Synthesis and pharmacological evaluation of novel γ-aminobutyric acid type B (GABAB) receptor agonists as gastroesophageal reflux inhibitors

Article abstract of DOI:10.1021/jm701425k

We have previously demonstrated that the prototypical GABA_B receptor agonist baclofen inhibits transient lower esophageal sphincter relaxations (TLESRs), the most important mechanism for gastroesophageal reflux. Thus, GABA_B agonists could be exploited for the treatment of gastroesophageal reflux disease. However, baclofen, which is used as an antispastic agent, and other previously known GABA_B agonists can produce CNS side effects such as sedation, dizziness, nausea, and vomiting at higher doses. We now report the discovery of atypical GABA_B agonists devoid of classical GABA_B agonist related CNS side effects at therapeutic doses and the optimization of this type of compound for inhibition of TLESRs, which has resulted in a candidate drug (R)-7 (AZD3355) that is presently being evaluated in man.

Full text of DOI:10.1021/jm701425k

J. Med. Chem. 2008, 51, 4315–4320
4315
Synthesis and Pharmacological Evaluation of Novel γ-Aminobutyric Acid Type B (GABAB)
Receptor Agonists as Gastroesophageal Reflux Inhibitors
Christer Alstermark,† Kosrat Amin,† Sean R. Dinn,‡ Thomas Elebring,*,† Ola Fjellstro¨m,† Kevin Fitzpatrick,‡ William B. Geiss,‡
Johan Gottfries,† Peter R. Guzzo,‡ James P. Harding,‡ Anders Holme´n,† Mohit Kothare,‡ Anders Lehmann,† Jan P. Mattsson,†
Karolina Nilsson,† Gunnel Sunde´n,† Marianne Swanson,† Sverker von Unge,† Alex M. Woo,‡ Michael J. Wyle,‡ and
Xiaozhang Zheng‡
AstraZeneca R&D Mo¨lndal, S-431 83 Mo¨lndal, Sweden, and Albany Molecular Research, Inc., 21 Corporate Circle,
P.O. Box 15098, Albany, New York 12212-5098
ReceiVed NoVember 12, 2007
We have previously demonstrated that the prototypical GABAB receptor agonist baclofen inhibits transient
lower esophageal sphincter relaxations (TLESRs), the most important mechanism for gastroesophageal reflux.
Thus, GABAB agonists could be exploited for the treatment of gastroesophageal reflux disease. However,
baclofen, which is used as an antispastic agent, and other previously known GABAB agonists can produce
CNS side effects such as sedation, dizziness, nausea, and vomiting at higher doses. We now report the
discovery of atypical GABAB agonists devoid of classical GABAB agonist related CNS side effects at
therapeutic doses and the optimization of this type of compound for inhibition of TLESRs, which has resulted
in a candidate drug (R)-7 (AZD3355) that is presently being evaluated in man.
Introduction
It has been demonstrated that the prototypical γ-aminobutyric
a
acid type B (GABAB ) receptor agonist baclofen inhibits
transient lower esophageal sphincter relaxations (TLESRs) in
ferrets, dogs and humans.1 TLESRs account for the majority
of reflux episodes in both healthy individuals and patients with
gastroesophageal reflux disease (GERD).1c,d However, high
doses of baclofen and other previously known GABAB agonists
produce central nervous system (CNS) related side effects like
Figure 1. Structure of the novel GABAB agonist (R)-7 together with
sedation, dizziness, nausea, and vomiting, which would not be
acceptable for a drug intended for treatment of GERD. Thus,
GABAB agonists effective in inhibiting TLESRs but devoid of
CNS side effects would have a major therapeutic potential. Since
GABAB receptors are expressed both within and outside the
CNS in the reflex pathway underlying TLESRs,2 one strategy
was to find agonists that would affect peripheral GABAB
receptors selectively.
Here, we report the syntheses and some pharmacological
characteristics of a series of GABAB agonists from which a
candidate drug (R)-7 (AZD33553) has been selected and
submitted to studies in humans (for structure, see Figure 1). In
contrast to baclofen, compound (R)-7 lacks CNS side effects at
doses effective in inhibiting TLESRs in animal models.
We decided to explore derivatives of 3-aminopropylphos-
phinic acid (1, CGP27492), since this compound has been
demonstrated to be a potent and selective GABAB agonist.4 As
compound 1 has been described in the literature as being inactive
in vivo (as a muscle relaxant), speculatively because of
metabolic lability of the P-H bond,4,5 we soon focused on
structures of some known GABAB agonists.
derivatives of 3-aminopropyl(methyl)phosphinic acid (2,
SK&F975414) to avoid this metabolic liability.6 However, as it
proved difficult to significantly improve the therapeutic window
for derivatives of 2, we eventually decided to re-evaluate the
alleged metabolic lability of 1 and its P-H congeners. Sur-
prisingly,4,5 we found that it was possible to design new P-H
phosphinic acids that showed a high metabolic stability (see
data for (S)-3 in Results and Discussion). Since 1 has been found
to be inactive at CNS GABAB receptors after systemic
administration, our hypothesis was that 3-aminopropylphos-
phinic acids may inhibit TLESR through a peripheral site of
action without carrying the burden of CNS side effects. In fact,
almost all efforts to find therapeutically useful GABAB agonists
have been done in the CNS field, which explains the lack of
interest in exploring analogues of 1. Therefore, we decided to
explore this compound class further, aiming at increased
therapeutic window and maintained potency.3
Chemistry
* To whom correspondence should be addressed. Phone: (+46) 31
7762116. Fax: (+46) 31 7763724. E-mail: Thomas.Elebring@
astrazeneca.com.
ThesynthesesoftheknownGABAB agonists1-3(CGP35583),
(S)-4 (CGP44532), (R)-4 (CGP44533), and 6 [(3-amino-2-
oxopropyl)(methyl)phosphinic acid] were all performed in
accordance with those procedures already being described in
the literature7 (for structures, see Table 1). All compounds that
were isolated as single enantiomers were synthesized by routes
using optically pure starting materials (single enantiomers of
serine or epichlorohydrin). Since the reactions are known to be
†
AstraZeneca R&D Mo¨lndal.
