JPET Introducing ALZET?ew Model 2006 Pump

Home Help [Feedback] [For Subscribers] [Archive] [Search] [Contents]
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow An erratum has been published
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Sennitt, M. V.
Right arrow Articles by Arch, J. R. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Sennitt, M. V.
Right arrow Articles by Arch, J. R. S.

Vol. 285, Issue 3, 1084-1095, June 1998

The Contribution of Classical (beta 1/2-) and Atypical beta -Adrenoceptors to the Stimulation of Human White Adipocyte Lipolysis and Right Atrial Appendage Contraction by Novel beta 3-Adrenoceptor Agonists of Differing Selectivities

Matthew V. Sennitt1 , Alberto J. Kaumann2 , Peter Molenaar, Lee J. Beeley, Paul W. Young, John Kelly, Helen Chapman, Sian M. Henson, John M. Berge, David K. Dean, Nikesh R. Kotecha, Helen K. A. Morgan, Harshad K. Rami, Robert W. Ward, Mervyn Thompson, Shelagh Wilson, Stephen A. Smith, Michael A. Cawthorne1, Michael J. Stock and Jonathan R. S. Arch

SmithKline Beecham Pharmaceuticals, Great Burgh, Epsom, Surrey, The Frythe, Welwyn, Herts and New Frontiers Science Park, Harlow, Essex, CM19 5AW, UK (L.J.B., P.W.Y., J.K., H.C., S.M.H., J.M.B., D.K.D., N.R.K., H.K.A.M., H.K.R., R.W.R., M.T., S.W., S.A.S., J.R.S.A.); St Georges Hospital Medical School, Tooting, London, SW17 0RE, UK (M.V.S., M.J.S.); The Babraham Institute, Babraham, Cambridge, CB2 4AT, UK (A.J.K.) and Department of Pharmacology, University of Melbourne, Victoria 3052, Australia (P.M., A.J.K.)


    Abstract
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References

The role of beta 3- and other putative atypical beta -adrenoceptors in human white adipocytes and right atrial appendage has been investigated using CGP 12177 and novel phenylethanolamine and aryloxypropanolamine beta 3-adrenoceptor (beta 3AR) agonists with varying intrinsic activities and selectivities for human cloned beta AR subtypes. The ability to demonstrate beta 1/2AR antagonist-insensitive (beta 3 or other atypical beta AR-mediated) responses to CGP 12177 was critically dependent on the albumin batch used to prepare and incubate the adipocytes. Four aryloxypropanolamine selective beta 3AR agonists (SB-226552, SB-229432, SB-236923, SB-246982) consistently elicited beta 1/2AR antagonist-insensitive lipolysis. However, a phenylethanolamine (SB-220646) that was a selective full beta 3AR agonist elicited full lipolytic and inotropic responses that were sensitive to beta 1/2AR antagonism, despite it having very low efficacies at cloned beta 1- and beta 2ARs. A component of the response to another phenylethanolamine selective beta 3AR agonist (SB-215691) was insensitive to beta 1/2AR antagonism in some experiments. Because novel aryloxypropanolamine had a beta 1/2AR antagonist-insensitive inotropic effect, these results establish more firmly that beta 3ARs mediate lipolysis in human white adipocytes, and suggest that putative `beta 4ARs` mediate inotropic responses to CGP 12177. The results also illustrate the difficulty of predicting from studies on cloned beta ARs which beta ARs will mediate responses to agonists in tissues that have a high number of beta 1- and beta 2ARs or a low number of beta 3ARs.


    Introduction
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References

beta 3ARs were first identified by functional studies in animal gastrointestinal and adipose tissues (for review see Arch and Kaumann, 1993). Subsequently, the beta 3AR gene was cloned from human genomic DNA, and expression of beta 3AR mRNA has now been detected in various human tissues, including adipose, gastrointestinal and myocardial tissues (Krief et al., 1993; Berkowitz et al., 1995; Gauthier et al., 1996). However, although there are numerous reports describing functional responses mediated by beta 3ARs in animal tissues, evidence for such responses in human tissues is more limited, occasionally contradictory and sometimes open to alternative interpretation.

Much of the published evidence for beta 3AR-mediated responses in human tissues derives from studies using CGP 12177. In most tissues this aryloxypropanolamine is a potent antagonist of beta 1- and beta 2ARs, but at concentrations about 1000-fold higher it is an agonist of beta 3ARs. It is able to stimulate human beta 1ARs when they are expressed in high density (Pak and Fishman, 1996) but its agonist activity in tissues is usually insensitive to standard beta 1- and beta 2AR antagonists, indicating that responses are not mediated by beta 1- or beta 2ARs. Thus both human colon (De Ponti et al., 1996) and taenia coli (Kelly et al., 1997) are relaxed by CGP 12177, and the taenia coli response has been shown to be resistant to nadolol (pA2 <=  6). Similarly, in human white adipocytes Arner, Lonnqvist and their coworkers have consistently demonstrated lipolytic responses to CGP 12177 that are poorly blocked by standard beta 1- and beta 2AR antagonists (Lonnqvist et al., 1993; Hoffstedt et al., 1996). Some reports support these findings (Sennitt et al., 1995; Portillo et al., 1995; Tavernier et al., 1996) although there are earlier reports that do not (Langin et al., 1991; Van Liefde et al., 1994). Again in human right atrial appendage, CGP 12177 elicits a small inotropic response that is resistant to blockade by propranolol. The response is antagonized by bupranolol but only at higher concentrations than those that block beta 1- and beta 2ARs (Kaumann, 1996).

