OUP user menu

Characterization and inhibition of β-adrenergic receptor kinase in intact myocytes

Karl-Ludwig Laugwitz , Kai Kronsbein , Manfred Schmitt , Katrin Hoffmann , Melchior Seyfarth , Albert Schömig , Martin Ungerer
DOI: http://dx.doi.org/10.1016/S0008-6363(97)00102-8 324-333 First published online: 1 August 1997

Abstract

Objectives: β-Adrenergic receptor kinase (βARK) phosphorylates and thereby inactivates agonist-occupied β-adrenergic receptors (βAR). βARK is thought to play an important role in the regulation of cardiac function. Therefore, we studied βARK activation and its inhibition in intact smooth muscle cells and in cardiomyoblasts. Methods and Results: βAR agonist-stimulated translocation of βARK was monitored by immunofluorescence labelling with specific antibodies and confocal laser scanning microscopy in DDT-MF 2 hamster smooth muscle cells and in H9c2 rat cardiomyoblasts. In unstimulated cells, βARK was mainly located in the cytosol. After βAR agonist stimulation, the βARK signal was partially translocated to the membranes. Liposomal gene transfer of the COOH-terminus of βARK (‘βARKmini’) as a βARK inhibitor led to functional expression of this protein in both cell lines with high efficiency. Western blots with βARK antibodies showed a gene concentration-dependent immunoreactivity of the ‘βARKmini’ protein. ‘βARKmini’-transfected myocytes demonstrated reduced membrane targeting of the βARK immuno-fluorescence signal. Additionally, the effect of ‘βARKmini’ on βAR-induced desensitization of myocytic cAMP accumulation was investigated. In control cells, desensitization with isoproterenol led to a subsequent reduction of βAR-induced cAMP accumulation. In ‘βARKmini’-transfected myocytes, this βAR-induced desensitization was significantly diminished, whereas normal βAR-induced cAMP accumulation was unaffected. A gene concentration of 2 μg ‘βARKmini’ DNA/100 000 cardiomyoblasts, and of 0.7 μg ‘βARKmini’ DNA/100 000 DDT-MF2 smooth muscle cells led to ≈5.9- and ≈5.6-fold overexpressions of ‘βARKmini’ vs. native βARK, respectively. These gene doses proved sufficient to attenuate β-adrenergic desensitization significantly. Conclusions: (1) βARK translocation was evidenced in DDT-MF2 smooth muscle cells and in cardiomyoblasts by confocal laser scanning microscopy. (2) Feasibility of ‘βARKmini’ gene transfer to myocytes was demonstrated, and necessary gene doses for βARK inhibition were titered. (3) Overexpression of ‘βARKmini’ functionally interacted with endogenous β-adrenergic signal transduction, leading to sustained cAMP accumulation after prolonged β-adrenergic stimulation.

Keywords
  • Myocytes
  • β-Adrenergic receptor kinase
  • Gene transfer
  • Adenylyl cyclase

Time for primary review 38 days.

1 Introduction

Several biochemical mechanisms contribute to β-adrenergic receptor desensitization [1]. Most importantly, agonist-occupied active receptors become phosphorylated by the specific β-adrenergic receptor kinase (βARK) [1, 2], facilitating the subsequent binding of the inhibitor protein, β-arrestin, to the receptors [3]. Thereby, receptor function is inhibited by up to 70% [4, 5]. Phosphorylated β-adrenergic receptors lose their capacity to activate subsequent second-messenger systems, such as adenylyl cyclase [1, 2]. Alternative mechanisms, such as the reduction of receptor number by down-regulation, become operative only after 4–12 h [1]. βARK is thought to play an important role in the regulation of cardiac function. In a previous study, we found a significant increase in cardiac βARK activity and mRNA levels in human heart failure [6]. This activation specifically involved the mRNA for βARK-1, but not for βARK-2 or for β-arrestin [7]. The activation of βARK seems to be a very early event during the development of heart failure, since it preceded the alterations of β-receptor density or of G-proteins in the model of pacing-induced heart failure in the pig [8]. Chronic exposure to β-blockers, in turn, decreased cardiac βARK activity in cytosol and membranes [9]. Also, cardiac ischemia increases βARK activity in cardiac membranes [10].

