The discovery of 3-(N-alkyl)aminopropylphosphonic acids as potent S1P receptor agonists

Article abstract of DOI:10.1016/j.bmcl.2004.04.069

3-(N-Alkyl)aminopropylphosphonic acids are potent agonists of four of the five known sphingosine-1-phosphate (S1P) receptor subtypes.

Full text of DOI:10.1016/j.bmcl.2004.04.069

Bioorganic & Medicinal Chemistry Letters 14 (2004) 3495–3499
The discovery of 3-(N-alkyl)aminopropylphosphonic acids as
potent S1P receptor agonists
Jeffrey J. Hale,a,* George Doherty,a,ꢀ Leslie Toth,a Zhen Li,a Sander G. Mills,a
Richard Hajdu,b Carol Ann Keohane,b Mark Rosenbach,b James Milligan,b
Gan-Ju Shei,b Gary Chrebet,b James Bergstrom,b Deborah Card,b Hugh Rosenb,ꢁ
and Suzanne Mandalab
aDepartment of Medicinal Chemistry, Merck Research Laboratories, Rahway, NJ 07065, USA
bDepartment of Immunology and Rheumatology Research, Merck Research Laboratories, Rahway, NJ 07065, USA
Received 12 March 2004; accepted 19 April 2004
Available online
Abstract—3-(N-Alkyl)aminopropylphosphonic acids are potent agonists of four of the five known sphingosine-1-phosphate (S1P)
receptor subtypes.
Ó 2004 Elsevier Ltd. All rights reserved.
Sphingosine-1-phosphate (S1P) has been demonstrated
to be an extracellular signaling factor for a family of
related G-protein coupled receptors, which are now
named for this ligand.1 S1P receptors (of which there are
five known subtypes, S1P1;2;3;4;5) have been implicated to
play roles in a variety of biological processes for mul-
tiple organ systems.2 It has been proposed that the
immunosuppressive actions of FTY720 (1) result from
the formation in vivo of an active phosphate ester
metabolite (2), which is a potent agonist of four of the
five known S1P receptors.3;4
S1P receptor affinity while an a-hydroxy phosphonate
group was demonstrated to be a suitable bioisostere for
the phosphate ester. A compound (3) identified as part
of this work had an S1P receptor profile similar to 2 and
was found to replicate the primary pharmacodynamic
effect (a dose-dependent lowering of circulating lym-
phocytes) and the immunosuppressive efficacy of 1 in
rodents. While this does not prove that the efficacy of 1
is due the actions of the phosphate ester metabolite, the
similarities observed between 2 and 3 suggest that S1P
receptor agonists could find utility in immunosuppres-
sive therapy.
In order to better understand the pharmacology of
FTY720 phosphate (2), we sought compounds that
would remove the complicating factor of the observed
in vivo equilibrium between 1 and 2 and directly target
S1P receptors.5 Nonhydrolyzable phosphonate analogs
of 2 were found to be S1P agonists; deletion of the
hydroxymethyl group from the quaternary center of
these compounds was found to have minimal effect on
R
1, R = -OH (FTY720)
2, R = -OPO3H2
NH2
HO
PO3H2
3
4
NH2 OH
* Corresponding author. Tel.: +1-732-594-2916; fax: +1-732-594-5966;
e-mail: jeffrey_hale@merck.com
ꢀ
Current address: Array BioPharma, 2620 Trade Center Avenue,
Longmont, CO 80503-7551, USA.
PO3H2
(+/-)
ꢁ
Current address: The Scripps Research Institute, ICND-118, 10550
N. Torrey Pines Road, La Jolla, CA 92037, USA.
NH2
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.04.069

