Autotaxin structure-activity relationships revealed through lysophosphatidylcholine analogs

Article abstract of DOI:10.1016/j.bmc.2009.03.030

Autotaxin (ATX) catalyzes the hydrolysis of lysophosphatidylcholine (LPC) to form the bioactive lipid lysophosphatidic acid (LPA). LPA stimulates cell proliferation, cell survival, and cell migration and is involved in obesity, rheumatoid arthritis, neuropathic pain, atherosclerosis and various cancers, suggesting that ATX inhibitors have broad therapeutic potential. Product feedback inhibition of ATX by LPA has stimulated structure-activity studies focused on LPA analogs. However, LPA displays mixed mode inhibition, indicating that it can bind to both the enzyme and the enzyme-substrate complex. This suggests that LPA may not interact solely with the catalytic site. In this report we have prepared LPC analogs to help map out substrate structure-activity relationships. The structural variances include length and unsaturation of the fatty tail, choline and polar linker presence, acyl versus ether linkage of the hydrocarbon chain, and methylene and nitrogen replacement of the choline oxygen. All LPC analogs were assayed in competition with the synthetic substrate, FS-3, to show the preference ATX has for each alteration. Choline presence and methylene replacement of the choline oxygen were detrimental to ATX recognition. These findings provide insights into the structure of the enzyme in the vicinity of the catalytic site as well as suggesting that ATX produces rate enhancement, at least in part, by substrate destabilization.

Full text of DOI:10.1016/j.bmc.2009.03.030

Bioorganic & Medicinal Chemistry 17 (2009) 3433–3442
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry
journal homepage: www.elsevier.com/locate/bmc
Autotaxin structure–activity relationships revealed through
lysophosphatidylcholine analogs
*
*
E. Jeffrey North, Daniel A. Osborne , Peter K. Bridson, Daniel L. Baker , Abby L. Parrill
Department of Chemistry and Computational Research on Materials Institute, The University of Memphis, Memphis, TN 38152, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Autotaxin (ATX) catalyzes the hydrolysis of lysophosphatidylcholine (LPC) to form the bioactive lipid
lysophosphatidic acid (LPA). LPA stimulates cell proliferation, cell survival, and cell migration and is
involved in obesity, rheumatoid arthritis, neuropathic pain, atherosclerosis and various cancers, suggest-
ing that ATX inhibitors have broad therapeutic potential. Product feedback inhibition of ATX by LPA has
stimulated structure–activity studies focused on LPA analogs. However, LPA displays mixed mode inhibi-
tion, indicating that it can bind to both the enzyme and the enzyme–substrate complex. This suggests
that LPA may not interact solely with the catalytic site. In this report we have prepared LPC analogs to
help map out substrate structure–activity relationships. The structural variances include length and
unsaturation of the fatty tail, choline and polar linker presence, acyl versus ether linkage of the hydrocar-
bon chain, and methylene and nitrogen replacement of the choline oxygen. All LPC analogs were assayed
in competition with the synthetic substrate, FS-3, to show the preference ATX has for each alteration.
Choline presence and methylene replacement of the choline oxygen were detrimental to ATX recognition.
These findings provide insights into the structure of the enzyme in the vicinity of the catalytic site as well
as suggesting that ATX produces rate enhancement, at least in part, by substrate destabilization.
Ó 2009 Elsevier Ltd. All rights reserved.
Received 29 December 2008
Revised 10 March 2009
Accepted 14 March 2009
Available online 21 March 2009
Keywords:
Autotaxin
Lysophosphatidylcholine
Lysophosphatidic acid
LPC
LPA
SAR
1. Introduction
homology models17,18 constructed using alkaline phosphatase19
and later a bacterial nucleotide pyrophosphatase/phosphodiester-
Autotaxin (ATX) has recently become an attractive target for
therapeutic development efforts. ATX is a 125 kDa glycoprotein
originally isolated from the human melanoma cell line A20581
and is upregulated in many tumor cell lines.2 ATX, a lysophospho-
lipase D enzyme, hydrolyzes lysophosphatidylcholine (LPC) to
form the bioactive lipid lysophosphatidic acid (LPA).3,4 ATX elicits
its biological activity through its product LPA.3,4 LPA induces many
ase (NPP) homolog.20 Additional indirect structural insights can
be obtained from substrates, substrate analogs, and inhibitors.
Until Parrill et al., Moulharat et al., and Saunders et al. recently
described non-lipid ATX inhibitors,21–23 LPA analogs and metal
chelators24 were the only known ATX inhibitors. LPA analogs
showing ATX inhibition include sphingosine 1-phosphate,25 phos-
phonates,26–28 thiophosphates,27 cyclic glycerothiophosphates,29
carba cyclic phosphatidic acids,30 alpha-substituted phosphonic
acids,31 and FTY720-phosphate.32 Since all published phospho-
lipid ATX inhibitors are analogs of LPA, which does not exhibit
simple competitive inhibition but rather shows mixed mode inhi-
bition,25 they may not interact solely with the active site. LPC
analogs, (i.e., analogs of the natural substrate) can provide struc-
ture–activity relationships that more directly reflect the ATX ac-
tive site.
biological events by activating specific G protein-coupled recep-
5–10
tors, LPA1–8
and a nuclear hormone receptor, PPARc 10,11
. The
effects of LPA include stimulation of cell proliferation, cell migra-
tion, and cell survival.2 LPA-induced cell motility is mediated
through the LPA1 receptor.12 These are detrimental cellular re-
sponses as it pertains to cancer cell biology. LPA is also implicated
in obesity,13 rheumatoid arthritis,14 neuropathic pain,15 and ath-
erosclerosis,16 a precursor to cardiovascular disease.
The ATX protein has yet to be crystallized; therefore little is
directly known about the three-dimensional structure of the en-
zyme. In contrast, indirect insights have been obtained from
In this study, multiple LPC analogs have been examined to help
define ATX substrate recognition. The structural features of LPC
which were varied are length and unsaturation of the fatty tail,
choline and polar linker presence, acyl versus ether linkage of the
hydrocarbon chain, and methylene and nitrogen replacement of
the choline oxygen (Fig. 1). All compounds were tested for their
ability to inhibit hydrolysis of a fluorescent synthetic ATX sub-
strate, FS-3,33 to show the preference ATX has for each
modification.
* Corresponding authors. Tel.: +1 901 678 2638/4178; fax: +1 901 678 3447.
E-mail addresses: dlbaker@memphis.edu (D.L. Baker), aparrill@memphis.edu
(A.L. Parrill).
Present address: Department of Physiology, The University of Tennessee Health
Science Center, Memphis, TN 38163, USA.
0968-0896/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2009.03.030

3434
E. J. North et al. / Bioorg. Med. Chem. 17 (2009) 3433–3442
Figure 1. LPC, LPA, and LPC analogs examined.
and LPA inhibition of ATX were linear over the time frame examined.
2. Results
Thus, LPC activities are due to LPC and not due to metabolism to LPA.
