ephedrine

Pharmacological Effects of Ephedrine Alkaloids onHuman �1- and �2-Adrenergic Receptor Subtypes

Guoyi Ma, Supriya A. Bavadekar, Yolande M. Davis, Shilpa G. Lalchandani,Rangaswamy Nagmani, Brian T. Schaneberg, Ikhlas A. Khan, and Dennis R. FellerThe National Center for Natural Products Research (G.M., R.N., B.T.S., I.A.K., D.R.F.), Department of Pharmacology(S.A.B., Y.M.D., S.G.L., D.R.F.), and Department of Pharmacognosy (I.A.K.), School of Pharmacy, University of Mississippi,University, Mississippi

Received January 29, 2007; accepted April 2, 2007

ABSTRACTEphedra species of plants have both beneficial and adverseeffects primarily associated with the presence of ephedrinealkaloids. Few reports have appeared that examine the directactions of ephedrine alkaloids on human subtypes of adrener-gic receptors (ARs). In the present study, ephedrine alkaloidswere evaluated for their binding affinities on human �1A-, �1B-,�1D-, �2A-, �2B-, and �2C-AR subtypes expressed in HEK andChinese hamster ovary cells. Cell-based reporter gene assayswere used to establish functional activity of ephedrine alkaloidsat �1A-, �2A-, and �2C-ARs. The data showed that ephedrinealkaloids did not activate �1- and �2-ARs and that they antag-onized the agonist-mediated effects of phenylephrine and me-detomidine on �1- and �2-ARs, respectively. As in the bindingstudies, 1R,2R- and 1R,2S-ephedrine showed greater func-

tional antagonist activity than the 1S,2R- and 1S,2S-isomers.The rank order of affinity for the isomers was 1R,2R � 1R,2S �1S,2R � 1S,2S. The rank order of potencies of alkaloids con-taining a 1R,2S-configuration was norephedrine � ephedrine�� N-methylephedrine. These studies have demonstrated thatorientation of the �-hydroxyl group on the ethylamino sidechain and the state of N-methyl substitution are important for�-AR binding and functional activity of the ephedrine alkaloids.In conclusion, the ephedrine isomers and analogs studied didnot exhibit any direct agonist activity and were found to pos-sess moderate antagonist activities on cloned human �-ARs.The blockade of presynaptic �2A- and �2C-ARs may have apharmacological role in the direct actions of Ephedra alkaloids.

The Ephedra or Ma huang species of plants are widely usedfor their medicinal properties. The principal active constitu-ent in the Ephedra species is ephedrine, which possesses twochiral centers, and can exist as four isomers designated as1R,2S- and 1S,2R-ephedrine and 1R,2R- and 1S,2S-pseudo-ephedrine (Griffith and Johnson, 1995). Naturally occurringephedrine alkaloids include mainly 1R,2S-ephedrine, 1S,2S-pseudoephedrine, 1R,2S-norephedrine, 1R,2S-N-methyl-ephedrine, and 1S,2S-norpseudoephedrine. These Ephedra/ephedrine alkaloids have been used as nasal decongestants,bronchodilators, and CNS stimulants (Kalix, 1991; Hoffman,

2001), and Ephedra has also been used for the treatment ofobesity (Arch et al., 1984; Liu et al., 1995). In recent years,there have been numerous reports of adverse reactions re-sulting from intake, especially in excessive doses, of herbalproducts containing extracts of Ephedra as weight loss aids(Josefson, 1996). Other effects such as hypertension, tremors,myocardial infarction, seizures, and stroke have resulted infatalities (Chua and Benrimoj, 1988; White et al., 1997;Haller and Benowitz, 2000). The United States Food andDrug Administration has recently prohibited the sale of di-etary supplements containing Ephedra. However, herbalproducts containing Ephedra remain in use in other coun-tries.

The beneficial and adverse effects of ephedrine and relatedanalogs are known to be mediated via the �- and �-adrenergicreceptors (ARs) and can be elicited by either direct interactionswith the receptors as agonists or antagonists or indirectly byeither causing a release of endogenous catecholamines and/orby preventing their neuronal reuptake (Trendelenburg, 1963;Trendelenburg et al., 1963; Patil et al., 1967; Vansal and Feller,

This work was supported in part by Grant R21AT00510 from the NationalCenter for Complementary and Alternative Medicine, and its contents aresolely the responsibility of the authors and do not necessarily represent theofficial views of the National Center for Complementary and AlternativeMedicine, National Institutes of Health. This study was also supported in partby United States Department of Agriculture-Agricultural Research ServiceAgreement 58-6408-2-0009) and the National Heart, Lung, and Blood InstituteUndergraduate Short-Term Training Grant 5T35HL07926 (to Y.D.).

Article, publication date, and citation information can be found athttp://jpet.aspetjournals.org.

doi:10.1124/jpet.107.120709.

ABBREVIATIONS: CNS, central nervous system; AR, adrenoceptor; HEK, human embryonic kidney; CHO, Chinese hamster ovary; TRE-LUC,phorbol ester-response element-luciferase gene; CRE-LUC, cAMP-response element-luciferase.

0022-3565/07/3221-214–221$20.00THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS Vol. 322, No. 1Copyright © 2007 by The American Society for Pharmacology and Experimental Therapeutics 120709/3216436JPET 322:214–221, 2007 Printed in U.S.A.

214

by guest on August 6, 2011

jpet.aspetjournals.orgD

ownloaded from

http://jpet.aspetjournals.org/

1999; Rothman et al., 2003; Wellman et al., 2003). The relativecontribution of these direct and indirect interactions to thepharmacological effects of the ephedrine isomers in vivo on ARshas remained controversial. Ephedra alkaloids of the 1R,2S-configuration have generally been shown to exert direct AReffects, whereas isomers of the 1S,2S- and 1S,2R-configurationhave indirect actions (Patil et al., 1965, 1967; LaPidus et al.,1967; Tye et al., 1967; Patil, 1968; Waldeck and Widmark, 1985;Kawasuji et al., 1996; Vansal and Feller, 1999; Liles et al.,2006). In studies using ephedrine isomers, indirect actions werepredominantly demonstrated by depletion of norepinephrinetissue stores using reserpine, and more recently, Rothman et al.(2003), using a battery of in vitro tests, reported that the mostpotent action of ephedrine and norephedrine analogs and iso-mers is as substrates for norepinephrine transporters.

