Lewin AH. Receptors of Mammalian Trace Amines. AAPS Journal. 2006; 8(1): E138-E145. DOI:  10.1208/aapsj080116

Correspondence to:
Anita H. Lewin
Tel: (919) 541-6691
Fax: (919) 541-8868
Email: ahl@rti.org

Received: August 9, 2005;  Accepted: December 21, 2005;  Published: March 10, 2006

Abstract

The discovery of a family of G-protein coupled receptors, some of which bind and are activated by biogenic trace amines, has prompted speculation as to the physiological role of these receptors. Observations associated with the distribution of these trace amine associated receptors (TAARs) suggest that they may be involved in depression, attention-deficit hyperactivity disorder, eating disorders, migraine headaches, and Parkinson’s disease. Preliminary in vitro data, obtained using cloned receptors, also suggest a role for TAARs in the function of hallucinogens.

Keywords: Trace amine associated receptor, TAAR, mammalian, G-protein coupled receptor, ADHD, hypothyroidism-associated depression, prepulse inhibition

Introduction

Endogenous amines such as tyramine, tryptamine, phenethylamine, and octopamine have long been known to be present in mammalian brain at potentially relevant physiological concentrations. In fact, some of these so-called trace amines have for over 25 years been thought to be associated with affective behavior,¹ paranoid chronic schizophrenia,² and depression.^3,4 Moreover, specific binding sites with unique pharmacology and localization for some tritium-labeled trace amines have been reported.^5-9 This mini-review provides a summary of the mammalian trace amine receptor literature up to the end of calendar year 2004.

Literature Review

Although trace amines have important functions in invertebrates, and particularly in insects, no role has been associated with these materials in mammals. The invertebrate receptors for trace amines, also referred to as octopamine receptors, are known to belong to the family of G-protein coupled receptors (GPCRs).¹⁰ Four of them have been distinguished using pharmacological tools, and they have been found to show different coupling to second messenger systems, including activation and inhibition of adenylyl cyclase, activation of phospholipase C, and coupling to a chloride channel. Recently, octopamine receptors from mollusks and insects have been cloned. In humans and other mammals the existence of functional receptors for trace amines had, until recently, only been hypothesized.

A mammalian receptor for trace amines was identified with the discovery of a G-protein coupled receptor capable of binding both tyramine and phenethylamine and coupling to the stimulation of adenylate cyclase through Gαs G-protein, leading to accumulation of cAMP. In 2001, scientists¹¹ at Synaptic Corp (Paramus, NJ) reported the identification of a phylogenetic tree for the human, rat, and mouse trace amine receptors; human 5-HT receptors; human α1a receptor (AR-_α1a); GPR57; GPR58; putative neurotransmitter receptor (PNR); 5-HT_4ψ; Drosophila receptors for octopamine; 5-HT; and tyramine receptors from Caenorhabditis elegans, bee and locust, and a snail octopamine receptor. For rat, 14 trace amine receptors had been identified while 4 human trace amine receptors were found.¹¹ The abbreviation TA was used by these authors for these new receptors. Human and rat trace amine receptors were also described by Bunzow et al¹² who used the abbreviation TAR for the same receptors.

These newly discovered receptors have prompted multiple speculations regarding their physiological and pathological relevance. Trace amines have been hypothesized to act as neurotransmitters or neuromodulators,¹³ as endogenous enhancer substances,¹⁴ as monoamine releasers,¹⁵ and even as vasoconstrictors.¹⁶ A recent review has suggested that they may heterodimerize, for example, with dopamine (DA) receptors, which may increase the intrinsic activity of DA receptors by increasing their affinity to agonist ligands.¹⁷ Indirect activation of dopamine autoreceptors by trace amines has also been proposed to be caused by an efflux of newly synthesized dopamine.¹⁸

Recently, a new nomenclature has been proposed for trace amine receptors.¹⁹ Part of the rationale justifying the need for new nomenclature is because the terms TA and TAR are both used in other contexts. For example, TA₅ has been used to refer to the human GPR 102, and TAR refers to Escherichia coli aspartate receptors. The new term trace amine associated receptor (TAAR) is proposed to avoid such ambiguity, as well as to include members of the trace amine receptor family that do not respond to trace amines (vide infra). Based on the sequential order of the receptor genes on the chromosomes as well as their phylogenetic relationships across species, a series of rules for the naming of TAARs has also been proposed. These rules stipulate that

• the term TAAR would be followed by a number identifying the specific ortholog;
• a letter suffix distinguishing genes that are paralogues would follow the number identifying the specific ortholog; and
• pseudogenes would be identified by the suffix P.

