Identification of Key Residues in Transmembrane 4 Responsible for the Secondary, Low-Affinity Conformation of the Human β₁-Adrenoceptor

Abstract

The β₁-adrenoceptor exists in two agonist conformations/states: 1) a high-affinity state where responses to catecholamines and other agonists (e.g., cimaterol) are potently inhibited by β₁-adrenoceptor antagonists, and 2) a low-affinity secondary conformation where agonist responses, particularly CGP12177 [(−)-4-(3-tert-butylamino-2-hydroxypropoxy)-benzimidazol-2-one] are relatively resistant to inhibition by β₁-adrenoceptor antagonists. Although both states have been demonstrated in many species (including human), the precise nature of the secondary state is unknown and does not occur in the closely related β₂-adrenoceptor. Here, using site-directed mutagenesis and functional measurements of production of a cyclic AMP response element upstream of a secreted placental alkaline phosphatase reporter gene and accumulation of ³H-cAMP, we examined the pharmacological consequences of swapping transmembrane (TM) regions of the human β₁- and β₂-adrenoceptors, followed by single point mutations, to determine the key residues involved in the β₁-adrenoceptor secondary conformation. We found that TM4 (particularly amino acids L195 and W199) had a major role in the generation of the secondary β₁-adrenoceptor conformation. Thus, unlike at the human β₁-wild-type adrenoceptor, at β₁-TM4 mutant receptors, cimaterol and CGP12177 responses were both potently inhibited by antagonists. CGP12177 acted as a simple partial agonist with similar K_B and EC₅₀ values in the β₁-TM4 but not β₁-wild-type receptors. Furthermore pindolol switched from a biphasic concentration response at human β₁-wild-type adrenoceptors to a monophasic concentration response in the β₁-TM4 mutant receptors. Mutation of these amino acids to those found in the β₂-adrenoceptor (L195Q and W199Y), or mutation of a single residue (W199D) in the human β₁-adrenoceptor thus abolished this secondary conformation and created a β₁-adrenoceptor with only one high-affinity agonist conformation.

Introduction

The β₁-adrenoceptor exists in two agonist conformations or states. Catecholamines activate the primary, high-affinity, catecholamine conformation, and these responses are readily inhibited by β-blockers (β-adrenoceptor antagonists). Much higher concentrations of β-blockers are required to inhibit secondary conformation agonist responses (e.g., response to CGP12177 [(−)-4-(3-tert-butylamino-2-hydroxypropoxy)-benzimidazol-2-one]), and this conformation is therefore termed the secondary, low-affinity, CGP12177 conformation (Pak and Fishman, 1996; Granneman, 2001; Molenaar, 2003; Arch, 2004).

Conventional partial agonists need to saturate receptors to stimulate their maximum response and, thus, the concentrations required for half receptor occupancy (K_B) should normally be equal to the concentration required to stimulate half maximum response (EC₅₀) value. However, early observations found that several β-blockers with partial agonist activity did not conform to this. The concentration of pindolol, for example, needed for the cardiostimulant effects in feline heart was 10 times greater than that needed to antagonize responses of the full agonist isoprenaline and thus was labeled as a nonconventional partial agonist (Kaumann, 1973; Kaumann and Blinks, 1980; Kaumann and Molenaar, 2008). CGP12177 (Staehelin et al., 1983) was also found to be a nonconventional partial agonist, with even larger differences between K_B and EC₅₀ (Kaumann, 1983), and, therefore, it has become an important tool in understanding β₁-adrenoceptor pharmacology (Kaumann and Molenaar, 2008). In 1996, Pak and Fishman (1996), using transfected cells, showed that CGP12177 inhibited isoprenaline responses at low concentrations (i.e., with high affinity), but required 100 times higher concentrations to stimulate responses. Furthermore, CGP12177 agonist responses required higher concentrations of conventional β-blockers to antagonize them. They therefore proposed a high-affinity and a low-affinity state or conformation of the β₁-adrenoceptor, and this terminology has now been widely adopted (Granneman, 2001; Kaumann et al., 2001).

Studies using transfected cells have determined the site of action of many β-ligands (e.g., Konkar et al., 2000; Baker et al., 2003; Joseph et al., 2004a; Baker, 2005a). Pindolol, alprenolol, and several other ligands have biphasic concentration-response relationships (Walter et al., 1984; Baker et al., 2003; Kaumann and Molenaar, 2008; Baker, 2010a). The high-affinity component (readily blocked by β-blockers) is thought to occur via the high-affinity catecholamine conformation, while the low-affinity component (resistant to β-blockade) occurs via the secondary low-affinity conformation (Walter et al., 1984; Baker et al., 2003).

These two β₁-adrenoceptor conformations have also been demonstrated in rat, mouse, ferret, and human heart (e.g., Kompa and Summers, 1999; Kaumann et al., 2001; Lowe et al., 2002; Joseph et al., 2003; Sarsero et al., 2003; Molenaar et al., 2007), human blood vessels (e.g., Mallem et al., 2004; Kozlowska et al., 2003, 2006), and in whole animals (e.g., Malinowska and Schlicker, 1996; Zakrzeska et al., 2005). However, the physiologic or clinical relevance of this conformation remains unknown, even though the plasma concentration of carvedilol (100 ng/ml = 300 nM) used in human cardiovascular diseases is sufficient to occupy this secondary conformation (Sawangkoon et al., 2000; Baker et al., 2003).

Limited mutagenesis studies have, in passing, noted secondary conformation effects and detected some overlap of the amino acids involved in the catecholamine conformation, although these studies are not all in agreement. D138 (transmembrane [TM] 3) and N363 (TM7) have been suggested by some (Baker et al., 2008), but not all (Joseph et al., 2004b) to be involved in the secondary conformation, as have mutations in TM5 (Kaumann and Molenaar, 2008) and TM6 (Baker et al., 2008). However, the precise nature of this secondary conformation remains unknown from its location within the receptor sequence to the physiologic relevance or potential clinical implications (Molenaar, 2003; Molenaar et al., 2007).

Finally, although the β₁-adrenoceptor of many species (rat, mouse, guinea pig, ferret, cat, and human) (Kaumann and Molenaar, 2008), and the human β₃-adrenoceptor (Baker, 2005b) have been shown to exist in at least two agonist conformations, this is not true for all β-adrenoceptors. CGP12177 is a conventional partial agonist at the human β₂ and turkey β₃C adrenoceptors, (i.e., with similar K_B values and EC₅₀ values) suggesting interaction via a single conformation (Pak and Fishman, 1996; Baker et al., 2002; Baker, 2010b).

This study aimed to discover the primary region important for the secondary conformation of the β₁-adrenoceptor. Using a chimeric receptor approach (Isogaya et al., 1999; Kaumann and Molenaar, 2008), amino acids in the β₁-adrenoceptor were mutated to those of the β₂-adrenoceptor, and the pharmacological consequences observed. Here, we locate TM4, and specifically L195 and W199 within the human β₁-adrenoceptor, as the major site responsible for the secondary conformation pharmacology.

Materials and Methods

Molecular biology reagents were from Promega (Madison, WI). Lipofectamine, OPTIMEM, pcDNA3.1, and Top 10F–competent cells were from Life Technologies (Paisley, UK). The QuikChange mutagenesis kit was from Stratagene (La Jolla, CA), and fetal calf serum was from PAA Laboratories (Teddington, Middlesex, UK). ³H-Adenine and ¹⁴C-cAMP were from Amersham International (Buckinghamshire, UK), and Ultima Gold XR scintillation fluid from PerkinElmer (Shelton, CT). Cimaterol, CGP 20712A, bisoprolol, and carvedilol were from Tocris Life Sciences (Avonmouth, UK). All other reagents were from Sigma-Aldrich (Poole, Dorset, UK). Racemic ligands were used throughout.

Molecular Biology

The cDNA sequence encoding the wild-type human β₁-adrenoceptor (β₁-WT) in pJG3.6 was a gift from Steve Rees (GlaxoSmithKline, Stevenage). This cDNA was subcloned as a HindIII/XbaI fragment into pcDNA3.1, and the sequence was confirmed by DNA sequencing. Mutations described in Table 1 were generated using QuikChange mutagenesis and BioLine PolyMate Additive for GC-rich templates (Baker et al., 2008). After subcloning in Top 10F–competent cells, each mutant β₁-adrenoceptor cDNA was excised on HindIII/XbaI and subcloned into native pcDNA3.1 containing a neomycin selection marker. All mutations and sequences were confirmed by DNA sequencing using the School of Biomedical Sciences’ Sequencing Facility. To detect the most important areas of the human wild-type β₁-adrenoceptor involved in the secondary conformation, we made point mutations in the transmembrane regions of the receptor such that the each TM region resembled that of the wild-type human β₂-adrenoceptor (β₂-WT). Prediction of the transmembrane regions was performed using ExPASy topology prediction tools (www.expasy.org). For example, six point mutations were made in β₁-WT (L63I, L64V, A66S, L71A, A74F, V81T) which effectively converted the TM1 region of this receptor to that of the β₂-WT. This chimeric receptor was called β₁-TM1 (i.e., β₁-adrenoceptor but with TM1 of the β₂-WT, Table 1). This was then replicated for each TM region such that we had eight β₁ receptor constructs; the β₁-WT and one with each of the seven TM regions in turn mutated to that of the β₂-adrenoceptor (Table 1). A similar set of mutations was made starting with the β₂-WT and creating seven chimeric receptors, each with a TM region of the human β₁-WT (Table 1). These constructs were expressed in Chinese hamster ovary (CHO) cells stably expressing a reporter construct with six cAMP response elements upstream of a secreted placental alkaline phosphatase reporter gene (CRE-SPAP) cells and in stable cell lines we generated (see below).

TABLE 1.

