Identification of novel allosteric modulators for the G-protein coupled US28 receptor of human cytomegalovirus

Article abstract of DOI:10.1016/j.bmcl.2011.06.120

The highly constitutively active G-protein coupled receptor US28 of human cytomegalovirus (HCMV) is an interesting pharmacological target because of its implication on viral dissemination, cardiovascular diseases and tumorigenesis. We found that dihydrois

Full text of DOI:10.1016/j.bmcl.2011.06.120

Bioorganic & Medicinal Chemistry Letters 21 (2011) 5446–5450
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Identification of novel allosteric modulators for the G-protein coupled
US28 receptor of human cytomegalovirus
Ana Kralj a, Alexander Wetzel b, Shohreh Mahmoudian c, Thomas Stamminger c, Nuska Tschammer a,
,
⇑
Markus R. Heinrich a,
⇑
a Department of Chemistry and Pharmacy, Emil Fischer Center, Friedrich-Alexander-Universität Erlangen-Nürnberg, Schuhstraße 19, D-91052 Erlangen, Germany
b Department of Chemistry and Biochemistry, Technische Universität München, Lichtenbergstraße 4, D-85747 Garching b. München, Germany
c Institute for Clinical and Molecular Virology, University Hospital Erlangen, Schlossgarten 4, D-91054 Erlangen, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
The highly constitutively active G-protein coupled receptor US28 of human cytomegalovirus (HCMV) is
an interesting pharmacological target because of its implication on viral dissemination, cardiovascular
diseases and tumorigenesis. We found that dihydroisoquinolinone and tetrahydroisoquinoline scaffolds
may be promising lead structures for novel US28 allosteric inverse agonists. These scaffolds were rapidly
synthesized by radical carboamination reactions followed by non-radical transformations. Our novel
US28 allosteric modulators provide valuable scaffolds for further ligand optimization and may be helpful
chemical tools to investigate molecular mechanisms of US28 constitutive signaling and its role in
pathogenesis.
Received 26 May 2011
Revised 27 June 2011
Accepted 28 June 2011
Available online 2 July 2011
Keywords:
US28
Human cytomegalovirus
G-Protein coupled receptor
Inverse agonist
Ó 2011 Elsevier Ltd. All rights reserved.
Radical carboamination
Human herpesviruses are widespread pathogens involved in
acute and chronic diseases.1 The omnipresent human cytomegalo-
virus (HCMV) causes a life-long latent infection in healthy hosts.1,2
In patients with immature or suppressed immune systems (e.g.,
neonates, AIDS [Acquired Immune Deficiency Syndrome], cancer
patients and transplant patients), HCMV can lead to severe and
life-threatening disease. HCMV infection is also associated with a
number of chronic diseases, including atherosclerosis, cancer and
transplant rejection.1,2 Viruses like HCMV encode viral G-protein
coupled receptors (vGPCRs) that are often highly homologous with
the host’s chemokine receptors; these vGPCRs couple very effi-
ciently to signaling networks of the host.3 Most vGPCRs signal
independently from a ligand. Constitutive signaling of vGPCRs aug-
ments virus survival, host invasion and, in some cases, oncogenesis
or cardiovascular disease, by exploiting preferred signaling cas-
cades.4 HCMV encodes US28, a highly constitutively active viral
GPCR receptor, which has the ability to bind different human CC-
chemokines (e.g., CCL5/Rantes) as well as the CX3C-chemokine
CX3CL1/fractalkine.1,2 Because of the potential role of US28 in viral
dissemination and persistence as well as in cardiovascular disease
and tumorigenesis, US28 remains an interesting pharmacological
target.2,5,6 The search for a potent inverse agonist that would re-
verse the constitutive activity of US28 is an ongoing challenge.
