Comparison of the Β-Adrenergic Receptor Antagonists Landiolol and Esmolol: Receptor Selectivity, Partial Agonism, and Pharmacoc

Home , Esmolol

1521-0103/359/1/73–81$25.00 http://dx.doi.org/10.1124/jpet.116.232884 THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS J Pharmacol Exp Ther 359:73–81, October 2016 Copyright ª 2016 by The American Society for Pharmacology and Experimental Therapeutics

Comparison of the b-Adrenergic Receptor Antagonists Landiolol and Esmolol: Receptor Selectivity, Partial Agonism, and Pharmacochaperoning Actions

Shahrooz Nasrollahi-Shirazi, Sonja Sucic, Qiong Yang, Michael Freissmuth, and Christian Nanoff Institute of Pharmacology, Center of Physiology and Pharmacology, Medical University of Vienna, Vienna, Austria Received February 15, 2016; accepted July 18, 2016 Downloaded from ABSTRACT

Blockage of b1-adrenergic receptors is one of the most effective was very low. Both landiolol and esmolol caused a very modest treatments in cardiovascular medicine. Esmolol was introduced rise in cAMP levels but a robust increase in the phosphorylation some three decades ago as a short-acting b1-selective antag- of extracellular signal regulated kinases 1 and 2, indicating that the onist. Landiolol is a more recent addition. Here we compared the two drugs exerted partial agonist activity with a signaling bias. If two compounds for their selectivity for b1-adrenergic receptors cells were incubated for $24 hours in the presence of $1 mM jpet.aspetjournals.org over b2-adrenergic receptors, partial agonistic activity, signaling esmolol, the levels of b1-adrenergic—but not of b2-adrenergic— bias, and pharmacochaperoning action by using human embry- receptors increased. This effect was contingent on export of onic kidney (HEK)293 cell lines, which heterologously express the b1-receptor from endoplasmic reticulum and was not seen each human receptor subtype. The affinity of landiolol for b1- in the presence of landiolol. On the basis of these observa- adrenergic receptors and b2-adrenergic receptors was higher tions, we conclude that landiolol offers the advantage of: 1) and lower than that of esmolol, respectively, resulting in an improved selectivity and 2) the absence of pharmacochaperon- improved selectivity (216-fold versus 30-fold). The principal ing activity, which sensitizes cells to rebound effects upon drug metabolite of landiolol (M1) was also b -selective, but its affinity discontinuation.

1 at ASPET Journals on October 2, 2021

Introduction which the affinity is quoted in reviews, (e.g., Plosker, 2013), but for which the original data are inaccessible. b Landiolol is a short-acting -adrenergic receptor antagonist G protein-coupled receptor antagonists have been catego- – that has a half-life (3 4 minutes) shorter than the reference rized traditionally according to receptor selectivity, binding compound esmolol (9 minutes) (Plosker, 2013). Cleavage of an affinity, and the pharmacokinetic properties of the com- ester bond generates the major landiolol metabolite, M1, pounds. Additional possible discriminators that have been which has a half-life (1.8 hours) substantially longer than appreciated more recently are their intrinsic activity and the parent compound (Murakami et al., 2005). During short- signaling bias (Kenakin, 2005) and their ability to act as term infusion, M1 reaches peak concentrations in the range of pharmacochaperones (Tao and Conn, 2014): m 3.24 M (Murakami et al., 2005). Because of its long half-life, 1) G protein-coupled receptors elicit signals not only the concentrations of M1 are predicted to increase further by recruiting their cognate G protein. They can activate upon prolonged infusion and to reach their steady state, with additional signaling pathways in a manner independent of continuous intravenous infusion, after about 7 hours. In fact, heterotrimeric G proteins. Among all receptors known, the b2- the ratio of half-lives predicts that under steady-state condi- adrenergic receptor has been most extensively investigated tions, the concentration of M1 may exceed that of landiolol by with respect to its ability to generate a second wave of several fold. It was therefore of interest to document the intracellular signals: Agonist occupancy triggers phosphor- b b affinity and selectivity of M1 for human 1- and 2-adrenergic ylation of several serine and threonine residues in the receptors and to compare it to that of the parent compound, for C-terminus of the b2-adrenergic receptor by G protein- coupled receptor kinases (Lefkowitz and Shenoy, 2005). This supports binding of b-arrestins, which serve as versatile This work was supported by a grant from AOP Orphan Pharmaceuticals AG, adapters: They recruit the endocytotic machinery and thus Vienna, Austria, which markets esmolol and has a commercial interest in landiolol. support the clathrin-dependent endocytosis of receptors. dx.doi.org/10.1124/jpet.116.232884. In addition, b-arrestins scaffold via their C-terminus the

ABBREVIATIONS: ANOVA, analysis of variance; BSA, bovine serum albumin; COPII, coatomer protein II; ER, endoplasmic reticulum; ERK, extracellular signal-regulated kinase (MAP kinase); FCS, fetal calf serum; GFP, green fluorescent protein; HEK, human embryonic kidney cells; M1, principal landiolol metabolite (3-{4-[(S)-2-hydroxy-3-(2-morpholinocarbonylamino) ethylamino]propoxy}phenylpropionic acid); MAP, mitogen- activated protein; PBS, phosphate-buffered saline; PMA, phorbol-12-myristat-13-acetate; PMSF, phenylmethylsulfonyl fluoride; siRNA, small- interfering RNA.

