at 0.1 μm there was only very little inhibition, whereas in pAd-A1AdoR-infected myocytes in the absence of GTP-γ-S this concentration resulted in activation of a current with saturating amplitude (see Fig. 3), underscoring that inhibition requires a higher concentration of Ado than activation. The major conclusion to be drawn from this series of experiments is that inhibition of GIRK current by overexpressed A1AdoR is independent of the GTP/GDP cycle of a G protein.

The similarity of GIRK current inhibition to acute desensitization has already been pointed out above. In a recent study we have shown that this type of desensitization is abolished in myocytes transfected with constructs encoding GIRK4 subunits or dimeric or tetrameric tandem constructs of this subunit, supporting the notion that acute desensitization represents a property of the GIRK channel associated with GIRK1 rather than that of an upstream signalling element (Bender et al. 2001). To address the question of whether the novel inhibition of GIRK current bears a relation to acute desensitization we double-infected myocytes with pAdA1AdoR and pAd-GIRK4. Recordings of currents evoked by Ado (10 μm) from two representative cells (out of a total of 18) are illustrated in Fig. 6A and B. The two traces were selected as they represent the range of responses that were obtained in this series of experiments. In the trace depicted in Fig. 6A, despite rapid activation (half-time 290 ms) there was no current decay in terms of acute desensitization in the pAd-GIRK4-infected cell; correspondingly, no rebound activation occurred. This was found in a total of seven cells. The trace shown in Fig. 6B represents the myocyte that showed the most pronounced inhibition/rebound activation. For comparison a current trace obtained from a cell infected with pAd-A1AdoR, but in a time-matched sister culture, is shown in Fig. 6C. In line with a previous study, the I-V relation of whole-cell GIRK current in myocytes overexpressing the GIRK4 subunit was characterized by less inward rectification compared with that of cells whose level of expression of GIRK subunits had not been manipulated (Bender et al. 2001). This is demonstrated in Fig. 6D, which shows superimposed normalized I-V curves. The summarized data in Fig. 6E clearly demonstrate that inhibition or acute desensitization was almost completely abolished in myocytes overexpressing the A1AdoR and the GIRK4 subunit (see figure legend for details). Double-infection with pAd-A1AdoR and pAd-GIRK1 resulted in currents that were not different from those in cells transfected with pAd-A1AdoR only (not shown). Thus, homomeric GIRK4 channels are less susceptible to both acute desensitization and the novel inhibition, supporting the notions that both phenomena: (i) might represent a common mechanism and (ii) are dependent on the subunit composition of the channel complex. If acute desensitization in terms of a partial rapid decay in current and inhibition/rebound activation share a common mechanism related to one of the channel subunits, one would expect that both phenomena could also be evoked by activation of the M2AChR that is assumed to be endogenously expressed at a much higher level than the A1AdoR. Indeed, in native cells current activated by high (saturating) concentrations of ACh is characterized by a distinct desensitizing component. Since the data obtained with overexpressed A1AdoR suggest that inhibition/rebound activation require (i) a high density of the activating agonist and (ii) rapid exposure to a strongly saturating concentration of the agonist, we tested whether a rebound activation can be observed upon stimulation of the M2AChR by a concentration of ACh (1 mm) in the extremely saturating range. In 10 cells tested, upon washout of this concentration of ACh a rebound activation of IK(ACh) was detected, though this was less pronounced than following washout of Ado in A1AdoR-overexpressing myocytes. Current recordings from a representative cell comparing the effects of ACh at 20 μm and 1 mm are shown in Fig. 7. Both responses were nearly identical regarding their amplitude and kinetics of activation, demonstrating that 20 μm is already saturating for these parameters. Desensitization appeared to be accelerated at the higher concentration. Upon washout of 20 μm ACh, IK(ACh) showed a monophasic de-activation, whereas rebound activation occurred following exposure to the high concentration (1 mm). This demonstrates that, in principle, rebound activation of IK(ACh) is not limited to A1AdoR overexpression and, furthermore, that it appears to be related to acute desensitization.