VUF2274 [5-(4-(4-chlorophenyl)-4-hydroxypiperidin-1-yl)-2,
2,-diphenylpentanenitrile], initially discovered as the CCR1 recep-
tor antagonist,7 has also been described as the first US28 inverse
agonist.8,9 VUF2274 and its derivatives behave as inverse agonists
in the PLCb pathway (measured as an accumulation of inositol tri-
phosphate). Despite significant efforts and detailed structure–
activity relationship (SAR) studies, the potency of this class of com-
pounds and some non-VUF2274-like compounds (e.g., methiothe-
pin) could not be improved beyond micromolar range.8,10,11
With the intention to identify novel molecular scaffolds that
would serve as a template for the development of US28 inverse
agonists, we dedicated our attention to the known chemokine
CXCR3 receptor ligands. The human CXCR3 receptor possesses
28% sequence homology with US28. As a comparison, US28 has
a 30% amino acid sequence homology with the human chemokine
CCR1 receptor,12 the primary target of VUF2274. Inspired by the
structure of the CXCR3 receptor antagonist, VUF5834,13 with no
reported activity at US28, we identified tetrahydroisoquinolines
and dihydroisoquinolinones as promising lead scaffolds for fur-
ther structural optimization. We applied radical carboamination
reactions starting from simple precursors to rapidly and effi-
ciently access the pharmaceutically important class of b-aryl-
amines, which were subjected to non-radical transformations to
afford dihydroisoquinolinone and tetrahydroisoquinoline deriva-
⇑
Corresponding authors. Tel.: +49 9131 852 3931; fax: +49 9131 852 2585.
E-mail addresses: nuska.tschammer@medchem.uni-erlangen.de (N. Tscham-
mer), markus.heinrich@medchem.uni-erlangen.de (M.R. Heinrich).
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.06.120

A. Kralj et al. / Bioorg. Med. Chem. Lett. 21 (2011) 5446–5450
5447
tives. Radical carboamination reactions starting from such simple
precursors as aryldiazonium salts and olefins offer fast and effi-
cient access to the pharmaceutically important class of b-aryl-
Because the structural designs of compounds 6 and 7 were in-
spired by VUF5834,13 we conducted several further experiments
to introduce a phenyl substituent at the nitrogen atom of the
isoquinolinone. These attempts to prepare compounds even more
closely related to the known CXCR3 ligands failed. To get access
to the N-phenylated dihydroisoquinolinones 8 and 10, the key
intermediates 4b and 4c were submitted to a copper-catalyzed
arylation protocol (step h).19 Esterification of 8 with decanoyl chlo-
ride gave test compound 9 (step g).
For the preparation of tetrahydroisoquinolines 11a and 11b the
two-step carboamination sequence shown in Scheme 2 was mod-
ified by a further reductive step with lithium aluminum hydride
(Scheme 2, steps a–c). Each of the two intermediates 11a and
11b was then reacted with benzoyl chloride, 2-furoyl chloride
and 2-phenacetyl chloride under reaction conditions leading to
double acylation at the nitrogen atom and the oxygen atom. In this
way, the six tetrahydroisoquinoline-derived test compounds 12–
17 were available.
Without a carboxylic ester functionality on the aromatic core of
the arenediazonium salt, b-arylamines of type 18 are obtained as
products from the carboamination sequence (Scheme 3, steps a
and b) instead of the isoquinolinones shown above (Scheme 1).
By applying a protocol reported by Ohwada20, we were able to
convert seven b-arylamines 18a–g into 1-phenyltetrahydroiso-
quinolines 19–25. For this purpose, the amines 18a–g were
condensed with benzaldehyde to yield imines (step c), which cy-
clized at elevated temperatures when treated with trifluorometh-
anesulfonic acid (step d). The reaction conditions for the
cyclization step led to a strong predominance of the cis-configured
products with a selectivity usually higher than 10:1 (cis:trans). In a
few cases we were also able to isolate the minor trans isomers,
such as compounds 200 and 230.
amines.14–16
In
combination
with
other
non-radical
transformations, a wide range of further products, especially het-
erocycles, becomes available in only a few steps. Compared to
organometallic methods for carboamination, which currently
have to be conducted in a partially intramolecular fashion (amine
equivalent usually attached to olefinic substrate), radical reac-
tions are not limited in this way.17,18 Herein, we present the first
application of this powerful synthetic strategy in the field of
medicinal chemistry. The HPLC purity of investigated compounds
was >95%.