Although there are reports of the direct effects of ephedrineisomers on human �-ARs (Vansal and Feller, 1999), no com-prehensive information exists on the direct interactions (ag-onist and antagonist potencies) of the ephedrine isomers andrelated Ephedra alkaloids at human �-ARs. Rothman et al.(2003) found that the ephedrine isomers and analogs werenot agonists on human �1A- or �2A-AR, but they did not testcompounds for either antagonist activity or on other �-ARsubtypes. In the present study, the direct effects of the ephed-rine isomers and naturally occurring ephedrine alkaloidshave been systematically examined as agonists and antago-nists on cloned human �1 (�1A, �1B, and �1D)- and �2 (�2A,�2B, and �2C)-AR subtypes expressed in host cells. A prelim-inary report of our work has appeared (Ma et al., 2004).

Materials and MethodsMaterials. Human embryonic kidney (HEK293) cells stably ex-

pressing homogeneous populations of the human �1A-, �1B-, and

�1D-AR subtypes were obtained from Dr. Kenneth Minneman(Emory University, Atlanta, GA). Human �2A-, �2B-, and �2C-ARsubtypes stably expressed in CHO cells were obtained from Drs.Marc Caron, Robert Lefkowitz (Duke University, Durham, NC), andStephen Liggett (University of Cincinnati, Cincinnati, OH). Theephedrine alkaloids used in this study were provided by Dr. Popat N.Patil (The Ohio State University, Columbus, OH), or purchased fromSigma-Aldrich (St. Louis, MO). The compounds were dissolved inwater or in a 1:5 mixture of dimethyl sulfoxide and water. Stocksolutions of 10 mM were prepared fresh daily and diluted in waterto appropriate concentrations for the studies. [3H]Rauwolscineand [3H]prazosin were obtained from PerkinElmer Life and Ana-lytical Sciences (Wellesley, MA). The phorbol ester (12-O-tetradeca-noylphorbol-13-acetate)-response element-luciferase gene (TRE-LUC) and cAMP-response element-luciferase reporter gene (6 CRE-LUC) were kindly provided by Dr. A. Himmler (BoehringerIngelheim Research and Development, Vienna, Austria). All cellculture reagents were obtained from Life Technologies (Carlsbad,CA). Other chemicals were purchased from Sigma-Aldrich.

Cell Culture. HEK293 cells stably expressing �1A-, �1B-, and�1D-AR subtypes were grown in 150 cm2 Corning culture flasks (Acton,MA) with Dulbecco’s modified Eagle’s medium supplemented with10% fetal bovine serum, 2 mM glutamine, penicillin G (100 U/ml),streptomycin (100 �g/ml), and G418 (Geneticin, 100 �g/ml). Upon con-fluence, the cells were detached by gentle scraping. CHO cells stablyexpressing �2A-, �2B-, and �2C-AR were grown in 150 cm2 Corningculture flasks with Ham’s F12 medium supplemented with 10% fetalbovine serum, 2 mM glutamine, penicillin G (100 U/ml), streptomycin(100 �g/ml), and G418 (100 �g/ml). All cells were cultured at 37°C in anatmosphere of 5% CO2 and 95% humidity. Media were changed every48 h until the cells were confluent. Upon confluence, the cells weredetached by trypsinization (0.25% trypsin EDTA) for 2 min.

Competitive Radioligand Binding Assays. Radioligand bind-ing assays were carried out on intact HEK293 cells stably expressing�1A-, �1B-, �1D-AR subtypes and intact CHO cells stably expressing�2A-, �2B-, and �2C-AR subtypes as described by Lalchandani et al.

Fig. 1. The chemical structures of ephedrineisomers and naturally occurring alkaloids.

Effects of Ephedrine Alkaloids on Human Adrenoceptors 215



ownloaded from


(2002). The detached cells were washed and centrifuged with Tris-EDTA buffer, pH 7.4, containing 50 mM Tris, 20 mM disodiumEDTA, and 154 mM NaCl, in which they were finally suspended. Theradioligand was used at a fixed concentration of 0.1 �Ci in theabsence and presence of various concentrations (the range was10�10–10�3 M or 10�11–10�4 M) of competing drugs. The drugs wereadded to the cells (50,000) in 50 mM Tris-EDTA buffer to a totalvolume of 2.0 ml and allowed to incubate at 37°C for 1 h. Nonspecificbinding was determined in the presence of 10 �M phentolamine.Reactions were terminated by rapid filtration through WhatmanGF/C filters using a Brandel 12R cell harvester followed by washeswith ice-cold buffer twice. Radioactivity on the dried filter discs wasmeasured using a liquid scintillation analyzer (Tri-Carb 2900TR;PerkinElmer Life and Analytical Sciences). The displacement curveswere plotted and the Ki values of the test ligands for the receptorsubtypes were determined using GraphPad Prism (GraphPad Soft-ware Inc., San Diego, CA). The percentage specific binding wasdetermined by dividing the difference between total bound (disinte-grations per minute) and nonspecific bound (disintegrations perminute) by the total bound (disintegrations per minute).

Functional Assays. A recently developed sensitive reporter geneassay (Lalchandani et al., 2002) was used to elucidate the effects ofthe ephedrine alkaloids on human �2-AR subtypes. In �1-AR studies,a TRE-LUC plasmid provided by Dr. A. Himmler (Stratowa et al.,1995) was used to study the functional effects of the ephedrinealkaloids. HEK293 cells stably expressing �1A-AR were transfectedwith the TRE-LUC plasmid (40 �g/ml) using electroporation (70 ms,single pulse, 150 V). The transfected cells were seeded at a density of50,000 cells/well in microtiter plates (Cultureplate; PerkinElmerLife and Analytical Sciences) in 200 �l of media and allowed to growfor 24 h with incubation at 37°C (5% CO2). After 24 h, the cells weretreated with varying drug concentrations for a period of 20 h, whichwas found to be optimum during time course analyses performedearlier (data not presented). When antagonist studies were per-formed, the compounds were added 15 min before the addition of anagonist, L-phenylephrine. After drug exposure, the cells were lysed,and luciferase activity was measured using the Luclite assay kit(PerkinElmer Life and Analytical Sciences).