Pivotal work by Lefkowitz and collaborators has shown that activation of βARK is closely related to a translocation of the enzyme from the cytosol to the membranes. βARK membrane targeting is specifically mediated by G-protein βγ-subunits. βγ-Subunits can bind to the pleckstrin-homology (PH) domain, which forms part of the COOH-terminus of βARK [11, 12]. Overexpression of the COOH-terminus of βARK-1 (amino acids 495–689, ‘βARKmini’) has subsequently been shown to exert a ‘scavenging’ effect on a variety of mechanisms mediated by βγ-subunits, such as α2-adrenergic receptor-mediated stimulation of adenylyl cyclase II [13], or muscarinic M2-stimulated inositol phosphate turnover [13]. ‘βARKmini’ also inhibits Gβγ-stimulated phosphorylation of rhodopsin in vitro [14]. Most importantly, overexpression of ‘βARKmini’ in the hearts of transgenic mice also improved cardiac contractility, as demonstrated by heart catheterization in vivo [14]. Up to now, however, direct evidence for the role of ‘βARKmini’ on βARK translocation in intact cells and in myocytes has not been provided. Generally, βARK translocation has been investigated in isolated, prepared cell fractions, instead of whole cells. Moreover, the inhibitory effect of ‘βARKmini’ on adenylyl cyclase activity has only been shown in reconstituted systems of recombinant proteins [15]. In contrast, direct effects of ‘βARKmini’ on endogenous β-adrenergic desensitization in myocytes have not been demonstrated, since adenylyl cyclase activity in myocardial membranes was not altered by ‘βARKmini’ [14].

Therefore, we set out to characterize βARK translocation directly in intact myocytes. Similarly, we intended to investigate more precisely the effects of ‘βARKmini’ on the endogenous β-adrenergic system in cardiomyocytes. Thereby, we sought to determine adequate gene concentrations of ‘βARKmini’ needed for relevant inhibition of β-adrenergic desensitization in smooth muscle cells and in cardiomyocytes.

2 Methods

2.1 Cell culture

DDT-MF-2 smooth muscle cells derived from hamster ductus deferens leiomyosarcoma were a kind gift of M. Lohse (Würzburg). The cells were grown as monolayers in Dulbecco's modified Eagle's medium (DMEM), supplemented with 10% fetal calf serum (FCS), non-essential amino acids, 2 mmol/l glutamine, penicillin (100 IU), and streptomycin (100 μg/ml) in 5% CO2 in water-saturated air at 37°C.

H9c2 cardiomyoblasts (ATCC CRL 1446, cardiac myoblasts from rat) were purchased from American Type Culture Collection (Rockville, MD, USA) and cultured as monolayers in DMEM, 10% FCS, 2 mmol/l glutamine, penicillin (100 IU/ml), and streptomycin (100 μg/ml) in 7% CO2 in water-saturated air at 37°C. Both cell lines were transfected transiently by liposomal transfer, using 20 μg/ml lipofectamine (Gibco/BRL, Gaithersburg, MD, USA) and varying concentrations of pβARKmini (as indicated), a kind gift of R.J. Lefkowitz (Duke University, Durham, NC, USA). Alternatively, the myocytes were transfected with control plasmid pRK which did not contain the specific insert. For transfer, cells were freshly plated and grown for 24 h. Then, medium was exchanged for preincubated liposome/plasmid mix in Optimem medium (Gibco/BRL, Gaithersburg, MD, USA) for 6 h. Cells were harvested 36 h later. The study conforms with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised in 1985).