Dihydroisoquinolinones are usually obtained when the two-
step carboamination sequence is carried out with aryldiazonium
salts bearing a carboxylic ester functionality in the ortho position
to the diazonium group. Initially (step a), azo compounds of type
3 are prepared by the iron(II)-mediated reaction of diazonium salt
1a and olefin of type 2 (Scheme 1). In the next step (b), which is
commonly carried out without isolation of azo compound 3, the
reductive cleavage of the N–N double bond is accompanied by lac-
tamization to give the desired target molecules 4. According to this
pathway, we prepared three dihydroisoquinolinones 4a–c, which
served as key intermediates for further transformations. The spiro-
cyclic ketone 5 was obtained from acetate 4a by hydrolysis (step c)
and Swern oxidation (step d). Reductive amination of ketone 5
with N,N-dimethyl ethylenediamine (steps e and f) and subsequent
coupling to decanoyl chloride provided the spirocyclic test com-
pound 6. Compound 4a was obtained as a 2:1 mixture of diastere-
oisomers (trans:cis of –NH- and –OAc on the cyclohexane ring).
Compound 6 was obtained as a single isomer. Its relative stereo-
chemistry could not be determined due to an overlap of signals
in the NMR spectrum. Comparable reagents and conditions (steps
d–g) were employed to convert dihydroisoquinolinone 4b into test
compound 7.
For the characterization of novel allosteric modulators of US28,
we used the luciferase based PathDetect Elk1 gene reporter assay.
Functional assays have the capability to detect ligands that
O
O
NH
e,f,g
NH
O
Me2N
N
C9H19
5
6
O
c,d from 4a
O
O
NH
CO2Me
N2
R1
NH
a,b
d,e,f,g
+
R1
R2
Me2N
from 4b
N
O
R2
C9H19
1a
2a-c
7
BF4
4a:
R1,R2 = (CH2)4-CH(OAc)
4b:
R
R
1 = Me, R2 = CH2OH
1 = Me, R2 = (CH2)2OH
4c:
O
h
N
from 4b
OR3
R = H
3 = C(O)C9H19
3
8:
9:
g
R
CO2Me
O
CO2Me
N
N
h
N
from 4c
R1
R2
OH
3
10
Scheme 1. Reagents and conditions: (a) FeSO4, H2O–DMSO, rt; (b) Raney-Ni/H2 (50 bar), MeOH or EtOH, rt or Zn, HCl, rt; (c) NaOH, MeOH; (d) (COCl)2, NEt3, DMSO, CH2Cl2,
ꢀ78 °C; (e) N,N-dimethyl ethylenediamine, CHCl3, 90 °C ; (f) NaBH4, MeOH, rt; (g) H19C9COCl, NEt3, 1,2-C2H4Cl2, 50 °C; (h) Ph–I, Cu, DMF, 150 °C.

5448
A. Kralj et al. / Bioorg. Med. Chem. Lett. 21 (2011) 5446–5450
O
CO2Me
N
R2
O
a,b,c
NH
R1
d
+
R1
2b,c
R2
from 11a
N2
BF4
11a: R1 = CH2OH
12
13
14
2
O
O
1a
: R2 = Ph
11b: R1 = (CH2)2OH
: R2 = 2-furyl
: R = CH2Ph
O
N
R2
d
from 11b
O
R2
2
15
16
17
: R2 = Ph
: R2 = 2-furyl
: R = CH2Ph
Scheme 2. Reagents and conditions: (a) FeSO4, H2O–DMSO, rt; (b) Zn, HCl, 50 °C; (c) LiAlH4, THF, reflux; (d) PhCOCl, furan-2-carboxylic acid chloride or PhCH2COCl, NEt3, 1,2-
C2H4Cl2, 50 °C.
R1
R1
R1
NH
R2
NH2
R2
BF4
N2
a,b
c,d
R2
+
1b-f
2d-f
18a-g
2
19-25
19
20
R1
R1
=
=
H , R = n-C4H9 :
18a
p-F , R2 = n-C4H9: 18b
20' (trans config.)
R1 = p-Cl, R2 = n-C4H9 :
18c
21
22
23
R1 = m-Cl, R2 = n-C4H9 : 18d
R
1 = o-Cl , R2 = n-C4H9 : 18e
23'
24
25
(trans config.)