Figure 6. Inhibition and rebound activation of GIRK current are absent or reduced in myocytes co-infected with pAd-A1AdoR and pAd-GIRK4.

Figure 6

Open in a new tab

A and B, representative recordings of Ado-induced current from two double-infected cells showing high intensity of GFP fluorescence. The two traces were selected since they are representative of the spectrum of responses that were obtained in this series of experiments. C, typical recording from a pAd-A1AdoR-infected myocyte. D, comparison of background-subtracted current–voltage relations of Ado-activated currents from A and C. Currents were normalized to current level at −100 mV a and b in the key refer to the labelled traces in A and C, respectively. E, summarized data; maximum current was divided by the current at 4 s after switching to Ado-containing solution (P < 0.001; n = 18 for pAd-A1AdoR infected cells; n = 11 for pAd-A1AdoR/pAd-GIRK4 double-infected cells).

Figure 7. Example of rebound activation upon washout of ACh.

Figure 7

Open in a new tab

Superimposed responses to 20 μm and 1 mm ACh recorded from a representative native myocyte. The arrow indicates the point when washout of 1 mm ACh started.

DISCUSSION

In this study we have presented evidence for a receptor-mediated G protein-independent inhibition of GIRK current in atrial myocytes. G protein-gated inwardly rectifying K+ channels are important regulators of neuronal and cardiac excitability. In different systems they respond to various transmitters and hormones that act on G protein-coupled receptors. Apart from the canonical activation by direct interaction of GIRK subunits with βγ dimers, which are as a rule, but not exclusively, released from pertussis toxin-sensitive G proteins (Sorota et al. 1999; Wellner-Kienitz et al. 1999), other modes of regulation have been described more recently. Activation of receptors coupling via Gq/11 to phospholipase C (PLC) has been shown to cause inhibition of GIRK current in atrial myocytes that results from depletion of phosphatidylinositol 2,4-bisphosphate (PIP2) in the inner leaflet of the membrane (Cho et al. 2001; Meyer et al. 2001; Bender et al. 2002). This phospholipid has been described as an essential co-factor for activity of various transport proteins and ion channels, including members of the Kir family (Levade et al. 2001; Nasuhoglu et al. 2002). The slow kinetics of inhibition of GIRK current in atrial myocytes via PIP2 depletion, with a half-life in the order of magnitude of some tens of seconds (Meyer et al. 2001), clearly argues against a contribution to the rapid inhibition and rebound activation described here. Moreover, receptor-mediated PLC activation is a G protein-dependent process and should become irreversibly stimulated in GTP-γ-S-loaded cells. In the Xenopus oocyte expression system an inhibition of GIRK channels by endothelin (ETA) receptors has been demonstrated, which appears to be mediated by arachidonic acid (AA) resulting from activation of phospholipase A2 (Rogalski & Chavkin, 2001). These authors suggested a binding site for AA related to the domain in GIRK subunits that confers sensitivity to PIP2 and Na+ ions. On the other hand, also in the oocyte expression system, an activation of GIRK channels by AA has been described that seemed to be G protein-independent (Lohberger et al. 2000). In a series of preliminary experiments using extracellular application as well as pipette loading of AA, neither desensitization in native cells nor inhibition/rebound activation were affected by this mediator (data not shown), suggesting that under the present experimental conditions it does not seem to play a regulatory role in GIRK channels in their native environment.

A novel signalling pathway that results in inhibition of voltage-dependent relaxation in atrial myocytes and reduction in channel activity in excised patches has been demonstrated more recently (Ishii et al. 2001, 2002). These authors suggested a regulator of G protein signalling (RGS)-mediated inhibition of GIRK channels that is blocked by PIP3. This block can be relieved via Ca2+/calmodulin, resulting in a voltage control of the G protein cycle. The persistence of the inhibition in GTP-γ-S-loaded cells described here, i.e. in a condition in which the G protein cycle is locked in the GTP-bound state, argues against the contribution of an RGS protein which acts on the rate of G protein cycling. Thus, all inhibitory modes of GIRK current described so far require intact G protein cycling, and therefore are most unlikely to contribute to the GIRK current inhibition/rebound activation described in the present investigation.