Albany Molecular Research, Inc.
‡
a Abbreviations: GABAB, γ-aminobutyric acid type B; TLESR, transient
lower esophageal sphincter relaxation; GERD, gastroesophageal reflux
disease; CNS, central nervous system; EC50, 50% effective concentration;
ED2, effective dose producing 2 °C decrease in body temperature; IC50,
50% inhibition concentration; FLIPR, fluorescence imaging plate reader.
10.1021/jm701425k CCC: $40.75
2008 American Chemical Society
Published on Web 06/25/2008

4316 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 14
Alstermark et al.
Scheme 2a
Table 1. Binding Affinities and Agonistic Properties of Compounds
1-8
GABAB affinity
Ki ( SEM
(n) (nM)
GABAB agonism
a,b
a
EC50
( SEM
compd
R1
R2
(n) (nM)
baclofen
1
2
220 ( 50 (6)
15 ( 3.2 (11)
33 ( 2.3 (87)
94 ( 32 (3)
750 ( 150 (6)
19 ( 3.0 (3)
41 ( 6.3 (7)
220 ( 29 (3)
130 ( 17 (7)
1100 (1)
150 ( 50 (3)
600 ( 200 (2)
81 ( 6.3 (3)
270 ( 25 (2)
15 ( 1.5 (5)
250 ( 10 (2)
8.64 ( 0.77 (4)
23 ( 1.9 (2)
1700 ( 240 (3)
14 ( 3.2 (3)
a Reagents: (i) LDA, THF, -78 °C, 1 h; (ii) Boc-Gly-OMe, -78 °C, 45
min; (iii) HOAc, -78 °C, then room temp, 45 min; (iv) 3 M HCl, room
temp, 14 h; (v) MeOH, propylene oxide, room temp, 12 h.
H
CH3
H
H
H
CH3
CH3
H
CH3
H
H
H
CH3
CH3
CH3
H
H
3
OH
OH
OH
OH
OH
dO
)O
F
F
F
F
F
(S)-3
(R)-3
(S)-4
(R)-4
5
6
7
(S)-7
(R)-7
8
(S)-8
(R)-8
50 ( 8.1 (11)
2100 ( 570 (2)
180 ( 20 (3)
840 ( 330 (3)
48 ( 8.8 (12)
200 ( 50 (11)
10 ( 1.9 (12)
70 ( 7.1 (5)
5.1 ( 1.2 (10)
14 ( 3.2 (9)
480 ( 30 (3)
4.3 ( 1.3 (3)
Scheme 3a
F
a Binding affinities (Ki) and agonistic properties (EC50) of compounds
at the GABAB receptor were measured as detailed in the Experimental
Section. Data are the mean ( SEM of n experiments. b All compounds
shown in the table had an intrinsic activity of 1.
Scheme 1a
a Reagents: (i) HMDS, reflux, 2 h; (ii) ethyl 2-fluoroacrylate, 60 °C, 3
days for 16 (room temp, 60 h for 17); (iii) NH4OH, EtOH, room temp,
16 h; (iv) BH3-THF, THF, reflux, 2.5 h; (v) 6 M HCl, reflux, 2.5 h; (vi)
DOWEX 50WX-8-200, H+ form.
to afford ketone 13, which then was completely deprotected by
a hydrolytic reaction to give 5.
The preparation of the racemic 2-fluoro analogues 7 and 8
was achieved as shown in Scheme 3. Ethyl (diethoxymeth-
yl)phosphinate (149) and ethyl (methyl)phosphinate (1510) were
both transformed to the corresponding trimethylsilylated P(III)
intermediates (structures not shown), which on addition to ethyl
2-fluoroacrylate11 afforded fluoropropanoates 16 and 17, re-
spectively. The carboxylic acid ester moiety of 16 and 17 was
converted to the corresponding amide, yielding 18 and 19,
respectively, which then were reduced and finally subjected to
acidic hydrolyses to afford 7 and 8, respectively.
a Reagents: (i) HMDS, reflux, 3 h; (ii) (R)- or (S)-epichlorohydrin, ZnCl2,
60 °C, 15 h; (iii) 1% HOAc/MeOH, 15 h; (iv) 9% NH4OH, room temp, 4
days, then 60 °C, 24 h; (v) concentrated HCl, reflux 2 h; (vi) propylene
oxide, MeOH, 5 h.
The syntheses of the enantiomerically pure 2-fluoro analogues
(S)-7, (R)-7, (S)-8, and (R)-8 employed either D-serine or L-serine
as starting materials (see Scheme 4). The stereospecific trans-
formations of the ꢀ-hydroxy-R-amino acids to the R-fluoro-ꢀ-
amino acid derivatives (S)-20 and (R)-20 were performed in
accordance to Somekh and Shanzer.12 Reduction of the ester
group was followed by removal of the benzyl groups, which
afforded amino alcohols (S)-21 and (R)-21, respectively. The
amino group was Boc-protected, which gave carbamates (S)-
22 and (R)-22, and then the hydroxy group was replaced by an
iodine atom to give the iodo intermediates (S)-23 and (R)-23,
respectively. Subsequent introduction of the phosphinic acid
head was accomplished by treating the alkyl iodides with 10
equiv of bis(trimethylsilyl) phosphonite.13 Boc deprotection
afforded the GABAB receptor ligands (S)-7 and (R)-7, respec-
tively, which optionally were P-methylated to afford (S)-8 and
(R)-8, respectively.
stereospecific, the absolute configuration of compounds (S)-3,
(R)-3, (S)-7, (R)-7, (S)-8, and (R)-8 could be determined from
the knowledge of stereochemistry of each starting material.