2.2 Western blot

For immunoblotting, cells were scraped from dishes, placed in 500 μl of TE buffer (20 mmol/l Tris-HCl, pH 7.5, 1 mmol/l EDTA) and sonicated for 3 min. Where indicated, cytosolic and particulate fractions were prepared as follows: Cells were homogenized for 30 s in a Polytron tissue mincer and centrifuged for 30 min at 100 000×g. The supernatant represented the cytosolic fraction. The pellet was resuspended in TE buffer to be used as membrane fraction, and homogenized again. The volume of both fractions was reduced by filtration in a microconcentrator Centricon 30 (Amicon, Düsseldorf, Germany) to yield 100 μl. Total protein levels in the concentrates were determined according to Bradford, as described [6]. Samples containing 30 μg of protein were boiled with Laemmli buffer for 5 min, and electrophoresed on 12% SDS-polyacrylamide gels. Proteins were then blotted onto Amersham Hybond-C membranes. Efficiency of transfer was verified by Ponceau red staining of the blots. The blots were blocked with 2% ovalbumin in TBS. Immunogenic signals of native βARK or ‘βARKmini’ were detected by probing with rabbit GRK-2 antibody directed against the COOH-terminus of βARK-1, purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA), followed by peroxidase-coupled secondary antibodies (from Sigma, Deisenhofen, Germany) using chemoluminiscent substrate (ECL, Amersham, Braunschweig, Germany) and Kodak XAR films. Densitometric analysis of the Western blots was carried out using NIH graphics software.

2.3 Confocal laser scanning microscopy

pβARKmini- or pRK-transfected myocytes were grown in 6-well tissue culture plates on microscope cover slips for 24 h. Stimulation of the cells was carried out with 1 μmol/l (−)-isoproterenol in PBS. After rinsing 3 times with PBS, cells were fixed for 15 min with 4% para-formaldehyde in PBS. Cells were then permeabilized with 0.1% saponin for 10 min. To avoid non-specific antibody binding, cells were treated with 10% FCS in DMEM for 30 min prior to labelling with GRK-2 antibody (1μg/ml). After incubation with goat anti-rabbit IgG FITC-conjugate (Sigma, Deisenhofen, Germany), the coverslips were mounted upside-down on microscope slides and sealed with nail polish. For microscopy, samples were visualized by differential interference contrast microscopy (Nomarski optics) and by excitation with a low-power argon laser at 488 nm.

Digitalized microscopic images of fluorescent myocytes were analyzed densitometrically. The signal intensity in representative membrane areas was quantified as arbitrary units.

2.4 FACS

Transfected myocytes were grown on 6-well tissue culture plates for 24 h. Cells to be used for fluorescence analysis were treated as described above for immunofluorescence microscopy and removed from the plates with trypsin-EDTA prior to FACS analysis. Samples were analyzed on a FACScan cytometer (Becton Dickinson) equipped with an argon laser. Cell signals were plotted in a forward versus side scatter, and 50 000 particles were analyzed at a rate of less than 100 events per second. A gate was defined in order to exclude all debris and aggregates. Logarithmic amplification was used for both, the fluorescence and light scatter signals. The mean intensity of immunofluorescence was used as a quantitative measure of antigen representation on cells.

2.5 cAMP accumulation in intact myocytes

pβARKmini- or pRK-transfected myocytes were grown in 6-well Falcon tissue culture plates to 80–90% confluence. They were incubated in PBS containing 1 mmol/l of the phosphodiesterase inhibitor, IBMX, for 30 min. For desensitization, cells were incubated in 1 μmol/l (−)-isoproterenol in PBS and IBMX for 15 min to cause maximal stimulation of the cells' β2-adrenergic receptors. Controls were treated equally without isoproterenol. Following aspiration of the buffer, the cells were rinsed 4 times with PBS and incubated with various concentrations of (−)-isoproterenol for 15 min. The incubation was stopped by addition of 0.5 ml of boiling TE buffer at pH 7.5. The cells were then scraped off the plates and were separated from soluble material by centrifugation at 5000×g for 10 min. The pellet was kept for protein determination.

The cAMP content in the supernatant was determined by a [3H]cAMP assay system purchased from Amersham (Braunschweig, Germany) according to instructions supplied by the manufacturer.