The obligatory paradigm that 7-helix receptors link extracellular stimuli to intracellular targets via heterotrimeric G proteins has been challenged recently in a number of systems (Hall et al. 1999; Heuss & Gerber, 2000; Brzostowski & Kimmel, 2001). The longest known examples of proteins, apart from Gα subunits, that interact with a G protein-coupled receptor (GPCR) are arrestins (e.g. Krupnick & Benovic, 1998; Bünemann et al. 1999 for reviews). Other examples are G protein-coupled receptor kinases (GRKs), small GTP-binding proteins and PDZ domain-containing proteins (e.g. Fanning & Anderson, 1999; Sheng & Sala, 2001). In cardiac muscle these include, for example, Na+-H+ exchange factor, homer proteins, calmodulin, eNOS or iNOS (see Heuss & Gerber, 2000 for review). From this incomplete list it is evident that a large array of signals can be potentially transmitted via GPCRs without contribution of heterotrimeric G proteins. The rapid onset and reversibility of GIRK current inhibition suggests a short direct signalling pathway. Inhibition and rebound activation were observed at a physiological level of expression of M2AChR upon exposure to a concentration of ACh that was highly saturating for Gβγ-mediated activation of GIRK channels and at an increased expression level of A1AdoR with saturating Ado concentrations. This clearly suggests that the unknown pathway leading to inhibition/rebound activation requires a higher total number or membrane density of agonist-activated receptors than Gβγ-mediated activation, which is in line with the finding that adenoviral transfer of the A1AdoR gene results in an increase in the number of transcripts by almost three orders of magnitude. Delineating the mechanism requires further work using pharmacological tools and manipulation of the expression levels of putative candidate proteins that transmit GPCR activation to a signal affecting GIRK current.

Desensitization, i.e. a decrease in the response of a cell response during prolonged exposure of a signalling system to a transmitter or hormone appears to be common to virtually all receptor-linked signalling pathways. The term is not restricted to a distinct mechanism and comprises conformational changes in ionotropic channels on a sub-millisecond time scale to changes in sensitivity on the time scale of hours or days in pathways linked to heptahelical receptors. Generally the term is used in the context of changes on the receptor side, although other targets of desensitization have been described. There is general agreement that acute desensitization of GIRK channel current does not reflect receptor desensitization. The major argument in support of this comes from the observation that the fast (‘acute’) component of desensitization of GIRK current is observed in two different forms, namely: (i) as a decay in current upon rapid activation of highly expressed endogenous receptor species such as the M2AChR by an agonist at a saturating concentration and (ii) during slow activation by either a low concentration of agonist in case of the M2AChR, or a high concentration of agonist in case of the A1AdoR, which is functionally expressed at a lower level. In these cases the heterologous nature of this type of desensitization becomes evident. Although we cannot exclude that the similarity between inhibition/rebound activation in myocytes overexpressing A1AdoR and acute desensitization of M2AChR-activated currents in native myocytes is coincidental, it appears conceivable that these are related if not identical phenomena. This is supported by the following observations. (i) In pAd-A1AdoR-infected myocytes, transition from acute desensitization without subsequent rebound activation to responses with a massive rebound current to Ado showed a graded dependence upon concentration. (ii) Both acute desensitization and inhibition/rebound activation were almost completely abolished in myocytes overexpressing the GIRK4 subunit. Moreover (iii) neither phenomenon was related to a single type of receptor. Whether or not the rapid inhibition associated with rebound activation plays a role in regulating GIRK channel activity in the heart cannot be answered at present. In case of the A1AdoR this seems unlikely, since its low level of functional expression prevents saturating responses to Ado. Moreover, rises in interstitial concentration of this compound as a consequence of increasing metabolic activity are supposedly slow, blunting desensitization in terms of a distinct kinetic component of current or hyperpolarization. This might be different for the M2AChR, whose activation does in principle result in current responses with a desensitising component or, as shown in the present study, in some rebound activation, though less pronounced than that via overexpressed A1AdoR. This observation is in line with a recent publication investigating kinetic aspects of GIRK current activation by different receptors in a reconstituted system (HEK 293 cell line, Benian et al. 2003). When activated via A1 receptors, the current showed pronounced rebound activation upon washout of the agonist. This was less pronounced for the α2A adrenergic receptor and was absent when D2B dopamine receptors or M4 muscarinic receptors were used.