There has been no sign of any racemization in the reaction steps,
and thus, the compounds are stated to be optically pure. In
addition, benzamide derivatives of compounds (S)-7 and (R)-7
have been analyzed by chiral HPLC and no opposite enantiomer
has been detected (see Experimental Section).
The optically pure 2-hydroxy analogues, (S)-3 and (R)-3, were
prepared as shown in Scheme 1. Ethyl (diethoxyethyl)phosphi-
nate (9)4 was transformed to the corresponding P(III) intermedi-
ate followed by reaction with (R)- and (S)-epichlorohydrin,
respectively. The two enantiomers of the formed trimethylsi-
lylated ethers were then hydrolyzed to yield chlorohydrins (R)-
10 and (S)-10, respectively, which then were transformed to
aminoalcohols (S)-11 and (R)-11, respectively. The phosphinic
acid ester moiety was finally subjected to acid hydrolysis to
afford (S)-3 and (R)-3, respectively.
Results and Discussion
The 2-oxo analogue 5 was prepared by the synthetic route
diagrammed in Scheme 2. Ethyl (diethoxyethyl)(methyl)phos-
phinate (12)8 was condensed with N-Boc-glycine methyl ester
Generally, only minor modifications are tolerated in the parent
3-aminopropyl(methyl)phosphinic and 3-aminopropylphosphinic

GABAB Receptor Agonists
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 14 4317
Scheme 4a
°C decrease in body temperature). In contrast, the corresponding
values for (S)-3 are 130 nM (Table 1), 7 µmol/kg, and >8000
µmol/kg, respectively. Consequently, if therapeutic index is
defined as the ratio of ED2 to ED50, 3-aminopropylphosphinic
acids with a small substituent or no substituent in the 2-position
have significantly better therapeutic index than the corresponding
P-Me analogues. Also 3-aminopropylsulfinic acids16 with a
small substituent or no substituent in the 2-position behave as
3-aminopropylphosphinic acids; i.e., they have considerably
better therapeutic index than baclofen and 3-aminopropyl-
(methyl)phosphinic acids. A more detailed SAR for 3-amino-
propylphosphinic acids and 3-aminopropyl(methyl)phosphinic
acids regarding effects on TLESR and CNS side effects will be
the subject of a forthcoming publication.15 Also, the mechanisms
underlying the in vivo differences between these two compounds
classes are beyond the scope of the current work and will be
published separately.15
Despite the fact that the molecules contain a P-H bond, it is
notable that the novel 3-aminopropylphosphinic acid derivatives
have metabolic properties making them useful as drugs even
when given perorally. This finding was unexpected because
contradictory results of the compound class have previously been
reported in the literature.4,5 The discovery is exemplified by
the fact that the bioavailability of compound (S)-3 is 88% in
dog and 78% in rat.
a Reagents: (i) SOCl2, MeOH; (ii) PhCH2Br, NaHCO3, DMSO, THF;
(iii) DAST, THF, room temp; (iv) LiBH4, THF, -15 °C, then room temp,
17 h; (v) aqueous NH4Cl, 0 °C; (vi) H2, Pd(OH)2-C, EtOH, 55 psi, room
temp, 6 h; (vii) (Boc)2O, K2CO3, H2O, dioxane, room temp, 17 h; (viii)
PPh3, I2, imidazole, CH2Cl2, 0 °C, then room temp, 17 h; (ix) HP(OTMS)2,
10 equiv, CH2Cl2, room temp, 18 h; (x) MeOH/H2O; (xi) DOWEX 50WX-
8-200, H+ form; (xii) HMDS, reflux, 16 h; (xiii) diglyme, reflux, 6 h; (xiv)
N,N-diisopropylethylamine, MeI, room temp, 24 h; (xv) DOWEX 50WX-
8-200, H+ form.
Conclusions
acids, in both of which the 2-position is the most accommodating
site. For example, as has been shown previously,4 introduction
of a hydroxy group in the 2-position is quite well-tolerated. As
for the aminopropyl(methyl)phosphinic acids, it was shown that
the S-isomer was more potent than the corresponding R-isomer.4
We have found that this is also true for the corresponding P-H
analogues. Thus, it seems that a hydroxy group in the 2-position
is involved in a stereospecific interaction with the GABAB
receptor. Interestingly, by introducing a fluorine atom in the
2-position of both of the compound classes, we discovered
GABAB agonists with potencies that exceed that of any
previously known GABAB agonist (see data for 7, (R)-7, and
(R)-8). In contrast to what has been established for the
2-hydroxy-substituted analogues, the R-isomer of the 2-fluoro
derivatives was unexpectedly found to be more potent than the
S-isomer (cf. data for (R)-7 and (S)-7, and (R)-8 and (S)-8,
respectively).
In summary, we have discovered that GABAB agonists can
be divided into two groups with respect to inhibition of TLESR
and production of CNS side effects. Tentatively, we have
attributed this differentiation to a difference in availability of
the compounds to CNS GABAB receptors. This discovery
enabled us to select (R)-7 as a candidate drug, devoid of typical
GABAB receptor related CNS side effects despite its high
affinity for the GABAB receptor. This compound is presently
being evaluated in man with respect to inhibition of TLESR
and reflux and may be exploited in the treatment of GERD.
Experimental Section
General Methods. Melting points are uncorrected. 1H NMR
spectra were recorded at 300 or 500 MHz, 19F NMR spectra at
282 MHz, and 31P NMR at 121 MHz. Elemental analyses were
performed at Quantitative Technologies Inc. (Whitehouse, NJ). All
reactions were performed under a dry N2 or argon atmosphere.
Analytical TLC was performed on silica gel GF (Analtech) or silica
gel 60 F254 (EM Science) plates. Flash column chromatography
was performed with silica gel 60 (230-400 mesh, EM Science).
Baclofen used in this work was racemic, i.e., (()-ꢀ-(aminomethyl)-
4-chlorobenzenepropanoic acid, and purchased from Aldrich-Sigma.