3 Results

3.1 Western blots

Western blotting of whole cell extracts of both DDT-MF2 smooth muscle cells and H9c2 cardiomyoblasts showed a weak, but clear signal at 80 kDa, the expected size of native βARK (Fig. 1A,B, lane 7). A second, non-specific band at a size of ≈60 kDa was obvious on cell extract blots of DDT-MF 2 cells, and was also present on blots exposed to preimmune serum instead of specific βARK antibody (Fig. 1A, lane 8). After liposomal transfection of 0.25–5 μg pβARKmini to 100 000 DDT-MF2 smooth muscle cells or of 0.3–3 μg pβARKmini to 100 000 cardiomyoblasts, a gene dose-dependent band appeared at a size of 27 kDa (Fig. 1A,B, lanes 2–6). Both native βARK signals and ‘βARKmini’ signals were competitively inhibited by increasing concentrations of the GRK-2 peptide. Lane 9 of Fig. 1A,B demonstrates the effect of 10 μg/mL GRK-2 peptide on the signal of the extracts shown in lane 4. Analysis of signal intensity by laser densitometry revealed that the transfer of 0.7 μg pβARKmini to 100 000 DDT-MF 2 smooth muscle cells led to ≈5.6-fold overexpression of ‘βARKmini’, compared to native βARK. Similarly, 2 μg pβARKmini transferred to 100 000 cardiomyoblasts led to ≈5.9-fold overexpression of ‘βARKmini’.

Fig. 1

Western blots of GRK-2 antibody immunoreactivity in whole cell extracts of DDT-MF2 smooth muscle cells (A) and cardiomyoblasts (B) transfected with increasing concentrations of pβARKmini (lanes 2–6), or control plasmid (lane 7). GRK-2 antibody is directed against the COOH-terminal region of βARK-1, and thus recognizes both native βARK (expected size of 80 kDa) and recombinant ‘βARKmini’ (expected size of 27 kDa). Blots prepared as in lane 4 were exposed to pre-immune serum instead of specific antibody (lanes 8, ‘pre’), or to GRK-2 peptide (lane 9, ‘pep’).

Western blotting of membrane extracts of both cell lines demonstrated a strong signal of the ‘βARKmini’ transgene in cytosol and a weaker signal in the membranes. Fig. 2A shows that in membranes from DDT-MF 2 smooth muscle cells transfected with control vector pRK, the signal for native βARK was increased after stimulation with 1 μmol/l isoproterenol for 30 min (lanes 3 and 4). In contrast, this increase was clearly diminished in smooth muscle cells transfected with 0.7 μg pβARKmini/100 000 cells (lanes 1 and 2). No detectable differences in the cytosolic immunosignal of native βARK occurred under any condition (lanes 5–8). Similarly, cardiomyoblasts transfected with 2 μg control vector pRK/100 000 cells showed an increased native βARK membrane signal after agonist exposure (Fig. 2B, lanes 3 and 4). This increase was markedly attenuated in cardiomyoblasts transfected with 2 μg pβARKmini/100 000 cells (lanes 1 and 2). Again, the cytosolic signal for native βARK was not altered under any condition (lanes 5–8).

Fig. 2

Western blots of GRK-2 antibody immunoreactivity in membrane fractions (left images) and in cytosol (right images) prepared from DDT-MF2 smooth muscle cells (A) and cardiomyoblasts (B). The myocytes were either transfected with 0.7 μg (A) or 2 μg (B) pβARKmini (‘+’), or control plasmid (‘−’). The cells were then exposed either to 1 μmol/l isoproterenol (‘+’) or to vehicle (‘−’).

3.2 Flow cytometry

Flow cytometric analysis was performed on 50 000 cells after being detached from culture flasks. Fig. 3A demonstrates the analysis of the fluorescence intensity of DDT-MF2 smooth muscle cells which had been incubated with the secondary antibody only, or with specific GRK-2 antibody, and of myocytes which had been transfected with 0.7 μg pβARKmini/100 000 cells. The smooth muscle cells which had been incubated with the specific GRK-2 antibody showed significantly higher fluorescence intensity than cells exposed to FITC antibody only. After gene transfer with pβARKmini, a second peak of nearly 10 scale units higher intensity appeared. The areas under the two curves were estimated by geometric analysis. Based upon this approximation, ≈40% of the DDT-MF2 smooth muscle cells exposed to the ‘βARKmini’ plasmid had gained additional immuno-fluorescence intensity, corresponding to the cytoplasmic expression of ‘βARKmini’ in addition to native βARK.