In contrast to these results for CGP 12177, demonstration of beta 3AR-mediated responses in human tissues using phenylethanolamines that are agonists of beta 1- and beta 2ARs as well as beta 3ARs has proved difficult. Isoproterenol, norepinephrine and the beta 3AR-selective agonist BRL-37344 each stimulate cyclic AMP accumulation in cells transfected with high numbers of human beta 3ARs, and they stimulate adenylyl cyclase in membranes from these cells (Emorine et al., 1994). However, responses to BRL-37344 (if they occur at all), and to isoproterenol and norepinephrine in human gut, adipose tissue and atrium are mediated primarily by beta 1- or beta 2- rather than beta 3ARs (MacLaughlin and MacDonald, 1991; Langin et al., 1991; Lonnqvist et al., 1993; Rosenbaum et al., 1993; Sennitt et al., 1995; Wang et al., 1996; Kaumann AJ and Sanders L, quoted in Arch and Kaumann, 1993), although responses to isoproterenol and norepinephrine in gut do involve a significant beta 3AR-mediated component (MacLaughlin and MacDonald, 1991; De Ponti et al., 1996; Kelly et al., 1997). Indeed, other than CGP 12177, only CL 316243, a phenylethanolamine with very low efficacy at beta 1- and beta 2ARs, has been shown to elicit a lipolytic response in human white adipocytes that is totally resistant to antagonism by 10-7 M propranolol (Hoffstedt et al., 1996); and only in human ventricle, for which a negative inotropic effect has been reported (Gauthier et al., 1996; but see Molenaar et al., 1997), and possibly platelets (Gill et al., 1991) is there evidence for a strong beta 3AR component in the response to BRL-37344 in a human tissue.

One possible explanation for the failure of BRL-37344 to elicit responses via beta 3ARs in human tissues is that its low efficacy and selectivity for the human (as compared with the rat or mouse) beta 3AR precludes it acting via these receptors in human tissues that express them in low numbers compared to beta 1- and beta 2ARs (Arch and Wilson, 1996; Wilson et al., 1996). Another possibility is that CGP 12177 in fact elicits its effects not via beta 3ARs but via a related, putative `beta 4AR' (Kaumann, 1997). Such an explanation has been used to interpret the pharmacology of cardiac responses in the rat, mouse, guinea pig, cat and ferret and has recently been extended to human atrium and ventricle(Kaumann, 1996; Kaumann, 1997; Kaumann and Molenaar, 1997; Molenaar et al., 1997). Thus cardiac pharmacology differs from colonic pharmacology not only by the failure of BRL-37344 and other beta 3AR-selective agonists to elicit positive inotropic responses via beta 3ARs, but also in a number of other respects (Kaumann and Molenaar, 1996; Kaumann, 1997). Finally, it is possible that in human tissues beta 3ARs adopt a conformation or associate with G proteins differently from in transfected cells, so that they are stimulated by CGP 12177 but not BRL-37344 (Kenakin, 1995; Arch, 1997).

Further insight into the role of atypical beta ARs in mediating functional responses in human tissues may come from studies on novel beta 3AR agonists with differing efficacies and relative potencies at beta 1-, beta 2- and beta 3ARs. We characterize a number of novel beta AR agonists of both the phenylethanolamine and aryloxypropanolamine type in terms of their pharmacology at human cloned beta 1-, beta 2- and beta 3ARs. We describe the lipolytic activities of these compounds in human white adipocytes, and in some cases their inotropic activities in human right atrial appendage, in the absence and presence of antagonists of beta 1- and beta 2ARs. The results not only support a role for beta 3ARs in human adipocytes, but also illustrate the difficulties of predicting the tissue pharmacology of agonists, especially phenylethanolamines, from their profiles in cells expressing cloned beta ARs.

The compounds studied may be broadly classified from their pharmacology and chemistry as: 1) the phenylethanolamines SB-215691 and SB-220646, which are similar to BRL-37344 in being agonists at beta 1- and beta 2- as well as beta 3ARs but have higher efficacy than BRL-37344 at human beta 3ARs; 2) the aryloxypropanolamines SB-226552, SB-229432 and SB-232602, which are similar to CGP 12177 in being primarily antagonists at beta 1- and beta 2ARs, but have far lower affinity than CGP 12177 at these receptors; 3) the aryloxypropanolamines SB-236923, SB-246982 and SB-248320 which are similar to the compounds in category 2) except that partial agonist activity at beta 1ARs was clearly demonstrated.

    Materials and Methods
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References

Cloned beta ARs and membrane preparation. Binding studies were conducted using membranes from CHO cells expressing 7100, 2300 and 3000 fmol/mg protein of human beta 1-, beta 2 and beta 3ARs, respectively. Adenylyl cyclase measurements were conducted using membranes from CHO cells expressing 210, 560 and 390 fmol/mg protein of human beta 1-, beta 2- and beta 3ARs, respectively. The CHO cells expressing the higher numbers of beta 1- and beta 2ARs were obtained from A D Strosberg (Tate et al., 1991) and those expressing the lower numbers from S B Liggett (Green et al., 1992). The CHO cells expressing beta 3ARs were derived in-house by transfecting cells with the short splice variant of the beta 3AR (Liggett, 1992) as described elsewhere (Wilson et al., 1996). We have found similar EC50 values and intrinsic activities (relative to isoproterenol) for isoproterenol, CGP 12177 and BRL-37344 for the short and long forms of the beta 3AR expressed at similar densities in CHO cells (Wilson S and Chambers JK, unpublished data).

All recombinant CHO cell lines were grown in alpha -MEM without nucleosides supplemented with 10% dialyzed serum. Production cultures were grown in 1 liter spinner flasks (Techne, Cambridge, UK) to approximately 106 cells/ml at which time the cultures were harvested by centrifugation (10 min at 1000 × g). The cells were washed twice (10 ml/109 cells) with ice-cold Dulbecco's phosphate-buffered saline without calcium or magnesium but including protease inhibitors (5 µg/ml benzamidine; 5 µg/ml leupeptin; 10 µg/ml soya bean trypsin inhibitor). The cell pellets were then resuspended and homogenized in 10 mM Tris, 1 mM EDTA, pH 7.4 at 4°C containing the above protease inhibitors for the cyclase assays, and 150 µg/ml bezamidine, 5 µg/ml leupeptin, 1 µg/ml aprotonin, 0.7 µg/ml pepstatin A and 5 µg/ml AEBSF for the other assays. Membranes were precipitated by centrifugation at 30,000 × g and washed using the same buffer prior to storage at -70°C. The protein content of each batch of membranes was determined using a commercial kit (Coomasie Plus Protein Assay Reagent, Pierce, Rockford, IL).