Ethyl (2R)-3-Chloro-2-hydroxypropyl(1,1-diethoxyethyl)phos-
phinate ((R)-10). A mixture of ethyl (diethoxyethyl)phosphinate (9)4
(15.0 g, 71 mmol) and toluene was evaporated to dryness. The
residue and 1,1,1,3,3,3-hexamethyldisilazane (HMDS) (13.2 g, 82
mmol) were heated to reflux for 3 h under an argon atmosphere.
The mixture was cooled to room temperature and evaporated. (R)-
Epichlorohydrin (6.6 g, 71 mmol) and anhydrous zinc chloride (2.5
g, 18 mmol) were added, and the reagents were heated to 60 °C
overnight under an argon atmosphere. The mixture was cooled to
room temperature and diluted with methylene chloride and water.
The organic layer was washed with water, dried over MgSO4,
filtered, and evaporated to give 20.7 g of a yellow oil. The residue
was dissolved in methanol (150 mL) containing 1% acetic acid,
and the solution was stirred overnight. The solvent was removed
to give 17.7 g (82%) of (R)-10 as a clear oil. 1H NMR (500 MHz,
CDCl3) δ 4.37 (m, 1H), 4.23 (m, 2H), 3.57-3.86 (m, 6H),
1.92-2.37 (m, 2H), 1.53 (dd, J ) 2.3, 11.4 Hz, 3H), 1.32-1.37
(m, 3H), 1.18-1.24 (m, 6H).
In general, as has also been shown previously,4 3-aminopropyl-
(methyl)phosphinic acids have in vitro potency in the same range
at the GABAB receptor compared to the corresponding 3-ami-
nopropylphosphinic acids. However, these two compound
classes behave very differently in animal studies with regard to
inhibition of TLESRs as well as to possible side effects. While
those 3-aminopropyl(methyl)phosphinic acids that are potent
GABAB agonists in vitro have an in vivo profile similar to that
of baclofen, the corresponding P-H analogues have an in vivo
profile that is considerably different with regard to TLESR
inhibition as well as to side effects. Thus, while both compound
classes act as inhibitors of TLESRs in a dose-dependent way,
where the in vivo potency correlates with the affinity for the
GABAB receptor in vitro, 3-aminopropyl(methyl)phosphinic
acids can produce CNS side effects such as sedation14 close to
therapeutic doses, whereas the P-H analogues are devoid of
side effects even at high doses.15 For instance, (S)-4 has an EC50
for stimulation of human recombinant GABAB receptors
amounting to 150 nM (Table 1), an ED50 for TLESR inhibition
in the dog of about 0.17 µmol/kg, and an ED2 in the mouse
approximating 5.4 µmol/kg (ED2, which is used to assess CNS
side effects, is defined as the subcutaneous dose producing a 2

4318 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 14
Alstermark et al.
Ethyl (2S)-3-Amino-2-hydroxypropyl(1,1-diethoxyethyl)phos-
phinate ((S)-11). A solution of (R)-10 (5.0 g, 17 mmol) in ethanol
containing 9% of ammonia was stirred in an autoclave at room
temperature for 4 days and at 60 °C for 1 further day. The solution
was evaporated, and the residue was purified by chromatography
on a wet-packed silica gel column eluting with methylene chloride/
methanol (5-8% MeOH) containing 5% triethylamine. The ap-
propriate fractions were combined, evaporated, and diluted with
methylene chloride and water. The aqueous layer was pH adjusted
to the alkaline side by the addition of a few milliliters of 10%
aqueous Na2CO3 and repeatedly extracted with methylene chloride.
The combined organic layers were dried over Na2SO4 and
MHz, D2O) δ 7.13 (d, J ) 551 Hz, 1H), 4.14 (s, 2H), 3.14 (d, J )
18 Hz, 2H); FABMS: m/z ) 138 (M + H)+. Anal. (C3H8NO3P) C,
H, N, P.
Ethyl 3-[(Diethoxymethyl)(ethoxy)phosphoryl]-2-fluoropro-
panoate (16). A mixture of ethyl (diethoxymethyl)phosphinate (14)9
(26.0 g, 133 mmol) and HMDS (28 mL, 133 mmol) was heated to
reflux for 2 h under an argon atmosphere. The mixture was cooled
to room temperature, and ethyl 2-fluoroacrylate11 (10.5 g, 89.0
mmol) was added. The reagents were heated to 60 °C for 3 days
under an argon atmosphere. The mixture was cooled to room
temperature, diluted with ethyl acetate (300 mL), and washed with
1 N HCl (2 × 150 mL) and saturated sodium chloride (100 mL).
The organic layer was dried over MgSO4, filtered, and evaporated
to give 32.0 g of a yellow oil. The residue was purified by
chromatography on a wet-packed silica gel column (6 cm × 30
cm) eluting with 97:3 methylene chloride/methanol. The appropriate
fractions were combined and evaporated to give 16.0 g (57%) of
1
evaporated to give 1.2 g (26%) of (S)-11 as a clear oil. H NMR
(300 MHz, CDCl3) δ 4.40-4.55 (m, 1H), 4.10-4.30 (m, 2H),
3.55-3.80 (m, 4H), 3.20-3.30 (m, 1H), 3.00-3.10 (m, 1H),
2.00-2.40 (m, 2H), 1.45-1.53 (dd, J ) 3.4, 11.7 Hz, 3H),
1.30-1.40 (m, 3H), 1.15-1.25 (m, 6H).
1
16 as a clear oil. H NMR (300 MHz, CDCl3) δ 5.32 (dm, 1H),
(2S)-3-Amino-2-hydroxypropylphosphinic Acid ((S)-3). A mix-
ture of (S)-11 (1.0 g, 3.5 mmol) and concentrated HCl (50 mL)
was heated to reflux for 2 h. The solution was cooled to room
temperature and evaporated. The residue was dissolved in methanol
(100 mL) and treated with propylene oxide (2 mL) at room
temperature. After the mixture was stirred for 5 h, the precipitated
solid was collected by decanting off the solvent. The solid was
dried with a stream of argon to give 220 mg (45%) of (S)-3 as a
4.67-4.77 (m, 1H), 4.18-4.32 (m, 4H), 3.58-3.91 (m, 4H),
2.30-2.62 (m, 2H), 1.20-1.41 (m, 12H).