The chimeric β₁/β₂-adrenoceptor constructs, the amino acid changes made, and where a stable cell line was made; receptor expression levels are given as fmol/mg protein

Constructs	Amino Acid Changes	Protein
		fmol/mg
β1-WT		556 ± 39
β1-TM1	L63I, L64V, A66S, L71A, A74F, V81T	372 ± 47
β1-TM2	M98T, S102C, L110A, T117A, I118H, V119I, V120L	567 ± 58
β1-TM3	L133F, V137I, L154V	1994 ± 214
β1-TM4	G177V, L178I, V179I, C180L, T181M, A184I, I185V, A187G, V189T, L195Q, W199Y	233 ± 18
β1-TM5	R222Q, V230I, C238V, A241V, L245S	359 ± 30
β1-TM6	V332T, L342I, A343V, V345I, K347H, A348V, F349I	1207 ± 100
β1-TM7	R357E, L358V, F359Y, V360I, F361L, F362L, L365I, A368V, A371G, I375L	2078 ± 133
β1-A187G	A187G
β1-V189T	V189T
β1-L195Q	L195Q
β1-W199Y	W199Y
β1-TM4 stage 1	G177V, L178I, V179I
β1-TM4 stage 2	G177V, L178I, V179I, C180L, T181M, A184I, I185V
β1-TM4 stage 3	G177V, L178I, V179I, C180L, T181M, A184I, I185V, A187G
β1-TM4 stage 4	G177V, L178I, V179I, C180L, T181M, A184I, I185V, A187G, V189T
β1-TM4 stage 5	G177V, L178I, V179I, C180L, T181M, A184I, I185V, A187G, V189T, L195Q
β1-V189T-L195Q	V189T, L195Q
β1-V189T-W199Y	V189T, W199Y
β1-L195Q-W199Y	L195Q, W199Y
β1-V189T-L195Q-W199Y	V189T, L195Q, W199Y
β1-W199A	W199A
β1-W199D	W199D
β1-W199F	W199F
β1-W199K	W199K
β1-W199L	W199L
β1-W199N	W199N
β2-WT		268 ± 25
β2-TM1	I38L, V39L, S41A, A46L, F49A, T56V	138 ± 8
β2-TM2	T73M, C77S, A85L, A92T, H93I, I94V, L95V	405 ± 41
β2-TM3	F108L, I112V, V129L	1037 ± 113
β2-TM4a	V152G, I153L, I154V, L155C, M156T, I159A, V160I, G162A, T164V, Q170L	458 ± 64
β2-TM5	Q197R, I205V, V213C, V216A, S220L	849 ± 83
β2-TM6	T281V, I291L, V292A, I294V, H296K, V297A, I298F	728 ± 58
β2-TM7	E306R, V307L, Y308F, I309V, L310F, L311F, I314L, V317A, G320A, L324I	67 ± 5

^aMaking the full change here (i.e., with Y174W) rendered the receptor nonfunctional (no binding and no functional responses). This chimera with 10 of the 11 amino acid substitutions was therefore used as the β₂-TM4 mutant.

Once important TM regions for the secondary site were identified, to determine which individual amino acids were involved, several single point mutations and chimeric receptors were made (Table 1) and these constructs expressed in CHO-CRE-SPAP cells, either as stable mixed populations of cells (CRE-SPAP assays) or in transiently transfected cells (³H-cAMP accumulation assays).

Cell Culture.

All CHO cells were grown in Dulbecco’s modified Eagle’s medium nutrient mix F12 containing 10% fetal calf serum and 2 mM L-glutamine in a 37°C humidified 5% CO₂ with 95% air atmosphere.

Generation of Stable Cell Lines.

CHO cells stably expressing a CRE-SPAP reporter gene were secondarily transfected with the wild-type human β₁- or β₂-adrenoceptor, or one of the full TM chimeric receptors (total 16 cell lines) using lipofectamine and OPTIMEM and selected for 3 weeks using resistance to geneticin (1 mg/ml for the receptor) and hygromycin (200 µg/ml for the CRE-SPAP reporter). Single clones were identified by dilution cloning and expanded to generate stable cell lines. These cell lines were used to identify the TM regions important for the secondary site. Receptor expression level in these stable cell lines was measured as previously described (Baker 2005b).

Generation of Stable Mixed Populations of Cells.

Here, the same parent CHO-CRE-SPAP cells were transfected with the wild-type human β₁- or β₂-adrenoceptor or a single point mutation or chimeric receptor and selected for 3 weeks using resistance to geneticin (1 mg/ml for the receptor) and hygromycin (200 µg/ml for the CRE-SPAP reporter). During this time the cells were passaged twice. The cells were then plated into 96-well plates for CRE-SPAP production experiments.

Generation of Transient Populations.

For transiently transfected cells used in the ³H-cAMP assays, the parent CHO-CRE-SPAP cells were transfected, as above, on day 1, the transfection reagents removed and replaced with media on day 2, the cells plated into 24-well plates on day 3, and the ³H-cAMP accumulation assay performed on day 4.

CRE-SPAP Production

Cells were grown to confluence in sterile, clear plastic, tissue culture–treated 96-well plates. The cells were then serum starved by removal of the media and addition of 100 µl serum-free media (sfm) per well for 24 hours before experimentation. At the start of each experiment, the sfm was removed from each well and 100 µl sfm or 100 µl sfm containing an antagonist at the final required concentration was added to each well and the cells incubated for 30 minutes at 37°C. Agonist in 10 μl sfm was then added to each well and the plates incubated for 5 hours. After 5 hours, all drugs and media were removed from each well and 40 µl sfm added to each well. The plates were then incubated for 1 hour at 37°C before being transferred to an oven preheated to 65°C and incubated for 30 minutes to destroy endogenous phosphatases. SPAP production was then measured by the addition of 100 µl 5 mM p-nitrophenyl phosphate per well (in diethanolamine buffer) and read on a Dynatech MRX plate reader at 405 nM.

³H-cAMP Accumulation

Cells were grown to confluence in sterile, clear plastic, tissue culture–treated 24-well plates. Cells were prelabeled with ³H-adenine by incubation for at least 2 hours with 2 μCi/ml ³H-adenine in media (0.5 ml per well). The cells were washed, then 1 ml sfm containing 1 mM 3-isobutyl-1-methylxanthine was added to each well and the cells incubated for 15 minutes at 37°C. Agonists (in 10 μl sfm) were added to each well and the plates incubated for 5 hours to maximize the responses (without altering the EC₅₀ values or % maximum isoprenaline response observed (Baker, 2010a). The assay was terminated by adding 50 μl concentrated HCl per well, the plates frozen, thawed, and ³H-cAMP separated from other ³H-nucleotides by sequential Dowex and alumina column chromatography (using ¹⁴C-cAMP to determine column efficiency). Isoprenaline (10 µM) was used to define the maximal response in each plate of each experiment.

Data Analysis

Agonist responses were best described by a one-site sigmoidal concentration response curve using the following equation:

where E_max is the maximum response, [A] is the agonist concentration, and EC₅₀ is the concentration of agonist that produces 50% of the maximal response.

The affinities of antagonists (K_B values) were calculated from the rightward shift of the agonist concentration response curve in the presence of a fixed concentration of antagonist using the following:

where DR (dose ratio) is the ratio of the agonist concentration required to stimulate an identical response in the presence and absence of a fixed concentration of antagonist [B].

When CGP12177 was used to antagonize the more efficacious agonists, clear partial agonism was seen (e.g., Fig. 2). Here, the affinity was initially calculated by the method of Stephenson (1956):

where [P] is the concentration of CGP12177, [A₁] is the concentration of the agonist at the point where CGP12177 alone causes the same agonist response, [A₂] is the concentration of agonist causing a given response above that achieved by CGP12177, and [A₃] is the concentration of the agonist, in the presence CGP12177, causing the same stimulation as [A₂].

Fig. 2.

CRE-SPAP production in response to cimaterol in the absence and presence of CGP12177 in stable cell lines expressing (A) β₁-WT, (B) β₁-TM4, and (C) β₂-WT. Bars represent basal CRE-SPAP production, that in response to 10 µM isoprenaline or that in response to 1–30 nM CGP12177 alone. Data points are mean ± S.E.M. of triplicate determinations. These single experiments are representative of (A) six, (B) eight, and (C) nine separate experiments. This figure shows that CGP12177 inhibits cimaterol responses with similar high affinity at β₁-WT, β₁-TM4, and β₂-WT.

The response to pindolol (and at times CGP12177) was best described by a two-site concentration response (e.g., Fig. 6B)

where N is the percentage of site 1, [A] is the concentration of agonist, and EC1₅₀ and EC2₅₀ are the respective EC₅₀ values for the two agonist sites.

Fig. 6.

³H-cAMP accumulation in response to CGP12177 (A, C, and E) and pindolol (B, D, and F) in transiently transfected cells expressing β₁-WT (A and B), β₁-TM4 (C and D), and β₂-WT (E and F). The CGP12177 responses were examined in the absence and presence of 3, 10, and 30 nM cimaterol (A and C); and 0.3, 1, and 3 nM cimaterol (E). Bars represent basal ³H-cAMP accumulation, that in response to 10 µM isoprenaline or that in response to 0.3–30 nM cimaterol alone. Data points are mean ± S.E.M. of triplicate determinations. These single experiments are representative of (A) six, (B) eight, (C) four, (D) three, (E) four, and (F) four separate experiments. This figure examines different evidence for the absence or presence of the secondary conformation. At β₁-WT, low concentrations of CGP12177 (e.g., 3 nM, −8.5) inhibit the response to cimaterol, whereas higher concentrations of CGP12177 cause a stimulatory response (e.g., 100 nM, −7 and above). This creates a “dip” in the curve (CGP12177 + fixed concentration of cimaterol) and is strongly suggestive of interaction at two different conformations. This is absent at β₁-TM4 and β₂-WT, suggesting single-conformation interaction. Similarly, the response to pindolol is biphasic at β₁-WT but monophasic at β₁-TM4 and β₂-WT, again suggesting two conformations at β₁-WT but only one conformation at β₁-TM4 and β₂-WT.

A two-site analysis was also used for the experiments e.g., Fig. 6A:

where basal is the response in the absence of agonist, Ag is the response to a fixed concentration of agonist, [P] is the concentration of partial agonist (e.g., CGP12177), IC₅₀ is the concentration of competing partial agonist that inhibits 50% of the response of the fixed agonist, PAg is the maximum stimulation by the competing partial agonist, and EC₅₀ is the concentration of competing agonist that stimulated a half maximal competing partial agonist response.

A 10 μM (maximal) isoprenaline concentration was included in each plate for each separate experiment for CRE-SPAP production and ³H-cAMP accumulation, to allow agonist responses to be expressed as a percentage of the isoprenaline maximum for each experiment. Data points in the figures are presented as mean ± S.E.M. of triplicate determinations from a single experiment. Data in the text and tables are mean ± S.E.M. of n separate experiments.

Results

Demonstration of the Two Agonist Conformations of the Human β₁-Adrenoceptor: CRE-SPAP Production in Stable Cell Lines.

Cimaterol, a primary catecholamine conformation agonist (Baker 2005a), stimulated a response in cells stably expressing the human wild-type β₁-adrenoceptor (log EC₅₀ = −8.25 ± 0.14) that was 95.8 ± 2.9% that of the response to 10-µM isoprenaline (top half Table 2; Fig. 1A). As expected, this β₁ response was readily inhibited by the β₁-selective antagonist CGP20712A [2-hydroxy-5-(2-[{hydroxy-3-(4-[1-methyl-4-trifluoromethyl-2-imidazolyl]phenoxy)propyl}amino]ethoxy)benzamide] (log K_B −9.18) and poorly inhibited by the β₂-selective antagonist ICI118551 [(−)-1-(2,3-[dihydro-7-methyl-1H-inden-4-yl]oxy)-3-([1-methylethyl]-amino)-2-butanol] (log K_B −6.96). CGP12177, a partial agonist, also inhibited this cimaterol response with high affinity indicating high affinity for the catecholamine conformation of the receptor (log K_B −9.55; Fig. 2A; Table 2). CGP12177 is however also known to activate the secondary conformation of the β₁-adrenoceptor. At the β₁-WT, CGP12177 stimulated a partial agonist response, log EC₅₀ −8.18 ± 0.08, which was 73.8 ± 5.6% that of the maximum to isoprenaline (Fig. 1B). This CGP12177 response was also inhibited by antagonists but in each case higher concentrations of antagonist were required, yielding lower log K_B values (e.g., log K_B for CGP20712A was −7.16; bottom half Table 2). The fact that 10–500-fold higher concentrations of antagonist were required to inhibit the CGP12177 rather than the cimaterol responses, and the discrepancy between the log K_B value and the log EC₅₀ value for CGP12177 itself, demonstrate the high affinity and the lower affinity conformations of the human β₁-adrenoceptor.