The6:0, 8:0, and10:0LPCcompoundsfailedto inhibitATX-catalyzed
2.1. Determination of acyl chain length and unsaturation
effects on ATX recognition
hydrolysis of FS-3 (Table 1). Beginning with LPC 12:0 (10
pression of ATX-mediated FS-3 hydrolysis (Table 1) was noted. LPC
12:0, 14:0, 16:0, 18:0, and 18:1 (10 M) inhibited FS-3 hydrolysis
lM), sup-
l
Initially, multiple LPC compounds were assayed to investigate
ATX preferences for hydrophobic tail length and degree of unsatura-
tion. The assayed LPC compounds had 6:0, 8:0, 10:0, 12:0, 14:0, 16:0,
18:0, and 18:1 acyl chains. All compounds were tested at 0.1, 1, and
by 34 4.3%, 64 3.9%, 64 16.9%, 41 17.8%, and 83 8.4%,
respectively (Table 1). There is no statistical difference in ATX inhi-
bition by LPC 12:0 and 18:0 (p = 0.7154) or by LPC 16:0 and 18:1
(p = 0.0907). LPC 18:1 inhibited ATX-mediated FS-3 hydrolysis the
10
lected from MD-MBA 435 breast cancer cell cultures as the source
of ATX, and 1 M FS-3 as the ATX substrate. Since LPC is converted
lM concentrations using concentrated conditioned media col-
greatest at both 10 and 1 lM concentrations, indicating a preference
by ATX for unsaturation in the hydrophobic tail. With respect to the
saturated chains, both LPC 14:0 and 16:0 inhibited ATX-mediated
hydrolysis of FS-3 similarly at 64 3.9% and 64 16.9%, respectively,
whereas LPC 12:0 and 18:0 were less efficacious (Table 1). The opti-
mal unbranched chain lengths for LPC as ATX inhibitors are 14:0,
16:0, and 18:1.
l
to LPA, a potent ATX inhibitor, the time courses of ATX activity were
carefully examined to verify that no changes in rate occurred prior to
the time point used to calculate inhibition. A decrease in rate during
the course of the assay would suggest that LPA was generated in suf-
ficient amounts to mask LPC effects. Figure 2 shows that both LPC
Figure 2. LPC and LPA time course. FS-3 inhibition remained linear during the course of the assay, indicating that LPA accumulation due to LPC hydrolysis was low and did not
affect ATX activity.

E. J. North et al. / Bioorg. Med. Chem. 17 (2009) 3433–3442
3435
Table 1
ATX percent activity
three cases, LPA inhibited ATX-mediated FS-3 hydrolysis, as well
as, or better than the corresponding LPC, indicating that the choline
functional group is detrimental to ATX recognition (Table 1).
To examine the effect of an ester versus an ether linkage to the
hydrocarbon chain we tested two commercially available lysoPAF
compounds, 16:0 and 18:0 (Fig. 1). There was no statistical differ-
ence in ATX inhibition for 16:0 lysoPAF and LPC (p = 0.1150) and
18:0 lysoPAF and LPC (p = 0.1058) (Table 1). This indicates that
the carbonyl group is not required for ATX recognition.
Percent activity SD
0.1
l
M
1
lM
10 lM
LPC
6:0 Acyl
8:0 Acyl
10:0 Acyl
12:0 Acyl
14:0 Acyl
16:0 Acyl
18:0 Acyl
18:1 Acyl
104 4.2
94 5.3
101 4.1
102 3.8
72 4.8
109 17.2
111 17.0
92 8.4
100 4.8
81 5.6
101 3.7
97 4.5
80 4.5
89 16.8
99 17.2
55 7.9
97 4.3
101 4.1
85 5.4
66 4.3
36 3.9
36 16.9
59 17.8
17 8.4
2.3. ATX recognition of the polar linker
Alkylphosphocholines were synthesized to show the impor-
tance of the substrate polar linker to ATX recognition (Fig. 1).
Scheme 1 shows the synthesis for these alkylphosphocholines.
Phosphorus oxychloride was reacted first with a long chain alcohol
(tetradecanol, hexadecanol, octadecanol, or oleyl alcohol), second
with 2-bromoethanol, and third with water all in the presence of
triethylamine. Finally, intermediates 1a–d were reacted with tri-
methylamine to produce final products 2a–d.
To compare an LPC, which is a glycerophospholipid, to an
alkylphosphocholine, the alkylphosphocholine must contain four
extra carbons in the hydrocarbon chain to compensate for the four
additional heavy atoms within the polar linker of LPC. For example,
LPC 14:0 has the same number of atoms extending from the phos-
phate as alkylphosphocholine 18:0. When assayed in the presence
LPA
16:0 Acyl
18:0 Acyl
18:1 Acyl
74 16.9
89 15.6
90 17.1
49 8.6
74 8.1
23 17.4
5
8.7
36 8.6
À2 16.9
LysoPAF
16:0 Alkyl
18:0 Alkyl
100 5.3
92 18.7
89 4.8
86 17.2
49 4.2
44 18.1
Alkylphosphocholine
2a (14:0 Alkyl)
2b (16:0 Alkyl)
2c (18:0 Alkyl)
2d (18:1 Alkyl)
110 11.2
114 10.1
110 10.0
105 11.1
112 10.2
113 10.9
96 9.7
91 10.4
92 13.2
44 10.0
59 10.5
8
9.9
Alkylglycero-phosphonocholine
6a (R 10:0 Alkyl)
6b (R 12:0 Alkyl)
6c (R 14:0 Alkyl)
6d (R 16:0 Alkyl)
6e (R 18:0 Alkyl)
6f (S 10:0 Alkyl)
6g (S 12:0 Alkyl)
6h (S 14:0 Alkyl)
6i (S 16:0 Alkyl)
6j (S 18:0 Alkyl)
101 4.7
95 9.8
96 10.0
85 9.4
82 4.1
89 5.5
96 6.1
95 6.6
95 5.5
86 4.7
77 6.7
103 4.2
89 6.6
76 6.0
72 4.7
76 4.2
89 6.6
87 6.2
80 6.2
76 4.7
68 6.6
99 10.5
98 6.7
83 4.5
81 5.1
97 6.4
98 6.4
97 5.7
82 6.1
78 5.6
of FS-3, the 10
lM concentrations of saturated alkylphosphocho-
lines showed similar ATX inhibition compared to the correspond-
ing LPC. LPC 10:0, 12:0, and 14:0 inhibited ATX-mediated
hydrolysis of FS-3 by 15 5.4%, 34 4.3%, and 64 3.9%, respec-
tively (Table 1). Compounds 2a, 2b, and 2c, with 14:0, 16:0, and
18:0 alkyl chains, inhibited ATX-mediated hydrolysis by
9
10.4%, 8 13.2%, and 56 10.0%, respectively (Table 1). Only
Alkylphosphonocholine
8a (14:0 Alkyl)
8b (16:0 Alkyl)
8c (18:0 Alkyl)
LPC 12:0 showed significantly greater ATX inhibition by an extra
26% compared to alkylphosphocholine 16:0 (p = 0.0097) (Table
1). The polar linker is therefore not detrimental to ATX recognition,
but also it is not universally beneficial.
94 17.7
97 18.0
114 16.9
101 18.0
90 14.3
94 17.6
116 20.7
100 16.3
71 14.9
91 22.3
102 24.4
79 16.3
8d (18:1 Alkyl)
Phosphonamidate
11a (14:0 Alkyl)
11b (16:0 Alkyl)
11c (18:0 Alkyl)
11d (18:1 Alkyl)
Alkylphosphocholine 18:1 (2d) inhibited ATX activity by
89 8.8
81 12.9
86 8.9
93 9.4
90 9.1
81 9.9
74 7.9
89 11.2
75 8.7
77 9.6
29 7.2
74 9.7
92 9.9% at 10
ence for unsaturation in a hydrophobic tail regardless of overall
chain length. LPC 18:1 inhibits ATX by 83 8.4% at 10 M (Table
lM (Table 1). This confirms that ATX has a prefer-
l
1). Comparison of LPC 18:1 to alkylphosphocholine 18:1 suggests
that the position of unsaturation is not critical.
2.2. ATX dependence on choline and ester carbonyl functional
groups
2.4. The effect of methylene replacement for the choline
oxygen on ATX recognition
After establishing the optimal LPC chain lengths, the effect of
the choline group was examined (Fig. 1). To do this, commercially
available LPA compounds with the same chain lengths as commer-
The necessity of the choline oxygen in the presence and absence
of the polar linker was examined via the synthesis and evaluation
of phosphonocholines and alkylglycerophosphonocholines, respec-
tively (Fig. 1). Such compounds would act as nonhydrolyzable ATX
inhibitors if the choline oxygen proved unnecessary. We have
shown that LPC 12:0 inhibits ATX-mediated FS-3 hydrolysis more
cially available LPC were tested. LPA 16:0, 18:0, and 18:1 (10 lM)
inhibited ATX by 95 8.7%, 64 8.6%, and 102 16.9%, respec-
tively, compared with inhibition by LPC 16:0, 18:0, and 18:1 of
64 16.9%, 41 17.8% and 83 8.4%, respectively (Table 1). In all
Scheme 1. Alkylphosphocholine synthesis.