Ethyl (3-Amino-2-fluoro-3-oxopropyl)(diethoxymethyl)phos-
phinate (18). To a solution of phosphinate 16 (16.0 g, 51.1 mmol)
in ethanol (22 mL) was added concentrated ammonium hydroxide
(14.8 N, 3.5 mL, 51.1 mmol). The solution was stirred for 16 h
and evaporated. The residue was purified by chromatography on a
wet-packed silica gel column (7 cm × 37 cm) eluting with 96.5:
3.5 methylene chloride/methanol. The appropriate fractions were
combined and evaporated to give 3.43 g (27%) of 18 as a clear oil.
1H NMR (300 MHz, CDCl3) δ 6.43 (s, 1H), 5.70 (s, 1H), 5.21-5.49
(dm, 1H), 4.7 (dd, 1H), 4.18-4.31 (m, 2H), 3.65-3.91 (m, 4H),
2.21-2.81 (m, 2H), 1.30-1.40 (m, 3H), 1.20-1.28 (m, 6H).
(3-Amino-2-fluoropropyl)phosphinic Acid (7). To an ice bath
cooled solution of amide 18 (3.43 g, 13.5 mmol) in THF (15 mL)
was added 1 M BH3-THF (8.7 mL, 8.7 mmol) while under an
argon atmosphere. After 10 min, the solution was heated to reflux
for 2.5 h. The solution was cooled to room temperature, and 6 N
HCl (200 mL) was added slowly. The THF was removed by
evaporation in vacuo and the aqueous layer refluxed for 2.5 h. The
solution was cooled and evaporated. The residue was purified by
ion exchange chromatography (DOWEX 50WX-8-200, H+ form,
3.5 cm × 4.0 cm). The ion-exchange resin was washed with
2:1 methanol/water (400 mL). The crude product dissolved in 1:1
methanol/water was loaded onto the column and washed with 1:1
methanol/water (400 mL). The eluent was changed to 3:1 methanol/
concentrated ammonium hydroxide. Two fractions (150 mL) were
combined and evaporated to give 645 mg (34%) of 7 as a white
solid: mp 203-207 °C; Rf ) 0.37 (60:40:1 CH3OH/CH2Cl2/
NH4OH); 1H NMR (300 MHz, D2O) δ 7.11 (d, J ) 528 Hz, 1H),
5.18 (dm, J ) 54 Hz, 1H), 3.28-3.45 (m, 2H), 1.65-2.23 (m,
2H); 31P NMR (121 MHz, D2O) δ 17.5 (d, J ) 21 Hz); 19F NMR
(282 MHz, D2O) δ -184 (d, J ) 22.6 Hz); APIMS m/z 142 [M +
H]+. Anal. (C3H9FNO2P·0.5H2O) C, H, N. H: calcd, 6.72; found,
6.28.
(2R)-3-(Amino)-2-fluoro-1-propanol ((R)-21). Lithium borohy-
dride (5.3 g, 0.24 mol) was suspended in THF (200 mL) under a
nitrogen atmosphere and cooled to -15 °C with stirring. Ester (R)-
2012 (56.6 g, 0.190 mol) was suspended in THF (250 mL) and
added dropwise to the mixture over 1 h; the internal temperature
was maintained below -10 °C during the addition. On completion
of addition, the reaction mixture was allowed to warm to room
temperature and stirred at this temperature for 17 h. TLC analysis
indicated complete consumption of starting material. The reaction
mixture was cooled to 0 °C and cautiously quenched with a
saturated aqueous solution of ammonium chloride (300 mL). The
mixture was extracted with ethyl acetate (2 × 200 mL), and the
organic phase was concentrated under reduced pressure. The crude
residue was dissolved in 2 N hydrochloric acid (200 mL, pH 2
approximately), and the aqueous phase was washed with ether (2
× 200 mL). The aqueous phase was basified (pH 10 approximately)
with 80% ammonium hydroxide in brine, extracted with ethyl
acetate (3 × 200 mL), dried over anhydrous sodium sulfate (10 g),
1
white solid: mp 220-225 °C; H NMR (300 MHz, D2O) δ 7.08
(dt, J ) 1.2, 522 Hz, 1H), 4.21 (m, 1H), 2.94-3.23 (m, 2H),
1.72-1.99 (m, 2H); 31P NMR (121 MHz, D2O) δ 24.2 (d, J )
522 Hz); FABMS m/z ) 140 (M + H)+; [R]D25 +8° (0.5% in 0.1
M HCl). Anal. (C3H10NO3P) C, H, N.
Ethyl [3-[N-(tert-Butoxycarbonyl)amino]-2-oxopropyl](1,1,-di-
ethoxyethyl)phosphinate (13). To a solution of diisopropylamine
(195 mL, 1.39 mol, 3.5 equiv) in THF (300 mL) at -10 °C was
added dropwise (about 3.5 h) n-BuLi (1.6 M in hexanes, 867 mL,
1.39 mol, 3.5 equiv). After 10 min, the mixture was cooled to -78
°C and a solution of ethyl (diethoxyethyl)(methyl)phosphinate8 (12,
178 g, 0.793 mol, 2 equiv) in THF (400 mL) was added dropwise
(about 30 min). After the addition, the solution was stirred at -78
°C for 1 h. A solution of N-Boc-glycine methyl ester (75.0 g, 0.396
mol, 1 equiv) in THF (450 mL) was added dropwise (about 45
min). After the addition was complete, the mixture was stirred for
45 min. Acetic acid (79.5 mL, 1.39 mol) was added, and the mixture
was warmed to room temperature. Brine (200 mL) was added to
the reaction mixture, and the organic layer was separated. The
aqueous layer was extracted once with ethyl acetate (300 mL). The
combined organic extracts were dried over MgSO4, filtered, and
evaporated to remove the solvent. The residue was purified by
chromatography on 5 kg of silica gel eluting with ethyl acetate/
hexanes (3/2). The appropriate fractions were collected to give 129 g
(85%) of 13 as an oil. 1H NMR (300 MHz, CDCl3) δ 5.48 (s, 1H),
4.10-4.30 (m, 2H), 4.17 (d, 2H), 3.60-3.80 (m, 4H), 3.01-3.30
(m, 2H), 1.52 (d, 3H), 1.43 (s, 9H), 1.32 (t, 3H), 1.19 (t, 6H).
