Design and synthesis of a functionally selective D3 agonist and its in vivo delivery via the intranasal route

Article abstract of DOI:10.1016/j.bmcl.2007.10.059

This paper reports the synthesis and biological activity of a novel series of aryl-morpholine dopamine receptor agonists. Several compounds show high levels of functional selectivity for the D3 over the D2 dopamine receptor. Compound 26 has >1000-fold functional selectivity and has been successfully progressed in vivo using an intranasal delivery route.

Full text of DOI:10.1016/j.bmcl.2007.10.059

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 17 (2007) 6691–6696
Design and synthesis of a functionally selective D3 agonist
and its in vivo delivery via the intranasal route
Julian Blagg,a,* Charlotte M. N. Allerton,a David V. J. Batchelor,a Andrew D. Baxter,a,
Denise J. Burring,a Christopher L. Carr,a Andrew S. Cook,a Carly L. Nichols,a
Joanne Phipps,a Vivienne G. Sanderson,a Hugh Verriera and Stephen Wongb
aPfizer Global Research & Development, Ramsgate Road, Sandwich, Kent CT13 9NJ, UK
bDepartment of Neurosciences Biology, Pfizer Global Research & Development, Groton, CT 06340, USA
Received 17 September 2007; revised 16 October 2007; accepted 16 October 2007
Available online 22 October 2007
Abstract—This paper reports the synthesis and biological activity of a novel series of aryl-morpholine dopamine receptor agonists.
Several compounds show high levels of functional selectivity for the D3 over the D2 dopamine receptor. Compound 26 has >1000-
fold functional selectivity and has been successfully progressed in vivo using an intranasal delivery route.
ꢀ 2007 Elsevier Ltd. All rights reserved.
The dopaminergic system is characterised by 5 receptor
subtypes, which cluster into 2 families. The D1-like
receptors (D1 and D5) and the D2-like receptors (D2,
D3 and D4). Non-selective dopamine agonists, includ-
ing Apomorphine and Pramipexole, have been investi-
gated for the treatment of male erectile dysfunction
(MED) (Table 1).1 Apomorphine has been shown to
facilitate penile erection in clinical studies in men.2 Upri-
maTM (apomorphine in the form of a sublingual tablet)
was marketed for the treatment of MED; however, poor
efficacy3 and competition from ViagraTM were cited
amongst the reasons for its withdrawal. The efficacy of
apomorphine is compromised by dose-limiting side
effects (e.g., nausea, dizziness and hypotension)4 and
by a relatively short duration of exposure in man
(T1/2 = 35 min). We set out to examine the hypothesis
that whilst both the D2 and D3 receptors may be in-
volved in the erectile process, D2 receptor activation is
responsible for the dose-limiting side effects of non-
selective dopamine agonists.
We hypothesised that a selective D3 agonist would
prove effective in the treatment of MED whilst reducing
the levels of nausea, emesis and cardiovascular effects
observed with non-selective agents. Dopamine agonists
exemplified in the literature demonstrated marginal
selectivity for D3 over D2 receptor activation in our
functional assays, whilst Apomorphine and Quinelorane
demonstrated essentially balanced activity (Table 1).
The marginal in vitro functional selectivity of Pramipex-
ole for the D3 receptor translated to an equivalent pro-
erectile efficacy in in vivo animal models of MED to
Apomorphine. In addition, there was some indication
of a better therapeutic ratio over nausea, emesis and
blood pressure effects.5 These preliminary findings
prompted our search for a tool compound with signifi-
cantly greater functional selectivity for activation of
the D3 over the D2 receptor. Such a tool compound
would enable definitive in vivo demonstration of efficacy
with a widened therapeutic index. Literature reports of
dopamine subtype selective agonists have been exten-
sively reviewed.6 However, many disclosures have been
supported by in vitro binding as opposed to in vitro
functional data. Herein we report the discovery of a ser-
ies of aryl-morpholines, several of which show impres-
sive levels of in vitro functional selectivity for D3 over
D2 receptor activation in the same assay format and
conditions where literature agents (Table 1) show mar-
ginal or no functional selectivity. In addition, the com-
pounds described here show good selectivity over a
range of other aminergic GPCR targets.
Keywords: Dopamine D3 receptor agonist; Arylmorpholine; Synthesis;
Intranasal delivery.
*
Corresponding author at present address: Cancer Research UK
Centre for Cancer Therapeutics, The Institute of Cancer Research,
15 Cotswold Road, Sutton, Surrey SM2 5NG, UK. Tel.: +44 (0)20
8722 4051; fax: +44 0208 722 4126; e-mail: julian.blagg@icr.ac.uk
Present address: Serentis Ltd, 23 Cambridge Science Park, Milton
Road, Cambridge CB4 0EY, UK.
0960-894X/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.10.059