This mini-review will use this new nomenclature. Table 1 shows the new nomenclature and its relationship to terms used in previous publications. In addition, Table 1 demonstrates the identification of 19 individual TAARs for rat, 9 each for human and chimpanzee, and 16 for mouse.

Considerable homology had been reported between the human and rat TAARs: hTAAR 1 and rTAAR 1 share 79% identity; hTAAR 9 and rTAAR 9 share 87% identity; and hTAAR 6 and rTAAR 6 share 88% identity.¹¹ In fact, hTAAR 1 has been called the human ortholog of rTAAR 1.¹² For rhesus monkey, TAAR 1 and TAAR 9 were reported to be >96% homologous to the human orthologs, with only a single amino acid residue in the extracellular N terminus of monkey TAAR 1 differing from hTAAR 1.²⁰ Subsequent work revealed that although the chimpanzee TAAR 1 and TAAR 9P genes had 99.1% and 97.3% overall sequence identity, respectively, only 3 of the chimpanzee genes (TAAR 1, TAAR 5, and TAAR 6) had intact open reading frames, while the other 5 TAAR genes were pseudogenes.¹⁹ Surprisingly, there are twice as many functional TAAR genes in human as in chimpanzee. An additional obvious interspecies difference is the absence of any functional counterpart of the rodent TAAR 7 orthologs in human.¹⁹

The reported phylogeny of the TAAR genes across species reveals the existence of 3 distinct subfamilies into which the orthologs can be grouped (Table 1).¹⁹ All 4 species examined to date (human, chimpanzee, rat, and mouse) have at least one functional TAAR gene for each of the 3 subgroups, suggesting that each subgroup may have physiological relevance. At present putative, potential endogenous ligands for TAAR 1 and TAAR 4 have been identified. Specifically, in humans, TAAR 1 responds to tyramine,^11,21 β-phenethylamine,^11,21 octopamie,¹¹ and dopamine,¹¹ and the TAAR 4 is activated by tyramine¹¹ and β-phenethylamine.¹¹ No ligands are currently known to activate other TAARs. Overall, it has been concluded that TAARs represent a well-defined, coherent gene family, and not an extension of an established, closely related family, such as the 5-HT receptors.¹⁹

The discovery of the TAARs has prompted speculations as to their physiological role(s). Specifically, it has been pointed out²² that although there do not appear to be any mammalian neurons using any of the trace amines, these molecules may function as traditional neuromodulators working through their own receptors. The fact that the mRNA for rTAAR 1 and rTAAR 4 receptor proteins is expressed in certain cells of the substantia nigra/ventral tegmental area, locus coeruleus, and dorsal raphe, which are all areas where cell bodies of the classic biogenic amines are found, further supports such a role for trace amines. Low levels of hTAAR 1, as well as mTAAR 1 mRNA, were found to be expressed in the amygdala. Only trace levels of hTAAR 1 were found in cerebellum, dorsal root ganglia, hippocampus, hypothalamus, medulla, and pituitary. hTAAR 6 mRNA at low level was also found to be expressed in amygdala, and hTAAR 8 mRNA was expressed in both amygdala and hippocampus.

It had been noted that hTAAR genes map to chromosome 6q23.2, close to SCDZ5, a susceptibility locus for schizophrenia, and it has been proposed that hTAAR 6 may be a susceptibility gene for schizophrenia.²³ Since the 6q chromosomal area has been linked to bipolar disorder, hTAAR 6 may be involved in both disorders. In addition, it has been suggested that¹³ TAARs may be involved in depression, attention-deficit hyperactivity disorder (ADHD), eating disorders, migraine headaches, and Parkinson’s disease.

Recently some specific data supporting the involvement of TAARs in attention-deficit hyperactivity disorder and in depression have been presented.²⁴ Specifically, it has been observed that phenethylamine levels may be deficient in ADHD brains, leading to the suggestion that ADHD may be associated with insufficient TAAR 1 activation. If this is the case, it would account for the effectiveness of inhibitors of the dopamine transporter as ADHD medications. Thus, these agents, which have been found to inhibit phenethylamine transport as well, will lead to increased phenethylamine levels, thereby ameliorating the symptoms of ADHD.