3436
E. J. North et al. / Bioorg. Med. Chem. 17 (2009) 3433–3442
Scheme 2. Alkylglycerophosphonocholine and alkylphosphonocholine synthesis.
effectively than alkylphosphocholine 16:0 suggesting that the po-
lar linker can be beneficial, therefore compounds with and without
the polar linker were evaluated.
strate leaving group oxygen atom and the chiral center
orientation. LysoPAF and LPC naturally occur as the R isomer. The
16:0 chain length enantiomers, 6d and 6i, inhibited ATX by
Intermediate alkylglycerols were stereospecifically synthesized,
according to Erukulla and colleagues, by the treatment of a long
chain alcohol with DIBAL-H and subsequent addition of glycidol.34
To synthesize the alkylphosphonocholines and alkylglycerophos-
phonocholines, 3-chloropropylphosphonic acid was synthesized
using the procedure of Kley and co-workers.35 Attachment of a
long chain alcohol or alkylglycerol a to 3-chloropropylphosphonic
acid (4a) was accomplished using DCC/DMAP (Scheme 2).35 Finally,
treatment of coupled products (5a–j and 7a–d) with trimethyla-
mine produced final products (6a–j and 8a–d).35 The alkylglycero-
phosphonocholines were isolated as a mixture of sn2 and sn3
coupled isomers.
The alkylphosphonocholines displayed little to no ATX inhibi-
tion (Table 1). Alkylphosphonocholines 8a and 8d inhibited ATX
by 29 14.9% and 21 16.3%, respectively (Table 1). The corre-
sponding alkylphosphocholines (2a and 2d) inhibited ATX-medi-
ated FS-3 hydrolysis by 9 10.4% and 92 9.9%, respectively
(Table 1). The apparently better ATX inhibition of 8a compared
to 2a was not statistically significant (p = 0.0778). The 18:0 and
18:1 alkylphosphonocholine/alkylphosphocholine (8c/2c and 8d/
2d) pairs did show a statistically significant difference in inhibition
(p = 0.0003 for 8d and 2d, p = 0.0044 for 8c and 2c). Alkylphospho-
nocholines (8b and 8c) failed to inhibit ATX. These alkylphospho-
nocholines (8a–d), when compared to the alkylphosphocholines
(2a–d), show that regardless of chain length and degree of unsatu-
ration, methylene substitution for the choline oxygen is detrimen-
tal to ATX recognition.
28 4.7% and 24 4.7%, respectively at 10
chain length enantiomers, 6e and 6j, inhibited ATX by 24 4.2%
and 32 6.6%, respectively at 10 M (Table 1). None of the enantio-
meric pairs showed statistically different inhibition (p = 0.2964 for
the 16:0 pair and p = 0.0868 for the 18:0 pair), suggesting that ATX
is not stereoselective in its recognition of glycerol-based phospho-
lipids. However these compounds failed to show statistically sig-
lM (Table 1). The 18:0
l
nificant
dose-dependence,
so
future
stereoselectivity
investigations with LPC or lysoPAF are warranted. LysoPAF 16:0
and 18:0 inhibited ATX by 51 4.2% and 56 16.8%, respectively,
at 10 lM. Compared to the naturally occurring lysoPAF com-
pounds, the alkylglycerolphosphonates were not as effective (Table
1), indicating that the substrate leaving group oxygen plays a role
in recognition. The alkylglycerophosphonates were isolated as a
mixture of two constitutional isomers. After testing the mixture
we decided not to pursue further purification due to their failure
to substantially inhibit ATX-mediated FS-3 hydrolysis. These re-
sults confirm that ATX has a preference for the choline oxygen over
the methylene group. Inclusion of a glycerol backbone, regardless
of stereochemistry, did not improve ATX recognition for the phos-
phonate head group.
2.5. The effect of nitrogen substitution for the choline oxygen
on ATX recognition
A phosphoramidate is electrostatically more similar to a phos-
phate thanis a phosphonate (Fig. 1). Scheme3 shows a modifiedsyn-
thesis by Garrido-Hernandez and co-workers for the phosphona-
midate series.36 Phosphorus oxychloride, in the presence of triethyl-
amine, was successively reacted with a long chain alcohol (tetradec-
anol, hexadecanol, octadecanol, and oleyl alcohol), phenol, and the
hydrogen bromide salt of 2-bromoethylamine. Finally, treatment
of 10a–d with trimethylamine35 afforded final products 11a–d.
Evaluation of the alkyl phosphonamide series gave varying re-
sults. Compounds 11a, 11b, 11c, and 11d inhibited ATX by
The alkylglycerophosphonocholines were also compared to the
alkylphosphonocholines (Fig. 1). None of the compounds, 6a–c/f–
h, significantly inhibited ATX-mediated FS-3 hydrolysis (Table 1).
By comparing compounds 8a–c to 6a–f, it is evident that the glyc-
erol backbone is not beneficial to a phosphonate head group for
ATX recognition.
When compared to lysoPAF, the enantiomeric alkylglycero-
phosphonocholine mixtures indicate the importance of the sub-
Scheme 3. Alkylphosphoramidate synthesis.

E. J. North et al. / Bioorg. Med. Chem. 17 (2009) 3433–3442
3437
25 8.7%, 23 9.6%, 71 7.2%, and 26 9.7%, respectively (Table
1). For the alkylphosphoramidate series, only the 18:0 chain length
at 10 lM was effective at inhibiting ATX. Comparing the 18:0 and
and colleagues for phosphonomethyl phenylalanine as a protein–
tyrosine phosphatase inhibitor.37 When the phosphonate methy-
lene was converted to a difluoromethylene group, the inhibitory ef-
fect was greatly enhanced.37 The fluorine atoms likely mimic the
partial negative charge of the phosphate oxygen and therefore more
tightly interact with the active site. Choline is the leaving group
when ATX hydrolyzes LPC and therefore leaves with a negative
charge. A difluoromethylene group would more closely imitate the
negative electrostatic potential of an oxygen atom, and might there-
fore be recognized more readily by ATX.
Various LPA phosphonate analogs have been tested by multiple
groups as ATX inhibitors. These phosphonates have the methylene
group on the opposite side of the phosphorus atom than do our LPC
phosphonate analogs. Our best LPC phosphonate analogs were (S)-
alkylglycerophosphonocholine 18:0 (6j), alkylphosphonocholine
14:0 (8a), and (R)-alkylglycerophosphonocholine 16:0 (6d) inhibit-
18:1 chain lengths for both series, the alkyl phosphonamide 18:0
(11c) and 18:1 (11d) inhibited ATX by 71 7.2% and 26 9.7%,
respectively, and alkylphosphocholine 18:0 (2c) and 18:1 (2d) by
56 10.0% and 92 9.9%, respectively (Table 1). Alkyl phosphona-
mide 18:0 (11c) inhibited ATX more effectively than the corre-
sponding alkylphosphocholine 18:0 (2c) by an extra 15%
(p = 0.0476). On the other hand, alkylphosphocholine 18:1 (2d)
inhibited ATX more effectively than the corresponding alkyl phos-
phonamide 18:1 (11d) by an additional 66% (p = 0.0001). This
shows that nitrogen recognition is optimal with long saturated
hydrocarbon chains among the alkyl phosphonamides.
3. Discussion
ing ATX at 10
tively. Cui and colleagues identified two LPA phosphonate
analogs that inhibit ATX by 87% and 79% at 10
M.26 Durgam
and colleagues also identified an LPA phosphonate analog that
inhibited ATX by approximately 72% at 10
M.27 These results sug-
gest that methylene replacement of the choline oxygen is detri-
mental to ATX inhibition, but methylene replacement on the
opposite side of the phosphodiester bond is better tolerated.
lM by 32 6.6%, 29 14.9%, and 28 4.7%, respec-
Multiple LPC compounds were tested to evaluate the effect of
chain length and monounsaturation on ATX recognition. Tokumura
and colleagues report that LPC 12:0 and 14:0 are optimal chains
lengths for LPC as an ATX substrate.4 We report optimal chain
lengths of 14:0 and 16:0 for the saturated chains when LPC is eval-
uated as an inhibitor, but ultimately the 18:1 chain length was
optimal. The difference we see here is due to the bioassay utilized.