Radioligand binding. Saturation binding analysis was performed using concentrations of [125I] iodocyanopindolol, (Amersham International plc, UK) ranging from 1 pM - 1 nM for beta 1- and beta 2ARs and 0.1 to 12 nM for beta 3ARs. Membranes were incubated in assay buffer (50 mM Tris, 12.5 mM MgCl2, 2 mM EDTA, pH 7.4), at 37°C for 60 min in tubes pretreated with Sigmacote (Sigma Chemical Co. Ltd., Poole UK). Nonspecific binding was determined in the presence of 100 µM (-)-propranolol, which was shown in preliminary experiments to give the same amount of nonspecific binding as 1 or 10 µM (-)-propranolol. Competition studies were carried out using similar procedures, except that membranes were incubated with a Kd concentration of [125I] iodocyanopindolol (30, 70 and 600 pM for beta 1-, beta 2- and beta 3ARs, respectively) and varying concentrations of competing ligands. The specific activity of the [125I] iodocyanopindolol was 2000 Ci/mmol for the beta 1- and beta 2AR experiments and 300 Ci/mmol for the beta 3AR experiments. Bound radioligand was separated from free by rapid filtration through GF/C filters (Whatman Ltd., Maidstone, UK). Ki values were calculated from IC50 values according to the method of Cheng and Prusoff (1973). No specific binding could be detected in untransfected CHO cells.

Adenylyl cyclase activity. Adenylyl cyclase activity was measured using a similar method to that described previously (Chambers et al., 1994). Membranes (60 -120 µg of protein) were incubated in the presence of 25 mM Tris, 1.8 mM EDTA, 5 mM MgSO4, 1.0 mM ATP, 1 µM GTP, 240 U/ml creatine phosphokinase, 20 mM phosphocreatine, 10 mM theophylline, 0.5 µCi [alpha 32P]ATP (Amersham International plc) and varying concentrations of agonists at 30°C in a final volume of 0.1 ml. Incubations were started by addition of membrane and terminated after 20 min by addition of 1 ml ice-cold stop solution [0.25% sodium laurylsulphate, 5 mM ATP, 175 µM [3H] cyclic AMP (0.01 µCi), 10 mM Tris, 2 mM EDTA]. Cyclic AMP was isolated by sequential chromatography on Dowex cation exchange resin and aluminium oxide (Salomon et al., 1974).

Lipolysis. Human breast and axillary adipose tissue was obtained from breast tissue excised from patients that had undergone surgery either for breast cancer or breast reduction. Perirenal adipose tissue was taken from patients with either carcinomas or renal infection. The mean age and weight ± S.D. of the 34 patients from which breast or axillary adipose tissue was taken was 56 ± 17y and 68 ± 12 kg. Prescription of tamoxifen and heparin was common in these patients. Of the 15 patients from which kidney adipose tissue was taken eight were male and seven female. Their mean age was 64 ± 14y and their weights were 76 ± 7 and 75 ± 9 kg for males and females, respectively. All procedures were carried out with the approval of St. Georges Hospital, Tooting, London Clinical Committee. Tissue was rapidly transferred to the laboratory and the preparation of isolated fat cells began within 15 min.

Collagenase digestion was conducted at 37°C using 4 ml of Krebs-Hensleit buffer containing 5.5 mM glucose, 40 mg/ml bovine serum albumin and 2 mg/ml (170-204 U/g) collagenase (Worthington, Lorne Laboratories, Twyford, UK) per gram of adipose tissue for 30 to 45 min. The digestate was filtered through a plastic gauze to remove undigested material and the cells were harvested by flotation under gravity, the infranatant being removed by aspiration. The cells were washed a further three times in Krebs-Henseleit buffer.

Lipolytic activity was measured at 37°C by incubating isolated adipocytes (10-25 mg of total cell lipid) in 1 ml Krebs-Henseleit buffer containing 4 mg/ml albumin and 5.5 mM glucose with the compounds to be tested for 90 min in an atmosphere of 95% O2 and 5% CO2. Lipolysis was terminated by addition of trichloroacetic acid to give a concentration of 10%. After precipitation of protein, glycerol was determined by the method of Boobis and Maughan (1983).

The choice of the albumin source and batch was found to be crucial in obtaining high sensitivity to isoproterenol, and good intrinsic activity to CGP 12177 and the other beta 3AR agonists. Therefore batches of albumin were screened for their ability to permit good responses to isoproterenol and CGP 12177. The results described for the novel agonists were obtained using Sigma fatty acid free albumin batch 122H9307, Boehringer fraction V albumin lot 14065725-76 or Pentax albumin low hormone grade. These batches produced similar data for isoproterenol and CGP 12177. Even with this precaution, patient to patient variability resulted in poor responses to CGP 12177 and the novel agonists in a proportion of the experiments. Those experiments (about 20%) in which the intrinsic activity of CGP 12177 was less than 0.14 relative to isoproterenol were discarded.

Atrial contraction. Right atrial appendages were obtained with ethical committee approval from patients undergoing surgery at Papworth-Everard (Cambridgshire, UK) and Royal Melbourne Public and Private Hospital (Australia) for coronary artery bypass grafts. All patients were males, their mean age ± S.D. being 64 ± 9y. The majority of these patients had been prescribed beta AR blocking drugs (atenolol, metoprolol or propranolol) for at least three months prior to surgery. Some of the patients had been prescibed some of the following drugs: nifedipine, diltiazem, ranitidine, analapril, digoxin, omeprazole, isosorbidemononitrate.

Right atrial appendages were obtained, transported and dissected as previously described (Gille et al., 1985).They were sectioned into two to three strips and set up to contract at 1 Hz in an apparatus with a 50-ml bath (Blinks, 1965) in 106 mM NaCl, 5 mM KCl, 2.25 mM CaCl2, 0.5 mM MgSO4, 1 mM Na2HPO4, 34 mM NaHCO3, 5 mM fumarate, 5 mM glutamate, 10 mM glucose, 0.04 mM EDTA, equilibrated with 95% O2 and 5% CO2 at 37°C. The water was deionized and double distilled. The tissues were attached to Swema 4-45 strain gauge transducers and force recorded on a Watanabe polygraph. The strips were driven with square-wave pulses of 5-msec duration and of just over threshold voltage. After the determination of a length-tension curve, the length of each strip was set to obtain 50% of the resting tension associated with maximum developed force.