R
R
1 = o-Cl , R2 = n-C3H7 : 18f
1 = o-Cl , R2 = n-C6H13:18g
Scheme 3. Reagents and conditions: (a) FeSO4, H2O–DMSO, rt; (b) Zn, HCl, MeOH, 50 °C; (c) PhCHO, MgSO4, Et2O, rt; (d) CF3SO3H, 120 °C.
allosterically alter the affinity and/or efficacy for any species that
interacts with the receptor, including cytosolic signaling pro-
teins.21,22 The US28 receptor that we used was cloned from the
clinically relevant and highly endotheliotropic HCMV strain
TB40/E (Genbank Accession No. ABV71518.1, Supplementary
data).23 Constitutive activity of US28 mediates the activation of
transcription factors through p42/p44 mitogen-activated protein
kinase (MAPK) and p38 MAPK-dependent pathways.2,3 The
transcription factor Elk-1 is phosphorylated and activated by
p42/p44 MAPK.24 The activation of Elk-1 results in an increased
expression of the reporter protein luciferase. For the quantitation
of the luciferase expression, BrightGlo reagent was used and lumi-
nescence was measured with a microplate reader. The increase in
the luminescence was directly proportional to the activation of
Elk-1 transcription factor. The constitutive activity of US28 in-
creased the basal luminescence by 20-fold compared to mock
transfected HEK cells (Supplementary data). The degree of inverse
agonism (maximal effect) was entirely dependent on the level of
constitutive activity of the system (i.e., the difference between
the constitutive and non-constitutive basal responses).25
the maximal efficacy of VUF2274 can be attributed to different as-
say conditions. In the reporter gene assay, the reporter gene is
transcribed and accumulated in the cell as soon as the constitutive
active US28 is expressed. Unless compounds are added immedi-
ately after transfection, the efficacy of the inverse agonists is likely
underestimated.9
The substituted dihydroisoquinolinone 7 and its spirocyclic
derivative 6 demonstrated an inverse agonism on US28, with the
EC50 values of 1.50 and 4.80
l
M and efficacies of ꢀ20 and ꢀ21%,
respectively (Table 1, Fig. 1). The N-phenylated dihydroisoquinoli-
none derivative 8 behaved as an agonist on US28. The exchange of
the hydroxymethyl group of 8 for hydroxyethyl group of 10 yielded
the inverse agonist with EC50 value 1.00 lM, and the efficacy of
ꢀ14 (Table 1, Fig. 1).
Further modifications of the dihydroisoquinolinone scaffold
resulted in a series of strong US28 inverse agonists. The best
compounds in the series were benzoyl substituted tetrahydroiso-
quinoline 12 with the EC50 value of 8.50
ꢀ62% and phenacetyl substituted tetrahydroisoquinoline 14 with
the EC50 value of 3.40
lM and the efficacy
l
M and the efficacy ꢀ37% (Table 1, Fig. 1).
Antipsychotic methiothepin that has been described as an ago-
nist on US28,26 barely showed any agonist effect on the US28 wild
type receptor in our test system. VUF2274 is a weak inverse
The exact position of ester acyl oxygen as a potential hydrogen
bond acceptor seems to be important for the binding affinity of
the inverse agonists we examined. The exchange of hydroxymethyl
with hydroxyethyl linker as in 15, 16 and 17 resulted in a three to
four fold loss of affinity. To confirm that inverse agonism of the de-
scribed compounds does not originate from their cytotoxicity, the
agonist on US28 (EC50 = 4.5
l
M, efficacy = ꢀ22%). VUF2274 was re-
ported as full inverse agonist in the PLCb pathway (measured as an
accumulation of inositol triphosphate).9 The observed difference in

A. Kralj et al. / Bioorg. Med. Chem. Lett. 21 (2011) 5446–5450
5449
Table 1
which behaved as weak allosteric agonists on the US28 wild type
(Table 1). Unfortunately, this group of compounds demonstrated
visible affinity and efficacy on the D2L receptor (Supplementary
data), indicating that this molecular scaffold might be used for
the development of biogenic amine receptor ligands. The
Functional characterization of compounds 6–25 employing the US28 receptora
Compound
EC50
(lM)
pEC50 SEM
Efficacy (% over basal)
VUF2274
Methiothepin
6
7
8
9
4.50
5.34 0.46
6.45 0.45
5.32 0.27
5.82 0.21
5.36 0.18
n.d.
5.99 0.40
5.07 0.02
5.13 0.09
5.46 0.22
4.87 0.07
n.d.
5.12 0.08
6.56 0.32
5.75 0.26
5.70 0.51
5.11 0.49
n.d.
6.89 0.21
5.85 0.36
5.81 0.26
5.64 0.39
ꢀ22
10
ꢀ22
ꢀ20
41
n.d.
ꢀ14
ꢀ62
ꢀ18
ꢀ37
ꢀ70
n.d.