3496
J. J. Hale et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3495–3499
Compound 4 was identified as part of our initial inves-
tigations of nonhydrolyzable phosphonate S1P receptor
agonists. While this analog was found to have 3–80-fold
lower affinity for S1P receptors than either 2 or 3, it was
structurally simpler than either compound, lent itself to a
more rapid preparation of analogs and was therefore the
basis for some early range finding regarding which
structural elements in 4 were required for S1P receptor
affinity. These investigations revealed that the lipophilic
tail of these compounds could be transposed from the
carbon bearing the amino group onto the amino group
itself and the resulting 3-(N-alkylamino)propyl-phos-
phonic acids were ligands for S1P receptors. Some details
of the structure–activity relationships for this new class
of S1P receptor agonists is the subject of this report.
good yield. The carboxylate ester of this compound was
converted to a protected amine (7) with a straightfor-
ward sequence of reactions featuring a Curtius rear-
rangement. Global deprotection of 7 was carried out
using iodotrimethylsilane affording target compound 8.
The racemic 3-aminoalkylphosphonic acid analogs 14–
18 were readily prepared by substituting the appropriate
alkyl halide in the first step of this sequence.
The 3-(N-alkylamino)propylphosphonic analogs 19–29
were most conveniently synthesized in 20–40% yield by
treating the appropriate aldehyde and equivalent
amounts of 3-aminopropylphosphonic acid (9) and tetra-
butylammonium hydroxide in methanol with sodium
cyanoborohydride at 50 °C. Neutralization of the reac-
tion mixtures to give the zwitterionic 3-(N-alkyl-
amino)propyl phosphonic acids occurred during
reverse-phase HPLCpurification. Some of the aldehydes
required for the reductive aminations were commercially
available and in other instances they were readily pre-
pared via straightforward functional group manipula-
tions. The preparation of aldehydes 11 and 13 from 10
and 12, respectively, is representative of the synthetic
chemistry that was used to prepare aldehyde interme-
diates (Scheme 1). The phosphinic and carboxylic acid
analogs 30–36 were prepared using chemistry analogous
to that described above substituting the appropriate
amino phosphinic or carboxylic acid for 3-aminopro-
pylphosphonic acid.
The preparation of compound 8 is an example of syn-
thetic chemistry used to prepare the ( )-3-(amino)-
alkylphosphonic acids described herein (Scheme 1).
Treatment of triethyl 4-(phosphono)butyrate (5) with
potassium bis(trimethylsilyl)amide at )78 °Cfollowed
by 1-iodododecane afforded monoalkylated product 6 in
11
a
PO(OCH2CH3)2
CO2CH2CH3
PO(OCH2CH3)2
H3C
CO2CH2CH3
5
6
11
b,c
PO(OCH2CH3)2
d
H3C
NHCO2CH2Ph
Ligand competition studies between [33P]-S1P and 2, 3
and all new compounds were carried out for each of the
7
PO3H2
Table 1. Inhibition (IC50, nM) of [33P]-S1P binding to S1P receptorsa
H3C
NH2
8
PO3H2
(+/-)
R
NH2
S1P2
1100
e
R
R (Compd)
S1P1
S1P3 S1P4
6.3
18
S1P5
0.8
4.5
100
N
H
PO3H2
H2N
PO3H2
CO2H
2
3
4
0.3
15
16
9
19 - 29
0.10
8.4
>10,000
>10,000 280
240
H3C
150
>10,000
>10,000
1200
40
10
14
47
2500
140
76
86
86
CHO
f,g,h
11
8
H3C
Br
H3C
6
25
19
10
11
12
14
H3C
830
450
COCl
13
CHO
i,j
15
H3C
H3C
H3C
85
>10,000
1400
8
8
14
12
13
16
H3C
1600
>10,000 130
>10,000 1500
Scheme 1. Reagents and conditions: (a) 1-iodododecane, KHMDS,
THF, )78 °Cto rt (37%); (b) NaOH, aq MeOH, 50 °C, then
CH3OCOCl, TEA, THF, 0 °Cthen NaN 3, aq THF; (c) benzyl alcohol,
toluene, 85 °C(56%, three steps); (d) iodotrimethylsilane, CH 2Cl2
17
17
15
2100
2
94
110
(95%); (e) R–CHO, Bu4NþOHÀ, Na(CN)BH3, C HOH (20–40%); (f)
3
18
10
iBuOCOCl, NMM, THF, 0 °C, then NaBH4, aq THF, 0 °C(95%); (g)
1-heptene, 9-BBN, THF, rt, then cat. (Ph3P)4Pd, KOH, aq THF, reflux
ꢀ
(92%); (h) cat. TPAP, NMMO, 4 A sieves, CH2Cl2, (63%); (i)
a Displacement of [33P]-labeled sphingosine-1-phosphate (S1P) by test
compounds from human S1P receptors expressed on CHO cell
membranes. Data are reported as mean for n ¼ 3 determinations. SD
were generally 20% of the average. See Ref. 3 for assay protocol.
CH3NHOCH3ÆHCl, toluene/aq NaOH (91%); (j) DIBALH, CH2Cl2,
)78 °C(75%).