73 74 Nasrollahi-Shirazi et al. assembly of kinase cascades, most prominently the mitogen- 100, 31, and 33 mM, respectively. Isoproterenol (33 mM) was dissolved activated protein (MAP) kinase cascade (Lefkowitz and She- in 0.1 M HCl. noy, 2005; Shenoy and Lefkowitz, 2011). Receptor ligands Cell Transfections, Cell Cultures, and Cell Membrane b b differ in their ability to stabilize the active conformations of Preparations. Plasmids encoding human 1-and 2-adrenergic the receptor. Full agonists are thought to stabilize all active receptors,whichweretaggedontheirN-terminiwithaFLAG- epitope, were a generous gift of Dr. Mark von Zastrow (University conformations, pure antagonists are thought to trap the of California at San Francisco). Human embryonic kidney receptor in the inactive conformation. However, there is a (HEK)293 cells (a fibroblast cell line) were transfected using the continuum with respect to agonistic or antagonistic activity polycationic TurboFect reagent (Fermentas/Thermo Fisher Scien- and with respect to the conformations, which are achieved tific) as follows: The plasmid encoding the human b1-orthehuman upon binding of individual ligands (Kenakin, 2005). Accord- b2-adrenergic receptor plasmids (3 mg) were diluted with empty m m ingly, in the case of the b2-adrenergic receptor, antagonists carrier plasmid pcDNA3 (7 g) and 20 l of the TurboFect reagent have been found to act as b-arrestin-biased ligands, i.e., they in 1 ml Dulbecco’s modified Eagle’s medium (DMEM; Sigma- Aldrich). This mixture was incubated for 20 minutes at room block the canonical signaling pathway (i.e., Gs-dependent stimulation of adenylyl cyclase), but they support the re- temperature (22°C) to allow for complex formation between the cruitment of b-arrestin and the resulting MAP kinase stimu- DNA and the polycationic TurboFect reagent. Thereafter, it was pipetted dropwise onto the layer of HEK293 cells (80% confluent in lation (Azzi et al., 2003; Whalen et al., 2011). a 10-cm dish). After 24 hours the medium was exchanged and the 2) Continuous exposure to cell-permeable ligands can in- cells propagated in a humidified atmosphere (95% air/5% CO2)at crease the surface levels of a receptor and result in exagger- 37°C in DMEM containing 10% fetal calf serum (FCS) and Downloaded from ated responses to endogenous agonists, if the treatment with 0.7 mg/ml geneticin (G418) for selection of stable transfectants. an antagonist is suddenly stopped. In fact, this was first In some instances, a plasmid (3 mg/2 Â 106 cells) encoding a green observed with the b-adrenergic antagonist propranolol (“pro- fluorescent protein (GFP)-tagged bovine b-arrestin-2 (arrestin-3) pranolol withdrawal rebound”; Alderman et al., 1974; Miller or an empty control plasmid was subsequently introduced into et al., 1975) and linked to an increase in surface receptor levels stably transfected HEK293 cells. (Aarons et al., 1980). This effect is currently thought to reflect For membrane preparation, stably transfected cells were har- – jpet.aspetjournals.org pharmacochaperoning by cell-permeable antagonists, i.e., vested from 15-cm dishes (80 90% confluent) as follows: The dishes were first rinsed with phosphate buffered saline (PBS) and then specific ligands can assist receptor folding in the ER (endo- mechanically detached from the dish with a plastic scraper in 5 ml of plasmic reticulum) by binding to and stabilizing conforma- ice-cold PBS containing 0.1 mM phenylmethylsulfonyl fluoride tional intermediates on the trajectory to the stable low-energy (PMSF). Cells were recovered by centrifugation (400g for 10 minutes state of the mature receptor (Morello et al., 2000; Nanoff and at 4°C). The cell pellet was resuspended in 1 ml hypotonic HME

Freissmuth, 2012). buffer (20 mM HEPES NaOH, pH 7.5, 1 mM EDTA, 2 mM MgCl2) The underlying chemistry behind the development of the containing 0.1 mM PMSF and the Complete Protease Inhibitor short-acting b-blocker landiolol resulted in a large molecular Cocktail (Roche Biochemical Reagents/Sigma-Aldrich) and subse- at ASPET Journals on October 2, 2021 structure. Given the chemical difference with conventional quently frozen in liquid nitrogen. After rapid thawing, the cells were antagonists, it is reasonable to assume that it occupies a larger further homogenized by ultrasonication (Sonifier cell disruptor B15, surface area on the b-receptor than do conventional and 12 pulses of 0.5 second duration at 50% intensity; Branson Ultra- smaller antagonist ligands. Experimental data to show that sonics, Danbury, CT). Membranes were recovered by centrifugation (15 minutes at 34,000g at 4°C) and resuspended in HME buffer. The landiolol—and similarly, esmolol—act as competitive antag- protein concentration (about 5 mg/ml) was determined by dye onists are not publicly available. A priori, it cannot be ruled binding (using the Coomassie Brilliant Blue kit provided by Bio- out that landiolol also occupies an allosteric site and for this Rad Laboratories, Hercules, CA). Membranes were frozen in liquid reason may stabilize a unique type of receptor conformation. nitrogen and stored at –80°C. We therefore explored whether landiolol elicits a biased signal Radioligand Binding. The binding reaction was conducted in a × by binding to the human b1- and b2-adrenergic receptors and final volume of 0.2 ml containing buffer (20 mM HEPES NaOH, pH compared its effects to those of esmolol. We also examined 7.5, 1 mM EDTA, 2 mM MgCl2, 1 mM ascorbic acid), with membranes – – m b b their ability to enhance expression of the receptors at the (0.4 0.7/assay and 0.22 0.4 g/assay for 1- and 2-adrenergic receptor-expressing membranes, respectively), competing ligands surface of cells, which had been subjected to prolonged 125 incubation with each drug. The most conspicuous difference at the concentrations indicated in Fig. 2, and [ I]CYP (in the range of 40 pM for competition experiments; concentrations as indicated in revealed by these experiments was the finding that landiolol b Fig. 1 for saturation experiments). The reaction was started by the was not only more 1-selective than esmolol but also lacked addition of the radioligand. After an incubation for 90 minutes at appreciable pharmacochaperoning activity. 30°C, the reaction was stopped by rapid filtration over glass-fiber filter mats [Skatron FilterMAT 11731 (Molecular Devices, Sunny- Materials and Methods vale, CA)]. The radioactivity trapped was counted by liquid scintil- lation counting at an efficiency of 75% (i.e., 3700 cpm/fmol). Materials. Landiolol and its metabolite M1 were obtained from Nonspecific binding was defined as binding not displaced by Mehta Api Pvt Ltd (Mumbai, India), esmolol from Amomed Pharma 0.1 mM isoproterenol or 5 mM propranolol (which gave equivalent (Vienna, Austria), [125I]CYP (iodo-cyanopindolol, specific activity results). This nonspecific binding was below 20% of total binding at 2200 Ci/mmol) and [3H]adenine were from Perkin Elmer (Shelton, the highest radioligand concentration employed (400 pM) and ,10% CT), (–)isoproterenol and the M2-anti-FLAG antibody were from at the radioligand concentration employed in the competition Sigma-Aldrich (St. Louis, MO). All other reagents were of analytical experiments. grade. Landiolol ((–)-[(S)-2,2-dimethyl-1,3-dioxolan-4-yl]methyl-3-{4- [3H]cAMP Accumulation. Untransfected and transfected [(S)-2-hydroxy-3-(2-morpholino-carbonylamino)ethylamino] propoxy} HEK293 cells (3 Â 105 cells/well) were seeded in 6-well dishes in- phenylpropionate), its principal metabolite M1 (3-{4-[(S)-2-hydroxy- cubated for 16 hours in DMEM containing 0.5 mg/l each of benzylpe- 3-(2-morpholinocarbonylamino)ethylamino]propoxy}phenylpropionic nicillin and streptomycin, and [3H]adenine (1 mCi/ml; PerkinElmer). acid), and esmolol were dissolved in water to yield stock solutions of The cells were then stimulated with the indicated compounds Comparison of Landiolol and Esmolol 75

(e.g., 20 mM forskolin, esmolol, and landiolol) for 20 minutes. There- after, the medium was removed and the pellet was lysed in ice-cold 2.5% perchloric acid (0.9 ml) containing 0.1 mM cAMP, held on ice for 15 minutes, and neutralized with 4.2 M KOH (0.1 ml). ATP and cAMP were separated by sequential chromatography on columns containing Dowex 50-X4 (Sigma-Aldrich) and neutral alumina (Johnson et al., 1994). Phosphorylation of p44/p42 MAP Kinases 1 and 2 (ERK1/2).