ꢀ51
22
15
15
12
n.d.
21
16
27
21
8
3
4
3
7
0.35
4.80
1.50
4.40
unfavorable D2L receptor affinity (2 lM) was reported also for the
US28 inverse agonist VUF2274.7 Methiothepin displayed agonist
activity on US28 and has an antipsychotic action on serotonin
and dopamine receptors.27 The cross-reactivity of small-molecu-
lar-weight chemokine ligands with biogenic amine receptors is
one of the main issues regarding the design of highly selective che-
mokine ligands.28
n.d.
10
12
13
14
15
16
17
19
20
20’
21
22
23
23’
24
25
1.00
8.50
7.40
3.40
13.50
n.d.
7.50
0.28
1.80
2.00
7.80
n.d.
3
2
2
6
5
Our attempts to design a novel inverse agonist for US28 re-
sulted in a series of allosteric modulators with the potency and
efficacy comparable or superior to the reference VUF2274. These
novel US28 allosteric modulators provide valuable chemical tools
to investigate molecular mechanisms of US28 constitutive signal-
ing and its role in the pathogenesis of viral infection, tumorigenesis
and development of cardiovascular disease. The further
development of novel US28 allosteric inverse agonists by synthesis
procedures that will yield enantiopure compounds and the deter-
mination of preferred receptor ligand interactions, which depend
on the absolute stereochemistry of the enantiomers, are in
progress.
6
3
3
5
3
0.13
1.40
1.50
2.30
2
3
5
5
a
Functional data were obtained on transfected HEK cells that transiently
expressed US28 as shown by the PathDetect trans Elk-1 reporter gene assay. Dose
response curves of 4–8 experiments performed in triplicates have been normalized
and pooled to get a mean curve from which the EC50 value and the maximum
intrinsic activity of each compound was obtained. n.d.—not detectable.
Acknowledgments
We thank to DFG (Deutsche Forschungsgemeinschaft) for the
financial support to A.W. and Stefanie Fehler for the syntheses con-
ducted during her internship. T.S. was supported by DFG (Deutsche
Forschungsgemeinschaft, GRK1071) and IZKF (Interdisziplinäres
Zentrum für Klinische Forschung) Erlangen.
Supplementary data
Supplementary data (experimental procedures, sequence
alignment, results of the cytotoxicity assay, the reporter gene assay
performed on mock transfected HEK cells, and synthesis proce-
dures) associated with this article can be found, in the online ver-
sion, at doi:10.1016/j.bmcl.2011.06.120.
References and notes
1. Maussang, D.; Vischer, H. F.; Leurs, R.; Smit, M. J. Mol. Pharmacol. 2009, 76, 692.
2. Vomaske, J.; Nelson, J. A.; Streblow, D. N. Infect. Disord. Drug Targets 2009, 9,
548.
3. Boomker, J. M.; van Luyn, M. J.; The, T. H.; de Leij, L. F.; Harmsen, M. C. Rev. Med.
Virol. 2005, 15, 269.
4. Sodhi, A.; Montaner, S.; Gutkind, J. S. Nat. Rev. Mol. Cell Biol. 2004, 5, 998.
5. Streblow, D. N.; Orloff, S. L.; Nelson, J. A. Curr. Drug Targets Infect. Disord. 2001,
1, 151.
6. Maussang, D.; Verzijl, D.; van Walsum, M.; Leurs, R.; Holl, J.; Pleskoff, O.;
Michel, D.; van Dongen, G. A. M. S.; Smit, M. J. Proc. Natl. Acad. Sci. U.S.A. 2006,
103, 13068.
7. Hesselgesser, J.; Ng, H. P.; Liang, M.; Zheng, W.; May, K.; Bauman, J. G.;
Monahan, S.; Islam, I.; Wei, G. P.; Ghannam, A.; Taub, D. D.; Rosser, M.; Snider,
R. M.; Morrissey, M. M.; Perez, H. D.; Horuk, R. J. Biol. Chem. 1998, 273, 15687.
8. Vischer, H. F.; Hulshof, J. W.; Hulscher, S.; Fratantoni, S. A.; Verheij, M. H.;
Victorina, J.; Smit, M. J.; de Esch, I. J.; Leurs, R. Bioorg. Med. Chem. 2010, 18, 675.