(3-Amino-2-oxopropyl)phosphinic Acid (5). Compound 13 (153
g, 400 mmol) was dissolved in 3 N HCl (2400 mL) that was
previously deoxygenated by bubbling argon through the solution.
The mixture was stirred for 14 h at room temperature and then
concentrated. The residue was coevaporated with methanol. The
residue was then dissolved in methanol (450 mL), and propylene
oxide was added (225 mL). The mixture was stirred for 12 h and
the resulting precipitate isolated by filtration. The solid was washed
with cold methanol and dried under vacuum at 50 °C to give 41.8 g
of solid. The solid was dissolved in water (350 mL), and the
insoluble impurities were removed by filtration. The filtrate was
diluted further with water (400 mL). To this aqueous solution,
acetone was added with stirring until the completion of the
precipitation (1600 mL of acetone). The resulting precipitate was
isolated by filtration. The solid was washed with acetone and dried
under vacuum at 50 °C to give 39.1 g of 5 (71%) as a tan solid:
mp 143-145 °C; Rf ) 0.45 (85:15 methanol/water); 1H NMR (300

GABAB Receptor Agonists
Journal of Medicinal Chemistry, 2008, Vol. 51, No. 14 4319
filtered, and concentrated under reduced pressure to afford 48 g of
a yellow oil of which 29 g was dissolved in ethanol (300 mL). An
amount of 10 wt % palladium(II) hydroxide on carbon (5.0 g) was
added and the mixture placed on a Parr shaker and shaken under a
hydrogen atmosphere (55 psi) for 6 h. When no further hydrogen
uptake was observed, the mixture was filtered through a pad of
Celite (20 g). A fresh batch of palladium(II) hydroxide (5 g) was
added to the ethanol mixture and resubjected to the hydrogenation
conditions described above for 17 h. The crude mixture was filtered
through a pad of Celite and concentrated under reduced pressure
to afford 9.6 g of (R)-21 (96%) as a pale-yellow oil: 1H NMR (300
MHz, CD3OD) δ 4.78-5.00 (br s, 3H), 4.49-4.62 (m, 0.5H),
4.32-4.46 (m, 0.5H), 3.54-3.70 (m, 2H), 2.70-2.96 (m, 2H).
saturated solution of hydrogen chloride gas in ethyl acetate (500
mL) was added and the mixture stirred for 3 h. The reaction mixture
was filtered, and the solids were washed with a mixture of methanol
and ethyl acetate (90:10). The filtrate was concentrated under
reduced pressure, and the crude product was passed through a
Dowex 50WX8-200 mesh H+ form (500 g, 8 cm × 15 cm) column,
eluting with 1:1 methanol/water until no further material was
detected by TLC analysis (chloroform, methanol, concentrated
ammonium hydroxide solution, 6:3:1, ninhydrin development,
Rf((R)-23) ) 0.9, Rf(R)-7) ) 0.3). The crude product was then eluted
with 1:3 concentrated ammonium hydroxide solution/methanol. The
product was further purified by silica gel column chromatography,
eluting with chloroform, methanol, concentrated ammonium hy-
droxide solution (6:3:1) to afford (R)-7 as a white solid (3.12 g,
tert-Butyl [(2R)-2-Fluoro-3-hydroxypropyl]carbamate ((R)-22).
Amino alcohol (R)-21 (4.6 g, 49 mmol) was dissolved in 25%
aqueous dioxane (160 mL). Potassium carbonate (7.1 g, 51 mmol)
was added and the mixture cooled to 0 °C. Di-tert-butyl dicarbonate
(11.6 g, 53.0 mmol) was added in two portions. The mixture was
then allowed to warm to room temperature overnight. The crude
mixture was concentrated to dryness. Then water (150 mL) was
added followed by saturated aqueous potassium hydrogen sulfate
(until pH 3 approximately). The organic material was extracted with
methylene chloride (2 × 150 mL), dried over Na2SO4, filtered, and
concentrated under reduced pressure to afford 9.5 g (100%) of (R)-
1
24%): mp ) 183-185 °C; H NMR (300 MHz, D2O) δ 7.90 (s,
0.5 H), 6.15 (s, 0.5 H), 5.12-5.29 (m, 0.5 H), 4.92-5.10 (m, 0.5
H), 3.12-3.42 (m, 2H), 1.74-2.26 (m, 2H); [R]2D5 -4.0° (c 1.0,
H2O); APIMS m/z 142 [M + H]+. Anal. (C3H9FNO2P·0.25H2O)
C, H, N.