HEK293 cells expressing human b1-andb2-adrenergic receptors were seeded into 6-cm dishes in DMEM containing 10% FCS, allowed to adhere for 8 hours, and then rendered quiescent by serum starvation for 16 hours. Cells were subsequently stimulated by addition of medium containing or lacking 1 mM esmolol, 1 mM landiolol, or 1 mM isoproterenol (as a positive control). Cells were maintained at 37°C for up to 40 minutes, the incubation was terminated by rapid rinsing with ice-cold PBS. Thereafter, the dish was immediately immersed in liquid nitrogen. After rapid thawing, cells were disrupted by addition of 80 ml of lysis buffer (in mM: 50 Tris, 40 b-glycerophosphate, 100 NaCl, 10 EDTA, 10 p-nitrophenol

phosphate, 1 PMSF, 1 Na3VO4, 10 NaF, pH adjusted to 7.4 with HCl), Downloaded from 1% Nonidet P-40, 0.1% SDS, 250 IU/ml aprotinin, and 40 mg/ml leupeptin. The cellular debris was removed by centrifugation at 10,000g for 10 minutes, and the total protein content was measured photometrically using bicinchoninic acid (Pierce Micro BCA kit; Thermo Fisher Scientific). Aliquots corresponding to 2.5–5 Â 104 cells (10–30 mg of protein) were dissolved in Laemmli sample

buffer containing 30 mM dithiothreitol and applied to SDS- jpet.aspetjournals.org polyacrylamide gels (monomer concentration 10–15% acrylamide, 0.26–0.4% bisacrylamide). Phosphorylation of p44/p42 MAP kinases 1 and 2 (ERK1/2) was assayed by incubating nitrocellulose blots with an antiserum that recognizes only the dually phosphorylated forms of ERK1/2. The blots were also probed with an antiserum recognizing both the unphosphorylated (inactive) and phosphorylated (active) forms of ERK1/2 to correct for loading. The immunoreactive bands were detected by using fluores- at ASPET Journals on October 2, 2021 cently labeled secondary antibodies (Donkey anti rabbit 680RD and 880RD). The immunoreactive bands were quantified on an Odyssey Clx (LI-COR Biosciences, Lincoln, NE) infrared fluorescent imaging system. Determination of Receptor Levels at the Cell Surface by Flow Cytometry. HEK293 cells were transiently transfected with

plasmids expressing FLAG-tagged human b1-andb2-adrenergic receptors (see above). One day after transfection, the cells were incubated for 16–24 hours with propranolol, esmolol, or landiolol. In some experiments HEK293 cells were first transfected with a combination of small-interfering (si)RNAs directed against the COPII (coatomer protein II) components SEC24A, SEC24B, SEC24C, and SEC24D or control-scrambled siRNAs as described previously (Sucic et al., 2011). Forty-eight hours later, these cells were transfected with the plasmid driving the expression of the

FLAG-tagged human b1-adrenergic receptor and then incubated for 16–24 hours in the absence and presence of esmolol (1 and 10 mM),