TABLE 2.

Log EC₅₀ values and % maximum response to isoprenaline for CRE-SPAP production for the agonists cimaterol and CGP12177 at the human β₁-WT and chimeric β₁/β₂-adrenoceptors

Mutations in the β₁-WT mean that each TM region in turn is mutated to that of the β₂-WT (see Table 1). Log K_B values for several antagonists for inhibition of the cimaterol and CGP12177 responses were determined. These data were obtained from stable cell lines, and n in the table refers to the number of separate experiments. The table shows that the most consistent and highly significant difference between the β₁-WT receptor and all of the TM-swap mutants is that obtained from β₁-TM4 when CGP12177 is the agonist. As CGP12177 agonist responses occur at the secondary low-affinity conformation of the β₁-WT receptor, this suggests that the β₁-TM4 receptor mutations alter this conformation in some way.

	Cimaterol Log EC50	% Isoprenaline	n	Log KB CGP20712A	n	Log KB Bisoprolol	n	Log KB ICI118551	n	Log KB Propranolol	n	Log KB Carvedilol	n	Log KB CGP12177	n
Cimaterol as agonist: β1 stable cell lines
β1-WT	−8.25 ± 0.14	95.8 ± 2.9	8	−9.18 ± 0.07	4	−8.48 ± 0.17	5	−6.96 ± 0.04	5	−8.52 ± 0.08	12	−9.89 ± 0.13	5	−9.55 ± 0.12	8
β1-TM1	−7.91 ± 0.05*	96.5 ± 3.3	11	−9.35 ± 0.10	9	−8.65 ± 0.12	5	−6.96 ± 0.05	9	−8.66 ± 0.07	16	−9.88 ± 0.08	5	−9.77 ± 0.07	15
β1-TM2	−8.19 ± 0.03	98.9 ± 3.0	15	−8.78 ± 0.12	8	−8.15 ± 0.07	7	−7.47 ± 0.03*	14	−8.52 ± 0.03	24	−9.53 ± 0.18	8	−9.90 ± 0.09	17
β1-TM3	−8.82 ± 0.03*	102.4 ± 2.5	9	−9.34 ± 0.12	9	−8.43 ± 0.07	4	−7.31 ± 0.04*	10	−8.53 ± 0.06	15	−9.92 ± 0.16	5	−9.49 ± 0.06	13
β1-TM4	−8.15 ± 0.02	99.2 ± 1.8	10	−9.50 ± 0.13	11	−8.81 ± 0.04	5	−8.07 ± 0.06*	9	−8.99 ± 0.04*	15	−9.81 ± 0.12	6	−9.63 ± 0.08	17
β1-TM5	−8.48 ± 0.05	103.0 ± 1.7	9	−9.09 ± 0.04	11	−7.91 ± 0.03	5	−7.01 ± 0.06	11	−8.71 ± 0.04	17	−9.96 ± 0.09	6	−9.68 ± 0.09	17
β1-TM6	−8.59 ± 0.03*	103.2 ± 2.2	10	−8.52 ± 0.07	8	−8.10 ± 0.15	5	−7.92 ± 0.05*	12	−8.90 ± 0.05*	18	−10.13 ± 0.09	6	−9.58 ± 0.06	14
β1-TM7	−8.54 ± 0.05	106.1 ± 2.5	10	−8.45 ± 0.08	7	−8.19 ± 0.06	5	−7.80 ± 0.05*	12	−8.95 ± 0.05*	18	−10.06 ± 0.03	3	−9.39 ± 0.05	13
	CGP12177 Log EC50	% Isoprenaline	n	Log KB CGP20712A	n	Log KB Bisoprolol	n	Log KB ICI118551	n	Log KB Propranolol	n	Log KB Carvedilol	n
CGP12177 as agonist: β1 stable cell lines
β1-WT	−8.18 ± 0.08	73.8 ± 5.6	10	−7.16 ± 0.06	14	−5.83 ± 0.14	4	−5.90 ± 0.15	9	−6.18 ± 0.06	13	−7.25 ± 0.18	7
β1-TM1	−7.83 ± 0.05	53.7 ± 3.9	10	−7.41 ± 0.13	11	−5.77 ± 0.18	4	−5.79 ± 0.09	3	−6.55 ± 0.12	13	−7.72 ± 0.13	7
β1-TM2	−8.27 ± 0.09	53.5 ± 3.9	14	−7.13 ± 0.08	13	−6.72 ± 0.18	7	−5.87 ± 0.05	6	−6.87 ± 0.15*	18	−8.52 ± 0.15*,**	7
β1-TM3	−8.07 ± 0.03	70.5 ± 3.6	9	−7.45 ± 0.05	12	−5.83 ± 0.13	4	−5.83 ± 0.09	7	−6.31 ± 0.05	16	−7.45 ± 0.09	6
β1-TM4	−9.27 ± 0.06*,**	29.2 ± 1.4*,**	10	−9.49 ± 0.08*,**	9	−8.83 ± 0.21*,**	7	−7.93 ± 0.11*,**	11	−8.85 ± 0.09*,**	15	−9.73 ± 0.21*,**	7
β1-TM5	−7.91 ± 0.03	60.1 ± 3.4	10	−6.68 ± 0.04*	11	−5.31 ± 0.13	5	−5.71 ± 0.15	8	−6.23 ± 0.09	15	−7.53 ± 0.16	7
β1-TM6	−7.83 ± 0.02*	63.5 ± 2.8	11	−5.97 ± 0.08*,**	12	−5.22 ± 0.14	3	−5.79 ± 0.04	11	−6.05 ± 0.06	19	−7.24 ± 0.18	7
β1-TM7	−7.08 ± 0.02*,**	68.5 ± 3.8	11	−6.86 ± 0.05	10	−5.67 ± 0.05	5	−6.67 ± 0.03*,**	13	−6.42 ± 0.06	16	−7.35 ± 0.12	6

^*P < 0.001; one-way ANOVA with post hoc Newman–Keuls comparing values from the mutant receptors with those obtained from the β₁-WT. Thus, the log EC₅₀ for CGP12177 at β₁-TM4 is different from that obtained from the β₁-WT with P < 0.001; **P < 0.001; one-way ANOVA with post hoc Newman–Keuls comparing each value with all other values in this set. Thus, the log EC₅₀ value for CGP12177 at β₁-TM4 is different from those obtained for β₁-WT, β₁-TM1, β₁-TM2, β₁-TM3, β₁-TM5, β₁-TM6, and β₁-TM7, with P < 0.001 in all cases.

Fig. 1.

CRE-SPAP production in response to cimaterol (A, C, and E) and CGP12177 (B, D, and F) in the absence and presence of bisoprolol and propranolol in stable cell lines expressing β₁-WT (A and B), β₁-TM4 (C and D), and human β₂-WT (E and F). Bars represent basal CRE-SPAP production, that in response to 10 µM isoprenaline or that in response to bisoprolol or propranolol alone. Data points are mean ± S.E.M. of triplicate determinations. These single experiments are representative of (A) five, (B) four, (C) four, (D) seven, (E) six, and (F) 10 separate experiments. This figure shows that the greater concentrations of antagonist are required to inhibit the secondary conformation CGP12177 responses in the β₁-WT, but that this is not the case for the β₁-TM4 or β₂-WT.

At the human β₂-adrenoceptor, agonist responses to cimaterol and CGP12177 were also observed; however, similar concentrations of antagonist were required to inhibit these responses, yielding similar log K_B values (Fig. 1, E and F; Table 4), and the log K_B and log EC₅₀ values for CGP12177 were similar, suggesting that cimaterol and CGP12177 are acting through the same conformation of the β₂-adrenoceptor.

TABLE 4.

Log EC₅₀ values and % maximum response to isoprenaline for CRE-SPAP production for the agonists cimaterol and CGP12177 at the human β₂-WT and chimeric β₂/β₁-receptors, where mutations in the β₂-WT mean that each TM region in turn is mutated to that of the β₁-WT (see Table 1)

Log K_B values for several antagonists for inhibition of the cimaterol and CGP12177 responses are also given. These data were obtained from stable cell lines; n refers to the number of separate experiments. Although there are some differences among the mutant receptors (e.g., β₂-TM2 has higher affinity for CGP20712A than β₂-WT), there is no highly significant consistent change in one mutant receptor compared with the others.

	Cimaterol Log EC50	% Isoprenaline	n	Log KB CGP20712A	n	Log KB Bisoprolol	n	Log KB ICI118551	n	Log KB Propranolol	n	Log KB CGP12177	n
Cimaterol as agonist: β2 stable cell lines
β2-WT	−9.05 ± 0.08	101.5 ± 3.2	13	−6.03 ± 0.12	8	−6.85 ± 0.10	8	−9.56 ± 0.07	8	−9.65 ± 0.05	14	−10.00 ± 0.05	9
β2-TM1	−9.45 ± 0.06	105.3 ± 2.8	9	−6.46 ± 0.10	8	−6.82 ± 0.09	10	−9.74 ± 0.08	9	−9.56 ± 0.06	19	−9.74 ± 0.11	19
β2-TM2	−9.64 ± 0.08*	102.1 ± 1.8	12	−7.20 ± 0.11*	8	−7.25 ± 0.10	7	−9.30 ± 0.11	8	−9.87 ± 0.03	14	−10.01 ± 0.06	9
β2-TM3	−9.21 ± 0.13	98.3 ± 4.1	7	−6.57 ± 0.12	7	−7.05 ± 0.13	6	−9.71 ± 0.10	7	−9.81 ± 0.07	11	−10.00 ± 0.10	10
β2-TM4	−9.66 ± 0.06*	103.3 ± 2.4	13	−6.09 ± 0.09	10	−6.75 ± 0.03	9	−8.87 ± 0.04*	10	−9.34 ± 0.06	14	−9.83 ± 0.05	9
β2-TM5	−8.97 ± 0.09	100.6 ± 2.4	13	−6.63 ± 0.12	10	−7.04 ± 0.09	9	−9.50 ± 0.09	9	−9.46 ± 0.04	13	−9.82 ± 0.07	4
β2-TM6	−8.74 ± 0.05	101.0 ± 4.7	12	−7.05 ± 0.15*	7	−7.07 ± 0.10	9	−8.86 ± 0.12*	9	−9.52 ± 0.06	11	−9.59 ± 0.13	8
β2-TM7	−8.77 ± 0.06	107.4 ± 2.9	10	−6.54 ± 0.04	14	−7.06 ± 0.09	11	−9.18 ± 0.07	12	−9.30 ± 0.06*	17	−9.72 ± 0.10	13
	CGP12177 Log EC50	% Isoprenaline	n	Log KB CGP20712A	n	Log KB Bisoprolol	n	Log KB ICI118551	n	Log KB Propranolol	n
CGP12177 as agonist: β2 stable cell lines
β2-WT	−9.35 ± 0.09	23.5 ± 3.9	14	−6.12 ± 0.09	3	−6.60 ± 0.14	11	−9.31 ± 0.17	11	−9.02 ± 0.12	19
β2-TM1	−9.48 ± 0.19	26.9 ± 1.7	9	−6.00 ± 0.16	9	−6.63 ± 0.12	12	−9.51 ± 0.17	12	−9.07 ± 0.08	20
β2-TM2	−9.20 ± 0.09	12.6 ± 1.9	11	−6.79 ± 0.11	8	−7.00 ± 0.09	9	−9.03 ± 0.11	10	−9.32 ± 0.10	14
β2-TM3	−9.19 ± 0.10	35.9 ± 3.3	9	−6.21 ± 0.29	4	−6.53 ± 0.12	8	−9.47 ± 0.16	8	−8.95 ± 0.09	21
β2-TM4	−9.20 ± 0.07	38.5 ± 2.4*	13	−6.05 ± 0.12	8	−6.34 ± 0.09	7	−8.42 ± 0.08*	11	−8.34 ± 0.08*	18
β2-TM5	−9.31 ± 0.07	28.8 ± 2.0	12	−6.47 ± 0.15	8	−6.98 ± 0.22	8	−9.31 ± 0.08	9	−8.75 ± 0.10	19
β2-TM6	−8.81 ± 0.10	22.4 ± 1.6	8	−7.31 ± 0.34	5	−6.74 ± 0.11	5	−8.39 ± 0.27	4	−9.11 ± 0.11	4
β2-TM7	−9.71 ± 0.12	41.3 ± 2.7*	9	−6.19 ± 0.15	10	−6.59 ± 0.16	10	−9.09 ± 0.14	9	−9.02 ± 0.08	22