In �2-AR studies, the cell-based CRE-LUC was conducted as de-scribed previously by Lalchandani et al. (2002). The CHO cells stablyexpressing �2A- and �2C-ARs were transfected with the 6 CRE-LUCplasmid (40 �g/ml) using electroporation (70 ms, single pulse, 150 V).The transfected cells were seeded at a density of 50,000 cells/well inmicrotiter plates (Cultureplate; PerkinElmer Life and Analytical Sci-ences) in 200 �l of media and allowed to grow for 24 h with incubationat 37°C (5% CO2). After 24 h, the cells were treated with varying drugconcentrations for 4 h. When antagonist studies were performed, thecompounds were added 15 min before the addition of the direct adenyl-ate cyclase activator forskolin (3 �M). After drug exposure, the cellswere lysed, and luciferase activity was measured using the Lucliteassay kit (PerkinElmer Life and Analytical Sciences).

For both �1A- and �2A- and �2C-AR cells, the luciferase activity(counts per second, changes in light production) was determinedusing a TopCount Microplate Scintillation and LuminescenceCounter (model B9904; PerkinElmer Life and Analytical Sciences).Data were normalized relative to luciferase changes of L-phenyleph-rine (3 � 10�4 M � 100%) and forskolin (3 � 10�6 M � 100%) in �1A-and �2A-/�2C-AR expressed in HEK293 and CHO cells, respectively.

Data Accumulation and Analysis. For binding studies in celllines, varying concentrations of each drug were added in duplicatewithin each experiment, and the individual IC50 values were deter-mined using GraphPad Prism software. The Ki value of each ligand wasdetermined according to the equation described by Cheng and Prusoff(1973), and final data are presented as pKi � S.E.M. of n � 6 experi-ments. The concentration-dependent reversal of forskolin-induced (3�M) luciferase activity changes in CHO cells by medetomidine andselected ephedrine analogs was used to assess agonist activity, and datafor medetomidine were expressed as EC50 values � S.E.M. of n � 6

experiments. Antagonist activities of the ephedrine analogs were deter-mined by their addition before incubation with a fixed concentration ofthe �2-AR agonist, medetomidine (10 �M). Data were expressed as IC50

values � S.E.M. of at least n � 6 experiments. Differences betweenmeans of binding affinities and functional responses of individual drugswere analyzed using a paired t test. Values were considered to bestatistically significant when P � 0.05.

ResultsReceptor Binding. To obtain a better understanding of

the direct effect of ephedrine isomers on �-ARs, the binding

Fig. 2. Binding displacement curves of ephedrine isomers for �1A (a)-, �2A(b)-, and �2C (c)-ARs expressed in HEK293 or CHO cells. Data are ex-pressed as means � S.E.M. (n � 6 experiments).f, L-phenylephrine; Œ,medetomidine; �, 1R,2S-ephedrine; E, 1R,2S-norephedrine; E, 1R,2S-N-methylephedrine.

216 Ma et al.



ownloaded from


affinity of ephedrine isomers and closely related analogswere compared on human �1- and �2-AR subtypes. The struc-tures of all of the ephedrine isomers are provided in Fig. 1.With selected Ephedra alkaloids of the 1R,2S-configuration,the rank order of competitive displacement of radioligands onboth �1A- and �2A-AR subtypes was norephedrine � ephed-rine � N-methylephedrine (Fig. 2). The comparative affini-ties of these Ephedra alkaloids as displacing ligands weremuch lower those for than the standard agonist analogs,L-phenylephrine and medetomidine. Calculated affinity val-ues of the standard agonists and of all of the ephedrinealkaloids on �1A- and �2A-AR subtypes are given in Tables 1and 2. 1R,2S-Norephedrine, lacking an N-methyl group as in1R,2S-ephedrine, showed increased binding affinities at thesubtypes tested. The presence of an additional N-methylgroup, as in 1R,2S-N-methylephedrine, showed decreasedbinding affinities compared with 1R,2S-ephedrine at the�1A-, �2A-, and �2C-AR. The rank order of affinity (Ki values)of ephedrine isomers on human �1-AR in HEK293 cells was1R,2R-pseudoephedrine � 1R,2S-ephedrine � 1S,2R-ephed-rine � 1S,2S-pseudoephedrine at �1A-AR; 1S,2R-ephed-rine � 1R,2R-pseudoephedrine � 1R,2S-ephedrine � 1S,2S-

pseudoephedrine at �1B-AR; and 1R,2R-pseudoephedrine �1R,2S-ephedrine � 1S,2R-ephedrine � 1S,2S-pseudoephed-rine at �1D-AR. The affinities of 1R,2S-norephedrine and1S,2S-norpseudoephedrine were similar to 1R,2S-ephedrineat the �1A-AR. The rank order of affinity (Ki values) of ephed-rine isomers on human �2-AR in CHO cells was 1R,2R-pseudoephedrine � 1R,2S-ephedrine � 1S,2R-ephedrine �1S,2S-pseudoephedrine at �2A-AR; 1R,2R-pseudoephed-rine � 1R,2S-ephedrine � 1S,2R-ephedrine � 1S,2S-pseudo-ephedrine at �2B-AR; and 1R,2R-pseudoephedrine � 1R,2S-ephedrine � 1S,2R-ephedrine � 1S,2S-pseudoephedrine at�2C-AR. As shown in Tables 1 and 2, the 1R,2R- and the1R,2S-ephedrine isomers had greater affinities than the1S,2R- and 1S,2S-isomers at all �-ARs, except at the �1B-AR.1S,2S-Pseudoephedrine displayed the lowest binding affinityat all �-ARs. Furthermore, the binding affinities of 1R,2S-norephedrine and 1S,2S-norpseudoephedrine were similar tothat of 1R,2S-ephedrine on these subtypes.