J. J. Hale et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3495–3499
Table 2. Inhibition (IC50, nM) of [33P]-S1P binding to S1P receptorsa
3497
H
N
PO3H2
R
R (Compd)
S1P1
30
S1P2
S1P3
13
S1P4
S1P5
19
H3C
>10,000
650
11
19
H3C
2.5
150
580
740
740
2.8
200
140
230
800
38
13
11
12
12
20
H3C
13
2.3
6.6
7.0
3.6
2.7
21
H3C
14
22
H3C
15
15
23
H3C
0.6
6.3
>10,000
>10,000
21
26
24
5.5
8
24
H3C
4.3
45
7
25
H3C
5.3
2200
2000
530
14
840
>10,000
140
42
590
59
6 26
H3C
360
1000
5
27
1.1
0.91
10
28
a Displacement of [33P]-labeled sphingosine-1-phosphate (S1P) by test compounds from human S1P receptors expressed on CHO cell membranes.
Data are reported as mean for n ¼ 3 determinations. SD were generally 20% of the average. See Ref. 3 for assay protocol.
five human S1P receptors stably expressed in Chinese
Hamster Ovary (CHO) cell membranes.3 S1P receptor
agonism by the test compounds was also determined by
measurement of ligand-induced [35S]-50-O-3-thiotri-
phosphate (GTPcS) binding; all of the compounds tes-
ted were found to be agonists of S1P receptors (data not
shown).6 The ( )-3-(amino)alkylphosphonic acid com-
pounds in Table 1 are analogs of 4 containing simple
n-alkyl side chains. The length of the alkyl chain does
not contribute to selectivity for any single S1P receptor
and all of the compounds in this series have minimal
S1P2 affinity. It is apparent that a side chain length of 13
or 14 carbon atoms (compounds 14 and 15) leads to
optimal affinity for the other four S1P receptors,
although none of these is greater than that of 4 for S1P1,
S1P3, S1P4, and S1P5. It is noteworthy that a correlation
between n-alkyl chain length in analogs of 1 and in vivo
immunosuppressive activity has been reported;7 the S1P
receptor data for the compounds in Table 1 suggests
that may be due in part to the monophosphate esters of
those 2-alkyl-2-amino-1,3-propanediols demonstrating
similar trends in S1P receptor affinity.
Transposition of the n-alkyl side chains of the analogs in
Table 1 from the carbon bearing the primary amino
group onto the amino group itself afforded the

3498
J. J. Hale et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3495–3499
3-(N-alkylamino)propyl phosphonic acid analogs in
Table 2. The relationship between n-alkyl chain length
and S1P receptor affinity for 19–23 appears similar to
that observed for the corresponding analogs in Table 1
with a length of 13 or 14 carbon atoms being optimal.
These same analogs (19–23) had increases in receptor
affinity of 5–10-fold for both S1P1 and S1P2 with more
modest increases generally seen for S1P3, S1P4, and S1P5
as compared to the corresponding compounds in Table
1. Insertion of a phenyl ring in the alkyl chain could
bring about further enhancements in potency. Com-
pounds 24–28 all maintain the same approximate alkyl
chain length of compounds 21 or 22; benzyl amine
analog 24 has a subnanomolar S1P1 IC50 and 10-fold or
greater enhanced affinity for S1P3, S1P4, and S1P5 as
compared to 4. Phenyldecyl analog 28 was found to
have high affinity for both S1P1 and S1P3.
for 30–32. Substitution of the c-amino butyric acid
analog 32 with either hydroxy groups (compounds 33
and 34) or fluorine atoms (compounds 35 and 36) had
little effect on S1P receptor affinities.
The ability of S1P receptor agonists to lower the number
of circulating lymphocytes in the mouse after iv
administration can be used as a marker of their immuno-
suppressive efficacy.5 Many of the 3-(N-alkyl-
amino)propyl phosphonic acids here were evaluated for
that ability, but intravenous (iv) dose–titration mea-
surements in Balb/c mice were often complicated by an
acute toxicity, apparently cardiovascular in nature that
was characterized by ataxia, labored breathing, ruffling,
or reduced activity in mild instances and unconscious-
ness, seizures, paralysis, or even death in more severe
cases. For example, 3 mpk iv doses of compounds 20,
22–25, and 27 were all found to be severely toxic. Re-
duced doses of tetradecyl analog 21 or 4-(nonyl)benzyl
analog 29 were also acutely toxic, but administration of
these compounds via the peritoneal cavity was found to
enhance tolerability presumably by blunting maximum
compound concentration. A 0.25 mpk ip dose of 29 gave
the maximal lymphocyte lowering response at 3 h after
In anticipation of the probable poor oral adsorption of
compounds 19–28, a series of zwitterionic compounds
bearing a 4-(nonyl)benzyl side chain was prepared
(Table 3). Phosphinic and carboxylic acids were found
to be poor replacements for the phosphonate based on
the significant drop in S1P receptor affinities observed
Table 3. Inhibition (IC50, nM) of [33P]-S1P binding to S1P receptorsa
H
N
R
R (Compd)
S1P1
0.2
S1P2
S1P3
S1P4
S1P5
PO3H
PO2H
750
2.7
40
0.7
29
30
12
2
>10,000
>10,000
>10,000
>10,000
1000
170
900
590
140
60
PO(OH)CH3
31
32
CO2H
CO2H
7.7
22
1200
370
>10,000
1200
200
99
OH
33
OH
CO2H
CO2H
14
44
41
>10,000
>10,000
>1000
170
390
350
1400
4000
600
66
200
57
34
F
F
35
F
CO2H
36
a Displacement of [33P]-labeled sphingosine-1-phosphate (S1P) by test compounds from human S1P receptors expressed on CHO cell membranes.
Data are reported as mean for n ¼ 3 determinations. SD were generally 20% of the average. See Ref. 3 for assay protocol.