^*P < 0.001 One-way ANOVA with post hoc Newman–Keuls comparing values from the mutant receptors with those obtained from the β₂-WT. Thus, the log K_B value for CGP20712A at β₂-TM2 obtained in the presence of cimaterol is different (P < 0.001) from that obtained in the β₂-WT receptor. In contrast to data on the β₁-receptor in Table 2, there are no values with P < 0.001 for one-way analysis of variance with post hoc Newman–Keuls comparing each value with all other values in this set.

Identification of Transmembrane Regions Important in the Secondary Conformation of the β₁-Adrenoceptor: CRE-SPAP Production in Stable Cell Lines.

To identify areas in the receptor important for the secondary conformation, β₁/β₂ chimeric receptors were examined. These receptors had mutations such that each TM was in turn swapped for that of the other receptor (e.g., β₁-TM1 is the β₁-adrenoceptor sequence containing mutations that make it equivalent to the TM1 region of the β₂-adrenoceptor) (Table 1). When the β₁-TM chimeric receptors were examined (each stably expressed in a different cell line), cimaterol stimulated a full agonist response at all β₁/β₂ receptors (β₁-TM1 through to β₁-TM7) (Fig. 1C; top half Table 2) and this response was readily inhibited by CGP20712A and poorly by ICI118551, confirming that despite changing each transmembrane in turn to that of the β₂-adrenoceptor, these chimeric receptors retained a β₁-selective nature. There were some differences in the affinity of antagonists (e.g., affinity of ICI118551 at the β₁-TM2, β₁-TM3, β₁-TM4, β₁-TM6, and β₁-TM7 receptors is substantially higher than at the β₁-WT; top half Table 2). This could be due to a general trend of these chimeras “losing” β₁-subtype selectivity as judged by a range of ligands or due to a specific interaction between the ligand ICI118551 and specific amino acids in these chimeras. To determine which, several antagonist ligands (bisoprolol a less selective but still a β₁-selective ligand; propranolol and carvedilol slightly β₂-selective ligands) were examined to generate a pattern of affinity (top half Table 2; Fig. 1). The affinity of CGP20712A and bisoprolol remained high (similar to β₁-WT, top half Table 2 and very different from values obtained at the β₂-WT; top half Table 4). Thus, the overall pattern for these chimeras is one of retaining β₁-selective properties (and indeed ICI118551 was later found to interact with a specific amino acid in TM4 rather than a general pattern across several ligands; see below).

When the CGP12177 responses were examined, the data achieved for β₁-TM4 were significantly different from those at the other chimeras (P < 0.001; bottom half Table 2; Fig. 1D). Firstly, the log EC₅₀ for CGP12177 was very different from that at the other receptors (bottom half Table 2). Secondly, the log K_B value and log EC₅₀ values for CGP12177 in the β₁-TM4 cells were similar (left column, Table 3, and indeed similar to that seen in the β₂-WT cells; Table 4). Finally, the affinities of antagonists in the presence of CGP12177 at β₁-TM4 were very similar to those obtained when cimaterol was used as the agonist in these cells (Table 2). This is most easily seen in Table 3, where the ratios of the antagonist affinities are compared. Whereas the ratio of antagonist affinities for CGP20712A at β₁-WT was log 2.02 (e.g., it has 100-fold different affinity for the two conformations), it was log 0.01 at β₁-TM4, suggesting that there is no difference in affinity and that the secondary conformation was not in existence. Thus, β₁-TM4 had a pharmacological profile consistent with CGP12177 acting as a partial agonist at a single high-affinity conformation of the β₁-adrenoceptor (i.e., in a similar manner to which CGP12177 acts at the single receptor conformation of the β₂-WT adrenoceptor).

TABLE 3.

The ratio of the log K_B value for CGP12177 (as an antagonist of cimaterol; thus, at the high-affinity conformation) and log EC₅₀ value for CGP12177 (as an agonist of the secondary low-affinity conformation) for responses seen at the β₁-WT and β₁-TM chimeric receptors (data taken from Table 2)

The ratios of the affinity for antagonists at the two conformations (i.e., log K_B for antagonism of cimaterol at high-affinity site/log K_B for antagonism of CGP12177 at low-affinity site) were determined. This table examines the ratio of the affinities of ligands at the two conformations of the β₁-adrenoceptor and thus allows for differences in absolute affinity. For example, ICI118551 has a higher affinity at the β₁-TM7 receptor (Table 2) than the β₁-WT receptor; however, the ratio of the affinity in the presence of cimaterol and CGP 12177 is similar to that at the WT receptor, suggesting that although the β₁-TM7 mutation alters the affinity of ICI118551, it does not alter the secondary conformation. The most consistent change is seen with the β₁-TM4 mutation, where the ratios are all nearly 0, suggesting no difference in affinity (i.e., no secondary conformation).

	Ratio of Log KB CGP12177/ Log EC50 CGP12177	Ratio of Log KB (Cimaterol as Agonist)/Log KB (CGP12177 as Agonist)
		CGP20712A	Bisoprolol	ICI118551	Propranolol	Carvedilol
β1-WT	1.37	2.02	2.65	1.06	2.34	2.64
β1-TM1	1.94	1.94	2.88	1.17	2.11	2.16
β1-TM2	1.63	1.65	1.43	1.60	1.65	1.01
β1-TM3	1.42	1.89	2.60	1.48	2.22	2.47
β1-TM4	0.36	0.01	−0.02	0.14	0.14	0.08
β1-TM5	1.77	2.41	2.60	1.30	2.47	2.43
β1-TM6	1.75	2.55	2.88	2.15	2.85	2.89
β1-TM7	2.31	1.59	2.52	1.13	2.53	2.71

When the full β₂-TM4 construct was expressed (11 amino acid changes), no function or binding was detectable despite multiple attempts. Therefore, a construct containing 10 amino acid changes (all except the last Y174W) was made, which was then used as β₂-TM4 (Tables 1, 4, and 5). The data obtained for the β₂/β₁ chimeras (Table 4) suggest that the agonist responses and antagonist affinities obtained change very little in the β₂-adrenoceptors with whole TM region changes.

TABLE 5.

The ratio of the log K_B value for CGP12177 (as an antagonist of cimaterol) and log EC₅₀ value for CGP12177 for responses seen at the β₂-WT and β₂-TM chimeric receptors (data taken from Table 4)

Ratios of the affinity for antagonists at the two conformations are also given (i.e., log K_B for antagonism of cimaterol/log K_B for antagonism of CGP12177). This table examines the ratio of the affinities of ligands at the β₂-adrenoceptor and thus allows for differences in absolute affinity. ICI118551 has lower affinity at the β₂-TM4 receptor (Table 4), but this occurred both in the presence of cimaterol and CGP12177; thus, the ratio is similar to that at the other receptors. There is no large consistent change with any of the receptors.

	Ratio of Log KB CGP12177/Log EC50 CGP12177	Ratio of Log KB (Cimaterol as Agonist)/Log KB (CGP12177 as Agonist)
		CGP20712A	Bisoprolol	ICI118551	Propranolol
β2-WT	0.65	−0.09	0.25	0.25	0.63
β2-TM1	0.26	0.46	0.19	0.23	0.49
β2-TM2	0.81	0.41	0.25	0.27	0.55
β2-TM3	0.81	0.36	0.52	0.24	0.86
β2-TM4	0.63	0.04	0.41	0.45	1.00
β2-TM5	0.51	0.16	0.06	0.19	0.71
β2-TM6	0.78	−0.26	0.33	0.47	0.41
β2-TM7	0.01	0.35	0.47	0.09	0.28

Examination of the ratios of antagonist affinities at the two conformations is again helpful for detecting the most important changes (Tables 3 and 5). Thus, although there are some differences for the chimeric receptors (e.g., EC₅₀/K_B ratio for β₁-TM7 and β₂-TM7), by far the most overwhelming differences are in the β₁-TM4 chimeric receptor; this is clearly shown when comparing the ratios of the log K_B and log EC₅₀ values at the two conformations, as shown in Tables 3 and 5. Also, here, although ICI118551 was found to have higher affinity at β₁-TM4 (and indeed β₁-TM7), the ratio of the affinities for the two conformations of the β₁-adrenoceptor clearly shows that β₁-TM4 is different from the other β₁-chimeric receptors.

Importantly, because the stable cell lines used here have different levels of receptor expression (Table 1), neither of these methods of assessing the presence of the secondary conformation relies on the receptor expression levels being the same. Antagonist affinity measurements are independent of receptor expression level and, because CGP12177 is a partial agonist in all of these cell lines, its K_B value in a given cell line should be similar to its EC₅₀ value in that cell line (i.e., ratio of zero) if the two are acting at a single conformation. A further piece of evidence can be obtained from comparing the log EC₅₀ values for CGP12177 over the cell lines, although this is more problematic, not least as it is likely to depend on receptor expression levels. Interestingly here, the log EC₅₀ for CGP12177 at β₁-TM4 is markedly more potent (left-shifted) even though its receptor expression level is the lowest of the β₁ cell lines and, thus, if anything, would be expected to have a less potent (i.e., right-shifted) log EC₅₀ value. Even this softer evidence is therefore pointing to β₁-TM4 as an important region for the secondary conformation.