Direct Agonist Effects of Ephedrine Alkaloids on�-AR Subtypes. Cell-based reporter gene assays were usedto establish functional activity at the �1A-, �2A-, and �2C-ARs.The direct agonist effects of ephedrine isomers on �1-ARs

TABLE 1Binding affinities (pKi values) of L-phenylephrine and ephedrine alkaloids on human �1A-, �1B-, and �1D-ARs in HEK293 cells�3H�Prazosin was used as the radioligand in equilibrium competition radioligand binding assays for the �1A-, �1B-, and �1D-ARs, and nonspecific binding was measured inthe presence of 10 �M phentolamine. pKi value � �log Ki (Ki was calculated according to the Cheng-Prusoff equation (Cheng and Prusoff, 1973) and the data are the means �S.E.M. of n � 6.

CompoundspKi � S.E.M.

�1A-AR �1B-AR �1D-AR

L-Phenylephrine (PE) 6.32 � 0.30 N.D. N.D.Oxymetazoline 8.11 � 0.05 6.32 � 0.001 6.07 � 0.031R,2S-Ephedrine 4.95 � 0.04a,b �3.00 (37.4%c)a 4.28 � 0.09a

1S,2R-Ephedrine 4.41 � 0.05a 4.11 � 0.06a 3.95 � 0.1a

1R,2R-Pseudoephedrine 5.01 � 0.11d 4.04 � 0.05d 4.54 � 0.21d

1S,2S-Pseudoephedrine 4.16 � 0.08d �3.00 (12.5%c)d �3.00 (27.3%c)d

1R,2S-Norephedrine 5.14 � 0.1b N.D. N.D.1S,2S-Norpseudoephedrine 4.57 � 0.20 N.D. N.D.1R,2S-N-Methylephedrine 4.20 � 0.19e N.D. N.D.

N.D., not determineda The mean pKi values for 1R,2S- and 1S,2R-ephedrine isomers were significantly different on the receptor subtype (P � 0.05 using the paired Student’s t test).b The mean pKi values for 1R,2S-ephendrine and 1R,2S-norephedrine isomers were significantly different on the receptor subtype (P � 0.05 using the paired Student’s

t test).c Maximal percent inhibition of specific binding observed at 1 mM.d The mean pKi values for 1R,2R- and 1S,2S-pseudoephedrine isomers were significantly different on the receptor subtype (P � 0.05 using the paired Student’s t test).e The mean pKi value for 1R,2S-N-methylephedrine was significantly different from the mean values of 1R,2S-ephedrine and 1R,2S-norephedrine on the �1A-AR subtype

(P � 0.05 using the paired Student’s t test).

TABLE 2Binding affinities (pKi values) of medetomidine and ephedrine alkaloids on human �2A-, �2B-, and �2C-ARs in CHO cells�3H�Rauwolscine was used as the radioligand in equilibrium competition radioligand binding assays for the �2A-, �2B-, and �2C-ARs, and nonspecific binding was measuredin the presence of 10 �M phentolamine. pKi value � �log Ki (Ki was calculated according to the Cheng-Prusoff equation (Cheng and Prusoff, 1973) and the data are means �S.E.M. of n � 6.

CompoundspKi � S.E.M.

�2A-AR �2B-AR �2C-AR

Medetomidine 6.15 � 0.05 7.50 � 0.03 7.65 � 0.031R,2S-Ephedrine 4.83 � 0.16a 4.78 � 0.02 4.75 � 0.121S,2R-Ephedrine 4.44 � 0.19a,b 4.77 � 0.05 4.65 � 0.021R,2R-Pseudoephedrine 4.95 � 0.20c 5.26 � 0.08c 4.88 � 0.10c

1S,2S-Pseudoephedrine 4.19 � 0.10c 4.36 � 0.10c 4.18 � 0.11c

1R,2S-Norephedrine 5.16 � 0.05b N.D. 4.96 � 0.201S,2S-Norpseudoephedrine 4.32 � 0.05 N.D. 4.50 � 0.131R,2S-N-methylephedrine 4.20 � 0.30d 4.29 � 0.09d 4.10 � 0.30d

N.D., not determined.a The mean pKi values for 1R,2S- and 1S,2R-ephedrine isomers were significantly different on the receptor subtype (P � 0.05 using the paired Student’s t test).b The mean pKi values for 1R,2S-ephedrine and 1R,2S-norephedrine were significantly different on the receptor subtype (P � 0.05 using the paired Student’s t test).c The mean pKi values for 1R,2R- and 1S,2S-pseudoephedrine isomers were significantly different on the receptor subtype (P � 0.05 using the paired Student’s t test).d The mean pKi value for 1R,2S-N-methylephedrine was significantly different from the mean values of 1R,2S-ephedrine and 1R,2S-norephedrine on the receptor subtype

(P � 0.05 using the paired Student’s t test).




ownloaded from


were studied. As shown in Fig. 3, luciferase assay studiesusing the TRE-LUC plasmid showed that 1R,2S-ephedrine,1S,2R-ephedrine, 1R,2R-pseudoephedrine, 1S,2S-pseudo-ephedrine, 1R,2S-norephedrine, 1R,2S-N-methylephedrineand 1S,2S- norpseudoephedrine had little effects on �1A-ARat the highest concentration tested (0.3 mM). These isomersgave a response that was �12% of the L-phenylephrine max-imum. L-Phenylephrine activated �1A-AR, giving an EC50

value of 2.01 � 0.39 �M.

The agonistic effects of the ephedrine isomers on �2A- and�2C-AR subtypes were examined for their abilities to reverseforskolin-induced cAMP elevation measured using the 6CRE-LUC reporter gene assay. The results in Fig. 3 showthat none of the ephedrine alkaloids (1R,2S-ephedrine,1S,2R-ephedrine, 1R,2R-pseudoephedrine, 1S,2S-pseudo-ephedrine, 1R,2S-norephedrine, 1R,2S-N-methylephedrine,and 1S,2S-norpseudoephedrine) reversed forskolin-inducedcAMP elevations in the �2A- and �2C-AR subtypes. However,the agonist medetomidine significantly reversed forskolin-induced cAMP elevations in the �2A- and �2C-AR subtypeswith EC50 values of 79.8 � 3.5 and 78.3 � 7.3 nM (Fig. 3).These data indicate that the ephedrine alkaloids do not act asdirect agonists on these AR subtypes.