6692
J. Blagg et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6691–6696
Table 1. D3/D2 Selectivity data for some literature dopamine agonists
Compound
Structure
Average functional EC50
(nM)
D3/D2
D2
D3
Apomorphine
Pramipexole
20
27
7
1
3
27
Ropinerole
83
4
1
21
Quinelorane
2
2
Our initial starting point was to carry out a mini-high
throughput functional screen versus the D3 receptor.
In vitro functional potency was determined by inhibition
of forskolin stimulated cAMP accumulation in recombi-
nant cell lines expressing either the human D3 receptor
(expressed in CHO cells) or the human D2 receptor (ex-
pressed in rat pituitary cells).5 The compounds screened
comprised a subset of the Pfizer compound file chosen
for their propensity to interact with aminergic GPCR
targets. Selectivity versus the D2 receptor was subse-
quently used to triage the hits. The most promising hit
matter discovered from this screening campaign con-
tained the aryl morpholine template and included the
racemic compounds 1, 2 and 3 (Table 2). The optimal
N-alkyl chain was quickly established as n-propyl 6 by
in-house screening of analogues bearing varied substitu-
ents on the morpholine nitrogen and by reference to lit-
erature SAR from N-propyl-3-hydroxyphenyl piperidine
(3-PPP)7 and tricyclic dopamine agonists8 where an N-
propyl chain favoured dopamine activity. Compounds
with longer or shorter alkyl chains had significantly re-
duced functional activity. Compounds 1, 2, 3 and 6 all
triggered a full agonist response at the human recombi-
nant D3 receptor. Compound 1 displayed weak agonism
at the D2 receptor; no evidence for switching to func-
tional antagonism at either the D3 or D2 receptors
was observed.
Synthesis of the racemic aryl morpholines was achieved
according to the route exemplified for compound 6
(Scheme 1). Aryl aldehyde 7 was converted to the ami-
no-alcohol 8 via intermediate cyanohydrin formation.
Preparation of the amide 9 using propionyl chloride
was followed by borane reduction to give the propyl
amino-alcohol intermediate 10. Cyclisation was carried
out with chloro-acetylchloride, followed by reduction
to the target morpholine ring system. Deprotection of
the phenol in refluxing HBr gave the racemate 6
(Scheme 1). The route depicted in Scheme 1 was also
used as the basis for synthesis of potential replacements
for the phenol (Table 3). These compounds were pre-
pared by incorporation of the appropriate aromatic
functionality into the starting aldehyde 7 or by func-
tional group manipulation of the phenol 6 (routes not
shown).9
Table 2. Mini-HTS results and subsequent SAR
Compound R group
Functional Functional
D3 EC50
D2 EC50 (lM )
1
2
3
H
Et
864 nM
547 nM
1312 nM
7.2
>30
>100
The phenol proved essential for potent and full agonism
at the D3 receptor. None of the replacements described
here that bear polar groups, H-bond donor or H-bond
acceptor functionality designed to mimic the phenol re-
tained significant functional activity.
Allyl
NHAc
4
>10 lM
>10
*
OMe
5
6
>3000 nM >10
27 nM >10
*
n-propyl
Separation of the racemate 6 by chiral HPLC (Chir-
alpak AD column)10 gave the enantiomers 26 and 27