Identification of the Amino Acids in TM4 Involved in the Secondary Conformation: CRE-SPAP Production in Stable Mixed Populations of Cells.

To determine which of the amino acids in TM4 were responsible for this secondary conformation, intermediate stages of the production of the chimeric receptors were examined that had several, but not all, of the amino acids changes required to alter the TM domain from β₁ to β₂ (see Table 1 for the amino acid changes made). These constructs were expressed in parent CHO-SPAP cells and experiments performed in stable mixed populations of cells (4–24 separate populations of cells were generated for each chimeric receptor). Whereas the affinity for antagonists remained very similar when cimaterol was the agonist (primary catecholamine site, top half of Table 6, with the exception of ICI118551), when the secondary site conformation was examined, (i.e., with CGP12177 as the agonist), the log EC₅₀ value for CGP12177 became more potent (approaching that for the log K_B value) in β₁-TM4 stage 5 (bottom half Table 6). Secondly, the affinity of the other antagonists (log K_B values) also became higher with β₁-TM4 stage 5 and full β₁-TM4 chimeras (bottom half Table 6). The affinity for ICI118551 was high at both conformations for β₁-TM4 stage 4, although the ratio (Table 7) suggests the retention of both conformations of the receptor. This suggests that amino acids L195 and W199 are the most important for the secondary conformation.

TABLE 6.

Log EC₅₀ values and % maximum response to isoprenaline for CRE-SPAP production for the agonists cimaterol and CGP12177 at the β₁-WT, the chimeric β₁-TM4 receptor (containing the full TM4 mutations), and chimeric β₁/β₂-adrenoceptors with several amino acids in TM4 mutated to that of the β₂-WT

Log K_B values for several antagonists for inhibition of the cimaterol and CGP12177 responses are also given. Table 1 lists the detail of each mutation. These data were obtained from between 4 and 24 stable mixed populations of cells for each chimera, and n in the table refers to the number of separate experiments. This table shows that within TM4, it is only the amino acid changes in stage 5 and the full TM4 (i.e., L195Q and W199Y) that are very important for the alteration of the secondary conformation (i.e., responses to CGP12177 and antagonist affinities obtained in the presence of CGP12177). Stage 4 (addition of V189T to the other changes) affects the affinity of ICI18551. There is no change in the pharmacology of either conformation, even when the majority of amino acids have been changed (up to and including stage 3).

	Cimaterol Log EC50	Isoprenaline %	n	Log KB CGP20712A	n	Log KB ICI118551	n	Log KB Propranolol	n	Log KB CGP12177	n
Cimaterol as agonist: stable mixed populations of cells
β1-WT	−8.29 ± 0.04	81.2 ± 2.1	24	−9.36 ± 0.07	35	−7.13 ± 0.06	31	−8.69 ± 0.08	30	−10.01 ± 0.06	33
β1-TM4 stage 1	−8.26 ± 0.11	85.6 ± 3.9	4	−9.10 ± 0.09	4	−7.04 ± 0.09	6	−8.64 ± 0.14	4	−9.96 ± 0.33	3
β1-TM4 stage 2	−8.20 ± 0.06	77.8 ± 1.5	12	−9.34 ± 0.10	18	−7.14 ± 0.06	14	−8.79 ± 0.11	12	−9.81 ± 0.09	14
β1-TM4 stage 3	−8.13 ± 0.07	78.2 ± 3.9	11	−8.95 ± 0.09	13	−7.00 ± 0.07	10	−8.63 ± 0.07	11	−9.93 ± 0.06	15
β1-TM4 stage 4	−8.08 ± 0.05	76.3 ± 1.5	12	−9.14 ± 0.08	18	−7.62 ± 0.05*	16	−8.59 ± 0.06	12	−9.77 ± 0.06	19
β1-TM4 stage 5	−8.14 ± 0.06	78.2 ± 3.2	11	−8.97 ± 0.09	16	−7.71 ± 0.07*	15	−8.81 ± 0.04	11	−9.93 ± 0.08	19
β1-TM4	−8.32 ± 0.05	77.3 ± 1.9	23	−9.43 ± 0.09	35	−7.94 ± 0.06*	31	−9.08 ± 0.07	29	−9.98 ± 0.09	35
	CGP12177 Log EC50	Isoprenaline %	n	Log KB CGP20712A	n	Log KB ICI118551	n	Log KB Propranolol	n
CGP12177 as agonist: stable mixed populations of cells
β1-WT	−8.12 ± 0.07	53.3 ± 1.5	21	−7.37 ± 0.08	37	−5.86 ± 0.08	22	−6.69 ± 0.08	27
β1-TM4 stage 1	−8.43 ± 0.06	50.4 ± 2.2	4	−7.32 ± 0.25	5	−6.21 ± 0.50	3	−6.17 ± 0.17	4
β1-TM4 stage 2	−8.46 ± 0.07	46.5 ± 1.9	11	−7.55 ± 0.12	17	−6.08 ± 0.11	10	−6.78 ± 0.09	20
β1-TM4 stage 3	−8.48 ± 0.08	45.9 ± 3.0	11	−7.13 ± 0.10	18	−5.90 ± 0.11	10	−6.88 ± 0.11	13
β1-TM4 stage 4	−7.93 ± 0.12	40.0 ± 1.6*	10	−7.26 ± 0.10	15	−6.57 ± 0.13*	10	−6.98 ± 0.13	18
β1-TM4 stage 5	−9.29 ± 0.06*	30.2 ± 2.3*	11	−8.74 ± 0.09*	19	−7.54 ± 0.09*	10	−8.18 ± 0.08*	25
β1-TM4	−9.43 ± 0.06*	33.9 ± 1.5*	19	−8.98 ± 0.07*	39	−7.64 ± 0.08*	32	−8.39 ± 0.06*	40

^*P < 0.001; one-way ANOVA with post hoc Newman-Keuls comparing values from the mutant receptors with those obtained from the β₁-WT; thus, the log EC₅₀ for CGP12177 at β₁-TM4 Stage 5 receptor is different from that obtained from the β₁-WT with P < 0.001

TABLE 7.

The ratio of the log K_B value for CGP12177 (as an antagonist of cimaterol; thus, at the high affinity conformation) and log EC₅₀ value for CGP12177 (as an agonist of the secondary low affinity conformation) for responses seen at the β₁-WT and β₁-TM4 chimeric receptors (data taken from Table 6)

Ratio of the affinity for antagonists at the two conformations is given (i.e., log K_B for antagonism of cimaterol at high affinity site/log K_B for antagonism of CGP12177 at low-affinity site). This table examines the ratio of the affinities of ligands and shows that the ratio of affinities between the two conformations is much less for β₁-TM4 stage 5 and beyond, suggesting that only the amino acids in stage 5 and beyond are important for the secondary conformation.

	Ratio of Log KB CGP12177/Log EC50 CGP12177	Ratio of Log KB (Cimaterol as Agonist)/Log KB (CGP12177 as Agonist)
		CGP20712A	ICI118551	Propranolol
β1-WT	1.89	1.99	1.27	2.00
β1-TM4 stage 1	1.53	1.78	0.83	2.47
β1-TM4 stage 2	1.35	1.79	1.06	2.01
β1-TM4 stage 3	1.45	1.82	1.10	1.75
β1-TM4 stage 4	1.84	1.88	1.05	1.61
β1-TM4 stage 5	0.64	0.23	0.17	0.63
β1-TM4	0.55	0.45	0.30	0.69

To investigate the role of these specific amino acids further, each of these individual single point mutations were examined alone (in stable mixed populations) rather than on the background of the other TM4 mutations (Tables 1, 8, and 9; Figs. 3 and 4). β₁-L195Q and β₁-W199Y had little effect on the cimaterol catecholamine conformation response (top half Table 8; Fig. 3). However, CGP12177 agonist responses were significantly more potent at β₁-L195Q and β1-W199Y (bottom half Table 8; Fig. 4). Examining antagonist affinity values, whereas β₁-V189T affected the affinity of ICI118551 at the catecholamine conformation, the affinities of CGP20712A, propranolol, and CGP12177 were largely unaffected by any of the mutations (top half of Table 8; Fig. 3). However, the secondary conformation CGP12177 responses were significantly more sensitive to inhibition with the L195Q and W199Y mutations (bottom half Table 8; Fig. 4).

TABLE 8.

Log EC₅₀ values and % maximum response to isoprenaline for CRE-SPAP production for the agonists cimaterol and CGP12177 at β₁-WT, the chimeric β₁-TM4 receptor (containing the full TM4 mutations), and chimeric β₁/β₂-adrenoceptors with single point mutations and then combinations of these mutations

Log K_B values for several antagonists for inhibition of the cimaterol and CGP12177 responses are given. These data were obtained from between 4 and 24 stable mixed populations of cells for each receptor, and n in the table refers to the number of separate experiments. This table shows that the mutations L195Q and W199Y, both alone and in combination, have a very significant effect on the responses to CGP12177 and affinity of antagonist measured in the presence of CGP12177 (i.e., the secondary conformation of the β₁-receptor). Every time the mutation V189T is present, the affinity of ICI118551 is increased, both at the catecholamine conformation (in the presence of cimaterol) and at the secondary conformation (presence of CGP12177).