Fig. 3

Fluorescence-activated cell sorter (FACS) histograms of permeabilized DDT-MF2 smooth muscle cells (A) and cardiomyoblasts (B), as described in Section 2. The myocytes were either transfected with control plasmid, or 0.7 μg (A) or 2 μg (B) pβARKmini. The cells were then permeabilized and incubated either with vehicle (smooth lines) or GRK-2 antibody (bold lines), followed by FITC-labelled second antibody. For each condition, relative fluorescence was detected in 50 000 cells by flow cytometry.

Similarly, cardiomyoblasts showed specific immunolabelling with the GRK-2 antibody (Fig. 3B). After gene transfer with 2 μg pβARKmini/100 000 cardiomyoblasts, ≈25% of total cells gained significantly higher specific fluorescence intensity, as estimated by geometric analysis of the surfaces under the two curves.

3.3 Confocal laser scanning microscopy

FITC-labelled cells were also subjected to confocal laser scanning microscopy. Fig. 4 shows representative light transmission and fluorescence images of DDT-MF2 smooth muscle cells transfected with either control plasmid pRK or pβARKmini and stimulated with 1 μmol/l isoproterenol or vehicle. Fig. 4A shows the absence of specific signals in cells exposed to FITC antibody only. Fig. 4B shows the largely homologous cytosolic distribution of native βARK in unstimulated cells. After exposure to isoproterenol, the βARK signal was significantly translocated to the membranes, whereas less signal was found in the cytosol (Fig. 4C). Translocation reached its maximum within 10 min and did not increase thereafter. After transfection with 0.7 μg pβARKmini/100 000 cells, approximately 30% of the cells showed a markedly higher fluorescence signal. Such a cluster of positive cells is shown in Fig. 4D. After stimulation of ‘βARKmini’-expressing cells with isoproterenol, markedly less translocation of the native βARK signal to the membranes occurred (Fig. 4E). Densitometric analysis of the membrane areas of ‘βARKmini’-expressing and ‘βARKmini’-negative cells after agonist stimulation showed significantly higher signal intensity in cells which did not express ‘βARKmini’ (Fig. 6A). DDT-MF2 smooth muscle cells transfected with more than 1.5 μg pβARKmini/100 000 cells changed their morphology, and a significant number did not survive the gene transfer. Lower concentrations of pβARKmini, however, did not change cell morphology detectably.

Fig. 6

Densitometric analysis of representative membrane areas of DDT-MF 2 smooth muscle cells (A, n = 10 cells) and cardiomyoblasts (B, n = 10 cells) after control incubation (solid bars) and after agonist stimulation (hatched bars). Digitalized images of fluorescence-activated cells were analysed for signal intensity in representative membrane areas. The cells were either transfected with pRK (white bars) or pβARKmini (grey bars). Results are shown as arbitrary units±s.e.m. *P≤0.01 isoproterenol vs. control, **P<0.05 pβARKmini vs. pRK.

Fig. 4

Representative fluorescence-activated and light-transmission images of permeabilized DDT-MF2 smooth muscle cells. The myocytes were incubated either with vehicle (A) or GRK-2 antibody (B–E), followed by FITC-labelled secondary antibody, and then subjected to confocal laser scanning microscopy. (B) Native βARK labelled in myocytes exposed to vehicle, only. (C) Membrane translocation of native βARK after exposure to isoproterenol (indicated with an arrow). (D) Marked increase in fluorescence in a cluster of myocytes transfected with 0.7 μg pβARKmini/100 000 cells. (E) In pβARKmini-transfected myocytes, membrane translocation of the native βARK immunosignal was attenuated. The following pseudocolour scale was used: white, high signal intensity; orange, medium signal intensity; red, low signal intensity; black, none.