J. Blagg et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6691–6696
6693
Figure 1. Profile of individual enantiomers 26 and 27.
Compound 26 met our potency and selectivity criteria.
It is a moderately lipophilic base (LogD7.4 octanol = 1;
pKa = 7.4), with high aqueous solubility (>1 g/ml),
good Caco-2 flux (Papp = 38 · 10À6 cm sÀ1 at 1 lM)
and low plasma protein binding (41%, 76% and 65%
in rat, dog and human plasma, respectively). Com-
pound 26 displayed minimal binding activity
(IC50 > 10 lM) against a wide panel of targets (Cerep
wide ligand profile) including dopamine D1 and D5
receptors. Weak binding activity was observed at the
human 5HT1A receptor (16% at 100 nM and 95% at
10 lM). The lack of significant 5HT1A activity was fur-
ther confirmed using human recombinant assays to de-
tect antagonism (IC50 > 20 lM) and agonist activity
(inactive at 1 lM). In addition, compound 26 displayed
weak binding activity at a range of other selectivity tar-
gets; for example: Rat 5HT2A receptor: Ki = 970 nM;
rat alpha1A receptor: Ki > 1 lM, human alpha2A recep-
tor: Ki > 1lM; human histamine H1 receptor:
Ki > 2 lM; human muscarinic M1 receptor: Ki > 1lM;
dofetilide binding: Ki > 4 lM. The promising pharma-
cology and physiochemical properties prompted
in vivo evaluation to explore the efficacy-toleration
profile. Multiple dose evaluation in animal models of
MED via subcutaneous administration gave an ED50
of approximately 0.2 to 0.5 nM free drug concentra-
tion. No nausea, emesis or hypotensive effects were
Scheme 1. Reagents and conditions: (a) 2 M HCl, Na2SO3, KCN,
THF, rt, 30 min; (b) BH3-THF (1 M in THF), reflux, 1.5 h under N2
69% combined yield for a and b; (c) Propionyl chloride, Et3N, DCM,
rt, 5 h 67%; (d) BH3.SMe2, THF, 93%; (e) chloro-acetyl chloride,
NaOH, H2O, DCM, rt, 30 min; then aqKOH, IPA, rt, 2 h, 90%; (f)
BH3-THF, (1 M in THF) reflux, 3 h 93%; (g) aqHBr, reflux, 1 h, 85%.
Table 3. D3 and D2 functional assay results for compounds 6–25
Compound
R group
Functional
D3 EC50 (nM)
Functional
D2 EC50
6
OH
27
>10 lM
>10 lM
>10 lM
>10 lM
>1000 nM
>10 lM
>10 lM
NTa
11
12
13
14
15
16
17
18
19
20
21
22
H
Cl
>1000
>2800
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
>1000
OMe
OCF3
SMe
SO2Me
CO2Me
CO2H
CH2OH
CN
>10 lM
NTa
NTa
CONH2
NH2
NTa
>1000 nM
23
24
NHSO2Me
>1000
>1000
>1000 nM
NTa
25
>1000
>10 lM
a Not tested.
which demonstrated that all activity resides in the R-
enantiomer (26), (Fig. 1). The absolute configuration
of the active enantiomer (crystallized as its HCl salt)
was confirmed by small molecule X-ray analysis
(Fig. 2).9,11
Figure 2. X-ray structure of active phenol enantiomer 26.11

6694
J. Blagg et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6691–6696
seen at these efficacious exposure levels or at signifi-
cantly higher exposures (up to 2.7 lM free drug con-
centrations following subcutaneous dosing in the dog)
indicative of a much wider therapeutic index than
non-selective dopamine agonists.5 However, compound
26 suffered from poor oral bioavailability due to clear-
ance at, or greater than, liver blood flow in rat and
dog. Unsurprisingly, glucuronidation of the phenolic
group was the major contributor to the rapid first pass
clearance. A programme was initiated to reduce the
rate of phenol conjugation by modulating the phenolic
pKa and by introduction of proximal steric bulk to dis-
rupt interaction of the phenol with the glucuronidase
active site.12,13 Maintaining functional D3 potency
whilst altering the pharmacokinetic profile proved diffi-
cult. Di-substituted compounds for example 30, 31 and
38 resulted in significant loss of potency (Table 4).
Compound 41, a potent full agonist; contains the
well-known ortho-hydroxyaniline toxicophore which
limited its potential for further investigation. Substitu-
tion ortho- to the morpholine ring (positions R1 and
R5) significantly reduced activity possibly due to
restriction of the torsional rotation between the aryl
moiety and the morpholine ring. However, compound
29 demonstrated that an R3 fluoro group is well toler-
ated and we further examined the effect of an ortho-flu-
oro substituent on the pharmacokinetic profile.
Figure 3. Data for separated enantiomers of compound 29.
There is precedent in the literature from Terbutaline15
and Phenylephrine16 (Fig. 4) that conjugative clearance
of a mono-phenol can be improved by addition of a sec-
ond phenolic group in the meta position. Application of
this strategy to compound 26 resulted in compound 32
with maintained activity (Table 4). Separation of 32 into
enantiomers 46 and 47 (Chiralpak AD column)17 again
showed that all activity resides in one isomer. In this
case improved rat and human hepatocyte stability was
achieved; however, in vivo clearance in rat significantly
exceeded rat liver blood flow indicating that the im-
proved in vitro stability was unlikely to translate
in vivo in human (Fig. 4).
The programme of phenol replacements and aryl ring
modifications described here did not solve the issue of
the high first pass clearance for compound 26. One solu-
tion was to opt for a non-oral delivery route following
the precedented bioavailability improvements observed
with, for example, intranasal Apomorphine.18 Advanta-
ges of the intranasal dosing route include the avoidance
of first pass metabolism, consequent improved bioavail-
ability and a rapid onset of action.19,20 Intranasal
Compound 29 was separated into enantiomers 44 and 45
(Chiralpak AD column),14 again all activity resided in
one isomer 44. However, the poor stability of 44 in
human hepatocytes, coupled with its stability in micro-
somal preparations, indicated that stability to glucuron-
idation or other Phase II conjugative pathways was
unlikely to have been improved. This compound was
not progressed further (Fig. 3).
Table 4. D3 functional assay results for compounds 28–43
Compound
R1
R2
R3
R4
R5
Functional D3 EC50
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
Me
H
H
H
H
H
H
H
H
H
H
H
H
H
H
OH
OH
H
H
H
H
H
H
H
H
H
Br
H
H
H
H
H
H
H
>1000 nM
35 nM
F
Cl
H
OH
OH
H
>688 nM
>5900 nM
37 nM
CH2OH
H
H
OH
OH
OH
CN
CONH2
H
H
>10 lM
>10 lM
>10 lM
>1000 nM
>1000 nM
427 nM
OH
OH
H
H
H
CN
OH
OH
OH
OH
OH
OH
OH
H
H
CONH2
COMe
CH2OH
NH2
NHSO2Me
H
H
>1000 nM
>1000 nM
40 nM
H
H
H
>1000 nM
43
H
OH
H
H
>1000 nM