	Cimaterol Log EC50	% Isoprenaline	n	Log KB CGP20712A	n	Log KB ICI118551	n	Log KB propranolol	n	Log KB CGP12177	n
Cimaterol as agonist: stable mixed populations of cells
β1-WT	−8.29 ± 0.04	81.2 ± 2.1	24	−9.36 ± 0.07	35	−7.13 ± 0.06	31	−8.69 ± 0.08	30	−10.01 ± 0.06	33
β1-A187G	−8.22 ± 0.08	78.9 ± 2.3	8	−9.23 ± 0.08	14	−7.02 ± 0.07	8	−8.72 ± 0.09	8	−9.99 ± 0.06	17
β1-V189T	−8.20 ± 0.04	80.6 ± 1.8	12	−9.26 ± 0.09	20	−7.73 ± 0.07*	15	−8.61 ± 0.05	14	−9.81 ± 0.08	27
β1-L195Q	−8.20 ± 0.05	80.6 ± 2.7	12	−8.97 ± 0.10	19	−7.12 ± 0.07	15	−8.86 ± 0.11	14	−9.85 ± 0.07	25
β1-W199Y	−8.46 ± 0.06	77.3 ± 2.4	12	−9.66 ± 0.08	19	−7.39 ± 0.07	16	−9.01 ± 0.12	15	−10.06 ± 0.05	26
β1-W199A	−8.63 ± 0.07*	85.4 ± 3.2	8	−9.51 ± 0.14	13	−7.26 ± 0.08	11	−9.12 ± 0.11	13	−10.07 ± 0.16	4
β1-W199D	−8.91 ± 0.07*	87.6 ± 3.5	8	−9.88 ± 0.08	12	−7.48 ± 0.11	10	−9.17 ± 0.10	12	−9.85 ± 0.12	8
β1-W199F	−8.71 ± 0.09*	80.0 ± 2.6	8	−10.14 ± 0.10*	14	−7.19 ± 0.12	12	−9.09 ± 0.13	14	−9.95 ± 0.16	11
β1-W199K	−7.64 ± 0.05*	86.4 ± 3.2	8	−9.46 ± 0.08	14	−6.41 ± 0.06*	9	−8.00 ± 0.12*	10	a
β1-W199L	−7.60 ± 0.06*	75.2 ± 2.9	8	−10.06 ± 0.11*	12	−6.90 ± 0.11	11	−8.72 ± 0.06	12	a
β1-W199N	−8.88 ± 0.06*	84.6 ± 5.4	8	−9.93 ± 0.09	11	−7.21 ± 0.08	12	−9.22 ± 0.10	13	−10.12 ± 0.12	12
β1-V189T-L195Q	−8.22 ± 0.04	87.0 ± 3.4	8	−9.49 ± 0.09	13	−7.67 ± 0.10*	11	−8.81 ± 0.11	14	−9.58 ± 0.09	12
β1-V189T-W199Y	−8.62 ± 0.07*	84.2 ± 2.1	8	−9.93 ± 0.08	12	−7.85 ± 0.08*	10	−8.80 ± 0.08	12	−10.10 ± 0.15	11
β1-L195Q-W199Y	−8.41 ± 0.06	79.6 ± 2.0	8	−9.68 ± 0.06	13	−7.24 ± 0.06	11	−9.04 ± 0.11	12	−9.82 ± 0.14	9
β1-V189T-L195Q-W199Y	−8.44 ± 0.04	82.7 ± 2.7	8	−9.85 ± 0.08	14	−7.92 ± 0.06*	11	−8.99 ± 0.09	14	−9.79 ± 0.11	12
β1-TM4	−8.32 ± 0.05	77.3 ± 1.9	23	−9.43 ± 0.09	35	−7.94 ± 0.06*	31	−9.08 ± 0.07	29	−9.98 ± 0.09	35
	CGP12177 Log EC50	% Isoprenaline	n	Log KB CGP20712A	n	Log KB ICI118551	n	Log KB propranolol	n
CGP12177 as agonist: stable mixed populations of cells
β1-WT	−8.12 ± 0.07	53.3 ± 1.5	21	−7.37 ± 0.08	37	−5.86 ± 0.08	22	−6.69 ± 0.08	27
β1-A187G	−8.21 ± 0.09	41.5 ± 3.1	7	−7.25 ± 0.11	12	−5.81 ± 0.08	7	−6.72 ± 0.12	10
β1-V189T	−8.01 ± 0.07	44.1 ± 2.3	11	−7.64 ± 0.08	21	−6.36 ± 0.09*	16	−6.98 ± 0.08	24
β1-L195Q	−9.00 ± 0.06*	32.9 ± 1.4*	11	−8.53 ± 0.09*	19	−6.64 ± 0.11*	13	−7.74 ± 0.12*	21
β1-W199Y	−9.16 ± 0.05*	40.2 ± 2.3*	11	−8.46 ± 0.07*	23	−6.69 ± 0.09*	17	−7.90 ± 0.09*	29
β1-W199A	−9.46 ± 0.11*	68.6 ± 4.0*	8	−8.79 ± 0.10*	12	−6.88 ± 0.11*	9	−7.96 ± 0.15*	9
β1-W199D	−9.48 ± 0.08*	45.3 ± 3.3	8	−8.94 ± 0.09*	17	−6.98 ± 0.16*	13	−8.87 ± 0.12*	20
β1-W199F	−8.95 ± 0.13*	58.6 ± 3.4	8	−8.10 ± 0.07*	13	−6.53 ± 0.10*	10	−7.61 ± 0.17*	12
β1-W199K	−9.40 ± 0.05*	85.1 ± 2.5*	8	−8.13 ± 0.10*	18	−6.34 ± 0.09*	13	−7.47 ± 0.08*	20
β1-W199L	−9.20 ± 0.06*	82.9 ± 3.2*	8	−8.12 ± 0.09*	20	−6.45 ± 0.08*	13	−7.56 ± 0.11*	20
β1-W199N	−9.27 ± 0.12*	47.7 ± 3.2	8	−8.26 ± 0.11*	12	−6.67 ± 0.10*	10	−8.19 ± 0.09*	17
β1-V189T-L195Q	−9.21 ± 0.06*	35.1 ± 2.6*	8	−8.64 ± 0.09*	15	−7.11 ± 0.08*	12	−8.17 ± 0.10*	14
β1-V189T-W199Y	−9.07 ± 0.11*	43.2 ± 2.9	8	−8.70 ± 0.08*	18	−7.42 ± 0.14*	13	−7.82 ± 0.10*	16
β1-L195Q-W199Y	−9.51 ± 0.06*	41.3 ± 2.3	8	−9.17 ± 0.12*	15	−6.94 ± 0.09*	12	−8.15 ± 0.11*	15
β1-V189T-L195Q-W199Y	−9.40 ± 0.07*	39.2 ± 2.4*	8	−9.21 ± 0.12*	14	−7.49 ± 0.07*	12	−8.66 ± 0.11*	16
β1-TM4	−9.43 ± 0.06*	33.9 ± 1.5*	19	−8.98 ± 0.07*	39	−7.64 ± 0.08*	32	−8.39 ± 0.06*	40

^aCGP12177 is very efficacious in these mutants, similar to cimaterol, and it is therefore not possible to measure a log K_B value for CGP12177 by a rightward shift of the cimaterol concentration response.

^*P < 0.001 One-way ANOVA with post hoc Newman–Keuls comparing values from the mutant receptors with those obtained from the β₁-WT, thus the log K_B for ICI118551 at β₁-V189T is different from that obtained from the β₁-WT with P < 0.001

TABLE 9.

The ratio of the log K_B value for CGP12177 (as an antagonist of cimaterol; thus, at the high-affinity conformation) and log EC₅₀ value for CGP12177 (as an agonist of the secondary low-affinity conformation) for responses seen at β₁-WT, β₁-TM4, and β₁-receptors with point mutations in TM4 (data taken from Table 8)

Ratio of the affinity for antagonists at the two conformations is also given (i.e., log K_B for antagonism of cimaterol at high affinity site/log K_B for antagonism of CGP12177 at low affinity site). This table examines the ratios of the antagonist affinities and shows that the difference in affinity between the antagonists at the catecholamine conformation (in the presence of cimaterol) and secondary conformation (presence of CGP12177) is substantially less with the L195 and W199 mutations, suggesting that these are important for the secondary conformation. Although V189T alters the affinity of ICI118551, the ratio of affinity does not change, suggesting that V189T does not have a major role in the secondary conformation.

	Ratio of Log KB CGP12177/ Log EC50 CGP12177	Ratio of Log KB (Cimaterol as Agonist)/Log KB (CGP12177 as Agonist)
		CGP20712A	ICI118551	Propranolol
β1-WT	1.89	1.99	1.27	2.00
β1-A187G	1.78	1.98	1.21	2.00
β1-V189T	1.80	1.62	1.37	1.63
β1-L195Q	0.85	0.44	0.48	1.12
β1-W199Y	0.90	1.20	0.70	1.11
β1-W199A	0.61	0.72	0.38	1.16
β1-W199D	0.37	0.94	0.50	0.30
β1-W199F	1.00	2.04	0.66	1.48
β1-W199K	a	1.33	0.07	0.53
β1-W199L	a	1.94	0.45	1.16
β1-W199N	0.85	1.67	0.54	1.03
β1-V189T-L195Q	0.64	0.85	0.56	0.64
β1-V189T-W199Y	1.03	1.23	0.43	0.98
β1-L195Q-W199Y	0.31	0.51	0.30	0.89
β1-V189T-L195Q-W199Y	0.39	0.64	0.43	0.33
β1-TM4	0.55	0.45	0.30	0.69

^aCGP12177 is very efficacious in these mutants, similar to cimaterol, and it is therefore not possible to measure a log K_B value for CGP12177 by a rightward shift of the cimaterol concentration response.

Fig. 3.

CRE-SPAP production in response to cimaterol in the absence and presence of 10 nM and 100 nM CGP20712A and 1µM ICI118551 in stable mixed populations of cell lines expressing (A) β₁-WT, (B) β₁-V189T, (C) β₁-L195Q, (D) β₁-W199Y, and (E) β₁-TM4. Bars represent basal CRE-SPAP production, that in response to 10 µM isoprenaline or that in response to 10 nM or 100 nM CGP20712A or 1 µM ICI118551 alone. Data points are mean ± S.E.M. of triplicate determinations. These single experiments are representative of (A) 24, (B) 12, (C) 12, (D) 12, and (E) 23 separate experiments. This figure shows that CGP20712A inhibits cimaterol responses in all of these receptors with similar ability; however, ICI118551 is able to inhibit β₁-V189T and β₁-TM4 to a greater extent than the others.

Fig. 4.

CRE-SPAP production in response to CGP12177 in the absence and presence of 100 nM and 1 µM CGP20712A and 10 µM ICI118551 in stable mixed populations of cell lines expressing (A) β₁-WT, (B) β₁-V189T, (C) β₁-L195Q, (D) β1-W199Y, and (E) β₁-TM4. Bars represent basal CRE-SPAP production, that in response to 10 µM isoprenaline or that in response to 100 nM or 1 µM CGP20712A or 10 µM ICI118551 alone. Data points are mean ± S.E.M. of triplicate determinations. These single experiments are representative of (A) 21, (B) 11, (C) 11, (D) 11, and (E) 20 separate experiments. This figure shows that secondary conformation CGP12177 responses are more readily inhibited by CGP20712A and ICI118551 at β₁-L195Q, β₁-W199Y, and β₁-TM4 than at β₁-WT, suggesting that the secondary conformation is compromised in the β₁-L195Q, β₁-W199Y, and β₁-TM4 mutants. β₁-V189T, while remaining more resistant to CGP20712A inhibition (i.e., demonstrating preservation of the secondary conformation), is once again more sensitive to ICI118551 inhibition.

We also examined the impact of these specific mutations in combination with each other to determine whether the effects of each change were additive or whether one single change was sufficient to account for the secondary site. A combination of L195Q and W199Y largely achieved the same effect as the whole TM4 swap, with the exception of the affinity of ICI118551 (Table 8). However, addition of the V189T mutation (thus making a triple β₁-V189T-L195Q-W199Y) effectively normalized the affinity of ICI118551 to that observed with the full TM4 swap (Fig. 5, A and B). This suggests that β₁-V189T-L195Q-W199Y has also lost the secondary conformation.

Fig. 5.