concentrations of [125I]CYP for 90 minutes at 30°C. The solid lines were drawn by fitting the data to a rectangular hyperbola. Data are from a representative experiment done in duplicate. A second experiment gave similar results. (B) Experiments were carried out as outlined under Materials and Methods. Specific binding was converted to percentage of control (i.e., specific binding in the absence of any competitor) to normalize for differences in absolute levels (fmol/assay) and thus to allow for Fig. 1. Saturation hyperbola for specific binding of [125I]CYP (A), comparison. Data are means from three independent experiments carried competition by landiolol, its metabolite M1, and esmolol for binding of out in duplicate; error bars (n = 3, means 6 S.E.M.). (C) Experiments were 125 [ I]CYP to human b1- and b2-adrenergic receptors in membranes carried out as outlined under Materials and Methods. Data are means prepared from stably transfected HEK293 cells (B), and landiolol- and from three independent experiments carried out in triplicate; error bars esmolol-induced cAMP accumulation in HEK293 cells, which were represent S.E.M. As a control, cells were also incubated in the presence of transfected to express the human b1-receptor and prestimulated with 1 mM propranolol (prop 1, right hand panel, where the median and the 20 mM forskolin. (A) Membranes (0.22 and 0.4 mg/assay for b1- and b2- interquartile range is indicated; whiskers represent the 90% confidence adrenergic receptors, respectively) were incubated with the indicated interval). 76 Nasrollahi-Shirazi et al. landiolol (1 and 10 mM), and propranolol (1 mM). Alternatively, TABLE 1 stably transfected cells were incubated for 24 hours with pro- Dissociation constants (Ki values) for binding of landiolol, its metabolite b b pranolol, esmolol, or landiolol, at concentrations ranging from M1, and esmolol to human 1-and 2-adrenergic receptors 0.1–100 mmol/l. After the incubation, the cells were washed with IC50 values were calculated from individual competition curves (the averages of PBS and detached using EDTA (0.02%) solution. The following which are shown in Fig. 1B) and corrected for the actual radioligand concentration and the KD of the radioligand for each receptor (see Fig. 1A) by using the Cheng- procedures were performed at 4°C: Cells were collected by centrifu- Prusoff approximation (Cheng and Prusoff, 1973). Data are means 6 S.D. gation (1000 rpm for 5 minutes) and taken up in PBS with 1% bovine Selectivity Ratio ∼ Human b1-Receptor Human b2-Receptor serum albumin (BSA) in a volume of 0.3 ml. Cell-surface receptors (Ki b2/Ki b1) were labeled with the monoclonal M2 anti-FLAG antibody (at a 6 6 m ∼ m m Landiolol 90 16 nM 19.4 5.5 M 216 concentration of 0.3 g/100 l) and a secondary anti-mouse IgG M1 3.8 6 1.2 mM2246 61 mM59 conjugated to the fluorescent dye AlexaFluor 488 (0.25 mg/100 mlof Esmolol 194 6 7 nM 5.8 6 2.1 mM30 cell suspension). Labeling was at 4°C for 20 minutes with the primary and an additional 20 minutes with the secondary antibody. An aliquot of cells was labeled only with secondary antibody. Cells were washed once, resuspended in PBS with 1% BSA, and subjected was also selective for human b1-adrenergic receptors albeit to measuring cell-bound AlexaFluor 488 fluorescence by flow to a lower extent than landiolol (open and closed columns in cytometry. Flow cytometry was performed on a FACSCanto in- Fig. 2 and Table 1). strument (BD Biosciences San Jose, CA). Fluorescence was recorded at 4°C on samples diluted in PBS. For the analysis a cell fraction was Effects of Landiolol and Esmolol on Forskolin- gated with forward scatter values higher than 10% of the scale Stimulated cAMP Accumulation. HEK293 cells express- Downloaded from maximum. This fraction was set off from a fraction of nonviable cells ing FLAG-tagged human b1-andb2-adrenergic receptors (which stained positive for propidium iodide). Ten thousand events were incubated in the presence of [3H]adenine to metaboli- were recorded per gated sample. Receptor-specific fluorescence was cally label their adenine nucleotide pool. The cells were then determined after excluding events that overlapped the range of stimulated with 20 mM forskolin to raise cAMP levels nonspecific fluorescence (assessed on cells labeled with secondary (Klinger et al., 2002). Forskolin sensitizes adenylyl cyclase antibody alone). A range of 4–8000 (out of 10,000) events per sample to the action of activated Gas (Sunahara et al., 1996; Kudlacek jpet.aspetjournals.org represented cells expressing FLAG-tagged receptors on their sur- et al., 2001). Hence, basal receptor activity is magnified in face. Analysis of flow cytometry recordings was performed with the presence of forskolin. Accordingly, it is possible to detect Flowing Software (Dr. Perttu Terho, Turku Centre for Biotechnol- ogy, Turku, Finland). Treatment-dependent change was quanti- inverse agonism in the presence of forskolin, because antag- fied by comparing the histogram distribution of receptor-specific onists with inverse activity lock the receptor in the inactive fluorescence events. conformation and hence produce a decline of cAMP accumu- Data Analysis. Data from saturation experiments and from lation (Klinger et al., 2002). Conversely, in the presence of concentration-response curves were subjected to nonlinear least forskolin, the weak intrinsic activity of antagonists is also squares curve fitting using a Levenberg-Marquardt algorithm to the magnified and detected as an increase in cAMP, which would at ASPET Journals on October 2, 2021 equation for a rectangular hyperbola according to the law of mass not be detected under basal conditions. In fact, in the absence action. Inhibition curves were fitted in a manner similar to the of forskolin, we failed to observe any change in cAMP equation for a monophasic displacement curve. IC50 values (i.e., accumulation (data not shown). However, in the presence of concentrations resulting in 50% inhibition of radioligand binding) forskolin, both esmolol and landiolol modestly increased were converted into Ki values using the Cheng-Prusoff approximation cAMP levels in cells expressing human b1-adrenergic recep- (Cheng and Prusoff, 1973): Ki 5 IC50/(1 1 L*/KD,L), where Ki is the dissociation constant of the inhibitor, L* is the radioligand concentra- tors (Fig. 1C). Concentrations of landiolol $1 mM were tion, and KD,L is the dissociation constant of the radioligand (de- required to observe a significant rise in cAMP. Under the termined from saturation experiments). Statistically significant conditions employed, where 10 mM landiolol and esmolol differences were examined by Student’s t test for paired comparisons elevated cAMP levels by roughly 1.5- and 2-fold, 1 mM and by analysis of variance (ANOVA) or Kruskal-Wallis test followed isoproterenol stimulated cAMP accumulation by 18.2 6 1.9- by a post hoc test for multiple comparisons. fold (means 6 S.D, n 5 3). Taken together, these findings suggests that esmolol and landiolol are not inverse agonists b Results but very weak partial agonists at the human 1-adrenergic receptor. High receptor occupancy by landiolol is required to Competition by Landiolol, Its Metabolite M1, elicit this partial agonism via Gs. and Esmolol for Binding of [125I]CYP to Heterolo- Stimulation by Landiolol and Esmolol of MAP Kinase gously Expressed Human b1-andb2-Adrenergic Phosphorylation. The weak partial agonist activity of land- Receptors. Stably transfected HEK293 cells were selected iolol and esmolol suggested that these compounds induced an that expressed the individual receptors at high density: Bmax active conformation in the human b1- and b2-adrenergic values were 2.7 pmol/mg and 6.9 pmol/mg for b1- and b2- receptors albeit at low probability. This activity may give rise adrenergic receptor-carrying membranes, respectively (Fig. to biased agonism (Azzi et al., 2003; Kenakin, 2005; Whalen 1A). The radioligand [125I]CYP had only a very modest—if any et al., 2011). This possibility was explored by assessing the —selectivity for b2-adrenergic receptors; KD values were 37 6 capacity of esmolol and landiolol to stimulate the phosphor- 8 pM and 30 6 5 pM for b1- and b2-adrenergic receptors, ylation of p42 and p44 MAP kinases (ERK2 and ERK1, respectively (mean 6 S.D.; n 5 2). Landiolol was substantially respectively) in HEK293 cells expressing human b1-orb2- more potent in displacing the radioligand [125I]CYP from hu- adrenergic receptors. man b1-adrenergic receptors than from human b2-adrenergic In cells expressing human b1-adrenergic receptors, esmolol receptors (open and closed circles in Fig. 1B and Table 1). and landiolol caused a time-dependent stimulation of MAP The inhibitory potency of the metabolite M1 was some two kinase phosphorylation (Fig. 2A). The time course showed the orders of magnitude lower (triangles in Fig. 1B). Esmolol typical biphasic response, i.e., an initial peak followed by a Comparison of Landiolol and Esmolol 77

statistically significant increase in MAP kinase phosphorylation (not shown). The peak activation, which was observed in HEK293 cells expressing the human b1-adrenergic receptor stimulated with esmolol and landiolol (Fig. 2A), was sufficiently large in magnitude to allow for determination of the concentration- response relation: Both compounds stimulated MAP kinase phosphorylation over a comparable concentration range with a half-maximum effect seen at about 100 nM (Fig. 3). The effect of esmolol and landiolol was lower than the effect elicited by isoproterenol (right hand lane and column in Fig. 2A). This is consistent with partial agonism of the two compounds at the G protein-independent, b-arrestin- dependent pathway of the b1-adrenergic receptor (Azzi et al., 2003; Whalen et al., 2011). If this was the case, raising the cellular levels of b-arrestin ought to enhance the ability of esmolol and landiolol to stimulate phosphorylation of ERK1/2 b via the 1-adrenergic receptor. This prediction was verified: Downloaded from We compared the response of control cells and cells overexpressing b-arrestin-2 (5 arrestin-3) to esmolol, landiolol, and isoproterenol (right and left hand panel in Fig. 4A) and used the effect elicited by the phorbol ester PMA as an internal reference: The rationale for this approach is the fact that protein kinase C-dependent activation of the MAP kinase jpet.aspetjournals.org at ASPET Journals on October 2, 2021