When LPC is assayed in competition with FS-3, as in the present
study, LPC 14:0 and 16:0 most effectively reduce FS-3 hydrolysis.
When formation of choline directly from added LPC is monitored,
as in the work reported by the Tokumura group, LPC 12:0 and
14:0 were optimal. This suggests that LPC 16:0 binds to ATX more
effectively than LPC 12:0, but is not as well positioned for enzy-
matic hydrolysis. Furthermore, the catalytic product of LPC hydro-
lysis is LPA, a known inhibitor of ATX. We considered that LPA may
have masked direct LPC effects by inhibiting ATX, since its concen-
tration would increase during the assay. However, the accumula-
tion of FS-3 hydrolytic products remained linear at the 1 h time
point utilized to compute inhibition (Fig. 2), indicating that the
concentration of LPA did not amass sufficiently to affect the re-
ported results.
There are two ways to lower the activation energy in an en-
zyme-catalyzed reaction. The activation energy barrier can be re-
duced either by transition-state stabilization or by substrate
destabilization. LPA lacks the choline group of the substrate ana-
logs which were examined and inhibited ATX most effectively,
indicating that the choline group is detrimental to ATX recognition
and may serve to destabilize the substrate complex in order to re-
duce the activation energy barrier of the enzymatic phosphate
hydrolysis reaction.
Oleyl phosphocholine (2d) showed the most ATX inhibitionof the
choline-containing analogs. This molecule has a phosphodiester
linkage and may be hydrolyzed by ATX. The resulting product, oleyl
phosphate, may be inhibiting ATX. Durgam and colleagues reported
the synthesis of a series of fatty alcohol phosphates, their effect on
the LPA receptors, and on ATX activity for selected compounds.27
Oleyl phosphate had been reported to have no effect on the LPA1–3
receptors, although it was not evaluated as an ATX inhibitor.27 The
related oleyl thiophosphate, however, was an effective ATX inhibi-
tor.27 The catalytic product of oleyl phosphocholine (oleyl phos-
phate), might inhibit ATX similarly to oleyl thiophosphate. As was
justified for LPA production from LPC above, oleyl phosphate did
not amass sufficiently to affect the reported results.
l
l
4. Conclusions
Multiple LPC analogs were evaluated to determine ATX substrate
structure–activity relationships. The alkylphosphocholines, alkyl
phosphonamides, alkylglycerophosphonates and alkyl phospho-
nates were synthetic LPC analogs. Choline has been proven detri-
mental to ATX recognition (Fig. 3). Methylene replacement of the
choline oxygen proved ineffective for ATX inhibition, with or with-
out a polar linker, which is not required for ATX recognition/inhibi-
tion (Fig. 3). These results indicate that substrate recognition
involves direct interaction between the leaving group oxygen atom
and the enzyme. The stereochemistry of the glycerol backbone has
been shown to be irrelevant to ATX recognition (Fig. 3). Nitrogen
substitution for the choline oxygen proved to be context dependent,
displaying improved ATX recognition coupled with some chain
lengths and impaired ATX recognition in other contexts (Fig. 3).
5. Experimental
5.1. Materials and Instrumentation
All reactions were carried out in clean glassware that was dried
in a 175 °C oven for at least 8 h. All solvents and chemicals were
purchased from either Aldrich (St. Louis, Missouri) or Fisher Scien-
Phosphonate isosteric replacement for phosphate has long been
usedto probeinhibitoryeffects forphosphatases. Therefore, we used
phosphonate LPC analogs to test their inhibitory effect on the phos-
phodiesterase activity of ATX. All LPC phosphonate analogs tested
possessed modest activity. Similar results have been shown by Chen
Figure 3. Functional group relevance to ATX recognition.

3438
E. J. North et al. / Bioorg. Med. Chem. 17 (2009) 3433–3442
tific (Pittsburgh, Pennsylvania) except for the lysophosphatidylch-
olines, lysophosphatidic acids, and lyso platelet activating factors,
which were purchased from Avanti Polar Lipids, Inc. (Alabaster,
Alabama). All long chain alcohols, 3-chloropropylphosphonic acid,
and DMAP were dried using magnesium sulfate or over phospho-
rus pentoxide in a vacuum desiccator for at least 8 h prior to use.
Mass spectra were acquired with a Thermoelectron LTQ XL LC–
MS equipped with an electrospray ionization (ESI) source and a lin-
ear ion trap mass analyzer. All 1H NMR spectra were obtained on a
JEOL 270 MHz spectrometer and all 13C NMR and 31P NMR were
obtained on a Varian 500 MHz spectrometer. All NMR spectra were
reported in parts per million (ppm) relative to solvent peak. Ele-
mental analysis was performed by Columbia Analytical Services,
Inc. (Tucson, AZ). Assay data were obtained using a BioTek Syn-
ergy-2 plate reader. TLC plates and silica gel columns (24 g and
40 g SupraFlash Cartridges) were purchased from Sorbent Technol-
ogies (Atlanta, Georgia). The TLC plates were sprayed with 20% sul-
furic acid and heated at 175 °C to visualize spots. The removal of
solvents was performed by evaporation under reduced pressure.
The student’s t test was used for all statistical analyses that
compare two sets of data, where p < 0.05 was considered signifi-
cant. GraphPad software was used to calculate statistical
significance.
58.58; H, 10.77; N, 3.25; P, 7.19. Found: C, 59.09; H, 11.36; N,
3.37; P, 7.50.
5.2.4. Octadecylphosphocholine (2c)
Yield: 21 mg of white solid (0.74%). Rf = 0.19 (chloroform/meth-
anol/ammonium hydroxide 60:40:7). 1H NMR (chloroform-d/
methanol-d4 3:2): d = 0.65 (t, 3H), 1.03 (br, 30H), 1.39 (m, 2H),
2.98 (s, 9H), 3.38 (m, 2H), 3.61 (dt, 2H), 3.99 (m, 2H). 31P NMR:
d = 4.93. MS (ESI): m/z = 436.43 [M]+. Elemental Anal. Calcd: C,
60.24; H, 10.99; N, 3.05; P, 6.75. Found: C, 59.82; H, 11.08; N,
3.52; P, 7.10.
5.2.5. Oleylphosphocholine (2d)
Yield: 47 mg of off-white waxy solid (1.7%). Rf = 0.31 (chloro-
form/methanol/ammonium hydroxide 60:40:7). 1H NMR (chloro-
form-d/methanol-d4 3:2): d = 0.64 (t, 3H), 1.03 (br, 22H), 1.39 (m,
2H), 1.72 (m, 4H), 2.98 (s, 9H), 3.37 (m, 2H), 3.61 (dt, 2H), 3.99
(m, 2H), 5.09 (m, 2H). 31P NMR: d = 5.02. MS (ESI): m/z = 434.43
[M]+. Elemental Anal. Calcd: C, 60.50; H, 10.60; N, 3.07; P, 6.78.
Found: C, 50.40; H, 9.20; N, 2.60; P, 5.00. Insufficient sample re-
mained for repeat CHN and P analysis.