Also in cardiomyoblasts, native βARK was translocated after agonist stimulation (Fig. 5B,C). Fig. 5C demonstrates that stimulation with isoproterenol led to a somewhat asymmetrical increase in the βARK signal to one side of the membrane. This was a typical finding in almost all stimulated cardiomyoblasts investigated. Additionally, the cytosolic βARK signal appeared more granular after agonist stimulation, possibly corresponding to an internal translocation of βARK to intracellular microsomal structures, as suggested by some authors [16, 17]. Again, translocation reached its maximum within 10 min. After transfection with 2 μg pβARKmini/100 000 cells, nearly 25% of the transfected cells gained markedly higher βARK immunofluorescence, due to the additional expression of ‘βARKmini’ (Fig. 5D). ‘βARKmini’-expressing cardiomyoblasts showed no detectable translocation of the native βARK signal after agonist stimulation (Fig. 5E). Densitometric analysis of the membrane areas of ‘βARKmini’-expressing and ‘βARKmini’-negative cells after agonist stimulation showed significantly higher signal intensity in cells which did not express ‘βARKmini’ (Fig. 6B). Higher doses of pβARKmini >5 μg/100 000 cardiomyoblasts were cytotoxic.

Fig. 5

Representative fluorescence-activated and light-transmission images of permeabilized cardiomyoblasts. The myoblasts were incubated either with vehicle (A) or GRK-2 antibody (B–E), followed by FITC-labelled secondary antibody, immobilized and then subjected to confocal laser scanning microscopy. (B) Native βARK labelled in myocytes exposed to vehicle only. (C) Membrane translocation of native βARK after exposure to isoproterenol. (D) Marked increase in fluorescence in a cardiomyocyte from a batch transfected with 2 μg pβARKmini/100 000 cells. (E) In pβARKmini-transfected cardiomyoblasts, membrane translocation of the native βARK immunosignal was clearly attenuated. The following pseudocolour scale was used: white, high signal intensity; orange, medium signal intensity; red, low signal intensity; black, none.

3.4 Adenylyl cyclase desensitization

cAMP concentrations were measured in intact myocytes after stimulation with isoproterenol. Fig. 7 demonstrates that isoproterenol in concentrations ranging from 0.001 to 10 μmol/l led to intracellular accumulation of cAMP. On average, cAMP levels were raised from 1 to 38 pmol/mg protein in DDT-MF2 smooth muscle cells, and from 13 to 50 pmol/mg protein in cardiomyoblasts. In DDT-MF2 smooth muscle cells transfected with 0.7 μg pβARKmini/100 000 cells (corresponding to ≈5.6-fold overexpression of ‘βARKmini’ vs. native βARK, as estimated from Western blots), basal and isoproterenol-induced maximal cAMP levels were unaltered, with EC50 values ranging from 100 to 150 nmol/l. When the cells were desensitized by preincubation with isoproterenol for 15 min, the response to a subsequent, repeated exposure to the β-agonist was markedly blunted. This desensitization reached significance only in pRK-transfected smooth muscle cells (P<0.02), whereas cAMP accumulation after agonist exposure did not differ significantly in pβARKmini-transfected cells. After prior exposure to isoproterenol, ‘βARKmini’-transfected cells accumulated significantly more cAMP (Fig. 7A, P<0.05). When cells were transfected with lower gene concentrations of pβARKmini (0.1 μg/100 000 cells, corresponding to equal expression of ‘βARKmini’ and βARK), they could still be subjected to significant desensitization, accumulating significantly less cAMP after pre-exposure to isoproterenol (P<0.05, Fig. 8A). pβARKmini doses exceeding 1.5 μg/100 000 cells (≈15-fold overexpression of ‘βARKmini’) led to a change in cell morphology, and to a decrease in cAMP response to isoproterenol (Fig. 8C).