Antagonistic Effects of Ephedrine Alkaloids on �-ARSubtypes. Studies were undertaken to examine the antag-onistic effects of ephedrine alkaloids on �1A-, �2A-, and�2C-AR subtypes in HEK293 and CHO cells (Table 3; Fig. 4).The results showed that ephedrine isomers and analogs an-tagonized the effects of the agonists, L-phenylephrine on �1-ARs (Fig. 4a) and medetomidine on �2-ARs (Fig. 4b). In Table3, it was noteworthy that the antagonistic potencies of ephed-rine isomers on the �2A- and �2C-AR subtypes were consid-erably higher (5- to 24-fold) than those for the inhibition ofthe �1-AR. Similar to the rank order found in binding studies,1R,2S- and 1R,2R-pseudoephedrine showed greater antago-nist activity than 1S,2R- and 1S,2S-pseudoephedrine (com-pare data in Tables 1 and 3). No difference was noted in theantagonist potencies of the primary and secondary amines of1R,2S-norephedrine and ephedrine, respectively. However,1S,2S-norpseudoephedrine (the primary amine analog) wasmore potent than 1S,2S-pseudoephedrine as an antagoniston the three �-AR subtypes, and its potency was the same asthat of the 1R,2S-isomers of ephedrine and norephedrine onthese ARs. It appears that the orientation of the 1R-hydroxylsubstituent and the methylation state (primary versus sec-ondary amine) in the ephedrine alkaloids are important for�-AR activity. The antagonistic activities of all ephedrineisomers were consistent with their binding affinities.

DiscussionChemically, ephedrine possesses two chiral centers, and

1R,2S-ephedrine is a long-studied stimulant available bothas a prescription and over-the-counter medication and is amajor ingredient in widely marketed herbal preparations.Another isomer, 1S,2S-pseudoephedrine is used as a nasaldecongestant and precursor for the illicit synthesis of meth-amphetamine. Standard pharmacology textbooks emphasizethe fact that ephedrine is both a direct and indirect actingdrug for the activation of ARs (Hoffman, 2001). Previousreports attributed the effects of ephedrine alkaloids to directagonist activity and by release of norepinephrine from pre-synaptic nerve terminals via a carrier-mediated exchangemechanism (Trendelenburg, 1963; Trendelenburg et al.,1963; Patil et al., 1967; Vansal and Feller, 1999; Rothman etal., 2003). In the present study, we characterized the directreceptor-mediated effects of the stereoisomers of ephedrineand closely related naturally occurring compounds (1R,2S-isomers of ephedrine, norephedrine, and N-methylephedrineand 1S,2S-isomers of pseudoephedrine and norpseudoephed-rine).

Fig. 3. a, direct effects of L-phenylephrine and ephedrine alkaloids on�1A-AR in HEK 293 cells. Control and L-phenylephrine (300 �M) lucif-erase activities (counts per second, mean � S.E.M., n � 4–6) were 1154and 7560, respectively. Normalized data (300 �M L-phenylephrine �100%) are expressed as the mean � S.E.M. of n � 4–6 experiments. A,1R,2S-ephedrine (300 �M); B, 1S,2R-ephedrine (300 �M); C, 1R,2R-pseudoephedrine (300 �M); D, 1S,2S-pseudoephedrine (300 �M); E,1R,2S-norephedrine (300 �M); F, forskolin (3 �M); G, 1R,2S-N-methyl-ephedrine (300 �M); H, 1S,2S- norpseudoephedrine (300 �M). b, concen-tration-dependent effects of medetomidine and ephedrine alkaloids forthe reversal of effects on forskolin-induced cAMP elevation on the �2A-and �2C-ARs expressed in CHO cells. Control and forskolin-induced lu-ciferase activities (counts per second, mean � S.E.M., n � 4–6) were 1560and 12,200 for �2A and 3069 and 34,576 for �2C, respectively. Normalizeddata (3 �M forskolin � 100%) are expressed as the mean � S.E.M. of n �4–6 experiments. A, 1R,2S-ephedrine (300 �M); B, 1S,2R-ephedrine (300�M); C, 1R,2R-pseudoephedrine (300 �M); D, 1S,2S-pseudoephedrine(300 �M); E, 1R,2S-norephedrine (300 �M); F, forskolin (3 �M); G, 1R,2S-N-methylephedrine (300 �M); H, 1S,2S- norpseudoephedrine (300 �M).

218 Ma et al.



ownloaded from


Radioligand binding studies of the ephedrine alkaloidsshowed that the 1R,2R- and the 1R,2S-isomers generally hadgreater affinities than their corresponding 1S,2R- and 1S,2S-isomers on all of the �-AR subtypes. An exception was at the�1B-AR. 1S,2S-Pseudoephedrine displayed the lowest bind-ing affinity at all �-ARs. 1R,2S-Norephedrine, lacking anN-methyl group as in 1R,2S-ephedrine, showed increasedbinding affinities at the subtypes tested. The presence of anadditional N-methyl group, as in 1R,2S-N-methylephedrine,showed decreased binding affinities compared with 1R,2S-ephedrine at the �1A-, �2A-, and �2C-AR. In summary, thesteric orientation of the hydroxyl group in the 1R-configura-tion and the presence of a primary or secondary amine on theside chain were associated with the highest binding affinitiesto �-AR subtypes.