5.2.6. General procedure for alkylglycerol synthesis
The long chain alcohol (17.8 mmol) was dissolved in methylene
chloride (6.5 mL) and cooled on an ice bath. DIBAL-H (1 M in hex-
ane, 7.7 mL) was added and the reaction mixture was removed
from the ice bath and stirred for 1 h. The appropriate glycidol
5.2. Compound syntheses
5.2.1. General procedure for alkylphosphocholine synthesis
A mixture of 2-bromoethanol (460
lL, 6.5 mmol), triethylamine
(500 lL, 7.7 mmol) was added to the reaction mixture and was
(2.27 mL, 16.3 mmol), and THF (6 mL) was added dropwise over
15 min to a solution of phosphorus oxychloride (1.031 g, 6.5 mmol)
and THF (5 mL) on an ice bath. A mixture of fatty alcohol (6.5 mmol),
triethylamine (2.27 mL, 16.3 mmol), and THF (6 mL) was added
dropwise to the reaction mixture. After stirring for 1 h, a mixture
stirred for 72 h at room temperature. Sodium potassium tartrate
(1.64 g, 7.8 mmol) was dissolved in water (3 mL) and added to
the reaction mixture. The reaction mixture was stirred for 30 min
and the solvents were removed. The residue was dissolved in ethyl
acetate (100 mL) and extracted with water (40 mL). The organic
phase was dried over magnesium sulfate and the solvent was then
removed. The crude products were purified by flash column chro-
matography starting with ethyl acetate/hexane (2:3) then switch-
ing to 100% ethyl acetate as the eluent to afford a white solid after
evaporation of solvents.
of water (362 lL, 20.1 mmol) and triethylamine (2.27 mL,
16.3 mmol) was added, the ice bath was removed, and the mixture
was stirred for 1 h. The white precipitate was removed by gravity fil-
tration and the solvent was evaporated. The residue was dissolved in
a mixture of chloroform (10 mL) and methanol (24 mL) and was ex-
tracted with water (20 mL). The aqueous phase was extracted twice
with a mixture of chloroform (10 mL) and methanol (2 mL). The sol-
vents were removed from the combined organic layers. The residue
was loadedontoa 24 g SupraFlashcolumnand elutedwith ethylace-
tate/hexane (2:3) then chloroform/methanol/ammonium hydrox-
ide (60:40:7) collecting 20 mL fractions to remove unreacted
alcohol. After evaporation of solvents, the residue was dissolved in
4.2 M trimethylamine (77 equiv) in ethanol, diluted with twice the
volume with methanol, and stirred at 50 °C for 72 h. The solvents
were evaporated and the final product was purified by flash column
chromatography starting with chloroform/methanol/ammonium
hydroxide (60:40:7) then switching to chloroform/methanol/
ammonium hydroxide (60:40:12).
5.2.7. (R)-Decylglycerol (3a)
Yield: 209 mg (11.7%). Rf = 0.23 (ethyl acetate/hexane 2:3). 1H
NMR (chloroform-d): d = 0.87 (t, 3H), 1.24 (br, 14H), 1.57 (m, 2H),
2.16 (dd, 1H), 2.59 (d, 1H), 3.48 (overlapping multiplets, 4H),
3.67 (dq, 2H), 3.84 (m, 1H).
5.2.8. (S)-Decylglycerol (3b)
Yield: 228 mg (12.7%). Rf = 0.22 (ethyl acetate/hexane 2:3). 1H
NMR (chloroform-d): d = 0.86 (t, 3H), 1.24 (br, 14H), 1.56 (m, 2H),
2.17 (dd, 1H), 2.59 (d, 1H), 3.47 (overlapping multiplets, 4H),
3.67 (dq, 2H), 3.85 (m, 1H).
5.2.9. (R)-Dodecylglycerol (3c)
5.2.2. Tetradecylphosphocholine (2a)
Yield: 638 mg (31.8%). Rf = 0.24 (ethyl acetate/hexane 2:3). 1H
NMR (chloroform-d): d = 0.86 (t, 3H), 1.25 (br, 18H), 1.57 (m, 2H),
2.18 (dd, 1H), 2.61 (d, 1H), 3.49 (overlapping multiplets, 4H),
3.68 (dq, 2H), 3.84 (m, 1H).
Yield: 70 mg of white solid (2.8%). Rf = 0.20 (chloroform/metha-
nol/ammonium hydroxide 60:40:7). 1H NMR (chloroform-d/meth-
anol-d4 3:2): d = 0.64 (t, 3H), 1.03 (br, 22H), 1.38 (m, 2H), 2.98 (s,
9H), 3.35 (m, 2H), 3.61 (dt, 2H), 3.97 (m, 2H). 31P NMR: d = 4.83.
MS (ESI): m/z = 380.36 [M]+. Elemental Anal. Calcd: C, 56.70; H,
10.52; N, 3.48; P, 7.70. Found: C, 56.92; H, 10.81; N, 3.68; P, 7.80.
5.2.10. (S)-Dodecylglycerol (3d)
Yield: 176 mg (8.8%). Rf = 0.25 (ethyl acetate/hexane 2:3). 1H
NMR (chloroform-d): d = 0.87 (t, 3H), 1.25 (br, 18H), 1.56 (m, 2H),
2.14 (dd, 1H), 2.57 (d, 1H), 3.48 (overlapping multiplets, 4H),
3.68 (dq, 2H), 3.84 (m, 1H).
5.2.3. Hexadecylphosphocholine (2b)
Yield: 136 mg of white solid (5.1%). Rf = 0.23 (chloroform/meth-
anol/ammonium hydroxide 60:40:7). 1H NMR (chloroform-d/
methanol-d4 3:2): d = 0.64 (t, 3H), 1.07 (br, 26H), 1.39 (m, 2H),
2.98 (s, 9H), 3.37 (m, 2H), 3.61 (dt, 2H), 3.99 (m, 2H). 31P NMR:
d = 4.82. MS (ESI): m/z = 408.41 [M]+. Elemental Anal. Calcd: C,
5.2.11. (R)-Tetradecylglycerol (3e)
Yield: 483 mg (21.7%). Rf = 0.27 (ethyl acetate/hexane 2:3). 1H
NMR (chloroform-d): d = 0.87 (t, 3H), 1.25 (br, 22H), 1.57 (m, 2H),

E. J. North et al. / Bioorg. Med. Chem. 17 (2009) 3433–3442
3439
2.15 (dd, 1H), 2.58 (d, 1H), 3.48 (overlapping multiplets, 4H), 3.67
(dq, 2H), 3.84 (m, 1H).
product was purified by flash column chromatography using
methylene chloride/methanol/water (65:35:6). Products were
isolated as oils after evaporation of solvent. Yields were gener-
ally low, in the range of 1–9%.
5.2.12. (S)-Tetradecylglycerol (3f)
Yield: 458 mg (20.6%). Rf = 0.28 (ethyl acetate/hexane 2:3). 1H
NMR (chloroform-d): d = 0.87 (t, 3H), 1.25 (br, 22H), 1.57 (m, 2H),
2.15 (dd, 1H), 2.58 (d, 1H), 3.48 (overlapping multiplets, 4H),
3.67 (dq, 2H), 3.84 (m, 1H).
5.2.19. (R)-1-Decyl-2-hydroxy-sn-glycero-3-phosphonocholine
(6a)
Rf = 0.09 (methylene chloride/methanol/water 65:35:6). 1H
NMR (chloroform-d/methanol-d4 1:1): d = 0.50 (t, 3H), 0.94 (br,
14H), 1.23 (overlapping multiplets, 4H), 1.63 (m, 2H), 2.75 (s,
9H), 2.97–3.40 (overlapping multiplets, 8H), 3.51 (late fraction,
m, 1H), 3.94 (early fraction, m, 1H). 31P NMR: d = 28.61 (early frac-
tion), 28.84 (late fraction). MS (ESI): m/z = 396.33 [M]+. Elemental
Anal. Calcd: C, 54.53; H, 10.12; N, 3.35. Found: C, 55.15; H,
10.81; N, 3.08 (Insufficient sample for P analysis).
5.2.13. (R)-Hexadecylglycerol (3g)
Yield: 513 mg (21%). Rf = 0.28 (ethyl acetate/hexane 2:3). 1H
NMR (chloroform-d): d = 0.86 (t, 3H), 1.24 (br, 26H), 1.56 (m, 2H),
3.48 (overlapping multiplets, 4H), 3.67 (dq, 2H), 3.84 (m, 1H).