It is generally assumed that in isolated atrial myocytes the distributions of receptors, channels and other membrane proteins are homogeneous, reflecting the situation in the intact tissue. Although in this and previous studies we did not detect any significant changes in kinetic properties of ACh-activated currents in freshly isolated (i.e. < 4 h in vitro) and cultured myocytes, rapid changes in the distribution of receptors and/or channels from a patchy to a homogeneous pattern cannot be excluded.

In neuronal systems, desensitization of responses to neurotransmitters and modulators represents an important factor contributing to shaping the time course of synaptic potentials, which in turn should have a profound impact on integrative functions of a neuron. Synaptic receptors are concentrated or clustered in subsynaptic membranes. This would favour the kinetic behaviour of transmitter-activated current found in A1AdoR-overexpressing myocytes, provided the rise in transmitter concentration approaches a saturating range. For certain synaptic systems, transmitter concentrations in the millimolar range have been described; for synapses using GIRK channels, no information regarding these issues is available at present. Similar to the situation in atrial myocytes, convergence of multiple receptors on GIRK channels has been described in neurons (e.g. Sodickson & Bean, 1998). It is conceivable therefore, analogous to the result described in Fig. 4B, that in certain conditions of simultaneous release of different transmitters, an inhibitory action might turn into an excitatory one. Finally, overexpression of receptors in the heart (Akhter et al. 1997), including the A1AdoR (e.g. Dougherty et al. 1998), is being conceptually discussed for therapeutic purposes. In this regard dramatic changes in signalling properties of a receptor have to be taken into consideration.

Acknowledgments

This work was supported by the Deutsche Forschungsgemeinschaft. We thank Anke Galhoff, Bing Liu and Gabriele Reimus for expert technical assistance. We are grateful to Dr A. Karschin, the late Dr E. Peralta, and Dr Y. Kurachi for supplying cDNA clones.