Rothman et al. (2003) reported that the ephedrine isomers,at 10 �M, lacked agonist activity in functional studies ofintracellular calcium changes in cells expressing human �1A-and �2A-ARs. No further studies were done to determinewhether ephedrine analogs were antagonists in these cells.In our study, using cell-based reporter gene assays to mea-sure functional effects at the �1A-, �2A-, and �2C-ARs, wedemonstrated that ephedrine isomers and related analogswere also inactive as direct agonists and showed that theephedrine alkaloids antagonized the effects of the agonists,L-phenylephrine on �1- and medetomidine on �2-ARs. Similarto the rank order found in binding studies, 1R,2S- and 1R,2R-ephedrine showed greater antagonist activity than 1S,2R-and 1S,2S-pseudoephedrine, which indicated that the �-hy-droxyl substituent on the side chain in the R-configuration isimportant for �-AR antagonist activity. The antagonistic po-tencies of the isomers were consistent with their bindingaffinities, both of which were in the micromolar range. Takencollectively, these findings suggest that the ephedrine iso-mers possess only direct antagonist activity in cloned human�-AR systems.

Early reports by Patil and colleagues (Lapidus et al., 1967;Patil et al., 1967; Tye et al., 1967) have demonstrated thatthe 1R,2S- and 1R,2R-isomers of ephedrine possess agonistor antagonist activity (dependent upon the tissue examined),whereas the remaining ephedrine isomers exhibit antagonistactivities or possessed an indirect mechanism of action onisolated tissues (lung, ileum, vas deferens, and vascular

preparations) that contained �- and/or �-ARs. In anotherstudy by Lee et al. (1974) using rat epididymal fat tissue andmeasurement of lipolysis, the ephedrine isomers were devoidof �-AR activity on lipolysis, and they reported that 1R,2S-ephedrine is more potent as an antagonist than 1S,2R-ephed-rine. In this regard, Wellman et al. (2003) recently reportedthat 1R,2S-ephedrine-induced hypophagia in rats was atten-uated by prazosin, which suggested that this in vivo ephed-rine action was mediated via the �1-AR. The �1-AR subtypethat mediated this CNS effect of ephedrine on hypophagiawas not established. It is apparent that the mechanism ofaction for Ephedra alkaloids is dependent upon the stereo-chemistry of the ephedrine analog(s) used, the abundanceand distribution of AR subtypes present, and whether nervesremain intact in the target tissue under study.

To date, only a few studies have been completed usinghuman ARs. On the three human �-AR subtypes expressedin CHO cells, the ephedrine isomers showed weak agonistactivities (Vansal and Feller, 1999), and the 1R,2S-ephedrineisomer was the most active agonist on the three subtypes,exhibiting the most potent activity on the �2-AR and weakestagonist activity on the �3-AR subtype. In contrast, Rothmanet al. (2003) reported that 1R,2S-ephedrine was not active asan agonist on the human �-AR subtypes (data not presented)and suggested that the results of the earlier study by Vansaland Feller (1999) may have been due to the use of cells withhigher receptor densities. Other work using fluorescent la-beling techniques to evaluate direct �2-AR interactions with1R,2S-ephedrine (Gether et al., 1995) have yielded more in-sight into conformational changes. In this study, only fullagonists but not weaker partial agonists such as ephedrine,produced significant reductions in fluorescence, providing aninteresting approach to investigate the direct AR effects (ag-onist or antagonist changes). Additional studies with theisomers of Ephedra alkaloids and related analogs may beuseful in probing ligand-specific sites of interaction on ARsubtypes.

Taken collectively, our present studies show only a weakpartial agonist activity (at �10�4 M) by the 1R,2S-isomers ofephedrine and norephedrine and that all isomers of testedEphedra/ephedrine alkaloids produce only moderate antago-nist activities on �-AR subtypes. Based upon our findings, itis likely that the in vivo actions of these Ephedra alkaloids

TABLE 3The pKB or pIC50 values of antagonist effects of ephedrine alkaloids on human �1A-, �2A-, and �2C-ARs in HEK293 and CHO cellsValues were determined as the effective concentration (IC50) and negative log (pIC50) of each ephedrine analog that reversed the effect of medetomidine on the maximumcAMP response of forskolin. Each value is the mean � S.E.M. of data from three experiments performed in duplicate.

Compounds�1A-Adrenoceptors �2A-Adrenoceptors �2C-Adrenoceptors

pKB KB pIC50 IC50 pIC50 IC50

�M �M �M

1R,2S-Ephedrine 4.78 � 0.05a 16.6 5.08 � 0.10b 8.3 4.97 � 0.18b 10.71S,2R-Ephedrine 4.70 � 0.08 19.9 4.81 � 0.30b 15.5 4.65 � 0.29b 22.41R,2R-Pseudoephedrine 5.05 � 0.29c 10.2 5.12 � 0.10c 7.6 5.11 � 0.11c 7.81S,2S-Pseudoephedrine 3.91 � 0.07c 22.0 4.32 � 0.10c 47.9 4.25 � 0.23c 56.21R,2S-Norephedrine 5.02 � 0.20a 10.5 5.08 � 0.10 8.3 4.84 � 0.31 14.51S,2S-Norpseudoephedrine 4.79 � 0.13 16.9 5.10 � 0.10 7.9 4.76 � 019 17.41R,2S-N-Methylephedrine 4.39 � 0.10d 42.7 4.60 � 0.00d 25.1 4.26 � 0.21d 55.0

a The mean pKB or pIC50 values for 1R,2S-ephedrine and 1R,2S-norephedrine were significantly different on the receptor subtype (P � 0.05 using the paired Student’st test).

b The mean pKB or pIC50 values for 1R,2S- and 1S,2R-ephedrine isomers were significantly different on the receptor subtype (P � 0.05 using the paired Student’s t test).c The mean pKB or pIC50 values for 1R,2R- and 1S,2S-pseudoephedrine isomers were significantly different on the receptor subtype (P � 0.05 using the paired Student’s

t test).d The mean pKB or pIC50 value for 1R,2S-N-methylephedrine was significantly different from the mean values of 1R,2S-ephedrine and 1R,2S-norephedrine on the receptor

subtype (P � 0.05 using the paired Student’s t test).