Fig. 2. Time-dependent phosphorylation of ERK1/2 (p42 and p44 MAP kinases) in HEK293 cells stably expressing the human b1-adrenergic (A) and the b2-adrenergic receptor (B). HEK293 cells were seeded into six-well dishes, allowed to adhere in the presence of 10% FCS for 8 hours and incubated in medium containing 0.5% FCS for 16 hours. Thereafter, the medium was exchanged for medium containing 1 mM (A) and 10 mM(B) esmolol or 1 mM (A), and 10 mM landiolol (B) for up to 40 minutes. Stimulation with 1 mM isoproterenol was done for 5 minutes and was used as a positive control to allow for comparison between blots. Bar diagrams: The immunoblots were probed with an antibody recognizing the dually phosphorylated active form of p42 (ERK2) and p44 (ERK1) MAP kinases (upper blot) and with an antibody directed against all forms of the proteins (lower blot). Phosphorylation was quantified by correcting the recorded phospho-specific intensity for loading (as shown in the respective immunoblots) and plotted to visualize the time course. Bars show means 6 S.E.M. from three independent experiments. In (A), the stimulation at 5 minutes differs from the control value in a statistically significant manner (P , 0.001, repeated-measures ANOVA followed by Bonferroni multiple comparisons). Fig. 3. Concentration-response curve for esmolol- and landiolol-induced phosphorylation of ERK1/2 (p42 and p44 MAP kinases) in HEK293 cells b sustained phase. In contrast, even when tested at 10 mM, stably expressing the human 1-adrenergic receptor. (A) Assay conditions were as outlined in the legend to Fig. 3. Cells were incubated for 5 minutes neither esmolol nor landiolol caused any appreciable stimu- in the absence and presence of esmolol or landiolol at the indicated lation of MAP kinase phosphorylation in HEK293 cells concentration. Stimulation with 1 mM isoproterenol (iso) was used as a positive control to allow for comparison between blots. Data are expressing the human b2-adrenergic receptor (Fig. 2B). In- representative for two additional independent experiments. (B) The levels cidentally, this experiment also showed that in HEK293 cells of phospho-MAP kinase were quantified as outlined in the legend to Fig. 2 expressing b1-adrenergic receptors, esmolol and landiolol and plotted as a concentration-response curve. Data are means 6 S.E.M. stimulated MAP kinase phosphorylation via their cognate Repeated measures ANOVA followed by Holm-Sidak multiple compari- , receptor rather than via an off-target effect. This was also sons verified that there was a statistically significant difference (P 0.02) between the phospho-ERK1/2 levels seen at saturating concentrations and independently verified by challenging untransfected HEK293 that observed in the ascending limb (i.e., 0.01, 0.3 and 0.1 mM) of the with 1 mM esmolol or 1 mM landiolol, which did not result in a concentration-response curve. 78 Nasrollahi-Shirazi et al.

was detected by monitoring the fluorescence resulting from binding of a monoclonal antibody to the N-terminal FLAG- epitope (middle histogram in Fig. 5A; see also box plot in Fig. 5B). The effect was comparable in magnitude to that elicited by propranolol (Fig. 5A, left hand histogram, and Fig. 5B). In contrast, landiolol was ineffective at 1 mM, and even at 10 mM the effect was so modest (Fig. 5A) that it did not reach the threshold of statistical significance (Fig. 5B). Esmolol and landiolol were also tested at 0.1 mM, which did notproduceanystatisticallysignificanteffectonthesurface levels of the b1-adrenergic receptor (not shown). We also compared the effect of esmolol and of landiolol on the cell surface levels of the human b2-adrenergic receptor to that of propranolol: Cells exposed to 1 mM propranolol for 24 hours showed a substantial increase in b2-adrenergic receptors on their surface (left hand histogram in Fig. 6A) that was statistically significant (box plot in Fig. 6B). In contrast, m m

neither an incubation in 10 Mesmololnorin10 M Downloaded from landiolol resulted in an appreciable elevation of surface receptor levels (Fig. 6, A and B). jpet.aspetjournals.org

Fig. 4. Overexpression of b-arrestin-2 enhances ERK1/2 activation by esmolol, landiolol, and isoproterenol in human b1-adrenergic receptor- containing HEK293 cells. (A) HEK293 cells stably expressing human b1-adrenergic receptors were transfected with a plasmid encoding GFP- tagged b-arrestin-2 (right hand panel) or maintained as controls (left hand panel). After 20 hours, the cells were exposed to fresh medium (ctr) or stimulated with medium containing 1 mM landiolol (lan), 1 mM esmolol (esmo), 1 mM isoproterenol (iso), or 1 mMPMAfor5minutes.The at ASPET Journals on October 2, 2021 immunoblots were probed with an antibody recognizing the dually phosphorylated active form of p42 (ERK2) and p44 (ERK1) MAP kinases (blot phosphoMAP kinase), with an antibody directed against all forms of the proteins (blot MAP kinase, loading control), an antibody against GFP to verify the expression of b-arrestin-2 (blot GFPb-arr2), and an antibody against the FLAG-epitope to visualize the expression of the b1-adrenergic receptor (blot FLAG b1-AR). Data are representative for three additional experiments. (B) Phosphorylation was quantified by correcting the recorded phospho-specific intensity for loading and normalized to the PMA-induced response (= 100 arbitrary units) in four independent experiments (means 6 S.E.M.). The stimulations induced by landiolol, esmolol, and isoproterenol in the control cells (open bars) and in the cells overexpressing b-arrestin-2 (closed bars) were compared for statistically significant differences by a paired t test (*P , 0.05).

cascade occurs at the level of or downstream from RAF and is therefore independent of b-arrestin (Schönwasser et al., 1998). As can be seen from Fig. 4B, the overexpression of b b-arrestin-2 enhanced the response to the b-adrenergic Fig. 5. Quantification of 1-adrenergic receptor surface levels by flow cytometry using an anti-FLAG antibody. (A) HEK293 cells stably ligands in a statistically significant manner. Overexpression b b b expressing 1-adrenergic receptors carrying a FLAG-epitope tag on the of -arrestin-2 did not affect the levels of the 1-adrenergic (extracellular) N-terminus were incubated in the absence (control, gray receptor (Fig. 4A). histograms) and presence (black histograms) of esmolol, landiolol, and Change in Receptor Surface Levels after Long-Term propranolol for 24 hours. The blue area corresponds to the histogram for nonspecific fluorescence observed after incubation of the nontreated cells Incubation of Cells in the Presence of Esmolol, Land- with the secondary antibody alone (no anti-FLAG antibody). The level of iolol, and Propranolol. Prolonged incubation of cells with receptors at the cell surface were quantified by detecting the FLAG- antagonists can increase the levels of their cognate receptors epitope as outlined under Materials and Methods. Data are representative for seven additional experiments. (B) Median fluorescence intensity was by binding to and stabilizing folding intermediates in the ER; quantified from the histograms of eight independent experiments carried this action is referred to as pharmacological chaperoning, out as outlined in (A). The box plot shows the median values and the pharmacochaperoning (Morello et al., 2000; Leidenheimer and interquartile range; whiskers indicate the maximum and minimum Ryder, 2014), or pharmacoperoning (Tao and Conn, 2014). If observed values. Statistically significant differences between drug-treated m and untreated control cells were verified by a Kruskal-Wallis test followed cells were incubated for 24 hours in the presence of 1 M esmolol, by Dunn’s multiple comparison (*P , 0.05 for 1 and 10 mM esmolol and an increase in surface levels of the human b1-adrenergic receptor 1 mM propranolol versus control). Comparison of Landiolol and Esmolol 79

receptor. More importantly, under these conditions, neither preincubation with esmolol nor with propranolol increased the cell surface levels of the human b1-adrenergic receptor (Fig. 7, A and B).