Fig. 8

Analysis of isoproterenol-induced cAMP formation in DDT-MF 2 myocytes. 100 000 myocytes were either transfected with 0.1 μg (A) or 0.7 μg (B) or 1.5 μg pβARKmini, resulting in 1-, 5.6- and 15-fold overexpressions of ‘βARKmini’ vs. native βARK. The cells were then grown in 6-well plates; 48 h later, the myocytes were exposed either to vehicle (circles) or to 1 μmol/l isoproterenol (squares). The cells were washed thoroughly, and formation of cAMP was measured in the presence of increasing concentrations of isoproterenol. Similar results were obtained in cardiomyoblasts. *P<0.02 desensitized vs. non-desensitized.

Fig. 7

Agonist-induced desensitization of β-adrenergic receptors in intact myocytes: analysis of isoproterenol-induced cAMP formation in DDT-MF 2 myocytes (A) and cardiomyoblasts (B). 100 000 myocytes were either transfected with 0.7 μg (A) or 2 μg (B) pβARKmini (black symbols), or control plasmid pBC (white symbols). The cells were then grown in 6-well plates; 48 h later, the myocytes were exposed either to vehicle (circles) or to 1 μmol/l isoproterenol (squares). The cells were washed thoroughly, and formation of cAMP was measured in the presence of increasing concentrations of isoproterenol. Data are means of 3 experiments, each carried out in triplicate, ±s.e.m. *P≤0.05 pβARKmini vs. pRK, **P<0.02 desensitized vs. non-desensitized.

Similarly, cAMP accumulation was identical in cardiomyoblasts transfected with 2 μg pβARKmini/100 000 cells (corresponding to ≈5.9-fold overexpression of ‘βARKmini’), and in cells transfected with 2 μg of pRK. Also in cardiomyoblasts, expression of ‘βARKmini’ did not affect regular β-adrenergic stimulation. After pre-exposure to isoproterenol, however, the cAMP response elicited by isoproterenol was significantly higher in pβARKmini-transfected cardiomyoblasts than in pRK-transfected cardiomyoblasts (Fig. 7B). Expression of ‘βARKmini’ led to a virtually complete reversion of desensitization. In contrast, cardiomyoblasts transfected with 0.1 or 0.3 μg of pβARKmini (corresponding to ≈0.7- and ≈1.1-fold overexpression of ‘βARKmini’, respectively) showed a pattern of desensitization which did not differ significantly from that of control cells (not shown).

4 Discussion

The present study characterizes βARK translocation in intact myocytes, as detected by immunofluorescence labelling. βARK translocation and β-adrenergic desensitization could be inhibited effectively by a plasmid-induced expression of the inhibitor protein, ‘βARKmini’, in intact myocytes and cardiomyoblasts. Transfection of more than 0.7 μg of pβARKmini to 100 000 DDT-MF 2 smooth muscle cells, and of 2 μg of pβARKmini to 100 000 cardiomyoblasts led to sufficient expression of this protein to significantly attenuate endogenous β-adrenergic desensitization in these myocytes.

βARK selectively phosphorylates and thereby inactivates agonist-occupied β-adrenergic receptors. This process has been linked to translocation of the enzyme from a cytosolic compartment to the membranes [1]. We have demonstrated that specific antibodies can be used to effectively label βARK in permeabilized smooth muscle cells and cardiomyoblasts. Confocal laser images of these cells suggest that βARK is actually translocated rapidly to the myocytic membrane after β-agonist stimulation. In DDT-MF2 smooth muscle cells, which display a narrow cytoplasm, translocation seemed to be directed to all sides of the myocyte, and the cytosol appeared to clear rapidly of the βARK immunosignal after agonist exposure. Since small cytoplasmic margins compromise the microscopic differentiation between cytosol and membranes, we compared these effects with another myocytic cell line with larger cytoplasm. Our data corroborate the results of an earlier publication on the translocation of βARK activity in DDT-MF2 smooth muscle cells after β-adrenergic stimulation [18].

In cardiomyoblasts, which possess a pronouncedly larger cytoplasm, translocation appeared asymmetrical in all cells investigated. The homogeneous cytosolic distribution of the βARK immunosignal seemed to assume a more granular appearance after the agonist challenge, possibly corresponding to an internal translocation to a microsomal compartment, as suggested by others [16, 17]. This issue can, however, not be decided with certainty, since the spatial resolution of fluorescence imaging is limited. Translocation reached a maximum within 10 min and was not further increased thereafter. This time course corresponds quite well with previous studies in which the capacity of cell extracts to phosphorylate rhodopsin or β2-adrenergic receptors was used to determine βARK kinetics [19].