ownloaded from


are mediated principally by an indirect action on the sub-types of human �-ARs. However, a mechanism of direct ac-tion for 1R,2S-ephedrine on ARs has been proposed to ex-plain the observed in vivo effects for the pharmacologicaltreatment of asthma (Griffith and Johnson, 1995; Hoffman,2001) and for its CNS stimulatory and/or cardiovascular(heart and stroke) adverse effects observed with its abuseand misuse in humans (Haller and Benowitz, 2000). Inter-estingly, our findings contrast with recent in vivo reports on

pressor effects (Liles et al., 2006) and hypophagic activity inrats (Wellman et al., 2003), in which they demonstrated that1R,2S-ephedrine exhibits its agonist activity via direct ef-fects on the �1-AR receptors. In this regard, an in vitro studyinvolving mutational changes (Waugh et al., 2000) in the�1A-AR demonstrated a 4-fold increase in functional potencyfor the 1R,2S-ephedrine isomer, whereas potency for epi-nephrine was decreased by 36-fold. Thus, the presence of thearomatic catechol group in epinephrine was more affected bythe mutation than a molecule lacking the catechol function-ality (ephedrine). Waugh et al. (2000) demonstrated partialagonist activity for only this ephedrine isomer on wild-typeand mutated cloned rat and hamster �1A-ARs. In summary,it is apparent that the 1R,2S-ephedrine isomer has directagonist activity, in vitro and in vivo. Our studies and those ofRothman et al. (2003) failed to demonstrate significant ago-nist activity for 1R,2S-ephedrine on human �1-ARs. The ob-served agonist responses to 1R,2S-ephedrine, as reported byWaugh et al. (2000), may be explained in part by an increaseddensity of receptors in the cells used in their experiments.Overall, these results provide evidence in support of both invivo and in vitro agonist activities for 1R,2S-ephedrine on�-ARs.

Ephedra alkaloids are expected to produce their pharma-cological actions by mixed mechanisms, involving indirectand direct actions on �-AR subtypes. In conclusion, our stud-ies have shown that the ephedrine isomers and analogs in-vestigated did not exhibit any significant direct agonist ac-tivity on human �1- and �2-AR subtypes and can be classifiedas antagonists. We propose that the observed in vivo func-tional effects of Ephedra/ephedrine alkaloids may be pro-duced, in part, by blocking the action of norepinephrine at�1A- and /or �2A-AR. The results of our studies, particularlywith the findings on human �2-AR, suggest that an inhibitionof AR subtypes by ephedrine analogs may be related to theirin vivo effects. In particular, it is plausible to suggest thatnaturally occurring ephedrine alkaloids may act as antago-nists of presynaptic �2A/2C-ARs present in nerve terminals.As such, ephedrine analogs may interfere with presynapticnorepinephrine uptake and enhance synaptic concentrationsand response to this neurotransmitter. In this regard, Roth-man et al. (2003) used in vitro assays to measure the effect ofphenylpropanolamines on the release or reuptake of norepi-nephrine in rat whole brain (minus caudate and cerebellum)preparations and proposed that ephedrine analogs act via anindirect mechanism leading to an enhanced release of nor-epinephrine. They also showed that ephedrine isomers andrelated analogs produce a similar pattern but with less po-tency for dopamine release and that ephedrine analogs alsohave affinity for interaction with serotonergic receptors. Thestudy of Ephedra alkaloid effects on dopamine and serotoninreceptor subtypes may also be worthwhile but is beyond thescope of the present work. Thus, it is also clear that inter-pretation of the in vivo and in vitro results of the pharma-cology of ephedrine alkaloids may be further complicated bycombination and relative abundance of �- and/or �-AR sub-types and other pharmacological receptors and their signal-ing pathways that are present in the target tissue. Morestudies will be required to establish the importance of theseand other receptors to assess the overall actions of thesesympathomimetic amines in vivo. Irrespective of the under-lying differences between in vitro and in vivo actions of

Fig. 4. a, displacement curves of antagonist effects of ephedrine alkaloidson L-phenylephrine-mediated agonistic effects on �1A-AR expressing inHEK 293 cells. Data are expressed as the mean � S.E.M. of n � 4–6experiments. f, L-phenylephrine (0.1–300 �M); �, 1R,2S-ephedrine (300�M) L-phenylephrine; F, 1R,2S-norephedrine (300 �M) L-phenyleph-rine; E, 1R,2S-N-methylephedrine (300 �M) L-phenylephrine. Controland L-phenylephrine (300 �M) luciferase activities (disintegrations persecond, mean � S.E.M., n � 4–6) were 1269 and 7980, respectively.Normalized data (300 �M L-phenylephrine � 100%) are expressed as themean � S.E.M. of n � 4–6 experiments. b, reversal of medetomidineinhibition of forskolin-induced cAMP elevation by ephedrine alkaloids onhuman �2A-AR versus �2C-AR (B). cAMP changes were assessed by mea-surement of luciferase activity. Control and forskolin-induced measure-ments of luciferase activity (counts per second, mean � S.E.M., n � 4–6)were 1483 and 12,780 for �2A and 3069 and 32576 for �2c, respectively.pIC50 and IC50 values (Table 3) were determined as the concentration ofephedrine isomers that reversed the inhibition effect of medetomidine onthe luciferase response to forskolin). A, 1R,2S-ephedrine (1, 10, and 100�M); B, 1R,2S-norephedrine (1, 10, and 100 �M); C, 1R,2S-N-methyl-ephedrine (1, 10, and 100 �M); M, medetomidine (0.01 �M); F, forskolin(3 �M).

220 Ma et al.



ownloaded from


Ephedra or its alkaloids, it is apparent that their pharmaco-logical effects on ARs will be dependent upon the dose andmay involve both direct and/or indirect mechanisms on nor-adrenergic-dependent pathways.

ReferencesArch JR, Ainsworth AT, Cawthorne MA, Piercy V, Sennittt MV, Thody VE, Wilson

C, and Wilson S (1984) Atypical �-adrenoceptor on brown adipocytes as target foranti-obesity drugs. Nature 309:163–165.

Cheng Y and Prusoff WH (1973) Relationship between the inhibition constant (Ki)and the concentration of the inhibitor which causes 50% inhibition (IC50) of anenzymatic reaction. Biochemical Pharmacol 22:3099–3108.

Chua SS and Benrimoj SI (1988) Non-prescription sympathomimetic agents andhypertension. Med Toxicol Adverse Drug Exp 3:387–414.