In support of the role of TAARs in depression, the observation that 3-iodothyronamine, an analog of tyramine and a metabolite of thyroid hormone, activates rat and mouse TAAR 1 heterologously expressed in HEK 293 cells in vitro has been interpreted to mean that TAAR 1 may play a role in depression associated with hypothyroidism.²⁵ Moreover, since synthetic 3-iodothyronamine injected intaperitoneally (mice) produces several physiological manifestations that are reminiscent of hypothyroidism-associated depression-like symptoms in humans (such as blocking the ability to thermoregulate and maintain normal cardiovascular tone at room temperature, depressing locomotor activity and metabolic rates, and elevating blood sugar levels),²⁵TAAR 1 may be involved in these regulatory processes.

Only a very limited amount of information regarding activation of TAARs is available. Obtaining this information is difficult since the level of TAAR expression in both the central nervous system (CNS) and peripheral tissue is low. For hTAAR 1, only 15 to 100 copies/ng cDNA are expressed in amygdala, and <15 copies/ng cDNA are found in cerebellum, dorsal root ganglia, hippocampus, hypothalamus, medulla, and pituitary. The highest levels (100 copies/ng cDNA) are present in stomach.¹¹ Message from hTAAR 9, hTAAR 5, and hTAAR 6 was found in kidney; the first was also detected in the hippocampus and the latter 2 were also expressed in amygdala. All are expressed at low (<15 copies/ng cDNA) levels.¹¹ Countermanding these low receptor densities required the development of expression systems for screening purposes. This was accomplished by the cloning of rat, mouse, and human TAAR 1^11,12,19; rTAAR 1 was stably expressed in HEK 293 cells,^12,19,26 as was mTAAR 1.¹⁹ Transient transfection has been reported for hTAAR 1.¹¹ Stable expression of hTAAR 1 in HEK 293 cells was achieved by modification of the coding sequence by the addition of an influenza hemaglutinin viral leader sequence and by replacement of selected regions with the corresponding rTAAR 1 sequences.¹⁹ A stable cell line expressing hTAAR 1 (no details given) has been reported.²¹

The rat clone was used to qualitatively screen several CNS-active compounds for activation of rTAAR 1. The results showed amphetamines and lysergic acid diethylamide (LSD)-related compounds to be agonists,¹² leading to the hope that TAARs may provide insight into the molecular mode of action of these drugs of abuse. Evidence for the involvement of TAARs in psychostimulant activity has resulted from observations made in a line of mice lacking the mTAAR 1. These animals demonstrated reproducible deficits in prepulse inhibition, a condition that has been significantly correlated with abnormal functional interactions between the muscarinic, cholinergic, and dopaminergic systems,²⁷ with no difference in baseline startle response.²⁸ In addition, the mice lacking the mTAAR 1 displayed enhanced, dose-dependent sensitivity to the psychomotor stimulating effects of amphetamine, compared with the wild-type littermates, as well as a larger increase in the release of both dopamine and norepinephrine in the dorsal striatum. These observations have been interpreted to suggest that activation of TAAR 1 may serve to dampen the stimulatory effects of amphetamine.²⁸

Use of the clones expressing mTAAR 1 and rTAAR 1 to screen thyronamine derivatives has demonstrated that both are activated by 3-iodothyronamine, a thyroid hormone derivative found in rodent brain.²⁶ Based on this observation, it has been suggested that a signaling pathway, stimulation of which leads to consequences opposite those associated with excess thyroid hormone, may exist. However, considering the significant differences in pharmacology observed¹⁹ for rat and human TAAR 1 (see below and Table 2) such interpretations must be viewed with caution.

Table 2. Potency Values for Phenethylamine Analogs

Entry No. Structure EC50 (μM) rTAAR 1 hTAAR 1 mTAAR 1 Bunzow12 Lindeman19 Borowsky11 Lindemann19 Lindemann19 1 0.24 0.9 0.324 0.3 0.66 2 0.069 0.21 0.214 1.07 1.37 3 5.4 4 5.9 5.14 6.7 15.78 11.76 5 1.3 2.13 4.03 10.29 19.71 6 >50 000 >50 000 >50 000 7 0.3 0.16 0.15 8 0.14 2.05 1.02 9 0.58 10 0.21 11 0.44 12 0.051 13 1.7 14 0.3 5.24 >6 46.87 1.99 15 >10 >50 000 >10 >50 000 >50 000