5.2.14. (S)-Hexadecylglycerol (3h)
Yield: 913 mg (35%). Rf = 0.27 (ethyl acetate/hexane 2:3). 1H
NMR (chloroform-d): d = 0.86 (t, 3H), 1.24 (br, 26H), 1.56 (m, 2H),
3.48 (overlapping multiplets, 4H), 3.68 (dq, 2H), 3.85 (m, 1H).
5.2.20. (R)-1-Dodecyl-2-hydroxy-sn-glycero-3-phosphono-
choline (6b)
Rf = 0.09 (methylene chloride/methanol/water 65:35:6). 1H
NMR (chloroform-d/methanol-d4 1:1): d = 0.50 (t, 3H), 0.90 (br,
18H), 1.21 (overlapping multiplets, 4H), 1.65 (m, 2H), 2.76 (s,
9H), 2.98–3.42 (overlapping multiplets, 8H), 3.52 (late fraction,
m, 1H), 3.96 (early fraction, m, 1H). 31P NMR: d = 28.26 (early frac-
tion), 28.52 (late fraction). MS (ESI): m/z = 424.38 [M]+. Elemental
Anal. Calcd: C, 56.48; H, 10.38; N, 3.14; P, 6.94. Found: C, 52.65;
H, 10.24; N, 2.81; P, 6.90.
5.2.15. (R)-Octadecylglycerol (3i)
Yield: 785 mg (30%). Rf = 0.26 (ethyl acetate/hexane 2:3). 1H
NMR (chloroform-d): d = 0.87 (t, 3H), 1.24 (br, 30H), 1.56 (m, 2H),
3.48 (overlapping multiplets, 4H), 3.67 (dq, 2H), 3.85 (p, 1H).
5.2.16. (S)-Octadecylglycerol (3j)
Yield: 298 mg (11%). Rf = 0.26 (ethyl acetate/hexane 2:3). 1H
NMR (chloroform-d): d = 0.87 (t, 3H), 1.24 (br, 30H), 1.56 (m, 2H),
3.48 (m, 4H), 3.67 (dq, 2H), 3.84 (m, 1H).
5.2.21. (R)-1-Tetradecyl-2-hydroxy-sn-glycero-3-phosphono-
choline (6c)
5.2.17. 3-Chloropropylphosphonic acid (4a)
Rf = 0.09 (methylene chloride/methanol/water 65:35:6). 1H
NMR (chloroform-d/methanol-d4 1:1): d = 0.51 (t, 3H), 0.88 (br,
22H), 1.23 (overlapping multiplets, 4H), 1.62 (m, 2H), 2.74 (s,
9H), 2.95–3.39 (overlapping multiplets, 8H), 3.49 (late fraction,
m, 1H), 3.92 (early fraction, m, 1H). 31P NMR: d = 28.27 (early frac-
tion), 28.55 (late fraction). MS (ESI): m/z = 452.42 [M]+. Elemental
Anal. Calcd: C, 58.21; H, 10.62; N, 2.95; P, 6.53. Found: C, 59.59;
H, 10.96; N, 3.13; P, 6.20.
Dimethylphosphite (4.5 mL, 49.0 mmol) was added dropwise to
a solution of potassium t-butoxide (5.150 g, 45.9 mmol) and THF
(18 mL) within 15 min. The resulting thick grey gel was added
dropwise to
a solution of 1-bromo-3-chloropropane (5.7 mL,
57.6 mmol) and THF (9 mL). The reaction mixture was heated un-
der reflux for 15 min. The white precipitate was filtered using grav-
ity filtration. The precipitate was washed twice with two portions
of diisopropyl ether (18 mL). The solvents were evaporated to pro-
duce a clear liquid. The resulting residue was dissolved in concen-
trated HCl (69 mL) and heated under reflux for 9 h. The solvent was
evaporated under reduced pressure to produce 4.908 g (67% yield)
of waxy amber-colored crude final product. 1H NMR (D2O):
d = 1.70–2.10 (overlapping multiplets, 4H), 3.64 (m, 2H).
5.2.22. (R)-1-Hexadecyl-2-hydroxy-sn-glycero-3-phosphono-
choline (6d)
Rf = 0.09 (methylene chloride/methanol/water 65:35:6). 1H
NMR (chloroform-d/methanol-d4 1:1): d = 0.53 (t, 3H), 0.99 (br,
26H), 1.21–1.52 (overlapping multiplets, 4H), 1.65 (m, 2H), 2.75
(s, 9H), 3.11–3.40 (overlapping multiplets, 8H), 3.55 (m, 1H). 31P
NMR: d = 29.38. MS (ESI): m/z = 480.45 [M]+. Elemental Anal. Calcd:
C, 59.74; H, 10.83; N, 2.79; P, 6.16. Found: C, 59.90; H, 9.35; N,
2.17; P, 5.10.
5.2.18. General procedure for 1-alkyl-2-hydroxy-sn-glycero-3-
phosphonocholine synthesis
All reactions were carried out with at least 2.0 mmol of alkyl-
glycerol. Alkylglycerol (1 equiv), 3-chloropropylphosphonic acid
(1.1 equiv), DCC (2.2 equiv), and DMAP (0.1 equiv) were dis-
solved in THF (9.7 mL/mmol alkylglycerol) and stirred for 48 h
at room temperature. Water (84 equiv) was added to the reac-
tion mixture and it was allowed to stir for an additional 24 h.
The precipitate was removed by gravity filtration and the sol-
vents were evaporated. The residue was dissolved in a mixture
of chloroform (3.8 mL/mmol alkylglycerol) and methanol
(4.5 mL/mmol alkylglycerol) and was extracted with water
(3.8 mL/mmol alkylglycerol). The aqueous phase was extracted
twice with mixtures of chloroform (3.8 mL/mmol alkylglycerol)
and methanol (1 mL/mmol alkylglycerol). The solvents were re-
moved from the combined organic layers. The residue was
loaded onto a 24 g SupraFlash column and eluted with ethyl ace-
tate then methanol collecting 20 mL fractions to remove unre-
acted alkylglycerol. After evaporation of solvents, the residue
was dissolved in 4.2 M trimethylamine (77 equiv) in ethanol
and diluted with twice the volume with methanol and stirred
at 50 °C for 72 h. The solvents were evaporated and the final
5.2.23. (R)-1-Octadecyl-2-hydroxy-sn-glycero-3-phosphono-
choline (6e)
Rf = 0.09 (methylene chloride/methanol/water 65:35:6). 1H
NMR (chloroform-d/methanol-d4 1:1): d = 0.50 (t, 3H), 0.90 (br,
30H), 1.21 (overlapping multiplets, 4H), 1.64 (m, 2H), 2.75 (s,
9H), 2.97–3.40 (overlapping multiplets, 8H), 3.51 (late fraction,
m, 1H), 3.95 (early fraction, m, 1H). 31P NMR: d = 28.83. MS (ESI):
m/z = 508.50 [M]+. Elemental Anal. Calcd: C, 61.10; H, 11.02; N,
2.64; P, 5.84. Found: C, 58.58; H, 10.83; N, 3.37; P, 5.70.
5.2.24. (S)-1-Decyl-2-hydroxy-sn-glycero-3-phosphonocholine
(6f)
Rf = 0.09 (methylene chloride/methanol/water 65:35:6). 1H
NMR (chloroform-d/methanol-d4 1:1): d = 0.52 (t, 3H), 0.92 (br,
14H), 1.22 (overlapping multiplets, 4H), 1.65 (m, 2H), 2.76 (s,
9H), 3.00–3.44 (overlapping multiplets, 8H), 3.54 (late fraction,
m, 1H), 3.96 (early fraction, m, 1H). 31P NMR: d = 33.63 (early frac-

3440
E. J. North et al. / Bioorg. Med. Chem. 17 (2009) 3433–3442
tion), 34.08 (late fraction). MS (ESI): m/z = 396.36 [M]+. Elemental
Anal. Calcd: C, 54.53; H, 10.12; N, 3.35; P, 7.40. Found: C, 50.42;
H, 10.20; N, 3.02; P, 7.00.
4.2 M trimethylamine (77 equiv) in ethanol and diluted with
twice the volume of methanol and stirred at 50 °C for 72 h.