Gether U, Lin S, and Kobilka BK (1995) Fluorescent labeling of purified �2 adren-ergic receptor: evidence for ligand-specific conformational changes. J Biol Chem270:28268–28275.

Griffith RK and Johnson EA (1995) Adrenergic drugs, in Principles of Medicinalchemistry (Foye WO, Lemke TL and Williams DA eds) pp 345–365, Williams &Wilkins, Baltimore.

Haller CA and Benowitz NL (2000) Adverse cardiovascular and central nervoussystem events associated with dietary supplements containing ephedra alkaloids.N Engl J Med 343:1833–1838.

Hoffman BB (2001) Catecholamines, sympathomimetic drugs and adrenergic recep-tor antagonists, in Goodman and Gilman’s Pharmacological Basis of Therapeutics(Hardman JG and Limbird LE eds) p. 237, McGraw Hill, New York.

Josefson D (1996) Herbal stimulant causes US deaths. BMJ 312:1441.Kalix P (1991). The pharmacology of psychoactive alkaloids from ephedra and catha.

J Ethnopharmacol 32:201–208.Kawasuji T, Koike K, and Saito H (1996) Chronotropic effects of optical isomers of

ephedrine and methylephedrine in the isolated rat right atria and in vitro assess-ment of direct and indirect actions on �1-adrenoceptors. Biol Pharm Bull 19:1423–1428.

Lalchandani SG, Lei L, Zhang W, Suin MM, Moore BM, Liggett SB, Miller DD, andFeller DR (2002) Yohimbine dimmers exhibiting selectivity for the human �2c-adrenoceptor subtype. J Pharmacol Exp Ther 303:979–984.

Lapidus JB, Tye A, and Patil PN (1967) Steric aspects of adrenergic drugs. VII.Certain pharmacological actions of D-(�)-pseudoephedrine. J Pharm Sci 56:1125–1130.

Lee OS, Bescak G, Miller DD, and Feller DR (1974) Influence of substituted phen-ethylamines on lipolysis, in vitro. III. Stereoselectivity. J Pharmacol Exp Ther190:249–259.

Liles JT, Dabisch PA, Hude KE, Pradhan L, Varner KJ, Porter JR, Hicks AR, CorllC, Baber SR, and Kadowitz PJ (2006) Pressor responses to ephedrine are mediatedby a direct mechanism in the rat. J Pharmacol Exp Ther 316:95–105.

Liu YL, Toubro S, Astrup A, and Stock MJ (1995) Contribution of �3-adrenergicactivation to ephedrine-induced thermogenesis in humans. Int J Obes Relat MetabDisord 19:678–685.

Ma G, Bavadekar SA, Davis MY, Lalchandani SG, Nagmani R, Khan I, and FellerDR (2004). Pharmacological effects of ephedra alkaloids on human alpha-adrenergic receptor subtypes, in Proceedings of the 25th Southeastern Pharmacol-ogy Society Meeting; 2004 November 4–5; The University of Mississippi, Univer-sity, MS. American Society for Pharmacology and Experimental Therapeutics,Bethesda, MD.

Patil PN (1968) Steric aspects of adrenergic drugs. VIII. Optical isomers of betaadrenergic receptor antagonists. J Pharmacol Exp Ther 160:308–314.

Patil PN, Lapidus JB, Molinari S, and Tye A (1967) Steric aspects of adrenergicdrugs. IV. Antagonism of norepinephrine effects by ephedrine isomers on theisolated rabbit ileum. J Pharm Sci 56:287–288.

Patil PN, Tye A, and Lapidus JB (1965) A pharmacological study of the ephedrineisomers. J Pharmacol Exp Ther 148:158–168.

Rothman RB, Vu N, Partilla JS, Roth BL, Hufeisen SJ, Compton-Toth BA, Birkes J,Young R, and Glennon RA (2003) In vitro characterization of ephedrine-relatedstereoisomers at biogenic amine transporters and the receptorome reveals selec-tive action as norepinephrine transporter substrates. J Pharmacol Exp Ther307:138–145.

Stratowa C, Machat H, Burger E, Himmler A, Schafer R, Spevak W, Weyer U,Wiche-Castanon M, and Czernilofsky AP (1995) Functional characterization of thehuman neurokinin receptors NK1, NK2, and NK3 based on a cellular assaysystem. J Recept Res 15:617–630.

Trendelenburg U (1963) Supersensitivity and subsensitivity to sympathomimeticamines. Pharmacol Rev 15:225–276.

Trendelenburg U, Delasierra BG, and Muskus A (1963) Modification by reserpine ofthe response of atrial pacemaker to sympathomimetic amines. J Pharmacol ExpTher 141:301–309.

Tye A, Baldeserger R, LaPidus JB, and Patil PN (1967) Steric aspects of adrenergicdrugs. VI. Beta adrenergic effects of ephedrine isomers. J Pharmacol Exp Ther157:56–362.

Vansal SS and Feller DR (1999). Direct effects of ephedrine isomers on humanbeta-adrenergic receptor subtypes. Biochem Pharmacol 58:807–810.

Waldeck B and Widmark E (1985) The interaction of ephedrine with beta-adrenoceptors in tracheal, cardiac and skeletal muscles. Clin Exp PharmacolPhysiol 12:439–442.

Waugh DJ, Zhao MM, Zuscik MJ, and Perez DM (2000) Novel aromatic residues intransmembrane domains IV and V involved in agonist binding at alpha1a-adrenergic receptors. J Biol Chem 275:11698–11705.

Wellman PJ, Miller DK, and Ho DH (2003) Noradrenergic modulation of ephedrine-induced hypophagia. Synapse 48:18–24.

White LM, Gardner SF, Curley BJ, Narx MA, Wang P, and Estes M (1997) Phar-macokinetics and cardiovascular effects of Ma Huang (Ephedra sinica) in normo-tensive adults. J Clin Pharmacol 37:116–122.

Address correspondence to: Dr. Dennis R. Feller, Department of Pharma-cology, School of Pharmacy, The University of Mississippi, University, MS38677. E-mail: [email protected]




ownloaded from


ephedrine

Documents