CRE-SPAP production in response to cimaterol (A and C) and CGP12177 (B and D) in the absence and presence of 10 nM, 100 nM, or 1 µM CGP20712A and 10 µM ICI118551 in stable mixed populations of cell lines expressing β₁-V189T-L195Q-W199Y (A and B) or β₁-W199D (C and D). Bars represent basal CRE-SPAP production, that in response to 10 µM isoprenaline or that in response to 10 nM, 100 nM, or 1 µM CGP20712A or 10 µM ICI118551 alone. Data points are mean ± S.E.M. of triplicate determinations. These single experiments are representative of eight separate experiments in each case. This figure shows that CGP20712A and ICI118551 are able to inhibit cimaterol and CGP12177 responses to a similar degree at both β₁-V189T-L195Q-W199Y and β₁-W199D, unlike that seen at β₁-WT (Figs. 3A and 4A). This suggests that while both are primarily β₁-adrenoceptors (given that both are more readily inhibited by CGP20712A than ICI118551) the secondary conformation is compromised in both of these mutant receptors.

In addition to the simple substitution of β₂-adrenoceptor residues for their equivalents in the β₁-adrenoceptor, we also examined the effects of different substitutions at position 199. Thus at position 199, as well as mutating the tryptophan (W) to tyrosine (Y, as in the human β₂-adrenoceptor) it was also mutated to another aromatic residue (phenylalanine, F), a hydrophobic residue (leucine, L), a basic residue (lysine, K), an amide (asparagine, N), an acid (aspartic acid, D) and the small amino acid alanine (A). Substitution of W for other residues had little effect on the nM EC₅₀ for CGP12177 observed with the tyrosine (Y) residue from the β₂-adrenoceptor, and most appeared to have lost, or at least compromised, the secondary conformation (Fig. 5, C and D; Table 8). Interestingly, the W199K and W199L mutations decreased the EC₅₀ of cimaterol (top half of Table 8) while increasing the efficacy of CGP12177 which then appeared as a virtual full agonist in these two mutations (bottom half of Table 8).

Confirmation of Loss of the Secondary Site with Mutations V189T, L195Q, and W199Y: ³H-cAMP Accumulation in Transiently Transfected Cells.

So far, identification of the role of amino acids at positions 195, 199, and possibly 189 in the secondary conformation pharmacology of the β₁-adrenoceptor has relied on observation of 1) different affinity values for antagonists in the presence of cimaterol and CGP12177, and 2) the comparison between the partial agonist log EC₅₀ value and log K_B value for CGP12177. To confirm that these amino acids were indeed important, evidence was sought using a different assay to explore two further strands of evidence: the stimulation resulting from CGP12177 in the presence of a fixed concentration of cimaterol and the effect on the response to pindolol. To examine whether cimaterol and CGP12177 were competing at the same site of the receptor, CGP12177 concentration responses were examined in the presence of different fixed concentrations of cimaterol. Low concentrations of CGP12177 competed with cimaterol causing a reduction in the stimulatory response. However, higher CGP12177 concentrations stimulated responses, creating a “dip” in the curve, for the two conformation β₁-WT adrenoceptor (Pak and Fishman, 1996; Fig. 6A). This “dip” was absent in the single conformation β₂-WT adrenoceptor and at β₁-TM4 (Fig. 6, C and E; Table 10). The β₁-V189T mutation alone had little effect on the two conformations (Fig. 7A; Table 10), although the “dip” appeared reduced in both L195Q and W199Y mutations alone. However, the “dip” was absent from β₁-V189T-L195Q-W199Y and β₁-W199D (Fig. 8, A and C; Table 10) in keeping with a single conformation receptor. Interestingly, in the transiently transfected cells, the β₁-WT ³H-cAMP accumulation response to CGP12177 was best described by a two-component response (Fig. 6A; Table 10). Hints that the CGP12177 response may, at times, be biphasic have been seen before (e.g., Fig. 4 of Baker et al., 2013) and are also potentially present in the stable β₁-WT cell lines and stable mixed β₁-WT population data (Figs. 1B and 4A, respectively). Interestingly, mutations L195Q and W199Y (or W199D) removed the biphasic response of CGP12177 (Figs. 7, C and E, and 8C; Table 10). Finally, concentration responses to the biphasic ligand pindolol were examined (Figs. 6, B, D and F, 7, B, D, and F, and 8, B and D; Table 10). Both the racemic and S-isomer of this ligand have previously been shown to stimulate agonist responses via both conformations of the β₁-adrenoceptor (Walter et al., 1984; Baker, 2010a). The data obtained with this agonist were consistent with the observations made with CGP12177 (apart from W199Y, which continued to exhibit two-site pharmacology with pindolol; Table 10). However, this was not the case with W199D or combinations of W199Y with L195Q and/or V189T (Table 10).

TABLE 10.

Log EC₅₀ values and % maximum response to isoprenaline for ³H-cAMP accumulation in response to CGP12177 and pindolol at β₁-WT, β₂-WT, and β₁-adrenoceptors with single point mutations or combinations of these mutations

These data were obtained from transiently transfected cells, and n refers to the number of separate experiments, each conducted in a separate transiently transfected population of cells. This table shows that the biphasic responses to CGP12177 and pindolol, and the “dip” in the curve seen when CGP12177 competed with a fixed concentration of cimaterol (see Figs. 6–8) observed in the β₁-WT, are lost with several β₁-TM4 mutant receptors, including the β₁-L195Q-W199Y, β₁-V189T-L195Q-W199Y, and β₁-W199D receptors.

	CGP12177 as Agonist: Transiently Transfected Cells						Pindolol as Agonist: Transiently Transfected Cells
	Log EC150	Log EC250	% Site 1	% Max Isoprenaline	Competition with Cimaterol (Dip)	n	Log EC150	Log EC250	% Site 1	% Max Isoprenaline	n
β1-WT	−9.28 ± 0.17	−7.36 ± 0.21	42.3 ± 7.2	53.0 ± 2.9	Yes	6	−9.10 ± 0.04	−5.93 ± 0.18	57.1 ± 3.3	30.0 ± 2.0	8
β1-V189T	−9.37 ± 0.12	−7.16 ± 0.06	45.1 ± 2.7	48.4 ± 4.6	Yes	6	−8.87 ± 0.05	−5.78 ± 0.14	67.1 ± 3.3	22.3 ± 2.9	5
β1-L195Q	−9.02 ± 0.09			37.4 ± 1.5	Slight	3	−9.14 ± 0.08			27.9 ± 1.6	3
β1-W199Y	−9.28 ± 0.11			41.0 ± 3.2	Slight	4	−9.02 ± 0.21	−6.49 ± 0.18	71.7 ± 1.4	32.4 ± 3.0	4
β1-W199D	−9.80 ± 0.06			49.9 ± 1.0	No	6	−8.96 ± 0.23			28.9 ± 1.8	4
β1-L195Q-W199Y	−9.62 ± 0.05			39.9 ± 3.7	No	4	−9.33 ± 0.08			26.2 ± 3.0	4
β1-V189T-L195Q-W199Y	−9.52 ± 0.05			35.3 ± 6.3	No	3	−9.11 ± 0.03			21.8 ± 1.4	4
β1-TM4	−9.54 ± 0.06			30.9 ± 1.8	No	4	−9.25 ± 0.08			23.0 ± 3.1	3
β2-WT	−9.81 ± 0.15			11.2 ± 1.6	No	4	−9.58 ± 0.06			11.6 ± 0.4	4

Fig. 7.

³H-cAMP accumulation in response to CGP12177 (A, C, and E), and pindolol (B, D, and F) in transiently transfected cells expressing β₁-V189T (A and B), β₁-L195Q (C and D), and β₁-W199Y (E and F). The CGP12177 responses were examined in the absence and presence of 3, 10, and 30 nM cimaterol. Bars represent basal ³H-cAMP accumulation, that in response to 10 µM isoprenaline or that in response to 3, 10, and 30 nM cimaterol alone. Data points are mean ± S.E.M. of triplicate determinations. These single experiments are representative of (A) six, (B) five, (C) three, (D) three, (E) four, and (F) four separate experiments. This figure shows that both the “dip” in the CGP12177 + cimaterol curve and biphasic pindolol response are still present at the β₁-V189T, suggesting conservation of the secondary conformation. However, both are reduced at β₁-L195Q and β₁-W199Y, suggesting some loss of the secondary conformation at β₁-L195Q and β₁-W199Y (see Fig. 6A for β₁-WT comparison).

Fig. 8.

³H-cAMP accumulation in response to CGP12177 (A and C) and pindolol (B and D) in transiently transfected cells expressing β₁-V189T-L195Q-W199Y (A and B) and β₁-W199D (C and D). The CGP12177 responses were examined in the absence and presence of 3, 10, and 30 nM cimaterol. Bars represent basal ³H-cAMP accumulation, that in response to 10 µM isoprenaline or that in response to 3, 10, and 30 nM cimaterol alone. Data points are mean ± S.E.M. of triplicate determinations. These single experiments are representative of (A) three, (B) four, (C) six, and (D) four separate experiments. This figure shows that the “dip” in the CGP12177 + cimaterol curve and biphasic pindolol response, both present at β₁-WT with two conformations (see Fig. 6A), are absent in β₁-V189T-L195Q-W199Y and β₁-W199D, suggesting that neither of these receptors has the secondary conformation.

Discussion

The human β₁-WT exists in at least two agonist conformations: a high-affinity catecholamine conformation and a low-affinity secondary (CGP12177) conformation (Pak and Fishman, 1996; Kaumann and Molenaar, 2008). This secondary conformation does not occur in the closely related human β₂-WT (Pak and Fishman, 1996; Baker et al., 2002). This study attempted to identify the key area responsible for this secondary conformation using a β₁/β₂-chimeric receptor approach (Isogaya et al., 1999; Kaumann and Molenaar 2008).

At the β₁-WT, the secondary conformation was clearly delineated from the primary catecholamine conformation by two pieces of evidence. Firstly, the cimaterol responses (catecholamine conformation agonist) were readily inhibited by antagonists, whereas CGP12177 responses required much greater antagonist concentrations, yielding lower log K_B values (Table 2). Secondly, CGP12177 readily inhibited the cimaterol responses at low concentrations (log K_B −9.55), but required high concentrations to produce its agonist response (log EC₅₀ −8.18), in keeping with previous studies (e.g., Pak and Fishman, 1996; Lowe et al., 2002; Baker et al., 2003; Joseph et al., 2004a; Baker, 2005a). At the β₂-WT, cimaterol and CGP12177 responses were inhibited by similar concentrations of antagonist and the EC₅₀/K_B discrepancy for CGP12177 was less marked, suggesting competition of the two ligands at the same conformation.