Single cumulative concentration-effect curves to the novel compounds were determined in the absence of antagonists, and in the presence of either 200 nM (-)-propranolol, or, for SB-220646, 300 nM CGP 20712A or 50 nM ICI 118551. Antagonists were present for 45 min before a curve was begun. Some strips were treated with 10 µM CGP 12177 in the presence of 200 nM (-)-propranolol to confirm previous evidence (Kaumann, 1996) for an inotropic effect of CGP 12177. The experiments were concluded by the administration of a beta -adrenoceptor saturating concentration of (-)-isoproterenol (200 µM) and after an equilibrium response to (-)-isoproterenol was established, by raising the Ca++ concentration to 6.75 mM.

Data analysis. pD2 values are expressed relative to each compound's own maximum effect or its effect at 10-4 M. Intrinsic activity values were measured relative to (-)-isoproterenol or, where indicated, (±)-CGP 12177. All results are given as means ± S.E. Where nadolol (10-6 M) failed to shift the lipolysis concentration-response curve a pKB value of <=  6 was used in compiling table 3.

Compounds. The novel compounds were synthesized in the laboratories of SmithKline Beecham Pharmaceuticals by the methods that are described in patent applications WO 95/07284, WO 95/25104, WO 96/04233 and PCT/EP97/01286 and in Beeley et al. (1997).Their structures are shown in figure 1. For the cloned receptor and lipolysis work the compounds were dissolved at a concentration of 10 mM using the minimum concentration possible of dimethylsulphoxide and were diluted further in water. Dimethyl sulphoxide did not affect the activities of the compounds at the concentrations used. For the atrial tissue work ethanol was used rather than dimethylsulphoxide. The maximum concentration of ethanol used (0.01%) did not affect atrial contraction.


View larger version (25K):
[in this window]
[in a new window]
  		Fig. 1.   Structures of the novel beta 3AR agonists.

    Results
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References

Cloned beta ARs

The potencies (pD2 values) of the beta AR agonists and their intrinsic activities relative to isoproterenol as stimulants of adenylyl cyclase in membranes from CHO cells expressing each of the three human cloned beta ARs, together with their pKi values for displacement of [125I] -iodocyanopindolol binding are given in table 1.

                              
View this table:
[in this window]
[in a new window]
  		TABLE 1
Binding affinities (pKi), potencies (pD2) and intrinsic activities (IA) as stimulants of adenylyl cyclase in membranes from CHO cells expressing human cloned beta ARsa

Standard compounds. (-)-Isoproterenol had similar pD2 and pKi values for beta 1- and beta 2ARs, but was less potent (by 4- to 8-fold for cyclase stimulation; 30-fold for binding) at beta 3- compared to beta 1- or beta 2ARs.

(±)-CGP 12177 was not an agonist at beta 1- or beta 2ARs. It bound to beta 1- and beta 2ARs with pKi values of about 9 and had about 1000-fold lower affinity for beta 3ARs. Its pKi and pD2 values at beta 3ARs were similar. It was a partial agonist as a stimulant of beta 3ARs.

BRL-37344 was a partial agonist relative to isoproterenol at all three receptors. It had a slightly lower intrinsic activity than CGP 12177 as a stimulant of beta 3ARs. It had greater binding affinity for beta 2- and beta 3ARs than for beta 1ARs. It did not exhibit greater affinity for beta 3- than beta 2ARs, but its pD2 value at beta 2ARs was, somewhat surprisingly, lower than its pKi and it was therefore a more potent stimulant of the beta 3ARs.

Novel phenylethanolamines. In contrast to BRL-37344, the novel phenylethanolamines SB-215691 and SB-220646 were both full agonists relative to isoproterenol as stimulants of beta 3ARs. Nevertheless, they had similar potency (pD2 ~ 6) to BRL-37344. SB-215691 was a partial agonist at beta 1- and beta 2ARs but it was about 10-fold less potent at these receptors than at beta 3ARs. SB-220646 displayed no agonist activity in membranes from CHO cells that expressed 210 or 560 fmol beta 1- or beta 2ARs, respectively/mg protein. However, in two experiments where it was evaluated in membranes from CHO cells that expressed higher numbers of beta 2ARs (2300 fmol/mg protein) it displayed intrinsic activities relative to isoproterenol of 0.20 and 0.24, and pD2 values of 4.5 and 3.7.

Novel aryloxypropanolamines. The other compounds are, like CGP 12177, aryloxypropanolamines. SB-226552 and SB-229432 had higher intrinsic activities than CGP 12177 at beta 3ARs and they lacked agonist activity at beta 1- and beta 2ARs. In sharp contrast to CGP 12177, their pKi values at beta 1- and beta 2ARs were much lower than their pD2 values at beta 3ARs. SB-232602 had a similar profile to SB-226552 and SB-229432, except that it had higher affinity for beta 1- and beta 2ARs and hence lower selectivity for beta 3ARs.

The remaining three compounds, SB-236923, SB-246982 and SB-248320, were all beta 3AR agonists with high intrinsic activities relative to isoproterenol, but like the phenylethanolamines BRL-37344 and SB-215691, they were partial agonists at beta 1- and/or beta 2ARs. SB-248320 and SB-246982 were especially potent beta 3AR agonists, the EC50 value for SB-246982 being 0.7 nM. The selectivity of these three compounds as beta 3AR agonists decreased as their potency as beta 3AR agonists decreased: SB-246982>SB-248320>SB-236923.

The pKi values for SB-226552 and SB-246982 at beta 3ARs were much lower than their pD2 values. This contrasts with the similar pKi compared to pD2 values for isoproterenol, CGP 12177 and BRL-37344.

SB-229432 (100 µM), SB-232602 (10 µM) and SB-236923 (10 µM) were evaluated in a single experiment as antagonists of the stimulation of adenylyl cyclase activity by isoproterenol in membranes that expressed beta 1ARs. Their pKB values of 4.44, 5.54 and 5.70 were broadly similar to their pKi values of 4.18, 5.79 and 5.33.

Lipolysis

Influence of albumin batch. Our previous work has shown that the pD2 value of isoproterenol and intrinsic activity of CGP 12177 as stimulants of human white adipocyte lipolysis varies between preparations from different patients (Sennitt et al., 1995). In addition, we found that incubation of human adipocytes with some batches of fraction V albumin invariably resulted in poor responses to CGP 12177 and reduced sensitivity to (-)-isoproterenol.