The potency of “trace amines” and their congeners to activate TAAR 1 are shown in Tables 2 and 3. The data in Table 2 emphasize the interspecies differences as well as the effects associated with expression systems. As pointed out in the literature¹⁹ there appears to be greater correspondence between the assay data for mouse and human TAAR 1, than between mouse and rat TAAR 1. However, it needs to be understood that the number of data points is very small. Perhaps more meaningful is the difference observed between data obtained using stably and transiently expressed receptors, particularly for tryptamine (entry 14). It should also be noted that it is not known what effect the use of “modified” hTAAR 1 to prepare a stable expression system¹⁹ has on potency to activate hTAAR 1. Despite these caveats, some trends are detectable. Thus, introduction of a p-hydroxy group provides for slightly increased potency (compare entries 1 and 2, 7 and 8, 10/11 and 12) whereas introduction of an m-hydroxy group decreases potency by about an order of magnitude (compare entries 1 and 3, 2 and 4, 5 and 6). Similarly, the replacement of one of the amine protons by a methyl group slightly enhances potency (compare entries 1 and 7, 2 and 8, 5 and 9). The effect of aromatic iodo substituents (Table 3) is intriguing. Again, interspecies differences are apparent. While a single m-iodo group increases potency by an order of magnitude in both mouse and rat receptors (compare entries 1 and 2), a second m-iodo group in the second aromatic ring (compare entries 2 and 4) has a modest effect in rTAAR 1, but a significantly larger effect in mTAAR 1.

Table 3. Potency Values for Thyronamine Derivatives*

Substituents EC50 (μM) R1 R2 R3 R4 rTAAR 1 mTAAR 1 H H H H 0.131 ~1 I H H H 0.014 0.112 H H I H ~1 >1 I H I H 0.041 ~1 I I H H 0.056 0.371 H H I I >1 >1 I I I H 0.087 >1 I H I I >1 >1 I I I I >1 >1 *Data adapted from Grandy and Scanlon.25

In invertebrates lipophilicity, dipole moment, and molecular shape have been correlated to the agonist and antagonist efficacy of 49 trace amines in locust thoracic nerve cord.²⁹ Subsequent application of a 3-dimensional molecular field analysis to the same data set and to a larger set of 70 analogs, combined with a genetic algorithm/partial least squares statistical analysis, provided useful information in the characterization and differentiation of (insect) receptor types and subtypes.³⁰ Recently, a 3D quantitative structure activity relationship (QSAR) for a new series of 59 agonists provided a good correlation using a pharmacophore consisting of a positive charge center, aromatic ring, and 3 hydrophobic sites.³¹ Clearly a significant amount of work remains to be done before any similar correlations can be performed for mammals and particularly humans.

At present the discovery of trace amine-associated receptors holds the promise of providing potentially important novel insights into the origins and treatments of CNS disorders. Although cloned membranes expressing hTAAR 1 have been prepared and used to screen a few compounds, the validity of these assays remains to be determined. In particular, the different results reported for the efficacy of ligands to activate hTAAR 1 must be resolved. Thus, the effects of modification of the coding sequence and replacement of certain segments with rTAAR 1 sequences to obtain a stable expression system for hTAAR 1, as well as the use of a transient expression system relative to stable expression system for hTAAR 1 must be ascertained before meaningful QSAR studies can be undertaken and before specific drug leads can be targeted for in vitro evaluation. Even more demanding will be the development of an animal model for in vivo testing. Homozygote TAAR 1 knockout mice are viable¹³ and have been used to study the physiological role of TAAR 1 in the CNS.²⁸ However, considering the significant differences in pharmacology observed for rodent and human hTAAR 1 and in responses to trace amines and their analogs, it may be necessary to develop a hTAAR 1 transgenic rodent in order to get meaningful results.

Conclusions

Three distinct subfamilies of TAARs, comprising as many as 21 individual receptors, have been fully identified and characterized in human, chimpanzee, rat, and mouse. Significant interspecies differences have been found. Only 2 TAARs have had functional ligands, which are associated with them, identified. Owing to their low density in tissue, TAARs must be cloned and expressed for in vitro assays. At present very few stable expression systems have been generated. The scant data available suggest significant differences between results obtained in different expression systems and point to important species differences. The TAARs are an important new target for investigation in light of their likely involvement in neuropsychiatric and neurodegenerative disorders, as well as in drug abuse.