REFERENCES

  1. Akhter SA, Skaer CA, Kypson AP, McDonald PH, Peppel KC, Glower DD, Lefkowitz RJ, Koch WJ. Restoration of β-adrenergic signaling in failing cardiac ventricular myocytes via adenoviral-mediated gene transfer. Proc Natl Acad Sci U S A. 1997;94:12100–12105. doi: 10.1073/pnas.94.22.12100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Bender K, Wellner-Kienitz M-C, Inanobe A, Meyer T, Kurachi Y, Pott L. Overexpression of monomeric and multimeric GIRK4 subunits in rat atrial myocytes removes fast desensitization and reduces inward rectification of muscarinic K+ current (IK(ACh). Evidence for functional homomeric GIRK4 channels. JBiol Chem. 2001;276:28873–28880. doi: 10.1074/jbc.M102328200. [DOI] [PubMed] [Google Scholar]
  3. Bender K, Wellner-Kienitz M-C, Pott L. Transfection of a phosphatidyl-4-phosphate 5-kinase gene into rat atrial myocytes removes inhibition of GIRK current by endothelin and α-adrenergic agonists. FEBS Lett. 2002;529:356–360. doi: 10.1016/s0014-5793(02)03426-9. [DOI] [PubMed] [Google Scholar]
  4. Brzostowski JA, Kimmel AR. Signaling at zero G: G-protein-independent functions for 7-TM receptors. Trends Biochem Sci. 2001;26:291–297. doi: 10.1016/s0968-0004(01)01804-7. [DOI] [PubMed] [Google Scholar]
  5. Bünemann M, Brandts B, Pott L. Downregulation of muscarinic M2 receptors linked to K+ current in cultured guinea-pig atrial myocytes. J Physiol. 1996a;492:351–362. doi: 10.1113/jphysiol.1996.sp021497. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bünemann M, Lee KB, Pals-Rylaarsdam R, Roseberry AG, Hosey MM. Desensitization of G-protein-coupled receptors in the cardiovascular system. Annu Rev Physiol. 1999;61:169–192. doi: 10.1146/annurev.physiol.61.1.169. [DOI] [PubMed] [Google Scholar]
  7. Bünemann M, Liliom K, Brandts B, Pott L, Tseng J-L, Desiderio GM, Sun G, Miller D, Tigyi G. A novel receptor with high affinity for lysosphingomyelin and sphingosine 1-phosphate in atrial myocytes. EMBO J. 1996b;15:5527–5534. [PMC free article] [PubMed] [Google Scholar]
  8. Bünemann M, Pott L. Down-regulation of A1 adenosine receptors coupled to muscarinic K+ current in cultured guinea-pig atrial myocytes. J Physiol. 1995;482:81–92. doi: 10.1113/jphysiol.1995.sp020501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Cho H, Nam G-B, Lee SH, Earm YE, Ho W-K. Phosphatidylinositol 4,5-bisphosphate is acting as a signal molecule in α1-adrenergic pathway via the modulation of acetylcholine-activated K+ channels in mouse atrial myocytes. J Biol Chem. 2001;276:159–164. doi: 10.1074/jbc.M004826200. [DOI] [PubMed] [Google Scholar]
  10. Dascal N. Ion-channel regulation by G proteins. Trends Endocrinol Metab. 2001;12:391–398. doi: 10.1016/s1043-2760(01)00475-1. [DOI] [PubMed] [Google Scholar]
  11. Dougherty C, Barucha J, Schofield PR, Jacobson KA, Liang BT. Cardiac myocytes rendered ischemia resistant by expressing the human adenosine A1 or A3 receptor. FASEB J. 1998;12:1785–1792. doi: 10.1096/fasebj.12.15.1785. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Fanning AS, Anderson JM. PDZ domains: fundamental building blocks in the organization of protein complexes at the plasma membrane. J Clin Invest. 1999;103:767–772. doi: 10.1172/JCI6509. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Hall RA, Premont RT, Lefkowitz RJ. Heptahelical receptor signaling: beyond the G protein paradigm. J Cell Biol. 1999;145:927–932. doi: 10.1083/jcb.145.5.927. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. He TC, Zhou SB, Da Costa LT, Yu J, Kinzler KW, Vogelstein B. A simplified system for generating recombinant adenoviruses. Proc Natl Acad Sci U S A. 1998;95:2509–2514. doi: 10.1073/pnas.95.5.2509. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Heuss C, Gerber U. G-protein-independent signaling by G-protein-coupled receptors. Trends Neurosci. 2000;23:469–475. doi: 10.1016/s0166-2236(00)01643-x. [DOI] [PubMed] [Google Scholar]