J. J. Hale et al. / Bioorg. Med. Chem. Lett. 14 (2004) 3495–3499
3499
compound challenge. Aside from 2 and 3,8 none of the
other compounds in Tables 1–3 were able to induce the
maximal lymphopenic response at this dose level.
4. Brinkmann, V.; Davis, M. D.; Heise, C. E.; Albert, R.;
Cottens, S.; Hof, R.; Bruns, C.; Prieschl, E.; Baumruker,
T.; Hiestand, P.; Foster, C. A.; Zollinger, M.; Lynch, K.
R. J. Biol. Chem. 2002, 277, 21453–21457.
5. Hale, J. J.; Neway, W.; Mills, S. G.; Hajdu, R.; Keohane,
C. A.; Rosenbach, M.; Milligan, J.; Shei, G.-J.; Chrebet,
G.; Bergstrom, J.; Card, D.; Koo, G. C.; Koprak, S. L.;
Jackson, J. J.; Rosen, H.; Mandala, S. Bioorg. Med. Chem.
Lett. 2004, 14, 3351–3355.
6. All test compounds were found to be agonists of human
S1P1, S1P2, S1P3, S1P4, and S1P5 receptors as evidenced
by their ability to induce levels of GTPcS binding
comparable to S1P. The magnitudes of the calculated
EC50 values from these assays were generally +/)5-fold of
the IC50 values.
7. Fujita, T.; Yoneta, M.; Hirose, R.; Sasaki, S.; Inoue, K.;
Kiuchi, M.; Hirase, S.; Adachi, K.; Arita, M.; Chiba, K.
Bioorg. Med. Chem. Lett. 1995, 5, 847–852.
8. Signs of mild toxicity were observed in mice for 15–30 min
after doses of 0.5 mpk iv of 2 or 0.25 mpk iv of 3.
Escalating the dose of either compound resulted in more
severe toxicity, but not lethality. Metabolic factors or
altered binding to plasma proteins of 2 and 3 as compared
to the 3-(N-alkylamino)propyl phosphonic acid analogs
may account for the observed differences.
In conclusion, 3-(N-alkylamino)propyl phosphonic
acids have been discovered to be a novel class of S1P
receptor agonists. The fact that these new S1P agonists
and those from other structural series (e.g., 2 and 3) all
can induce a lowering of circulating lymphocytes indi-
cates that ligands for S1P receptors indeed have the
potential to be novel immunomodulators.9 The acute
toxicity seen in mice with S1P agonists may be an ex-
treme manifestation of the bradycardia that has been
reported in the clinic with 1.10 The varied distributions
of S1P receptor subtypes11 suggests that the separation
of efficacy and acute toxicity may be possible with
receptor-selective ligands; reports of the exploitation
of these structurally simple 3-(N-alkylamino)propyl
phosphonic acids to this end will follow shortly.
References and notes
9. Chiba, K.; Yanagawa, Y.; Masubuchi, Y.; Kataoka, H.;
Kawaguchi, T.; Ohtsuki, M.; Hoshino, Y. J. Immunol.
1998, 160, 5037–5044.
1. Kluk, M. J.; Hla, T. Biochim. Biophys. Acta 2002, 1582,
72–80.
2. Hla, T. Pharm. Res. 2003, 47, 401–407.
3. Mandala, S.; Hajdu, R.; Bergstrom, J.; Quackenbush, E.;
Xie, J.; Milligan, J.; Thornton, R.; Shei, G.-J.; Card, D.;
Keohane, C.; Rosenbach, M.; Hale, J.; Lynch, C. L.;
Rupprecht, K.; Parsons, W.; Rosen, H. Science 2002, 296,
346–349.
10. Budde, K.; Schmouder, R. L.; Brunkhorst, R.; Nashan,
B.; Lucker, P. W.; Mayer, T.; Choudhury, S.; Skerjanec,
A.; Kraus, G.; Neumayer, H. H. J. Am. Soc. Nephrol.
2002, 13, 1073–1083.
11. Fukushima, N.; Ishii, I.; Contos, J. J. A. Ann. Rev.
Pharmacol. Toxicol. 2001, 41, 507–534.