To investigate this problem, adipocytes were prepared using Boehringer Fraction V albumin and were then incubated with isoproterenol or CGP 12177 in the presence of the same source of albumin or one of two different batches of Sigma fatty acid free albumin. Table 2 demonstrates that responsiveness to CGP 12177 and sensitivity to isoproterenol was obtained with only one of the Sigma batches of albumin.

                              
View this table:
[in this window]
[in a new window]
  		TABLE 2
Influence of albumin source and batch on responses to isoproterenol and CGP 12177a

In a second experiment cells were prepared in the presence of a batch of Boehringer fraction V albumin that was known to give responsive cells, and they were then incubated with isoproterenol or CGP 12177 in the presence of other sources of albumin. Sensitivity to isoproterenol and responsiveness to CGP 12177 was greatest in the presence of Pentex low hormone grade albumin (table 2). The low hormone grade albumin and the less effective cell culture grade albumin used in these experiments had been derived from the same fraction V albumin batch.

Effects of standard beta AR agonists. Analysis of variance showed no difference between the tissue sources of adipocyte in sensitivities to isoproterenol (P = 0.32) or CGP 12177 (P = 0.14), although for each agonist the trend was for the potency to decrease in the order: breast>kidney>axillary adipose tissue (table 3). The intrinsic activity of CGP 12177 was also not different between tissues (P = 0.50). Because the source of adipocytes had no significant effect on the potencies of the standard agonists, results for the novel compounds were not distinguished according to the tissue. However, results for the novel compounds are given in table 3 together with results for isoproterenol and CGP 12177 obtained in the same experiments. The three least potent stimulants of lipolysis, SB-226552, BRL-37344 and SB-220646, were among the least potent stimulants of cloned beta 3AR-mediated adenylyl cyclase activity, but, other than this, there was no clear relationship between the two activities, even correcting for the high potencies of isoproterenol and CGP 12177 in the lipolysis experiments that included SB-232602, SB-248320 and SB-236923 (table 3). In particular, the high potencies SB-248320 and SB-246982 in the adenylyl cyclase assay were not reflected in the lipolysis experiments.

                              
View this table:
[in this window]
[in a new window]
  		TABLE 3
Stimulation of human white adipocyte lipolysisa

Isoproterenol was by far the most potent stimulant of lipolysis, despite its relatively low potency as a stimulant of cloned beta 3AR-mediated adenylyl cyclase activity. The high pKB value for nadolol (7.8) as an antagonist of the lipolytic response to isoproterenol, indicates that this is because its lipolytic activity was mediated primarily via beta 1- and beta 2ARs. In contrast, the low pA2 value for nadolol as an antagonist of the lipolytic response to CGP 12177 demonstrates that this compound acts exclusively via non beta 1/2ARs, possibly beta 3ARs. (The pKB value for nadolol as an antagonist of isoproterenol was not significantly different at 1 and 0.1 µM nadolol:7.82 ± 0.10 and 8.02 ± 0.04 respectively.)

Novel phenylethanolamines. In our previous work (Sennitt et al., 1995) we have shown that the lipolytic effect of BRL-37344 is sensitive to propranolol. Thus despite its selectivity as a stimulant of human cloned beta 3ARs (table 1), it stimulates human white adipocyte lipolysis primarily via beta 1- or beta 2ARs. Because this may be in part attributable to the low efficacy of BRL-37344 at beta 3ARs (Wilson et al., 1996), we evaluated SB-215691, which is a full agonist as a stimulant of human cloned beta 3ARs (table 1).

SB-215691 was 23-fold more potent than BRL-37344 as a stimulant of lipolysis, contrasting with its similar potency as a stimulant of cloned beta 3ARs. Its greater potency may be due to its higher efficacy at beta 3ARs. The lipolytic effect of SB-215691 appeared partially resistant to nadolol. Thus in three experiments exemplified by figure 2A, 0.1 µM nadolol failed to shift the concentration-response curve for SB-215691, except at concentrations of SB-215691 that elicited a greater maximal effect than that to CGP 12177. In these experiments the intrinsic activity of CGP 12177 was equal to or greater than 0.44. In three experiments in which the intrinsic activity of CGP 12177 was equal to or less than 0.26 there was some indication that 0.1 µM nadolol did not shift the foot of the SB-215691 curve in a parallel fashion (fig. 2B). In the other experiments the SB-215691 curve was biphasic in the presence of 1 µM nadolol, inflecting near the CGP 12177 maximal response. This is exemplified in figure 2C. The pKB value for nadolol (7.60) for shifts of the top part of the concentration response curves was similar to that for shifts of the isoproterenol curve (7.84). These results suggest that low concentrations of SB-215691 elicit a lipolytic response via beta 3ARs, but that when beta 3ARs are fully occupied further responses are mediated by beta 1- or beta 2ARs. The contribution of beta 3ARs to the lipolytic effect of SB-215691 varied widely between patients.


View larger version (17K):
[in this window]
[in a new window]
  		Fig. 2.   Stimulation of lipolysis in human white adipocytes by SB-215691 in comparison with isoproterenol and CGP 12177 in the same experiments. Closed symbols refer to experiments in the absence and open symbols to those in the presence of nadolol. A, An experiment in breast adipocytes in which 0.1 µM nadolol shifted the SB-215691 curve at concentrations of SB-215691 that elicited a greater maximal effect than that of CGP 12177. B, An experiment in breast adipocytes in which 0.1 µM nadolol antagonized all concentrations of SB-215691, although the shift of the curves was not quite parallel. C, An experiment in kidney adipocytes in which the SB-215691 curve was biphasic in the presence of 1 µM nadolol, inflecting near the CGP 12177 maximal response.

Although SB-220646 was similar to SB-215691 in being a full agonist at cloned beta 3ARs, no evidence was found for a beta 3AR-mediated component in its lipolytic activity, the pKB values for nadolol being high in every experiment.

Novel aryloxypropanolamines. The aryloxypropanolamines SB-226552 and SB-229432, which had no agonist activity at cloned beta 1- and beta 2ARs, behaved similarly to CGP 12177 in that their lipolytic activity was resistant to antagonism by nadolol. They also had similar intrinsic activities to CGP 12177, even though their intrinsic activities at the cloned beta 3AR had been higher. Concentration-response curves for SB-229432 in breast adipocytes and SB-226552 in kidney adipocytes are shown in figures 3A and B respectively. Results in figure 3A are expressed relative to the maximum effect of CGP 12177 to expand the scale. Curves for CGP 12177 are omitted from figure 3A for clarity because they overlap the curves for SB-229432.


