Introduction

The β-adrenergic pathway is activated upon an increase in catecholamine levels. Binding of catecholamines to the β-adrenergic receptors (βARs) increases contractility of the cardiomyocytes via protein-kinase-A-mediated phosphorylation of downstream target proteins involved in calcium handling (Hasenfuss and Fieske 2002) and myofilament contraction (Solaro et al. 1976). In the human myocardium, at least three types of βARs are present, including β1AR, β2AR, and β3AR (Brodde 2007). Analysis of βARs in crude membrane fractions using radioligand binding studies revealed a ratio of 80:20 for β1AR and β2AR in nonfailing human donor tissue (Brodde 2007), with β1AR as the predominant receptor. In failing idiopathic dilated cardiomyopathy (IDCM) myocardium, a downregulation of β1AR has been observed, leading to a relative increase in β2AR, as illustrated by the altered ratio of β1AR and β2AR to 60:40 in the failing heart (Brodde 2007; Bristow et al. 1986, 1989). Diverse alterations in myocardial βAR density have been reported in IDCM and ischemic heart disease (ISHD) patients (Bristow et al. 1991; Steinfath et al. 1991). In failing IDCM myocardium, a downregulation of β1AR has been observed (Bristow et al. 1991), while β2AR remained unchanged (Steinfath et al. 1991). In contrast, a reduction of both β1AR and β2AR was reported in ISHD patients (Steinfath et al. 1991; Bristow et al. 1986, 1989).

Apart from radioligand binding studies, Western blotting is often used to determine expression levels of βAR. The present study was designed to investigate specific binding of antibodies directed against β1AR and β2AR in different human heart samples. Left ventricular (LV) tissue from patients with IDCM and ISHD was compared with nonfailing donor myocardium. Moreover, specific binding of βAR antibodies was investigated in Chinese hamster ovary (CHO) cells expressing human βARs (Niclauss et al. 2006).

Methods and materials

Human ventricular tissue

LV transmural tissue samples were obtained during heart transplantation surgery from end-stage heart failure patients (NYHA III–IV) with IDCM (n = 21) and ISHD (n = 19). LV tissue from donor hearts (n = 10) served as reference for nonfailing myocardium. Tissue was collected in cardioplegic solution and stored in liquid nitrogen. Samples were obtained after informed consent and with approval of the local Ethical Committee (St Vincent’s Hospital Human Research Ethics Committee: File number: H03/118; Title: Molecular Analysis of Human heart Failure).

Cell culture

Our experiments are based on CHO cells expressing either the human β1AR, β2AR, or β3AR at densities (B max) of 118 ± 28, 202 ± 27, and199 ± 59 fmol/mg proteins as previously described (Niclauss et al. 2006). CHO cells were stably transfected with human βAR subtypes using a pSW104 vector and grown in an atmosphere of 5% CO2/95% air at 37°C in F-12 HAM medium supplemented with 10% heat-inactivated fetal calf serum, 1 mM glutamine, 100 U/ml penicillin, and 0.1 mg/ml streptomycin. Subconfluent cells were subcultured every 3–4 days with a 2.5 g/l trypsin and 0.2 g/l Na4 EDTA solution. To maintain selection pressure, the antibiotic G418 (400 mg/ml) was added to all growing cells in regular intervals but was not present during the experiments. For all experiments, the cells were cultured in the absence of serum for 24 h preceding the experiments to avoid interference of serum factors with cell growth and related signal transduction.

Detection of βARs

Antibodies against β1AR and β2AR

Two different primary rabbit polyclonal antibodies directed against the β1AR and β2AR were used for Western blotting. Both antibodies were from Santa Cruz (β1AR (V-19): catalog: sc-568; β2AR (H-73): catalog: sc-9042). The antibody directed against β1AR was raised against amino acids 420t–470 at the C-terminus of β1AR of mouse origin, while the β2AR antibody was raised against amino acids 338–413 mapping at the C-terminus of the β2AR of human origin.

Western blotting

Protein expression levels of β1AR and β2AR were analyzed by one-dimensional 15% sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (1D-PAGE) and subsequent Western blotting. Cardiac tissue homogenates (in sample buffer containing 15% glycerol, 62.5 mmol/l Tris (pH 6.8), 1% (w/v) SDS, and 2% (w/v) dithiothreitol) were applied in a concentration, which was within the linear range of detection: 20 μg protein per lane for β1AR and β2AR. After 1D separation, proteins were transferred to Hybond-ECL nitrocellulose membranes. Blots were preincubated with 0.5% milk powder in Tween-Tris-buffered saline (TTBS; 10 mmol/l Tris–HCl pH 7.6, 75 mmol/l NaCl, 0.1% Tween) for 1 h at room temperature. The blots were incubated overnight at 4°C with primary rabbit polyclonal antibodies against β1AR (dilution 1:200) and β2AR (dilution 1:200). After washing with TTBS, primary antibody binding was visualized using a secondary horseradish-peroxidase-labeled goat antirabbit antibody (dilution 1:2,000; DakoCytomation) and enhanced chemiluminescence (ECL plus Western blotting detection, Amersham Biosciences). All signals were normalized to actin (dilution 1:1,000; clone KJ43A; Sigma) stained on the same blots.