Acknowledgment

The author is grateful to Dr David K. Grandy for discussions and review of the manuscript.

References

1. Sabelli HC, Mosnaim AD.  Phenylethylamine hypothesis of affective behavior. Am J Psychiatry. 1974;131:695-699.
PubMed 

2. Potkin SG, Karoum F, Chuang LW, Cannon-Spoor HE, Phillips I, Wyatt RJ.  Phenylethylamine in paranoid chronic schizophrenia. Science. 1979;206:470-471.
PubMed 

3. Davis BA, Boulton AA.  The trace amines and their acidic metabolites in depression: an overview. Prog Neuropsychopharmacol Biol Psychiatry. 1994;18:17-45.
PubMed  DOI: 10.1016/0278-5846(94)90022-1

4. Sandler M, Ruthven CR, Goodwin BL, Coppen A.  Decreased cerebrospinal fluid concentration of free phenylacetic acid in depressive illness. Clin Chim Acta. 1979;93:169-171.
PubMed  DOI: 10.1016/0009-8981(79)90261-4

5. Altar C, Wasley A, Martin L.  Autoradiographic localization and pharmacology of unique 3Htryptamine binding sites in rat brain. Neuroscience. 1986;17:263-273.
PubMed  DOI: 10.1016/0306-4522(86)90241-1

6. Hauger R, Skolnick P, Paul S.  Specific 3Hbeta-phenylethylamine binding sites in rat brain. Eur J Pharmacol. 1982;83:147-148.
PubMed  DOI: 10.1016/0014-2999(82)90301-6

7. Kellar KJ, Cascio CS.  3HTryptamine: high affinity binding sites in rat brain. Eur J Pharmacol. 1982;78:475-478.
PubMed  DOI: 10.1016/0014-2999(82)90492-7

8. Perry DC.  3Htryptamine autoradiography in rat brain and choroid plexus reveals two distinct sites. J Pharmacol Exp Ther. 1986;236:548-559.
PubMed 

9. Ungar F, Mosnaim A, Ungar B, Wolf M.  Tyramine-binding by synaptosomes from rat brain: effect of centrally active drugs. Biol Psychiatry. 1977;12:661-668.
PubMed 

10. Roeder T.  Octopamine in invertebrates. Prog Neurobiol. 1999;59:533-561.
PubMed  DOI: 10.1016/S0301-0082(99)00016-7

11. Borowsky B, Adham N, Jones KA, et al.  Trace amines: identification of a family of mammalian G protein-coupled receptors. Proc Natl Acad Sci USA. 2001;98:8966-8971.
PubMed  DOI: 10.1073/pnas.151105198

12. Bunzow JR, Sonders MS, Arttmagangkul S, et al.  Amphetamine, 3,4-methylenedioxymethamphetamine, lysergic acid diethylamide, and metabolites of the catecholamine neurotransmitters are agonists of a trace amine receptor. Mol Pharmacol. 2001;60:1181-1188.
PubMed 

13. Branchek TA, Blackburn TP.  Trace amine receptors as targets for novel therapeutics: legend, myth and fact. Curr Opin Pharmacol. 2003;3:90-97.
PubMed  DOI: 10.1016/S1471-4892(02)00028-0

14. Shimazu S, Miklya I.  Pharmacological studies with endogenous enhancer substances: β-phenethylamine, tryptamine, and their synthetic derivatives. Prog Neuropsychopharmacol Biol Psychiatry. 2004;28:421-427.
PubMed  DOI: 10.1016/j.pnpbp.2003.11.016

15. Schmidt N, Ferger B.  The biogenic trace amine tyramine induces pronounced hydroxyl radical production via monoamine oxidase dependent mechanism: an in vivo microdialysis study in mouse striatum. Brain Res. 2004;1012:101-107.
PubMed  DOI: 10.1016/j.brainres.2004.03.036

16. Davenport AP.  Peptide and trace amine orphan receptors: prospects for new therapeutic targets. Curr Opin Pharmacol. 2003;3:127-134.
PubMed  DOI: 10.1016/S1471-4892(03)00003-1

17. Berry MD.  Mammalian central nervous system trace amines: pharmacologic amphetamines, physiologic neuromodulators. J Neurochem. 2004;90:257-271.
PubMed  DOI: 10.1111/j.1471-4159.2004.02501.x