  16. Himmel HM, Meyer zu Heringsdorf D, Graf E, Dobrev D, Kortner A, Schuler S, Jakobs KH, Ravens U. Evidence for Edg-3 receptor-mediated activation of I(KACh) by sphingosine-1-phosphate in human atrial cardiomyocytes. Mol Pharmacol. 2000;58:449–454. doi: 10.1124/mol.58.2.449. [DOI] [PubMed] [Google Scholar]
  17. Ishii M, Inanobe A, Fujita S, Makino Y, Hosoya Y, Kurachi Y. Ca2+ elevation evoked by membrane depolarization regulates G protein cycle via RGS proteins in the heart. Circ Res. 2001;89:1045–1050. doi: 10.1161/hh2301.100815. [DOI] [PubMed] [Google Scholar]
  18. Ishii M, Inanobe A, Kurachi Y. PIP3 inhibition of RGS protein and its reversal by Ca2+/calmodulin mediate voltage-dependent control of the G protein cycle in a cardiac K+ channel. Proc Natl Acad Sci U S A. 2002;99:4325–4330. doi: 10.1073/pnas.072073399. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Krupnick JG, Benovic JL. The role of receptor kinases and arrestins in G protein-coupled receptor regulation. Annu Rev Pharmacol Toxicol. 1998;38:289–319. doi: 10.1146/annurev.pharmtox.38.1.289. [DOI] [PubMed] [Google Scholar]
  20. Kurachi Y, Nakajima T, Sugimoto T. On the mechanism of activation of muscarinic K+ channels by adenosine in isolated atrial cells: involvement of GTP-binding proteins. Pflugers Arch. 1986;407:264–274. doi: 10.1007/BF00585301. [DOI] [PubMed] [Google Scholar]
  21. Leaney JL, Milligan G, Tinker A. The G protein α subunit has a key role in determining the specificity of coupling to, but not the activation of, G protein-gated inwardly rectifiying K+ channels. J Biol Chem. 2000;275:921–929. doi: 10.1074/jbc.275.2.921. [DOI] [PubMed] [Google Scholar]
  22. Lesage F, Guillemare E, Fink M, Duprat F, Heurteaux C, Fosset M, Romey G, Barhanin J, Lazdunski M. Molecular properties of neuronal G-protein-activated inwardly rectifying K+ channels. J Biol Chem. 1995;270:28660–28667. doi: 10.1074/jbc.270.48.28660. [DOI] [PubMed] [Google Scholar]
  23. Levade T, Auge N, Veldman RJ, Cuvillier O, Negre-Salvayre A, Salvayre R. Sphingolipid mediators in cardiovascular cell biology and pathology. Circ Res. 2001;89:957–968. doi: 10.1161/hh2301.100350. [DOI] [PubMed] [Google Scholar]
  24. Lohberger B, Groschner K, Tritthart H, Schreibmayer W. IK,ACh activation by arachidonic acid occurs via a G-protein- independent pathway mediated by the GIRK1 subunit. Pflugers Arch. 2000;441:251–256. doi: 10.1007/s004240000405. [DOI] [PubMed] [Google Scholar]
  25. Luscher C, Jan LY, Stoffel M, Malenka RC, Nicoll RA. G protein-coupled inwardly rectifying K+ channels (GIRKs) mediate postsynaptic but not presynaptic transmitter actions in hippocampal neurons. Neuron. 1997;19:687–695. doi: 10.1016/s0896-6273(00)80381-5. [DOI] [PubMed] [Google Scholar]
  26. Meyer T, Wellner-Kienitz M-C, Biewald A, Bender K, Eickel A, Pott L. Depletion of phosphatidylinositol 2,4-bisphosphate by activaton of PLC-coupled receptors causes slow inhibition but not desensitisation of GIRK current in atrial myocytes. J Biol Chem. 2001;276:5650–5658. doi: 10.1074/jbc.M009179200. [DOI] [PubMed] [Google Scholar]
  27. Milligan G, White JH. Protein-protein interactions at G-protein-coupled receptors. Trends Pharmacol Sci. 2001;22:513–518. doi: 10.1016/s0165-6147(00)01801-0. [DOI] [PubMed] [Google Scholar]