(2R)-(3-Amino-2-fluoropropyl)(methyl)phosphinic Acid ((R)-
8). A suspension of compound (R)-7 (1.0 g, 7.1 mmol) in HMDS
(7.47 mL, 35.4 mmol) was heated to reflux for 16 h. The mixture
was cooled to room temperature, treated with diglyme (8.0 mL),
and heated to reflux for 6 h. After the mixture was cooled to room
temperature, N,N-diisopropylethylamine (1.23 mL, 7.10 mmol) was
added followed by dropwise addition of methyl iodide (1.32 mL,
21.2 mmol). The reaction mixture was stirred for 24 h, and then it
was diluted with methylene chloride and extracted with 2 N HCl
solution. The aqueous layer was washed with methylene chloride
and diethyl ether and then evaporated under reduced pressure. The
crude product was passed through a Dowex 50WX8-200 mesh H+
form column, eluting with 1:1 methanol/water until no further
material was detected by TLC analysis (chloroform, methanol,
concentrated ammonium hydroxide solution, 6:3:1, ninhydrin
development, Rf((R)-7) ) 0.3, Rf((R)-8) ) 0.2). The requisite crude
(R)-8 was then eluted with 1:3 concentrated ammonium hydroxide
solution/methanol. The crude product (R)-8 was further purified
by column chromatography, eluting with methylene chloride,
methanol, concentrated ammonium hydroxide solution (6:3:1) to
afford (R)-8 as a white solid (720 mg, 65%): mp ) 210-214 °C;
1H NMR (300 MHz, D2O) δ 5.20 (m, 0.5H), 5.03 (m, 0.5H),
3.20-3.42 (m, 2H), 1.80-2.22 (m, 2H), 1.30 (d, J ) 14.0 Hz,
3H); CIMS m/z 156 [M + H]+.
1
22 as a colorless oil: H NMR (300 MHz, CDCl3) δ 4.82-5.04
(br s, 1H), 4.62-4.72 (m, 0.5H), 4.48-4.58 (m, 0.5H), 3.62-3.72
(m, 2H), 3.32-3.62 (m, 2H), 3.20-3.44 (br s, 1H), 1.48 (s, 9H).
tert-Butyl (2R)-2-Fluoro-3-iodopropylcarbamate ((R)-23). Imi-
dazole (26.6 g, 0.39 mol) was dissolved in methylene chloride (400
mL) at room temperature. Iodine (99 g, 0.39 mol) was added, and
the reaction mixture was stirred for 10 min at room temperature
and then cooled to 0 °C. Triphenylphosphine (102 g, 0.39 mol)
was added portionwise over 10 min such that the internal temper-
ature remained below 10 °C. A solution of (R)-22 (60.4 g, 0.31
mol) in methylene chloride (100 mL) was added dropwise. Upon
completion of addition, additional methylene chloride (200 mL)
was added. The reaction mixture was allowed to warm to room
temperature, and stirring was continued for 17 h. The mixture was
filtered through a pad of Celite (50 g) and washed with additional
methylene chloride. The filtrate was concentrated under reduced
pressure and purified by silica gel column chromatography, eluting
with methylene chloride. This procedure afforded 64.7 g (68%) of
(R)-23 as a white solid: 1H NMR (300 MHz, CDCl3) δ 4.80-5.10
(br s, 1H), 4.58-4.72 (m, 0.5H), 4.42-4.56 (m, 0.5 H), 3.48-3.70
(m, 1H), 3.20-3.46 (m, 3H), 1.48 (s, 9H).
In Vivo Studies. The potency of GABAB agonists as inhibitors
of TLESRs was determined using a dog model described by
Lehmann et al.1a Hypothermic effects in mice were assessed as
reported by Que´va et al.17
(2R)-(3-Amino-2-fluoropropyl)phosphinic Acid ((R)-7). Ammo-
nium hypophosphite (73.8 g, 0.89 mol) was added to a three-necked
2 L flask equipped with a mechanical stirrer, thermometer, addition
funnel, and an argon bubbler. The flask was placed in a water bath
at room temperature, and N,O-bis(trimethylsilyl)acetamide (215 mL,
0.87 mol, BSA) was added at such a rate that the internal
temperature was maintained below 38 °C (30 min approximately)
using ice cooling. During the addition ammonia gas was evolved
(during the first 15 min this was rather rapid). Upon completion of
the addition of BSA, the reaction mixture was heated to 45-48 °C
and maintained at this temperature for 1 h (during this time the
reaction became viscous). The mixture was cooled to room
temperature, and a solution of iodide (R)-23 (27.3 g, 0.090 mol) in
methylene chloride (300 mL) was added to the reaction mixture.
The mixture was then allowed to stir at room temperature for 18 h.
The reaction mixture was cooled to 0 °C and was cautiously
quenched with methanol (275 mL) and then with water (32 mL).
The reaction mixture was stirred for 30 min after which the mixture
was filtered and the solids were washed with methanol. The filtrate
was concentrated and the residue placed under high vacuum (0.1
mmHg) overnight. The crude residue was triturated with methylene
chloride, methanol, and concentrated ammonium hydroxide solution
(80:20:1) and was filtered. The filtrate was concentrated under
reduced pressure, and the trituration was repeated. The crude
concentrate was transferred to a 2 L flask, dissolved in methanol
(375 mL), and placed in a water bath at room temperature. A
Supporting Information Available: Experimental details for
synthesis of (R)-3, (S)-7, 8, (S)-8, and the N-benzamide derivatives
of (S)-7 and (R)-7; procedure for determination of the optical purity
of (S)-7 and (R)-7; description of measurement of [Ca2+]i in a
fluorescence imaging plate reader (FLIPR); elemental analysis
results of compounds (S)-3, 5, 7, (S)-7, (R)-7, and (S)-8. This
material is available free of charge via the Internet at http://
pubs.acs.org.
References
(1) (a) Lehmann, A.; Antonsson, M.; Bremner-Danielsen, M.; Fla¨rdh, M.;
Hansson-Bra¨nde´n, L.; Ka¨rrberg, L. Activation of the GABAB receptor
inhibits transient lower esophageal sphincter relaxations in dogs.
Gastroenterology 1999, 117, 1147–1154. (b) Blackshaw, L. A.;
Staunton, E.; Lehmann, A.; Dent, J. Inhibition of transient LES
relaxations and reflux in ferrets by GABA receptor agonists. Am. J.
Physiol.: Gastrointest. LiVer Physiol. 1999, 277, G867-G874. (c)
Lidums, I.; Lehmann, A.; Checklin, H.; Dent, J.; Holloway, R. H.