When the TM-chimeras were examined, antagonism of cimaterol responses suggested that all receptors (β₁-TM1 to β₁-TM7 and β₂-TM1 to β₂-TM7; Tables 2–5) retained their respective pharmacological parent β₁ or β₂-subtype, even when one entire transmembrane region was altered (i.e., high affinity for CGP20712A at β₁-adrenoceptors, high affinity for ICI118551 at β₂-adrenoceptors). Significant changes, however, were seen in the secondary conformation in β₁-TM4. Here, antagonists inhibited the cimaterol and CGP12177 responses to give similar log K_B values (Tables 4 and 5), and the concentration of CGP12177 required to inhibit the cimaterol response (−9.63) approached that required to stimulate the agonist response (−9.27). Taken together this strongly suggests that the secondary conformation is absent in β₁-TM4. Although direct comparison of partial agonist efficacy across cell lines is difficult, the log EC₅₀ for CGP12177 (−9.27) is substantially more potent (left-shifted) at β₁-TM4 than any other β₁-TM chimeric receptor (−7.83 to −8.27; Tables 2 and 3) even though the response is less efficacious (% isoprenaline maximum response achieved) and the receptor expression level is the lowest of the β₁ cell lines. If CGP12177 were simply more efficacious at β₁-TM4, the left-shifted EC₅₀ would be accompanied by an increase in the maximum response obtained. The increase in potency, but decrease in efficacy, suggests that something different is happening at β₁-TM4, and indeed these values are similar to those obtained for CGP12177 at the β₂-WT and β₂-chimeric receptors (Tables 2–5). This differs from the comment in Kaumann and Molenaar (2008) regarding TM5.

Thus, β1-TM4 appears to be a β₁-adrenoceptor without the secondary conformation. When the β₂-chimeras were examined (Tables 4 and 5), the antagonist affinities (whether cimaterol or CGP12177 was the agonist), log EC₅₀, and log K_B values for CGP12177 were similar for all β₂-chimeric receptors to those of β₂-WT and β₁-TM4. Thus, although the secondary site was lost in β₁-WT, it was not recreated in β₂-WT by the TM4 swap, and, therefore, other areas of the receptor must contribute to this secondary conformation.

To locate the specific amino acids involved in the secondary conformation, intermediate stages in the generation of β₁-TM4 were examined. The log EC₅₀ for CGP12177 was more potent and the antagonist affinities were higher for β₁-TM4 stage 5 and the full β₁-TM4 chimeric receptor (Tables 1, 6, and 7). ICI118551 had a slightly different pattern with increased affinity at both conformations from β₁-TM4 stage 4 onwards. This suggests that amino acids L195, W199, and maybe V189 are important for the secondary conformation. These mutations were therefore examined alone. For β₁-L195Q and β₁-W199Y, the antagonist affinities (log K_B) in the presence of CGP12177 were higher, and the log EC₅₀ for CGP12177 more potent, than at β₁-WT, suggesting that both of these residues are important in the secondary conformation. ICI118551 had higher affinity for β₁-V189T in the presence of both cimaterol and CGP12177, suggesting all-around (not conformation specific) higher affinity. This was seen in all receptors containing a V189T mutation (i.e., β₁-TM4, β₁-V189T, β₁-V189T-L195Q, β₁-V189T-W199Y, and β₁-V189T-L195Q-W199Y). ICI118551 also showed lower affinity at β₂-TM4 (containing the reverse T164V mutation). V189T is therefore specifically involved in the affinity of ICI118551. CGP20712A and propranolol both had slightly higher affinity in the presence of CGP12177 at β₁-V189T than β₁-WT, giving rise to lower ratios (Table 9) and suggesting a potential minor role for V189T in the secondary conformation.

However, neither L195Q nor W199Y alone gave K_B or EC₅₀ values that matched those of the β₁-TM4 (Tables 8 and 9). Double and triple mutations were therefore made. At β₁-V189TL-195Q-W199Y, antagonist affinities and log EC₅₀ values for CGP12177 were the same as for the β₁-TM4, suggesting complete abolishment of the secondary conformation. β₁-L195Q-W199Y also achieved this (with the exception of ICI118551 affinity because this lacks the V189T mutation). β₁-V189T-L195Q and β₁-V189T-W199Y both had the higher ICI118551 affinity, but the secondary conformation affinity for CGP20712A and propranolol did not reach that of β₁-TM4. This suggests that while L195Q and W199Y are each important, their effects are additive and that both are needed to completely abolish the secondary conformation. The effect of V189T, while important for ICI118551 affinity, was minor for the secondary conformation.

The effect of charge at the W199 site was then assessed. The addition of a negatively charged group (W199D) greatly increased CGP20712A and propranolol affinity in the presence of CGP12177 to that of β₁-TM4, β₁-L195Q-W199Y, and β₁-V189T-L195Q-W199Y, suggesting that this single change could remove the secondary conformation (although ICI118551 affinity does not reach that of β₁-TM4 or β₁-V189T-L195Q-W199Y because V189T is missing). Thus, the single amino acid change of W199D appeared to create a β₁-adrenoceptor with β₁ pharmacological characteristics (i.e., 100-fold higher affinity for CGP20712A than ICI118551) but that is completely devoid of the secondary conformation.

Finally, to confirm L195Q, W199Y, (and maybe V189T) involvement in the secondary conformation, ³H-cAMP accumulation was investigated (Figs. 6–8). At low concentrations, CGP12177 inhibited the stimulation of the catecholamine conformation by cimaterol, while at higher concentrations CGP12177 stimulated the secondary conformation, resulting in a “dip” in the response to cimaterol in the presence of CGP12177 (Fig. 6A) (Pak and Fishman, 1996; Konkar et al., 2000; Baker et al., 2003; Baker 2005a). This “dip” was present for β₁-V189T (Fig. 7A) but reduced at β₁-L195Q and β₁-W199Y (Fig. 7, C and E). However, there was no “dip” in β₂-WT, β₁-TM4, β₁-L195Q-W199Y, β₁-V189T-L195Q-W199Y, or β₁-W199D (Figs. 6, C and E, and 8, A and C). At all these receptors, CGP12177 competed with cimaterol to inhibit the cimaterol response at the same as (or just to the right of) that of the EC₅₀ value, strongly suggesting competition at the same conformation.

Pindolol has a biphasic concentration response curve at the β₁-adrenoceptor. The first (higher affinity) component is thought to occur via the catecholamine conformation, while the second component via the secondary conformation (Walter et al., 1984; Baker et al., 2003). Therefore, the pindolol response should become a single-component response at β₁-TM4 and the other receptors lacking the secondary conformation. Indeed, pindolol was found to have a two-component concentration response curve at β₁-WT and β₁-V189T, but a single-component response curve at β₂-WT, β₁-TM4, β₁-L195Q-W199Y, β₁-T189V-L195Q-W199Y, and β₁-W199D, confirming the crucial roles of L195 and W199 in the secondary conformation. Interestingly, for all β-adrenoceptors with a known sequence and two pharmacologically confirmed conformations, V, L, and W, or similar, are found at equivalent positions for 189, 195, and 199, respectively, while those of one-conformation β-adrenoceptors have T, Q, and Y (Table 11).

TABLE 11.

Amino acids occurring at positions 189, 195, and 199 or equivalent positions in other β-adrenoceptors from several species

	UniProtKB/TrEMBL Accession Number	Position 189	Position 195	Position 199
Two-conformation receptors
Human β1	P08588	V 189	L 195	W 199
Rat β1	P18090	V 189	L 195	W 199
Mouse β1	P34971	V 189	L 195	W 199
Guinea pig β1	B0FL73	V 189	L 195	W 199
Cat β1	Q9TST6	V 189	L 195	W 199
Turkey β1	P07700	V 172	M 178	W 182
Turkey β4C	P43141	I 154	M 160	W 164
Human β3	P13945	V 168	M 174	W 178
One-conformation receptors
Human β2	P07550	T 164	Q 170	Y 174
Turkey β3C	a	T 161	Q 167	Y 171

^aSequence from Fig. 1 in Baker, 2010b.

A recent turkey β₁-adrenoceptor crystal structure suggests that TM4 is involved in a β₁-dimer interface with TM5 (Huang et al., 2013). Figure 9A shows that the equivalent key turkey TM4 residues (V172, M178, and W182; Table 11) are largely outward facing, away from the binding pocket, but lie within the TM4/TM5 dimer interface (Fig. 9B). However, the key binding residues in TM3 (D138) and TM7 (N363), that are essential for both the catecholamine and secondary conformation responses (Baker et al., 2008), are both inward-facing residues within the ligand binding pocket (Fig. 9A). It is therefore interesting to speculate that both binding site activation and receptor dimerization are involved in the secondary conformation. However, more work is required to understand how and where molecules bind to create the distinct pharmacological patterns observed.

Fig. 9.

Location of equivalent key residues (V172, M178, W182) in the crystal structure of the turkey β₁-adrenoceptor described by Huang et al. (2013). This structure was obtained in the ligand-free state and in a lipid membrane–like environment and suggested that TM4 was involved in a dimer interface with TM5. Views of one protomer of the oligomeric structure (PDB ID 4GPO) reported by Huang et al. (2013) were generated with Cn3D (National Center for Biotechnology Information). (A) Residues V172, M178, and W182 (equivalent to the human V189, L195, and W199) and D121 and N329 (equivalent to D138 and N363, identified in a previous study as essential for both catecholamine and secondary conformation functional responses) (Baker et al., 2008) are highlighted in yellow. (B) View of V172, M178, and W182 (highlighted in red) with the key residues for dimerization in TM5 (R205, A206, A210, I218, R229), TM4 (L171), and the proximal region of extracellular loop 2 (W181, R183) identified by Huang et al. (2013) highlighted in purple. Interestingly, V172 and W182 are on the same outward facing surface as the key dimerization residues, with W182 sandwiched between the two residues (W181, R183), and V172 is next to another (L171) identified by Huang et al. (2013).

In conclusion, the β₁-adrenoceptor exists in two agonist conformations, and the key residues responsible for the secondary low-affinity conformation are in TM4. The mutation β₁-V189T has an effect on the affinity of ICI118551 and may have a minor role in secondary conformation pharmacology. However, the amino acids at positions 195 and 199 in TM4 both have a major role in the secondary conformation, and together L195Q and W199Y (or indeed W199D alone), abolish this secondary conformation and create a β₁-adrenoceptor with only one high-affinity agonist conformation. From Huang et al. (2013) data, it is tempting to speculate that oligomerization involving TM4 may underlie the secondary β₁-adrenoceptor conformation.

Abbreviations

CHO

Chinese hamster ovary

CGP12177

(−)-4-(3-tert-butylamino-2-hydroxypropoxy)-benzimidazol-2-one

CGP20712A

2-hydroxy-5-(2-[{hydroxy-3-(4-[1-methyl-4-trifluoromethyl-2-imidazolyl]phenoxy)propyl}amino]ethoxy)benzamide

CRE-SPAP

cAMP response element upstream of a secreted placental alkaline phosphatase reporter gene

ICI118551

(−)-1-(2,3-[dihydro-7-methyl-1H-inden-4-yl]oxy)-3-([1-methylethyl]-amino)-2-butanol

K_B

half receptor occupancy

sfm

serum free media

TM

transmembrane

WT

wild-type

Authorship Contributions

Participated in research design: Baker.

Conducted experiments: Baker, Proudman.

Contributed new reagents or analytic tools: Baker, Proudman, Hill.

Performed data analysis: Baker.

Wrote or contributed to the writing of the manuscript: Baker, Hill.