  28. Nasuhoglu C, Feng S, Mao Y, Shammat I, Yamamato M, Earnest S, Lemmon M, Hilgemann DW. Modulation of cardiac PIP2 by cardioactive hormones and other physiologically relevant interventions. Am J Physiol Cell Physiol. 2002;283:C223–234. doi: 10.1152/ajpcell.00486.2001. [DOI] [PubMed] [Google Scholar]
  29. Nicoll RA. The coupling of neurotransmitter receptors to ion channels in the brain. Science. 1988;241:545–551. doi: 10.1126/science.2456612. [DOI] [PubMed] [Google Scholar]
  30. Pfaffl MW, Horgan GW, Dempfle L. Relative expression software tool (REST(C)) for group-wise comparison and statistical analysis of relative expression results in real-time PCR. Nucleic Acids Res. 2002;30:e36. doi: 10.1093/nar/30.9.e36. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Rogalski SL, Chavkin C. Eicosanoids inhibit the G-protein-gated inwardly rectifying potassium channel (Kir3) at the Na+/PIP2 gating site. J Biol Chem. 2001;276:14855–14860. doi: 10.1074/jbc.M010097200. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Sheng M, Sala C. PDZ domains and the organization of supramolecular complexes. Annu Rev Neurosci. 2001;24:1–29. doi: 10.1146/annurev.neuro.24.1.1. [DOI] [PubMed] [Google Scholar]
  33. Sodickson DL, Bean BP. GABAB receptor-activated inwardly rectifying potassium current in dissociated hippocampal CA3 neurons. J Neurosci. 1996;16:6374–6385. doi: 10.1523/JNEUROSCI.16-20-06374.1996. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Sodickson DL, Bean BP. Neurotransmitter activation of inwardly rectifying potassium current in dissociated hippocampal CA3 neurons: Interactions among multiple receptors. J Neurosci. 1998;18:8153–8162. doi: 10.1523/JNEUROSCI.18-20-08153.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Sorota S, Rybina R, Yamamoto A, Du X-Y. Isoprenaline can activate the acetylcholine-induced K+ current in canine atrial myocytes via Gs-derived by subunits. J Physiol. 1999;514:413–423. doi: 10.1111/j.1469-7793.1999.413ae.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Takano M, Noma A. Development of muscarinic potassium current in fetal and neonatal rat heart. Am J Physiol. 1997;272:H1188–1195. doi: 10.1152/ajpheart.1997.272.3.H1188. [DOI] [PubMed] [Google Scholar]
  37. Takigawa T, Alzheimer C. G protein-activated inwardly rectifying K+ (GIRK) currents in dendrites of rat neocortical pyramidal cells. J Physiol. 1999;517:385–390. doi: 10.1111/j.1469-7793.1999.0385t.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Wellner-Kienitz M-C, Bender K, Brandts B, Meyer T, Pott L. Antisense oligonucleotides against receptor kinase GRK2 disrupt target selectivity of β-adrenergic receptors in atrial myocytes. FEBS Lett. 1999;451:279–283. doi: 10.1016/s0014-5793(99)00594-3. [DOI] [PubMed] [Google Scholar]
  39. Wellner-Kienitz M-C, Bender K, Meyer T, Bünemann M, Pott L. Overexpressed A1 adenosine receptors reduce activation of acetylcholine-sensitive K+ current by native muscarinic M2 receptors in rat atrial myocytes. Circ Res. 2000;86:643–648. doi: 10.1161/01.res.86.6.643. [DOI] [PubMed] [Google Scholar]
  40. Wickman K, Karschin C, Karschin A, Picciotto MR, Clapham D. Brain localization and behavioral impact of the G-protein-gated K+ channel subunit GIRK4. J Neurosci Res. 2000;20:5608–5615. doi: 10.1523/JNEUROSCI.20-15-05608.2000. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Yamada M. The role of muscarinic K+ channels in the negative chronotropic effect of a muscarinic agonist. J Pharmacol Exp Ther. 2002;300:681–687. doi: 10.1124/jpet.300.2.681. [DOI] [PubMed] [Google Scholar]
  42. Yamada M, Inanobe A, Kurachi Y. G protein regulation of potassium ion channels. Pharmacol Rev. 1998;50:723–757. [PubMed] [Google Scholar]

Articles from The Journal of Physiology are provided here courtesy of The Physiological Society

RESOURCES