Considerable interest has evolved in inhibiting cardiac βARK activation since overexpression of the inhibitor protein ‘βARKmini’ has been shown to increase cardiac contractility in transgenic mice [14]. These animals, however, developed high constitutive expression of ‘βARKmini’, since the transgene was coupled to the α-myosin heavy chain promotor which was active during the entire life of the animals. Our study shows that also external plasmid-derived gene transfer of ‘βARKmini’, which leads to less expression of the inhibitor protein, can effectively inhibit β-adrenergic desensitization in intact myocytes. Our results demonstrate that the effect of ‘βARKmini’ can be tested biochemically by estimating the amount of immunolabelled βARK which translocates after agonist stimulation, and by quantifying intramyocytic cAMP accumulation after previous desensitization. This finding is of some relevance, since measurement of adenylyl cyclase activity in membranes prepared from ‘βARKmini’-transgenic and control hearts did not reveal any differences [14]. Normal cAMP accumulation was not affected by ‘βARKmini’, but sustained cAMP formation after prolonged agonist stimulation occurred. Therefore, ‘βARKmini’ might interact specifically with the mechanisms which lead to rapid β-adrenergic refractoriness in heart failure. We conclude that the analysis of single myocytes contributes to the understanding of the biological effects of ‘βARKmini’ in cardiomyocytes.

The amount of plasmid needed for the expression of ‘βARKmini’ in biologically relevant concentrations was titered to be 0.7 μg for 100 000 DDT myocytes, and 2 μg for 100 000 cardiomyoblasts, corresponding to similar overexpressions of ‘βARKmini’ vs. native βARK (5.6- and 5.9-fold, respectively). Since several reports have described the possibility of cardiac gene expression after direct plasmid injection [20, 21], the direct application of a plasmid by intracardiac injection might suffice to attenuate β-adrenergic desensitization in vivo. This information about gene effects in cardiomyocytes should be useful for further studies, since several investigators have described a gene concentration-dependency of plasmid-derived expression of recombinant proteins in vivo. Such an example was given recently by the successful transfer of a plasmid encoding VEGF into the popliteal artery of a patient, leading to enhanced collateralisation and better clinical outcome [22]. In this study, effects started at a plasmid dose of 2000 μg. An even higher biological effect of ‘βARKmini’ can be expected from future experiments with viral transfer systems which will allow even more efficient gene transfer to a higher percentage of cells.

4.1 Limitations of the study

In the present study, we distinguished the signals of ‘βARKmini’ and of native βARK by their sizes on Western blots, and by their localisation by confocal laser microscopy. On Western blots, the bulk of ‘βARKmini’ was found in the cytosolic fraction. A small amount of ‘βARKmini’ in the particulate fraction is probably due to massive overexpression of the protein, and to minor contamination during preparation. This is in line with previous studies [13], since ‘βARKmini’ is known to be a strictly cytosolic protein which lacks all components necessary for membrane translocation, namely the site of isoprenylation. Therefore, the translocation observed by means of immunofluorescence (i.e., the gain in membrane signal after agonist stimulation) can be assumed to be due to translocation of native βARK. In contrast, the cytosolic signal for native βARK could not be studied in ‘βARKmini’-expressing cells by confocal microscopy. This signal could, however, be differentiated on Western blots of cytosolic fractions.

5 Note added in proof

During the review process of the present manuscript, an extensive study on the effect of adenoviral transfer of ‘βARKmini’ to cardiomyocytes was published. This paper corroborates our data on the effects of ‘βARKmini’ on β-adrenergic desensitization (Drazner et al., J Clin Invest 1997;99:288–296).

Acknowledgements

We thank Ms. Claudia Rokitta for skilful help with the FACS analysis. This study was supported by a grant of the Deutsche Forschungsgemeinschaft (Un 103/1-1).

References

View Abstract