View larger version (24K):
[in this window]
[in a new window]
  		Fig. 3.   Stimulation of lipolysis in human perirenal adipocytes by SB-226552 (A) and in human breast adipocytes by SB-229432 (B) in comparison with CGP 12177 or isoproterenol in the same experiments. Note that in A results are expressed as a percentage of the maximum response to CGP 12177 rather than isoproterenol to expand the scale, the mean intrinsic activity of CGP 12177 relative to isoprenaline being 0.42. Closed symbols refer to experiments in the absence and open symbols to those in the presence of 1 µM nadolol. A: n = 3; B. n = 6.

The lipolytic activity of the aryloxypropanolamine SB-232602 was totally resistant to blockade by 1 µM nadolol in three experiments, but pKB values of 7.1 and 6.7 were obtained in the other two experiments. These latter results raise the possibility that SB-232602 had some agonist activity at beta 1- and beta 2ARs, although such activity was not detected in the cloned receptor experiments. However, the poor selectivity of SB-232602 would ensure that any marginal beta 1- or beta 2AR agonist activity would be seen in the lipolysis experiments. The intrinsic activity of SB-232602 was higher than that of CGP 12177 (paired t test; P < 0.05; table 3). Again, this might be because SB-232602 had slight agonist activity at beta 1- or beta 2ARs.

The aryloxypropanolamines SB-236923, SB-246982 and SB-248320 all showed some intrinsic activity at human cloned beta 1ARs, but only in the case of SB-248320 was this apparent in its lipolytic activity (fig. 4A-C). Thus nadolol antagonized the lipolytic activity of SB-248320 in every experiment with similar pKB values to those achieved against isoproterenol. This may be because SB-248320 had the highest intrinsic activity of these compounds at beta 1ARs (table 1). Both SB-246982 (P < 0.05) and SB-248320 (P = 0.02) had higher intrinsic activities than CGP 12177 in the same lipolysis experiments (paired t tests). The higher intrinsic activity of SB-236923 compared to CGP 12177 was not statistically significant.


View larger version (21K):
[in this window]
[in a new window]
  		Fig. 4.   Stimulation of lipolysis in human axillary adipocytes by SB-236923 (A) in human perirenal adipocytes by SB-246982 (B) and in human breast adipocytes by SB-248320 (C) in comparison with isoproterenol and CGP 12177 in the same experiments. Closed symbols refer to experiments in the absence and open symbols to those in the presence of 1 µM nadolol. n = 3

Atrial Contraction

Previous work (see table 4) has shown that CGP 12177 has a small inotropic effect via an atypical (non-beta 1, non-beta 2) beta AR (Kaumann, 1996), although BRL-37344 has a larger effect mediated via beta 1- or beta 2ARs (Kaumann AJ and Sanders L, quoted in Arch and Kaumann, 1993). These compounds were therefore not included in our study, except that some strips were treated with 10 µM CGP 12177 to confirm the previous evidence of an inotropic response to this compound (table 4).

                              
View this table:
[in this window]
[in a new window]
  		TABLE 4
Stimulation of the force of contraction of human right atrial strips by beta AR agonists

Novel phenylethanolamines. The phenylethanolamine SB-220646 had a higher intrinsic activity than BRL-37344 at cloned beta 3ARs and a lower intrinsic activity at beta 1- and beta 2ARs (table 1). It was therefore thought that it might stimulate inotropic activity via atypical beta ARs, but this proved not to be so. It was a full agonist in the right atrial appendage (table 4), but the effect of a maximal concentration (60 µM) was totally reversed by 200 nM (-)-propranolol. Moreover, preincubation with either 300 nM CGP 20712A (selective beta 1AR blocker) or 50 nM ICI 118,551 (selective beta 2AR blocker) shifted the SB-220646 concentration-response curve about fivefold to the right (fig. 5). Because these concentrations of CGP 20712 and ICI 118,551 are insufficient to antagonize atypical beta ARs (Arch and Kaumann, 1993; Kaumann, 1997), it would appear that the inotropic, like the lipolytic, effect of SB-220646 was mediated via both beta 1- and beta 2ARs.


View larger version (16K):
[in this window]
[in a new window]
  		Fig. 5.   Stimulation of the force of contraction of human right atrial appendage by SB-220646 in the absence or presence of 300 nM CGP 20712A or 50 nM ICI 118,551. n = 5

Novel aryloxypropanolamines. In contrast to the phenylethanolamines SB-220646 and BRL-37344 acting via beta 1- and beta 2ARs, and the aryloxypropanolamine (-)-CGP 12177 acting via an atypical beta AR, the novel aryloxypropanolamine SB-226552 (0.2-60 µM) had no effect at all on atrial contraction, either in the absence of (-)-propranolol or when the tissues were preincubated with 200 nM (-)-propranolol (table 4). Similarly, in the presence of 200 nM (-)-propranolol the aryloxypropanolamine SB-229432 (2-60 µM) had no effect on atrial contraction (data not shown).

                              
View this table:
[in this window]
[in a new window]
  		TABLE 5
Summary of the pharmacological profiles of the beta AR agonists at human beta AR subtypes

The aryloxypropanolamines SB-232602 (1 nM -30 µM) and SB-236923 (0.06-6 µM) had slight inotropic activity that was largely blocked by preincubation of the tissues with 200 nM (-)-propranolol (table 4). The slight beta 1- or beta 2AR-mediated inotropic activity of SB-236923 had been predicted from its activity at cloned beta 1- and beta 2ARs (table 1). The slight beta 1- or beta 2AR-mediated inotropic effect of SB-232602 was not predicted from the cloned receptor data, but was consistent with the lipolysis results for this compound. SB-232602 or SB-226923 again failed to elicit an atypical beta AR-mediated response.