Control of transient lower esophageal sphincter relaxations and reflux
by the GABAB agonist baclofen in normal subjects. Gastroenterology
2000, 118, 7–13. (d) Zhang, Q.; Lehmann, A.; Rigda, R.; Dent, J.;
Holloway, R. H. Control of transient lower oesophageal sphincter
relaxations and reflux by the GABAB agonist baclofen in patients with
gastro-oesophageal reflux disease. Gut 2002, 50, 19–24.

4320 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 14
Alstermark et al.
(2) Blackshaw, L. A.; Page, A. J.; Smid, S. D.; Dent, J.; Lehmann, A.
GABAB receptors - peripheral and central targets for controlling
gastroesophageal reflux. Curr. Opin. Cent. Peripher. NerV. Syst. InVest.
Drugs 2000, 2, 333–343.
(3) Elebring, T.; Guzzo, P.; Holmen, A.; Olsson, T.; Swanson, M.; von
Unge, S. New Aminopropylphosphinic Acids. PCT Int. Appl. WO
0142252, Priority Date Dec 9, 1999; 48 pp.
(4) Froestl, W.; Mickel, S. J.; Hall, R. G.; von Sprecher, G.; Strub, D.;
Baumann, P. A.; Brugger, F.; Gentsch, C.; Jaekel, J.; Olpe, H.-R.;
Rihs, G.; Vassout, A.; Waldmeier, P. C.; Bittiger, H. Phosphinic acid
analogues of GABA. 1. New potent and selective GABAB Agonists.
J. Med. Chem. 1995, 38, 3297–3312.
(5) Froestl, W.; Mickel, S. J. In The GABA Receptors, 2nd ed.; Enna,
S. J., Bowery, N. J., Eds.; Humana Press: Totowa, NJ, 1997; Chapter
10, p 281.
(6) Amin, K.; Elebring, T.; Guzzo, P.; Olsson, T.; Swanson, M.; von Unge,
S. New (Aminopropyl)methylphosphinic Acids. PCT Int. Appl. WO
0141743, Priority Date Dec 9, 1999; 43 pp.
(7) (a) For the synthesis of 1 and 3, see the following: Dingwall, J. G.;
Ehrenfreund, J.; Hall, R. G. Diethoxymethylphosphonites and phos-
phinates. Intermediates for the synthesis of R,ꢀ- and γ-aminoalky-
lphosphonous acids. Tetrahedron 1989, 45, 3787–3808. (b) For the
synthesis of 2, (S)-4, and (R)-4, see ref 4. (c) For the synthesis of 6,
see the following: Substituted Aminoalkylphosphinic Acids. EP
0399949, Priority Date May 13, 1989.
(8) Baylis, E. K. 1,1-Diethoxyethylphosphinates and phosphonites. In-
termediates for the synthesis of functional phosphorus acids. Tetra-
hedron Lett. 1995, 36, 9385–9388.
(9) Gallagher, M. J.; Honegger, H. Organophosphorus intermediates. VI.
The acid-catalyzed reaction of trialkyl orthoformates with phosphinic
acid. Aust. J. Chem. 1980, 33, 287–294.
(11) (a) Victor, M. W.; Saffariannour, M.; Reynolds, J. R. Structural
characterization of poly(R-fluoroacrylonitrile) and poly(ethyl R-flu-
oroacrylate). J. Macromol. Sci., Pure Appl. Chem. 1994, A31, 721–
736. (b) Tolman, V.; Spronglova, P. Synthesis of 2-fluoropropenoic
acid derivatives. Collect. Czech. Chem. Commun. 1983, 48, 319–
326.
(12) Somekh, L.; Shanzer, A. Facile stereospecific synthesis of R-fluoro-
ꢀ-amino acids. J. Am. Chem. Soc. 1982, 104, 5836–5837.
(13) Boyd, E. A.; Boyd, M. E. K.; Kerrigan, F. Facile synthesis of
phosphorus-containing heterocycles. Tetrahedron Lett. 1996, 37, 5425–
5426.
(14) Hypothermia is an indicator of CNS side effects that can be monitored
in experimental animals.
(15) Lehmann, A. Manuscript in preparation; see ref 17.
(16) (a) Carruthers, N. I.; Spitler, J. M.; Wong, S.-C.; Blythin, D. J.; Chen,
X.; Shue, H.-J.; Hoyan, S. S.; Lee, J. F.; Rizzo, C.; Ting, P. C.; West,
R. E., Jr. Synthesis of a series of sulfinic acid analogs of GABA and
evaluation of their GABAB receptor affinities. Bioorg. Med. Chem.
Lett. 1998, 3059–3064. (b) Fitzpatrick, K.; Geiss, W.; Lehmann, A.;
Sunde´n, G.; von Unge, S. New Compounds Useful in Reflux Disease.
PCT Int. Appl. WO 02100823, Priority Date June 8, 2001; 34 pp. (c)
Lehmann, A.; Aurell-Holmberg, A.; Bhatt, U.; Bremner-Danielsen,
M.; Bra¨nde´n, L.; Elg, S.; Elebring, T.; Fitzpatrick, K.; Geiss, W. B.;
Guzzo, P.; Jensen, J.; Jerndal, G.; Mattsson, J. P.; Nilsson, K.; Olsson,
B.-M. Effects of (2R)-(3-amino-2-fluoropropyl)sulphinic acid (AFP-
SiA) on transient lower esophageal sphincter relaxation in dogs and
mechanism of hypothermic effects in mice. Br. J. Pharmacol. 2005,
146, 89–97.
(17) Que´va, C.; Bremner-Danielsen, M.; Edlund, A.; Ekstrand, A. J.; Elg,
S.; Erickson, S.; Johansson, T.; Lehmann, A.; Mattsson, J. P. Effects
-/-
of GABA agonists on body temperature regulation in GABAB(1)
(10) Petrov, K. A.; Bliznyuk, N. K.; Studnev, Y. N.; Kolomiets, A. F.
Monoalkoxymethylthiophosphonates and monoalkoxymethylphospho-
nites. Zh. Obshch. Khim. 1961, 31, 179–184.
mice. Br. J. Pharmacol. 2003, 140, 315–322.
JM701425K