The solvents were evaporated and the final product was purified
by flash column chromatography starting with chloroform/meth-
anol/ammonium hydroxide (60:40:7) then switching to chloro-
form/methanol/ammonium hydroxide (60:40:12).
5.2.25. (S)-1-Dodecyl-2-hydroxy-sn-glycero-3-
phosphonocholine (6g)
Rf = 0.09 (methylene chloride/methanol/water 65:35:6). 1H
NMR (chloroform-d/methanol-d4 1:1): d = 0.50 (t, 3H), 0.88 (br,
18H), 1.20 (overlapping multiplets, 4H), 1.62 (m, 2H), 2.72 (s,
9H), 2.92–3.38 (overlapping multiplets, 8H), 3.58 (late fraction,
m, 1H), 3.91 (early fraction, m, 1H). 31P NMR: d = 28.25 (early frac-
tion), 28.53 (late fraction). MS (ESI): m/z = 424.38 [M]+. Elemental
Anal. Calcd: C, 56.48; H, 10.38; N, 3.14; P, 6.94. Found: C, 56.97;
H, 11.11; N, 3.33; P, 7.20.
5.2.30. Tetradecylphosphonocholine (8a)
Yield: 47 mg (1.2% yield) of white solid. Rf = 0.15 (chloroform/
methanol/ammonium hydroxide 60:40:7). 1H NMR (chloroform-
d/methanol-d4 2:1): d = 0.65 (t, 3H), 1.03 (br, 22 H), 1.26–1.44
(overlapping multiplets, 4H), 1.75 (m, 2H), 2.88 (s, 9H), 3.19 (m,
2H), 3.62 (m, 2H). 31P NMR: d = 27.60. MS (ESI): m/z = 378.36
[M]+. Elemental Anal. Calcd: C, 59.97; H, 11.07; N, 3.50; P, 7.73.
Found: C, 54.68; H, 18.81; N, 4.02; P, 5.00 (insufficient sample re-
mained for repeat CHN and P analysis).
5.2.26. (S)-1-Tetradecyl-2-hydroxy-sn-glycero-3-
phosphonocholine (6h)
Rf = 0.09 (methylene chloride/methanol/water 65:35:6). 1H
NMR (chloroform-d/methanol-d4 1:1): d = 0.50 (t, 3H), 0.90 (br,
22H), 1.22 (overlapping multiplets, 4H), 1.65 (m, 2H), 2.76 (s,
9H), 2.98–3.42 (overlapping multiplets, 8H), 3.54 (late fraction,
m, 1H), 3.94 (early fraction, m, 1H). 31P NMR: d = 28.22 (early frac-
tion), 28.61 (late fraction). MS (ESI): m/z = 452.40 [M]+. Elemental
Anal. Calcd: C, 58.21; H, 10.62; N, 2.95; P, 6.53. Found: C, 54.62;
H, 11.40; N, 2.72; P, 6.10.
5.2.31. Hexadecylphosphonocholine (8b)
Yield: 15 mg (0.4% yield) of white solid. Rf = 0.21 (chloroform/
methanol/ammonium hydroxide 60:40:7). 1H NMR (chloroform-
d/methanol-d4 2:1): d = 0.65 (t, 3H), 1.03 (br, 26 H), 1.22-1.42
(overlapping multiplets, 4H), 1.71 (m, 2H), 2.88 (s, 9H), 3.20 (m,
2H), 3.60 (m, 2H). 31P NMR: d = 27.51. MS (ESI): m/z = 406.38
[M]+. Elemental Anal. Calcd: P, 7.23. Found: P, 6.10 (insufficient
sample available for initial CHN analysis).
5.2.27. (S)-1-Hexadecyl-2-hydroxy-sn-glycero-3-
5.2.32. Octadecylphosphonocholine (8c)
phosphonocholine (6i)
Yield: 7 mg (0.2% yield) of white solid. Rf = 0.13 (chloroform/
methanol/ammonium hydroxide 60:40:7). 1H NMR (chloroform-
d/methanol-d4 2:1): d = 0.65 (t, 3H), 1.05 (br, 30 H), 1.22–1.42
(overlapping multiplets, 4H), 1.71 (m, 2H), 2.90 (s, 9H), 3.22
(m, 2H), 3.62 (m, 2H). 31P NMR: d = 27.19. MS (ESI): m/
z = 434.43 [M]+. Elemental Anal. Calcd: C, 63.13; H, 11.48; N,
3.07; P, 6.78. Found: C, 55.40; H, 10.60; N, 2.70; P, 5.30 (Limited
sample availability resulted in larger reporting limits and analyt-
ical error).
Rf = 0.09 (methylene chloride/methanol/water 65:35:6). 1H
NMR (chloroform-d/methanol-d4 1:1): d = 0.58 (t, 3H), 0.98 (br,
26H), 1.16–1.43 (overlapping multiplets, 4H), 1.71 (m, 2H), 2.83
(s, 9H), 3.07–3.49 (overlapping multiplets, 8H), 3.61 (late fraction,
m, 1H), 4.04 (early fraction, m, 1H). 31P NMR: d = 28.57 (early frac-
tion), 28.88 (late fraction). MS (ESI): m/z = 480.45 [M]+. Elemental
Anal. Calcd: C, 59.74; H, 10.83; N, 2.79; P, 6.16. Found: C, 50.31;
H, 10.72; N, 5.11; P, 6.10.
5.2.28. (S)-1-Octadecyl-2-hydroxy-sn-glycero-3-
5.2.33. Oleylphosphonocholine (8d)
phosphonocholine (6j)
Yield: 145 mg (6.6% yield) of white solid. Rf = 0.15 (chloroform/
methanol/ammonium hydroxide 60:40:7). 1H NMR (chloroform-d/
methanol-d4 2:1): d = 0.56 (t, 3H), 0.95 (br, 22 H), 1.25–1.45 (over-
lapping multiplets, 4H), 1.65–1.85 (overlapping multiplets, 6H),
2.88 (s, 9H), 3.10 (m, 2H), 3.50 (m, 2H), 5.00 (t, 2H). 31P NMR:
d = 27.66. MS (ESI): m/z = 432.42 [M]+. Elemental Anal. Calcd: C,
63.41; H, 11.09; N, 3.08; P, 6.81. Found: C, 38.60; H, 7.00; N,
10.10; P, 5.00 (insufficient sample remained for repeat CHN and
P analysis).
Rf = 0.09 (methylene chloride/methanol/water 65:35:6). 1H
NMR (chloroform-d/methanol-d4 1:1): d = 0.51 (t, 3H), 0.89 (br,
30H), 1.12–1.34 (overlapping multiplets, 4H), 1.65 (m, 2H), 2.76
(s, 9H), 2.98–3.42 (overlapping multiplets, 8H), 3.53 (late fraction,
m, 1H), 3.96 (early fraction, m, 1H). 31P NMR: d = 28.69 (early frac-
tion), 28.95 (late fraction). MS (ESI): m/z = 508.50 [M]+. Elemental
Anal. Calcd: C, 61.10; H, 11.02; N, 2.64; P, 5.84. Found: C, 52.00;
H, 10.32; N, 5.01; P, 5.40.