Discussion Of the more than 800 human G protein-coupled receptors, about 70 are targets for currently approved drugs (Gruber et al., 2010). Among these, the b1- and b2-adrenergic receptors are the ones with the richest pharmacology: There is a long list of ligands that have been explored for their agonistic and antagonistic properties. A surprisingly small number of antagonists are useful for intravenous administration. Be- cause of their short half-lives, esmolol and landiolol are of particular interest in the prevention and management of cardiac arrhythmias. Our comparison revealed the follow-

ing salient differences: Downloaded from jpet.aspetjournals.org

Fig. 6. Quantification of b2-adrenergic-receptor surface levels by flow cytometry using an anti-FLAG antibody. (A) HEK293 cells expressing stably b2-adrenergic receptors carrying a FLAG-epitope tag on the (extracellular) N-terminus were incubated in the absence (control, gray histograms) and presence (black histograms) of esmolol, landiolol and propranolol for 24 hours at the indicated concentrations. The blue area corresponds to the histogram for nonspecific fluorescence observed after at ASPET Journals on October 2, 2021 incubation of the nontreated cells with the secondary antibody alone (no anti-Flag antibody). The level of receptors at the cell surface were quantified by detecting the FLAG-epitope as outlined under Materials and Methods. Data are representative for seven additional experiments. (B) Median fluorescence intensity was quantified from the histograms of eight independent experiments carried out as outlined in (A). The box plot shows the median values and the interquartile range, and whiskers indicate the observed maximum and minimum values. Statistically significant differences between drug-treated and untreated control cells were verified by a Kruskal-Wallis test followed by Dunn’s multiple comparison (*P , 0.01 for propranolol 1 mM versus control).

Pharmacochaperones exert their actions on folding intermediates. During their synthesis, like all other integral Fig. 7. The pharmacochaperoning effects of b-adrenergic antagonists is membrane proteins, G protein-coupled receptors are in- absent in HEK293 cells pretreated with siRNAs directehd against serted into the ER membrane and they must reach their SEC24A–D. (A) HEK293 cells were transfected with scrambled control siRNAs or the mixture of siRNAs directed against all SEC24 isoforms folded state prior to recruiting the COPII machinery to (SEC24A–D). After 48 hours, the cells were transfected with a plasmid exit the ER (Nanoff and Freissmuth, 2012). Export of in- encoding the human b1-adrenergic receptors carrying a FLAG-epitope tag tegral membrane proteins from the ER is contingent on on the (extracellular) N-terminus and incubated in the absence and their incorporation into the nascent COPII-coated vesicles; presence of esmolol, landiolol, and propranolol for 24 hours at the indicated concentrations. Shown are representative original histograms this is accomplished via the interaction with the SEC24- for the quantification of b1-adrenergic receptor surface levels by flow component, which acts as the cargo receptor (Gillon et al., cytometry using an anti-FLAG antibody (as in Figs. 5 and 6). The blue 2012; Zanetti et al., 2011). In mammalian cells, there are four area corresponds to the histogram for nonspecific fluorescence observed after incubation of the nontreated cells with the secondary antibody alone isoforms of SEC24 (termed SEC24A, SEC24B, SEC24C, and (no anti-FLAG antibody). The gray line indicates the quantification of the SEC24D). The isoform required for ER export of the human FLAG-specific immunoreactivity in cells transfected with the mixture of siRNAs directed against all SEC24 isoforms (SEC24A–D). The black lines b1-adrenergic receptor is not known. Accordingly, HEK293 correspond to the FLAG-specific immunoreactivity in cells transfected cells were depleted of all isoforms by a combination of siRNAs with the mixture of scrambled siRNAs. (B) The data from (A) was that targeted SEC24A-D. This was previously shown to quantified by measuring the median fluorescence intensity of the specific reduce their cellular levels by 80–90% and to decrease ER area under the curve and plotted to visualize the effect of landiolol, b export of membrane proteins (Sucic et al., 2011, 2013). As esmolol, and propranolol on HEK293 cells expressing human 1-adrenergic – receptors transfected with mixture of siRNAs directed against all SEC24 expected, this siRNA-mediated knockdown of SEC24A D isoforms (SEC24A–D). Bars show means 6S.E.M. from three independent also reduced the surface levels of the human b1-adrenergic experiments. 80 Nasrollahi-Shirazi et al.

The affinity of landiolol for human b1-adrenergic receptors folding is assisted by luminal and cytosolic proteinaceous exceeded that of esmolol by about 2-fold. Esmolol had a chaperones. These engage folding intermediates and lower the substantially higher affinity (i.e., more than 3-fold higher) energy barrier, which separates individual conformational for human b2-adrenergic receptors than landiolol. Accord- states of the folding trajectory (Nanoff and Freissmuth, 2012). ingly, the selectivity of landiolol for human b1-adrenergic Their action can be phenocopied by small molecules, which act receptors is superior to that of esmolol, i.e., 216- versus 30-fold, as chemical chaperones if they promote folding in a general which translates into a roughly 7-fold increase in selectivity. manner [e.g., dimethylsulfoxide], or as pharmacochaperones, Hydrolysis of both esmolol and landiolol generates an active if they promote folding of their cognate target (Morello et al., metabolite. The affinity of the esmolol acid metabolite (gen- 2000; Leidenheimer and Ryder, 2014; Tao and Conn, 2014). In erated by releasing methanol) was reported to have a 400-fold fact, pharmacochaperoning of G protein-coupled receptors was lower affinity than the parent compound and to lack selectiv- shown to occur in the endoplasmic reticulum (Málaga-Diéguez ity for b1- versus b2-adrenergic receptors (Jahn et al., 1995). et al., 2010). We verified that an intact ER export machinery Here we show that the landiolol metabolite M1 still discrim- was required to observe the increase in surface expression of inated between human b1- and b2-adrenergic receptors with a b1-adrenergic receptors, which was induced by esmolol and selectivity ratio that exceeds that of esmolol. However, and propranolol: Export of proteins from the ER relies on the more importantly, the affinity of the M1 metabolite for b1- COPII coat, where SEC24 acts as the cargo receptor (Zanetti adrenergic receptors was substantially lower than that of the et al., 2011; Gillon et al., 2012). The combined knockdown of all