9. Casarosa, P.; Menge, W. M.; Minisini, R.; Otto, C.; van Heteren, J.; Jongejan, A.;
Timmerman, H.; Moepps, B.; Kirchhoff, F.; Mertens, T.; Smit, M. J.; Leurs, R. J.
Biol. Chem. 2003, 278, 5172.
Figure 1. Functional characterization of reference compounds VUF2274 and
methiothepin, and representative novel compounds 7, 8 and 14. HEK cells were
transiently transfected with US28 and components of PathDetect Elk-1. Normalized
curves from 3 to 6 experiments, each performed in triplicate, are shown. The error
bars represent the SEM.
10. Hulshof, J. W.; Vischer, H. F.; Verheij, M. H.; Fratantoni, S. A.; Smit, M. J.; de
Esch, I. J.; Leurs, R. Bioorg. Med. Chem. 2006, 14, 7213.
cell viability was investigated. The compounds with the highest
rate of inverse agonism (compounds 12, 15 and 17) had minor
but significant cytotoxicity, which could account for the pro-
nounced inverse agonist effect (Supplementary data). The other
novel inverse agonists (7, 10, 13 and 14) had no cytotoxicity. The
observed inverse agonism was thus mediated by US28.
11. Hulshof, J. W.; Casarosa, P.; Menge, W. M.; Kuusisto, L. M.; van der Goot, H.;
Smit, M. J.; de Esch, I. J.; Leurs, R. J. Med. Chem. 2005, 48, 6461.
12. Gao, J. L.; Murphy, P. M. J. Biol. Chem. 1994, 269, 28539.
13. Storelli, S.; Verdijk, P.; Verzijl, D.; Timmerman, H.; van de Stolpe, A. C.; Tensen,
C. P.; Smit, M. J.; De Esch, I. J. P.; Leurs, R. Bioorg. Med. Chem. Lett. 2005, 15, 2910.
14. Heinrich, M. R.; Blank, O.; Wolfel, S. Org. Lett. 2006, 8, 3323.
15. Heinrich, M. R.; Blank, O.; Wetzel, A. J. Org. Chem. 2007, 72, 476.
16. Blank, O.; Wetzel, A.; Ullrich, D.; Heinrich, M. R. Eur. J. Org. Chem. 2008, 2008,
3179.
Structural modifications of the dihydroisoquinolinone core re-
sulted in a series of 1-phenyltetrahydroisoquinolines (19–25)

5450
A. Kralj et al. / Bioorg. Med. Chem. Lett. 21 (2011) 5446–5450
17. Bertrand, M. B.; Leathen, M. L.; Wolfe, J. P. Org. Lett. 2007, 9, 457.
18. Sherman, E. S.; Fuller, P. H.; Kasi, D.; Chemler, S. R. J. Org. Chem. 2007, 72, 3896.
19. Renaud, J.; Bischoff, S. F.; Buhl, T.; Floersheim, P.; Fournier, B.; Halleux, C.;
Kallen, J.; Keller, H.; Schlaeppi, J. M.; Stark, W. J. Med. Chem. 2003, 46, 2945.
20. Nakamura, S.; Tanaka, M.; Taniguchi, T.; Uchiyama, M.; Ohwada, T. Org. Lett.
2003, 5, 2087.
23. Sinzger, C.; Hahn, G.; Digel, M.; Katona, R.; Sampaio, K. L.; Messerle, M.; Hengel,
H.; Koszinowski, U.; Brune, W.; Adler, B. J. Gen. Virol. 2008, 89, 359.
24. Cruzalegui, F. H.; Cano, E.; Treisman, R. Oncogene 1999, 18, 7948.
25. Kenakin, T. Mol. Pharmacol. 2004, 65, 2.
26. Schall, T. J.; McMaster, B. E.; Dairaghi, D. J. U.S. Patent US2002/0127544 A1,
2002.
21. Kenakin, T. P. Trends Pharmacol. Sci. 2009, 30, 460.
22. Kenakin, T. P. J. Biomol. Screen. 2010, 15, 119.
27. Burt, D. R.; Creese, I. A. N.; Snyder, S. H. Mol. Pharmacol. 1976, 12, 631.
28. Allegretti, M.; Bertini, R.; Bizzarri, C.; Beccari, A.; Mantovani, A.; Locati, M.
Trends Pharmacol. Sci. 2008, 29, 280.