18. Geracitano R, Federici M, Prisco S, Bernardi G, Mercuri NB.  Inhibitory effects of trace amines on rat midbrain dopaminergic neurons. Neuropharmacology. 2004;46:807-814.
PubMed  DOI: 10.1016/j.neuropharm.2003.11.031

19. Lindemann L, Ebeling M, Kratochwil NA, Bunzow JR, Grandy DK, Hoener MC.  Trace amine associated receptors from structurally and functionally distinct subfamilies of novel G protein-coupled receptors. Genomics. 2005;85:372-385.
PubMed  DOI: 10.1016/j.ygeno.2004.11.010

20. Miller GM, Madras BK. A trace amine receptor (TAR1) is a novel amphetamine receptor in primate brain poster. Paper presented at: Sixty-fifth Annual Meeting of the College on Problems of Drug Dependence (CPDD), June 15-19, 2003; Bal Harbour, FL.

21. Yin T, Tu Y, Johnstone EM, Little SP. A Characterization of the Trace Amine 1 Receptor (Program No. 961.5). Paper presented at: 2004 Abstract Viewer/Itinerary Planner, 2004 Online; Washington, DC: Society for Neuroscience.

22. Premont RT, Gainetdinov RR, Caron MG.  Following the trace of elusive amines. Proc Natl Acad Sci USA. 2001;98:9474-9475.
PubMed  DOI: 10.1073/pnas.181356198

23. Duan J, Martinez M, Sanders AR, et al.  Polymorphisms in the trace amine receptor 4 (TRAR4) gene on chromosome 6q23.2 are associated with susceptibility to schizophrenia. Am J Hum Genet. 2004;75:624-638.
PubMed  DOI: 10.1086/424887

24. Madras BK, Verrico C, Jassen A, Miller GM. Attention Deficit Hyperactivity Disorder (ADHD): New Roles for Old Trace Amines and Monoamine Transporters poster. Paper presented at: The American College of Neuropsychopharmacology (ACNP) 43rd Annual Meeting, December 12-16, 2004; San Juan, Puerto Rico.

25. Grandy DK, Scanlan TS. Thyroid Hormone Metabolites and Depression: A New Twist on an Old Tale poster. Paper presented at: The American College of Neuropsychopharmacology (ACNP) 43rd Annual Meeting, December 12-16, 2004; San Juan, Puerto Rico.

26. Scanlan TS, Suchland KL, Hart ME, et al.  3-Iodothyronamine is an endogenous and rapid-acting derivative of thyroid hormone. Nat Med. 2004;10:638-642.
PubMed  DOI: 10.1038/nm1051

27. Jones CK, Eberle EL, Shaw DB, McKinzie DL, Shannon HE.  Pharmacologic interactions between the muscarinic cholinergic and dopaminergic systems in the modulation of prepulse inhibition in rats. J Pharmacol Exp Ther. 2005;312:1055-1063.
PubMed  DOI: 10.1124/jpet.104.075887

28. Wolinsky TD, Swanson CJ, Zhong H, Smith KE, Branchek TA, Gerald CP. Deficit in Prepulse Inhibition and Enhanced Sensitivity to Amphetamine in Mice Lacking the Trace Amine-1 Receptor poster. Paper presented at: The American College of Neuropsychopharmacology (ACNP) 43rd Annual Meeting; December 12-16, 2004; San Juan, Puerto Rico.

29. Hirashima A, Pan C, Shinkai K, et al.  Quantitative structure-activity studies of octopaminergic agonists and antagonists against nervous system of Locusta migratoria. Bioorg Med Chem. 1998;6:903-910.
PubMed  DOI: 10.1016/S0968-0896(98)00024-8

30. Hirashima A, Nagata T, Pan C, Kuwano E, Taniguchi E, Eto M.  Three-dimensional molecular field analyses of octopaminergic agonists and antagonists for the locust neuronal octopamine receptor class 3. J Mol Graph Model. 1999;17:198-218.
PubMed  DOI: 10.1016/S1093-3263(99)00031-5

31. Hirashima A, Morimoto M, Kuwano E, Taniguchi E, Eto M.  Three-dimensional common-feature hypotheses for octopamine agonist 2-(arylimino)imidazolidines. Bioorg Med Chem. 2002;10:117-123.
PubMed  DOI: 10.1016/S0968-0896(01)00247-4