parent compound landiolol (by about 42-fold). the SEC24 isoforms abolished the ability of esmolol and of Downloaded from Esmolol and landiolol are not inverse agonists, but are propranolol to raise the surface levels of b1-adrenergic recep- very weak partial agonists at b1-receptors, when examined tors. This observation proves that the pharmacochaperoning for their ability to induce cAMP accumulation. Landiolol is action of esmolol and propranolol was contingent on ER more efficacious. However, the differences between the two export. compounds are modest and presumably of little clinical Esmolol and landiolol differ in their ability to act as significance: Even in cells with heterologous overexpression pharmacochaperones. This difference can be rationalized as jpet.aspetjournals.org of b1-receptors, partial agonism of esmolol and landiolol was follows: to act as a pharmacochaperone, a ligand must be able only detectable in the presence of a sensitizing concentration to permeate cell membranes. A commonly used metric to of forskolin. When tested, neither compound had any detect- predict membrane permeation is the polar surface area of the able intrinsic effect on basal cAMP accumulation. In contrast, molecule, e.g., a polar surface area .90 Å2 correlates with esmolol and landiolol are robust partial agonists in the weak permeability (Hitchcock and Pennington, 2006). The noncanonical signaling pathway, which links the b1-adrenergic polar surface areas of landiolol and esmolol are 128 and receptor to stimulation of ERK1/2 (Azzi et al., 2003; Whalen 67.8 Å2, respectively, and predict that esmolol distributes into et al., 2011): Esmolol and landiolol elicited about 40% of the the cell interior sufficiently to make it an effective pharma- at ASPET Journals on October 2, 2021 response observed in the presence of isoproterenol (Fig. 4). cochaperone, whereas landiolol is unlikely to reach effective This suggests that esmolol and landiolol have a signaling intracellular concentrations. The volume of distribution— bias at b1-adrenergic receptors: In fact, landiolol stimulated assessed in human subjects—reflects the difference in ERK1/2 phosphorylation at lower b1-receptor occupancy than permeation. Its size was found to be more than an order of it caused an accumulation of cAMP in the presence of forskolin magnitude larger for esmolol (VD5 3.43 l/kg; Sum et al., 1983) (Figs. 3 and 1). Depending on the cell type, elevations of than for landiolol (VD5 0.16 – 0.24 l/kg; Murakami et al., 2005; cAMP can result in both inhibition and stimulation of Plosker, 2013). ERK1/2 phosphorylation (Stork and Schmitt, 2002). How- It is worth pointing out that this difference in pharma- ever, in HEK293cells, cAMP generated in response to Gs- cochaperoning activity is probably of clinical relevance: coupled receptors does not affect the MAP kinase cascade Continuous infusion of therapeutically relevant doses of (Daaka et al., 1997; Seidel et al., 1999). In addition, esmolol landiolol results in concentrations in plasma of about 1–4 mM and landiolol did not affect cAMP levels in the absence of (Murakami et al., 2005; Plosker, 2013). Given its small forskolin. Thus, changes in cAMP levels probably do not apparent volume of distribution, landiolol is not enriched confound the response of ERK1/2 to esmolol and landiolol. It within the intracellular compartment. Landiolol is thus un- is currently difficult to interpret the relevance of this biased likely to have any pharmacochaperoning action in vivo be- partial agonism of landiolol and esmolol in particular because cause at concentrations up 10 mM it failed to increase surface in the heart, b1-adrenergic receptors are weakly susceptible to levels of human b1-adrenergic receptors. In contrast, approved partial agonist activity, which is related to receptor baseline doses of esmolol lead to steady-state plasma levels of esmolol downregulation (Nanoff et al., 1989, 1990). Although currently in the range of 1.2–2.4 mg/ml (i.e., 4–8 mM, Menkhaus et al., considered of major interest for drug development (Azzi et al., 1985). As the volume of distribution of esmolol is about six 2003; Kenakin, 2005; Whalen et al., 2011), the clinical signif- times larger than total body water, intracellular concentration icance of biased agonism is poorly understood. It is however of esmolol must be higher than those in plasma. Within this clear that activation of ERK is cardioprotective during clinically relevant concentration range (i.e., 1–10 mM), esmolol ischemia (Lips et al., 2004; Rose et al., 2010) and contributes was found to be an effective pharmacochaperone. to the cardioprotective actions of b-adrenergic antagonists The current observations predict that: 1) prolonged infusion (Kovacs et al., 2009). of esmolol may be associated with a loss in b1-receptor Esmolol acts as a pharmacochaperone: Long term exposure blocking activity and an exaggerated response to endogenous of cells to esmolol raised the surface levels of b1-adrenergic catecholamines upon cessation of therapy, but 2) that there receptors. This action was not shared by landiolol. Integral should not be any loss of efficacy resulting from continu- membrane proteins fold in the endoplasmic reticulum. Their ous administration of landiolol. In fact, it has been noted Comparison of Landiolol and Esmolol 81 previously that, after prolonged infusion of esmolol (at indi- Kovacs K, Hanto K, Bognar Z, Tapodi A, Bognar E, Kiss GN, Szabo A, Rappai G, Kiss T, Sumegi B, et al. (2009) Prevalent role of Akt and ERK activation in car- vidual doses set to achieve a target heart rate), its effect faded dioprotective effect of Ca(21) channel- and b-adrenergic receptor blockers. Mol Cell gradually with time (to a pronounced extent after more than Biochem 321:155–164. Kudlacek O, Mitterauer T, Nanoff C, Hohenegger M, Tang WJ, Freissmuth M, 48 hours) and that withdrawal of esmolol resulted in an and Kleuss C (2001) Inhibition of adenylyl and guanylyl cyclase isoforms by the increase in heart rate (. 33% over baseline) in the majority of antiviral drug foscarnet. J Biol Chem 276:3010–3016. Leidenheimer NJ and Ryder KG (2014) Pharmacological chaperoning: a primer on patients (Harwood et al., 1999). The time course of diminish- mechanism and pharmacology. Pharmacol Res 83:10–19. ing beta-blocking activity is consistent with an enhanced Lefkowitz RJ and Shenoy SK (2005) Transduction of receptor signals by b-arrestins. b Science 308:512–517. sensitivity of the myocardium to 1-adrenergic receptor Lips DJ, Bueno OF, Wilkins BJ, Purcell NH, Kaiser RA, Lorenz JN, Voisin L, Saba- stimulation (Nanoff and Schütz, 1991) and the effect of El-Leil MK, Meloche S, Pouysségur J, et al. (2004) MEK1-ERK2 signaling pathway stopping the infusion reminiscent of the propranolol with- protects myocardium from ischemic injury in vivo. Circulation 109:1938–1941. Málaga-Diéguez L, Yang Q, Bauer J, Pankevych H, Freissmuth M, and Nanoff C drawal syndrome (Alderman et al., 1974; Miller et al., 1975; (2010) Pharmacochaperoning of the A1 adenosine receptor is contingent on the Aarons et al., 1980). Hence on the basis of our observations, we endoplasmic reticulum. Mol Pharmacol 77:940–952. Menkhaus PG, Reves JG, Kissin I, Alvis JM, Govier AV, Samuelson PN, Lell WA, conclude that landiolol has three potentially interesting Henling CE, and Bradley E (1985) Cardiovascular effects of esmolol in anes- thetized humans. Anesth Analg 64:327–334. features: it is a biased antagonist, it is highly b1-selective, Miller RR, Olson HG, Amsterdam EA, and Mason DT (1975) Propranolol-withdrawal and it lacks pharmacochaperoning activity. The latter two rebound phenomenon. Exacerbation of coronary events after abrupt cessation of features distinguish it from esmolol. antianginal therapy. N Engl J Med 293:416–418. Morello JP, Petäjä-Repo UE, Bichet DG, and Bouvier M (2000) Pharmacological chaperones: a new twist on receptor folding. Trends Pharmacol Sci 21:466–469. Acknowledgments Murakami M, Furuie H, Matsuguma K, Wanibuchi A, Kikawa S, and Irie S (2005) Downloaded from The authors thank Mark von Zastrow for generously providing the Pharmacokinetics and pharmacodynamics of landiolol hydrochloride, an ultra b short-acting b1-selective blocker, in a dose escalation regimen in healthy male plasmids encoding FLAG-tagged human -adrenergic receptors. volunteers. Drug Metab Pharmacokinet 20:337–344. Nanoff C and Freissmuth M (2012) ER-bound steps in the biosynthesis of G protein- Authorship Contributions coupled receptors. Subcell Biochem 63:1–21. Nanoff C, Freissmuth M, Tuisl E, and Schütz W (1989) A different desensitization Participated in research design: Nasrollahi-Shirazi,Sucic,Freissmuth, pattern of cardiac beta-adrenoceptor subtypes by prolonged in vivo infusion of Nanoff. isoprenaline. J Cardiovasc Pharmacol 13:198–203. Nanoff C and Schütz W (1991) Characterization of the beta-adrenoceptor blocking