J. Blagg et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6691–6696
6695
quired for erectile efficacy in the same species, indicates
that these adverse events may be linked to D2 receptor
activation. The efficacy of compound 26 in animal mod-
els of male erectile dysfunction indicates that activation
of the D3 receptor may play a significant role in male
erectile dysfunction. However, the hypothesis that selec-
tive activation of the D3 receptor in man will demon-
strate efficacy in the absence of dose limiting side
effects remains to be proven.
References and notes
1. (a) Giuliano, F.; Allard, J.; Rampin, O.; Droupy, S.;
Benoit, G.; Alexandre, L.; Bernabe, J. Int. J. Impot. Res.
2001, 13, 110; (b) Melis, M. A.; Argiolas, A. Neurosci.
Biobehav. Rev. 1995, 19, 19; (c) Rosen, R. C.; Ashton, A.
K. Arch. Sex. Behav. 1993, 22, 521.
2. Dula, E.; Bukofzer, S.; Perdok, R.; George, M. Eur. Urol.
2001, 39, 558.
3. MacLennan, K. M.; Boshier, A.; Wilton, L. V.; Shakir, S.
A. W. Brit. J. Urol. Int. 2006, 98, 125.
4. Montorsi, F. Int. J. Impot. Res. 2003, 15, S7.
5. Van der Graaf, P. H.; Wayman, C. P.; Baxter, A. D.;
Cook, A. S.; Wong, S. K. Chem. Abstr. 2003, 139, 63348,
WO 03/051370, 2003.
Figure 4. The effect of an additional meta-phenol on clearance.
delivery is greatly facilitated for highly soluble com-
pounds and the aqueous solubility of compound 26
(>1 g/ml) was very favourable in this regard. Rat and
dog intranasal pharmacokinetics (Table 5) gave encour-
agement that the intranasal route would indeed be a via-
ble development option.21
6. Boeckler, F.; Gmeiner, P. Pharmacol. Therapeut. 2006,
112, 281.
7. Cannon, J. G. Progress in Drug Research. In Chapter 9:
Dopamine Agonists: Structure–activity Relationships; Juc-
ker, E., Ed.; Birkhauser Verlag: Basel, 1985; vol. 29, p 381.
8. Van Vliet, L. A.; Rodenhuis, N.; Dijkstra, D.; Wikstrom,
H.; Pugsley, T. A.; Serpa, K. A.; Meltzer, L. T.; Heffner,
T. G.; Wise, L. D.; Lijiness, M. E.; Huff, R. M.; Svensson,
K. J. Med. Chem. 2000, 43, 2871.
9. Allerton, C. M. N.; Baxter, A. D.; Cook, A. S.; Hepworth,
D.; Wong, S. K. Chem. Abstr. 2004, 141, 71567, WO 04/
052372, 2004.
10. The enantiomers were separated by chiral column chro-
matography at room temperature (Chiralpak AD250,
20 mm column) eluting with hexane: isopropyl alcohol:
diethylamine (70:30:0.05) to give R(-)-3-(4-propylmorph-
olin-2-yl)phenol (26) (e.e. >99.5%) followed by S(+)- 3-(4-
propylmorpholin-2-yl)phenol (27) (e.e. >99%).
11. Crystallographic data (excluding structure factors) for the
structures in this paper have been deposited at the
Cambridge Crystallographic Data Centre for small mol-
ecules and allocated the deposition number CCDC
661368. Copies of data can be obtained free of charge
on application to CCDC, 12 Union Road, Cambridge,
CB2 1EZ, UK.
12. Yin, H.; Bennet, G.; Jones, J. P. Chem-Biol. Interact. 1994,
90, 47.
13. Soffers, A. E. M. F.; Veeger, C.; Rietjens, I. M. C. M.
Xenobiotica 1994, 24, 759.
14. The enantiomers were separated by chiral column chro-
matography at room temperature (Chiralpak AD250,
20 mm column) eluting with hexane: isopropyl alcohol
(90:10) to give R-(+)-2-Fluoro-5-(4-propylmorpholin-2-
yl)phenol (44) (e.e. >99.5%) followed by S-(À)-2-Fluoro-
5-(4-propylmorpholin-2-yl)phenol (45) (e.e. >99%).
15. The Pharmacological Basis of Therapeutics (IX edition),
Gilman, A. Ed.; McGraw-Hill, 1996, 1785.
Based upon these in vivo animal data, the predicted hu-
man intranasal bioavailability was an encouraging 25%
compared to a predicted 0% oral bioavailability. The
disappearance half-life values in rat, dog and human
hepatocytes for compound 26 were 28, 18 and 61 min,
respectively, indicative of high hepatic extraction in all
species. In vivo pharmacokinetic studies in rat and dog
demonstrated plasma clearance close to hepatic blood
flow as expected. Thus, compound 26 was predicted to
exhibit plasma clearance values approximating to liver
blood flow in humans. The volume of distribution in hu-
mans was predicted to be in the range 1.4–6.6 L/kg. To-
gether these figures predict a half-life in humans in the
range of 0.6–3 h.
In conclusion, we have described the synthesis and prop-
erties of the first potent, full D3 agonist with high func-
tional selectivity over D2 receptor agonism. High first
pass clearance of the lead compound, 26, could not be
overcome by the structural modifications presented
here. However, intranasal administration enabled signif-
icant systemic exposure to be achieved in rat and dog.
The absence of nausea, emesis and hypotensive effects
in preclinical animal models with compound 26 at free
drug concentrations, significantly exceeding those re-
Table 5. Intranasal pharmacokinetics of compound 26
16. The Pharmacological Basis of Therapeutics, (X edition),
Gilman, A. Ed.; McGraw-Hill, 2001, 1994.
17. The enantiomers were separated by chiral column chro-
matography at room temperature (Chiralpak AD250,
20 mm column) eluting with hexane: isopropyl alcohol
Rat 2 mg/kg
Dog 0.3 mg/kg
Tmax
T1/2
5 min
0.3 h
20%
5 min
1.6 h
33%
Intranasal bioavailability