    Discussion
Top
Abstract
Introduction
Materials & Methods
Results
Discussion
References

Lipolysis. Much of the evidence that atypical beta ARs can mediate lipolysis in human white adipocytes rests on studies with CGP 12177 (Lonnqvist et al., 1993; Hoffstedt et al., 1996; Portillo et al., 1995; Tavernier et al., 1996; Sennitt et al., 1995), but the literature is confused by some reports that CGP 12177 does not stimulate lipolysis in human adipocytes (Langin et al., 1991; Van Liefde et al., 1994). These opposing findings may have arisen because some sources of albumin prevent beta 3AR-mediated responses. Thus in the course of our work, conducted over 5 yr, there was a period of about 6 mo when almost no response was obtained to CGP 12177, which was included along with isoproterenol in every experiment. After this problem was tracked to a change in albumin batch (table 2) we obtained responses to CGP 12177 in the majority (about 80%) of experiments, variation in responsiveness probably reflecting physiological or pathological variation in beta 3AR number or their coupling to lipolysis. We have not identified why some sources of albumin prevented beta 3AR-mediated lipolysis. However, we have found that treatment of tissues with trypsin prevents the binding of a human beta 3AR-specific monoclonal antibody to beta 3ARs (Jennings KH, unpublished data) raising the possibility that a proteolytic enzyme present in the albumin destroyed the small numbers of beta 3ARs in human white adipocytes.

The profiles of the beta 3AR agonist are summarized in table 5. Of the aryloxypropanolamines, SB-226552, SB-229432, SB-232602, as well as CGP 12177, elicited nadolol-insensitive lipolytic responses. It seems probable that these responses were mediated by beta 3ARs, first because beta 3AR mRNA is present in human white adipocytes (Krief et al., 1993; Revelli et al., 1993; Berkowitz et al., 1995; Tavernier et al., 1996) and secondly because all these compounds stimulated human cloned beta 3ARs. There was no clear correlation between the potencies of the compounds as stimulants of the cloned receptor and lipolysis, but such a relationship would have been disrupted if some of the compounds bound more avidly than others to the albumin present in the lipolysis experiments.

It might be suggested that the aryloxypropanolamines stimulate lipolysis not via beta 3ARs but via a putative `beta 4AR` or perhaps a mixture of beta 3- and beta 4ARs. Pharmacological evidence for beta 4ARs has been obtained by one of us (A.J.K.) in cardiac tissue (Kaumann and Molenaar, 1996), including human right atrial appendage (Kaumann, 1997; Kaumann and Molenaar, 1997; Molenaar et al., 1997). Therefore, if SB-226552, SB-229432 and SB-236923 stimulate lipolysis via `beta 4ARs,` it is difficult to explain why they had no inotropic activity in human right atrial appendage. We cannot rule out the possibility that there is a ` beta 4AR` in human white adipocytes, but these novel compounds appear to stimulate lipolysis via beta 3- and not beta 4ARs. It seems probable that CGP 12177 also acts primarily via beta 3ARs to stimulate lipolysis.

The role of beta 3ARs in human white adipocytes might also be questioned by the failure of some phenylethanolamine selective beta 3AR agonists such as BRL-37344 either to stimulate lipolysis at all or to stimulate lipolysis via beta 3ARs (Hollenga et al., 1991; Langin et al., 1991; Hoffstedt et al., 1996; Sennitt et al., 1995). Three factors may combine to explain this finding, at least in the case of BRL-37344. First, BRL-37344 is only a marginally selective stimulant of human beta 3- compared to beta 2ARs (table 1). Second, BRL-37344 has very low efficacy at human beta 3ARs: as many as 250 beta 3ARs must be occupied by BRL-37344 to give the same response as one receptor occupied by isoproterenol (Wilson et al., 1996). Third, some workers have failed to identify beta 3AR mRNA in human white adipose tissue, although they could detect beta 1- and beta 2AR mRNA (Thomas and Liggett, 1993; Deng et al., 1997), whilst others have found the beta 3-mRNA is less abundant than beta 1- or beta 2AR mRNA (Tavernier et al., 1996). This indicates that there are few beta 3- compared to beta 1- or beta 2ARs in this tissue. It might therefore be expected that BRL-37344 and similar selective agonists of the rodent beta 3AR would act via beta 1- and beta 2ARs in human white adipocytes. Of the phenylethanolamines, only CL 316243, which has almost no efficacy at beta 1- and beta 2ARs, has been shown to stimulate lipolysis in a propranolol-resistant manner (Hoffstedt et al., 1996).

The importance of efficacy was addressed using SB-215691. Although we do not have a precise measure of the efficacy of this compound, it was clearly greater than that of BRL-37344 because it was a full agonist, whereas BRL-37344 was a partial agonist, as a stimulant of human cloned beta 3ARs (table 1). A nadolol-resistant component was detected in the lipolytic response to SB-215691 in some experiments (fig. 2A, B), indicating that phenylethanolamines that are agonists at all three beta ARs can elicit part of their lipolytic activity via beta 3ARs, provided they have sufficient efficacy at (and selectivity for) the beta 3AR.

The nadolol-resistant component of the action of SB-215691 was only apparent when its lipolytic activity was equal to or less than the maximal activity of CGP 12177; lipolysis elicited by higher concentrations of SB-215691 was always sensitive to nadolol (fig. 2A). This might suggest that beta 3AR stimulation can never produce a greater lipolytic effect than CGP 12177. This appears not be the case, however, because SB-246982, which had a greater intrinsic activity than isoproterenol as a stimulant of human cloned beta 3ARs (table 1), did elicit a greater nadolol-resistant maximal lipolytic effect than CGP 12177 (table 3).

Our results therefore support the view that stimulation of beta 3ARs can cause lipolysis in human white adipocytes. They also show, however, that it is difficult to predict from studies on cloned receptors whether a selective beta 3AR agonist will stimulate lipolysis via beta 3ARs. Slight efficacy at cloned beta 1- or beta 3ARs may translate to lipolytic activity being mediated via these receptors, especially if a compound is little more potent as a stimulant of beta 3- than beta 1- or beta 2ARs. It also seems that, for similar pharmacological profiles, aryloxypropanolamines (e.g., SB-236923, SB-246982) are less likely than phenylethanolamines to stimulate lipolysis via beta 1- or beta 2ARs and more likely to act via beta 3ARs.

Atrial contraction. Slight agonist activity at cloned beta 1- or beta 2ARs may also translate to marked inotropic activity in human right atrial appendage. It was particularly surprising that the phenylethanolamine SB-220646, which stimulated the activity of adenylyl cyclase only in CHO cells that expressed high numbers of beta 2ARs, was a full agonist via beta 1- or beta 2ARs in right atrial appendage.