Conducted experiments: Nasrollahi-Shirazi, Yang, Freissmuth, jpet.aspetjournals.org property of diprafenone in rats: stereoselective interaction, subtype specificity, and Nanoff. sensitization. J Cardiovasc Pharmacol 18:837–842. Performed data analysis: Nasrollahi-Shirazi, Sucic, Freissmuth, Nanoff C, Ströher M, Haschkowitz H, Schütz W, and Pittner H (1990) Desensitization Nanoff. pattern of cardiac beta-adrenoceptor subtypes by prolonged in vivo infusion of pindolol and celiprolol in rats. Basic Res Cardiol 85:88–95. Wrote or contributed to the writing of the manuscript: Nasrollahi- Plosker GL (2013) Landiolol: a review of its use in intraoperative and postoperative Shirazi, Freissmuth, Nanoff. tachyarrhythmias. Drugs 73:959–977. Rose BA, Force T, and Wang Y (2010) Mitogen-activated protein kinase signaling in References the heart: angels versus demons in a heart-breaking tale. Physiol Rev 90: – Aarons RD, Nies AS, Gal J, Hegstrand LR, and Molinoff PB (1980) Elevation of 1507 1546. b-adrenergic receptor density in human lymphocytes after propranolol adminis- Schönwasser DC, Marais RM, Marshall CJ, and Parker PJ (1998) Activation of the at ASPET Journals on October 2, 2021 tration. J Clin Invest 65:949–957. mitogen-activated protein kinase/extracellular signal-regulated kinase pathway by Alderman EL, Coltart DJ, Wettach GE, and Harrison DC (1974) Coronary artery conventional, novel, and atypical protein kinase C isotypes. Mol Cell Biol 18: – syndromes after sudden propranolol withdrawal. Ann Intern Med 81:625–627. 790 798. Azzi M, Charest PG, Angers S, Rousseau G, Kohout T, Bouvier M, and Piñeyro G Seidel MG, Klinger M, Freissmuth M, and Höller C (1999) Activation of mitogen- (2003) b-arrestin-mediated activation of MAPK by inverse agonists reveals distinct activated protein kinase by the A(2A)-adenosine receptor via a rap1-dependent and ras – active conformations for G protein-coupled receptors. Proc Natl Acad Sci USA 100: via a p21( )-dependent pathway. J Biol Chem 274:25833 25841. b 11406–11411. Shenoy SK and Lefkowitz RJ (2011) -Arrestin-mediated receptor trafficking and – Cheng Y and Prusoff WH (1973) Relationship between the inhibition constant (K1) signal transduction. Trends Pharmacol Sci 32:521 533. and the concentration of inhibitor which causes 50 per cent inhibition (I50) of an Sum CY, Yacobi A, Kartzinel R, Stampfli H, Davis CS, and Lai CM (1983) Kinetics of enzymatic reaction. Biochem Pharmacol 22:3099–3108. esmolol, an ultra-short-acting beta blocker, and of its major metabolite. Clin Pharmacol Ther 34:427–434. Daaka Y, Luttrell LM, and Lefkowitz RJ (1997) Switching of the coupling of the b2- adrenergic receptor to different G proteins by protein kinase A. Nature 390:88–91. Stork PJ and Schmitt JM (2002) Crosstalk between cAMP and MAP kinase signaling – Gillon AD, Latham CF, and Miller EA (2012) Vesicle-mediated ER export of proteins in the regulation of cell proliferation. Trends Cell Biol 12:258 266. and lipids. Biochim Biophys Acta 1821:1040–1049. Sucic S, El-Kasaby A, Kudlacek O, Sarker S, Sitte HH, Marin P, and Freissmuth M Gruber CW, Muttenthaler M, and Freissmuth M (2010) Ligand-based peptide design (2011) The serotonin transporter is an exclusive client of the coat protein complex – and combinatorial peptide libraries to target G protein-coupled receptors. Curr II (COPII) component SEC24C. J Biol Chem 286:16482 16490. Pharm Des 16:3071–3088. Sucic S, Koban F, El-Kasaby A, Kudlacek O, Stockner T, Sitte HH, and Freissmuth M Harwood TN, Butterworth J, Prielipp RC, Royster RL, Hansen K, Plonk G, and Dean R (2013) Switching the clientele: a lysine residing in the C terminus of the serotonin (1999) The safety and effectiveness of esmolol in the perioperative period in patients transporter specifies its preference for the coat protein complex II component – undergoing abdominal aortic surgery. J Cardiothorac Vasc Anesth 13:555–561. SEC24C. J Biol Chem 288:5330 5341. Hitchcock SA and Pennington LD (2006) Structure-brain exposure relationships. Sunahara RK, Dessauer CW, and Gilman AG (1996) Complexity and diversity of – J Med Chem 49:7559–7583. mammalian adenylyl cyclases. Annu Rev Pharmacol Toxicol 36:461 480. Jahn P, Eckrich B, Schneidrowski B, Volz-Zang C, Schulte B, Mutschler E, and Palm Tao YX and Conn PM (2014) Chaperoning G protein-coupled receptors: from cell biology to therapeutics. Endocr Rev 35:602–647. D (1995) b 1-adrenoceptor subtype selective antagonism of esmolol and its major metabolite in vitro and in man. Investigations using tricresylphosphate as red Whalen EJ, Rajagopal S, and Lefkowitz RJ (2011) Therapeutic potential of b – blood cell carboxylesterase inhibitor. Arzneimittelforschung 45:536–541. -arrestin- and G protein-biased agonists. Trends Mol Med 17:126 139. Johnson RA, Alvarez R, and Salomon Y (1994) Determination of adenylyl cyclase Zanetti G, Pahuja KB, Studer S, Shim S, and Schekman R (2011) COPII and the – catalytic activity using single and double column procedures. Methods Enzymol regulation of protein sorting in mammals. Nat Cell Biol 14:20 28. 238:31–56. Kenakin T (2005) New concepts in drug discovery: collateral efficacy and permissive antagonism. Nat Rev Drug Discov 4:919–927. Address correspondence to: Dr. Michael Freissmuth, Medical University of Klinger M, Kuhn M, Just H, Stefan E, Palmer T, Freissmuth M, and Nanoff C (2002) Vienna, Institute of Pharmacology, Center of Physiology and Pharmacology, Removal of the carboxy terminus of the A2A-adenosine receptor blunts constitutive Waehringer Straße 13A, A-1090 Vienna, Austria. E-mail: michael.freissmuth@ activity: differential effect on cAMP accumulation and MAP kinase stimulation. meduniwien.ac.at Naunyn Schmiedebergs Arch Pharmacol 366:287–298.