Results

Detection of βARs in human myocardium

Figure 1a shows Western blots of failing and nonfailing cardiac samples. The antibodies against β1AR and β2AR recognized multiple bands at different molecular weights (Fig. 1a). A similar pattern of protein bands was obtained when we replaced milk by albumin (bovine serum albumin 2%) as blocking agent (not shown). The molecular weight of the detected protein bands was calculated on the basis of the molecular weight marker (Precision Plus Protein Standards Dual Color).

Fig. 1
figure 1

Western blot staining a of failing (F) and nonfailing (NF) cardiac samples separated by 1D-PAGE. Figure a illustrates protein bands (indicated by arrows) detected by antibodies against β1AR and β2AR. Molecular weight of the protein bands was calculated on the basis of the molecular weight marker (mwm). The predicted molecular weight for the βARs is shown in bold. b Western blots incubated with β1AR antibody without blocking peptide (left panel) and with blocking peptide (right panel) of F and NF samples. c Analysis of βARs revealed no significant differences between IDCM, ISHD, and NF donor myocardium

Full size image

The most prominent band visualized upon β1AR labeling was observed at 33 kDa, and two more faint bands of higher molecular weight were found. All of these were no longer visible in the presence of blocking peptide against β1AR (dilution 1:200; sc-568p; Santa Cruz; Fig. 1b).

The β2AR antiserum recognized a large number of bands among which those at 32, 48, and 75 kDa were most prominent. None of the detected protein bands corresponded with the predicted molecular weight of the β2AR, which is ∼65 kDa.

Immunoreactivity levels of different protein bands detected by the βAR antibodies were expressed relative to actin expression. Bar graphs are shown in Fig. 1c for the mean data (sum of all signals obtained with antibodies directed against βAR) obtained in IDCM and ISHD samples in comparison to donor myocardium, which was set to 1. One-way analysis of variance revealed no significant differences in the immunoreactivity of β1AR and β2AR between all groups.

Analysis of βARs in CHO cells

Subsequently, specificity of antibodies directly against β1AR and β2AR was tested in CHO cells expressing β1AR, β2AR, and β3AR (Fig. 2). In addition, failing and nonfailing samples were used as controls. The antibodies directed against β1AR and β2AR all recognized several protein bands with molecular weights ranging between 25 and 120 kDa in CHO cells expressing β1AR, β2AR, and even β3AR, while no band was observed at the expected molecular weight of ∼65 kDa. All protein bands stained with the β1AR antibody disappeared in the presence of the blocking peptide against β1AR (not shown).

Discussion

Most of the antibodies known to interact with βARs described in the literature did not exhibit subtype selectivity (Moxham et al. 1986; Dixon et al. 1986; Kaveri et al. 1987). Moxham et al. (1986) investigated β1ARs and β2ARs in membrane preparations from rat lung and rat heart, respectively, and observed that antibodies directed against β1AR and β2AR interacted with both receptor subtypes. In these previous studies, antibodies were raised against isolated receptor subtypes, while the antibodies used in our study were directed against a peptide sequence, which might increase their selectivity.

Our study shows that antibodies directed at a specific part of the βARs are still not specific and do not allow determination of βAR expression levels since both antibodies detected a range of protein bands at different molecular weights in failing and nonfailing human cardiac samples. It is noteworthy that no bands were observed at the predicted molecular weight of 65 kDa. Several studies have shown that the 65-kDa subunit exists in a dimeric form in the cell membrane, with a molecular size of 109 kDa as determined by immunoaffinity chromatography using monoclonal autoantibodies and SDS-polyacrylamide gels (Boivin et al. 2006; Fraser and Venter 1982). We did not find evidence for a dimeric form in our study using the β1AR and β2AR antibodies.

It has been proposed that the different bands detected upon βAR antibody binding represent different forms of the βARs resulting from complex formation and/or posttranslational modifications such as glycosylation as described by Rybin et al. (2000). Both antibodies against β1AR and β2AR detected a band with a molecular weight of 75 kDa in all samples. Glycosylation may alter molecular weight of βARs by 8 to 11 kDa (George et al. 1986), and hence it is possible that the band detected at 75 kDa is the putative βAR.

Fig. 2
figure 2

Western blots illustrating the binding of antibodies directed against β1AR and β2AR in failing and nonfailing human samples and CHO cells expressing β1AR, β2AR, and β3AR

Full size image

Based on the assumption that all signals represent different forms of βARs, we averaged the sum of all βAR signals for the different cardiac samples. No significant difference was observed in βAR expression level (sum of all βAR signals relative to actin) between failing and nonfailing samples. It should be noted that tissue preparation might have interfered with analysis of βAR expression. In the present study, expression of β1AR and β2AR was determined in whole cardiac homogenates and therefore lacks information about receptor density at the membranes (i.e., functionally available receptors for ligand binding). However, the fact that multiple protein bands were stained with both β1AR and β2AR antibodies in CHO cells expressing βARs, even in cells with the β3AR, indicates that the antibodies are not specific. Moreover, all protein bands stained with the β1AR antibody disappeared in the presence of the blocking peptide against β1AR, both in tissue and CHO cells (not shown), illustrating that the Western blot analysis using antibodies directed against βAR in combination with a blocking peptide is invalid and cannot be used for quantitative analysis of βAR densities.

In conclusion, Western blot analysis cannot be used for quantitative analysis of βAR receptors because they do not show specific binding to their target protein. Further investigations have to be performed to clarify the selectivity and specificity of antibodies against βAR receptor proteins.