Slight inotropic activity was also detected with the aryloxypropanolamine SB-232602 despite it lacking activity at the cloned beta 1- or beta 2AR. However, beta 1- and beta 2AR agonist activity in aryloxypropanolamines seems to translate less readily than similar activity in phenylethanolamines into inotropic activity (just as it translates less readily into lipolytic activity). Thus the partial beta 1- and beta 2AR agonist SB-236923 had, like SB-232602, very low inotropic activity.

Our work provides no support for a role for beta 3ARs in human right atrial appendage, since none of the novel beta 3AR agonists had inotropic activity in this preparation. The simplest explanation for these findings is that CGP 12177 elicits its inotropic activity via a putative `beta 4AR' (Kaumann, 1997; Kaumann and Molenaar, 1997).

Variable beta 3AR pharmacology. The proposed existence of beta 4ARs cannot account for all the unexplained features of atypical beta AR pharmacology, however. Thus there are a number of reports that the potency or potency orders of beta 3AR agonists vary according to the measurement made in both beta 3AR-transfected CHO cells (Emorine et al., 1994; Arch, 1995; Wilson et al., 1996) and rat ileum (Hoey et al., 1996), neither of which have been suggested to express beta 4ARs. In rodent skeletal muscle the atypical beta AR does not show a typical beta 3AR pharmacology, but differs from cardiac atypical beta AR pharmacology in being highly sensitive to BRL-37344 (Liu et al., 1996 and see Arch 1997 for review). It is difficult to escape the conclusion that beta 3ARs can behave differently in different environments. Kenakin (1995) has explained how the G protein environment of a receptor might differentially affect its sensitivity to agonists and it appears that even a disruption of cells can have such an effect. A further peculiarity noted in our study was that SB-226552 and SB-246982 had much lower affinity for the cloned beta 3AR in binding studies than might have been expected from their potencies as stimulants of cloned beta 3ARs (table 1). A possible explanation for this finding is that there were far more beta 3ARs expressed in the CHO cells used for the binding studies than there were G proteins to which they could couple. The radioactive antagonist would bind indiscriminately to all the receptors, but only a proportion of these would be able to bind to G proteins and form a high affinity complex with the nonradioactive agonist (Kenakin, 1997). (The converse finding that pKi was greater than pD2 for BRL-37344 in CHO cells expressing beta 2ARs is more difficult to explain.)

In conclusion, studies on novel phenylethanolamine and aryloxypropanolamine beta 3AR agonists support previous work on CGP 12177 and CL 316243 (Hoffstedt et al., 1996) in showing that the beta 3AR in human white adipocytes can mediate a lipolytic response. No support was found for a beta 3AR-mediated inotropic response in human right atrial appendage, consistent with the argument that the previously reported inotropic activity of CGP 12177 is mediated by a putative `beta 4AR.` Stimulation of human cloned beta 3ARs does predict that compounds will stimulate human white adipocyte lipolysis. However, it is not straightforward to predict from studies on cloned receptors whether beta AR agonists, especially phenylethanolamines, will stimulate lipolysis via beta 1-, beta 2- or beta 3ARs, or enhance contractile activity via beta 1- or beta 2ARs.

    Footnotes

Accepted for publication February 6, 1998.

Received for publication October 13, 1997.

1 Current address: University of Buckingham, Hunter Street, Buckingham, MK18 1EG, UK.

2 Supported by the British Heart Foundation.

Send reprint requests to: Dr. J. R. S. Arch, SmithKline Beecham Pharmaceuticals, New Frontiers Science Park, Harlow, Essex CM19 5AW, UK.

    Abbreviations

beta AR, beta -adrenoceptor; CHO, Chinese hamster ovary; MEM, minimal essential medium; CGP12177, (±)-4-(3-t-butylamino-2-hydroxypropoxy)benzimidazol-2-one; ICI 118551, D-(±)-1-(7-methylylindan-4-yloxy)-3-isopropylaminobutan-2-ol; CGP 20712A, (±)-[2-(3-aminocarbamoyl-4-hydroxyphenoxy)ethylamino]-3-[4-(1-methyl-4-trifluoromethyl-2-imidazolyl)phenoxy]-2-propanol hydrochloride ; CL 316243, disodium (RR) -5-[2-[[2-(3-chlorophenyl)-2-hydroxyethyl]-amino]propyl]-1,3-benzodioxazole-2,2-dicarboxylate ; BRL-37344, (RR+SS)-(±)-4-[2-(2-(3-chlorophenyl)-2-hydroxyethyl)amino)propyl]phenoxyacetate; SB-215691, (RS+RR) 3-hydroxypropyl 4-[2-[2-(3,4-dihydroxyphenyl)-2-hydroxyethylamino]propyl]phenoxymethylphosphonate ethyl ester, dihydrate ; SB-220646, (R, R)-5-{2-[2-(3,4-dihydroxyphenyl)-2-hydroxyethylamino]propyl}-1,3-benzodioxole-2,2-dicarboxylate ; SB-226552, (S)-4-{2-[2-hydroxy-3-(4-hydroxyphenoxy)propylamino]ethyl}phenoxymethylcyclohexylphosphinic acid lithium salt ; SB-229432, (SR)-4-{2-[2-hydroxy-3-(4-hydroxy-3-hydroxymethylphenoxy)propylamino]propyl}phenoxymethylphenylphosphinic acid, lithium salt ; SB-232602, (S)-N-{2-hydroxy-5-[2-hydroxy-3-(4-methoxyphenyl)ethylamino]propoxy}methanesulfonamide hydrochloride ; SB-236923, (SR) 4-{2-[2-(2-hydroxy-3-(4-hydroxy-3-methanesulfonylaminophenoxy)-propyl)amino]propyl}phenoxymethylphenylphosphinic acid, hydrobromide; SB-246982, (S)-2-(4-(2-(2-(3-(3-N-phenylsulfonylamino-4-hydroxyphenoxy)-hydroxypropyl)amino)ethyl)benzyloxy)benzoic acid, trifluoroacetate ; SB-248320, (S)-diphenyl-4-{2-[2-hydroxy-3-(4-hydroxy-3-iso-propylsulfonyl-aminophenoxy)propylamino]ethyl}phenoxymethyl phosphine oxide .

    References
Top
Abstract
Introduction