6696
J. Blagg et al. / Bioorg. Med. Chem. Lett. 17 (2007) 6691–6696
(80:20) to give R-5-(4-propylmorpholin-2-yl)benzene-1,3-
rats and to male (n = 2) and female (n = 2) dogs.
Compound 26 was freshly prepared in saline solution. In
rat, the intravenous dose was administered as a bolus dose
of 1 mg/kg. The intranasal dose was administered via a
Gilson pipette (20 ll) into one nostril to give a final dose
of 0.5 and 2 mg/kg. In dog, the intravenous dose was
administered as an intravenous infusion over a 10-min
period at a rate of 3 ml/kg/h of a 0.2 mg/ml solution, to
give a final dose of 0.1 mg/kg. The intranasal dose was
administered via a Pfeiffer device (100 ll) into one nostril
to give a final dose of 0.1 and 0.3 mg/kg.
diol (46) (e.e. = 99%) followed by S-5-(4-propylmorpho-
lin-2-yl)benzene-1,3-diol (47) (e.e. = 96%).
18. Sam, E.; Jeanjean, A. P.; Maloteaux, J. M.; Verbeke, N.
Eur. J. Drug Metab. Ph. 1995, 20, 27.
19. Arora, P.; Garg, S.; Sharma, S. Drug Discov. Today 2002,
7, 967.
20. Behl, C. R.; Pimplaskar, H. J.; Sileno, A. P.; DeMeireles,
J.; Romeo, V. D. Adv. Drug Delivery Rev. 1998, 29, 89.
21. The pharmacokinetics of compound 26 were determined
following single intravenous and intranasal doses to male