5.2.29. General procedure for alkyphosphonocholine synthesis
All reactions were carried out with at least 8.8 mmol of long
chain alcohol. Alcohol (1 equiv), 3-chloropropylphosphonic acid
(1.1 equiv), DCC (2.2 equiv), and DMAP (0.1 equiv) were dis-
solved in THF (9.3 mL/mmol alcohol) and stirred for 48 h at
room temperature. Water (84 equiv) was added to the reaction
mixture and it was allowed to stir for an additional 24 h. The
precipitate was removed by gravity filtration and the solvents
were evaporated. The residue was dissolved in a mixture of chlo-
roform (41 mL) and methanol (49 mL) and was extracted with
water (40 mL). The aqueous phase was extracted twice with
mixtures of chloroform (40 mL) and methanol (10 mL). The sol-
vents were removed from the combined organic layers. The or-
ganic phase was evaporated and the residue was loaded onto a
24 g SupraFlash column and eluted with ethyl acetate/hexane
5.2.34. General procedure for phenyl-O-
alkylphosphoromonochloridate synthesis
Triethylamine (841
(2.3 mL) was added dropwise over 15 min to a freshly prepared
solution of phosphorus oxychloride (500 L, 5.5 mmol), alcohol
lL, 5.9 mmol) in methylene chloride
l
(5.5 mmol), and methylene chloride (13.6 mL) on an ice bath. Once
all of the triethylamine solution was added, the reaction mixture
was stirred for 30 min. Phenol (5.5 mmol) was added, followed
by dropwise addition of triethylamine (841 lL, 5.9 mmol) in meth-
ylene chloride (2.3 mL). The reaction mixture was stirred for 1 h at
0 °C, then quenched by addition of 20 mL of saturated ammonium
chloride. The mixture was extracted with three portions of methy-
lene chloride (11 mL ea.). The combined organic layers were
washed with 20 mL of brine and dried over magnesium sulfate.
The solvents were removed to afford crude monochloro products.
Partial purification of monochloro products was achieved using
flash column chromatography and ethyl acetate/hexane (9:1) as
the eluent.
(2:3)
then
chloroform/methanol/ammonium
hydroxide
(60:40:7), collecting 20 mL fractions to remove unreacted alco-
hol. After evaporation of solvents, the residue was dissolved in

E. J. North et al. / Bioorg. Med. Chem. 17 (2009) 3433–3442
3441
5.2.35. Phenyl-O-tetradecylphosphoromonochloridate (9a)
5.2.42. N-(2-Trimethylaminoethyl)-O-octadecyl
phosphoramidate (11c)
Rf = 0.15 (chloroform/methanol/ammonium
Yield: 1.222 g (58% yield). Rf = 0.59 (ethyl acetate/hexane 9:1).
1H NMR (chloroform-d): d = 0.84 (t, 3H), 1.30 (br, 22H), 1.72 (m,
2H), 4.31 (tq, 2H), 7.26 (m, 3H), 7.36 (m, 2H).
hydroxide
60:40:7). 1H NMR (chloroform-d/methanol-d4 2:1), d = 0.74 (t,
3H), 1.13 (br, 30 H), 1.47 (m, 2H), 3.06 (s, 9H), 3.14–3.38 (overlap-
ping multiplets, 4H), 3.63 (q, 2H). 31P NMR: d = 11.60. MS (ESI): m/
z = 435.45 [M]+. Elemental Anal. Calcd: C, 60.37; H, 11.23; N, 6.12;
P, 6.77. Found: C, 58.38; H, 11.16; N, 6.12; P, 6.20.
5.2.36. Phenyl-O-hexadecylphosphoromonochloridate (9b)
Yield: 607 mg (27% yield). Rf = 0.45 (ethyl acetate/hexane 9:1).
1H NMR (chloroform-d): d = 0.87 (t, 3H), 1.24 (br, 26H), 1.77 (m,
2H), 4.30 (tq, 2H), 7.24 (m, 3H), 7.35 (m, 2H).
5.2.43. N-(2-Trimethylaminoethyl)-O-oleyl Phosphoramidate
5.2.37. Phenyl-O-octadecylphosphoromonochloridate (9c)
Yield: 671 mg (28% yield). Rf = 0.48 (ethyl acetate/hexane 9:1).
1H NMR (chloroform-d): d = 0.87 (t, 3H), 1.24 (br, 30H), 1.77 (m,
2H), 4.30 (tq, 2H), 7.24 (m, 3H), 7.34 (m, 2H).
(11d)
Rf = 0.16
(chloroform/methanol/ammonium
hydroxide
60:40:7). 1H NMR (chloroform-d/methanol-d4 2:1), d = 0.66 (t,
3H), 1.05 (br, 22 H), 1.37 (m, 2H), 1.77 (overlapping multiplets,
4H), 2.97 (s, 9H), 3.04–3.28 (overlapping multiplets, 4H), 3.54 (q,
2H), 5.10 (t, 2H). 31P NMR: d = 11.60. MS (ESI): m/z = 433.43 [M]+.
Elemental Anal. Calcd: C, 60.63; H, 10.84; N, 6.15; P, 6.80. Found:
C, 56.10; H, 10.64; N, 5.68; P, 6.60.
5.2.38. Phenyl-oleylphosphoromonochloridate (9d)
Yield: 829 mg (35% yield). Rf = 0.51 (ethyl acetate/hexane 9:1).
1H NMR (chloroform-d): d = 0.87 (t, 3H), 1.25 (br, 22H), 1.79 (m,
2H), 2.00 (m, 4H), 4.29 (tq, 2H), 5.34 (m, 2H), 7.24 (m, 3H), 7.36
(m, 2H).
5.3. Cell culture
5.2.39. General procedure for N-(2-trimethylaminoethyl)-O-
alkyl phosphoramidate synthesis
MDA-MB-435 cells were cultured at 37 °C and 5% CO2 in Dul-
becco’s Modified Eagle Medium (DMEM) (MediaTech, Herndon,
VA) containing 5% fetal bovine serum (Hyclone, Logan, UT),
100 U/ml penicillin–streptomycin (Hyclone, Logan, UT), and
All reactions were carried out with at least 1.46 mmol of phe-
nyl-alkylphosphoromonochloridate. 2-Bromoethylamine hydrogen
bromide (1 equiv) was dissolved in a minimal amount of DMF and
diluted with THF (0.355 mL/mmol phenyl-alkylphosphoromo-
nochloridate). Phenyl-alkylphosphoromonochloridate (1 equiv)
was added to the DMF/THF mixture. Triethylamine (2.3 equiv)
was added dropwise to the reaction mixture at room temperature.
Once all of the triethylamine had been added, the reaction mixture
was stirred for 80 min. The precipitate was removed using gravity
filtration and the solvents were evaporated. The residue was dis-
solved in chloroform (20 mL) and extracted with water (10 mL).
The organic phase was evaporated and the residue was loaded onto
a 24 g SupraFlash column and eluted with chloroform/ethyl ace-
tate (10:1) collecting 20 mL fractions to remove unreacted phe-
nyl-alkylphosphoromonochloridate. After evaporation of solvents,
the residue was dissolved in 4.2 M trimethylamine (77 equiv) in
ethanol and diluted with twice the volume with methanol and stir-
red at 50 °C for 72 h. The solvents were evaporated and the final
product was purified by flash column chromatography starting
with chloroform/methanol/ammonium hydroxide (60:40:7) then
292 lg/ml L-glutamine (Hyclone, Logan, UT). Cells were grown to
ꢀ80% confluence at which time the cells were washed twice with
sterile phosphate buffered saline prior to the addition of serum free
DMEM containing L-glutamine. Conditioned medium was collected
after 24–30 h, supplemented with 20% ethylene glycol and was
clarified by centrifugation at 3000g and 4 °C for 10 min. The med-
ium was concentrated ꢀ10-fold and buffer exchanged into Tris
(50 mM, pH 7.4) containing 20% ethylene glycol using an Amicon
8050 cell fitted with a PM30 filter (Millipore, Billerica, MA). Ali-
quots of concentrated conditioned media were stored at 4 °C until
needed.
5.4. ATX inhibition assay
ATX inhibition was determined utilizing the synthetic sub-
strate FS-3 (Echelon Biosciences, Inc., Salt Lake City, Utah,
USA), concentrated conditioned media, and LPC analog, each
comprising one third of the total volume. Final concentrations
of FS-3 and charcoal-stripped fatty acid free bovine serum albu-
changing
to
chloroform/methanol/ammonium
hydroxide
(60:40:12). Products were isolated as white solids after evapora-
tion of solvents. Yields were generally low, in the range of 2–8%.
min were 1
MgCl2, 1 mM CaCl2, 5 mM KCl, 140 mM NaCl, 50 mM Tris, pH
8). LPC analogs had final concentrations of 0.1, 1, and 10
lM and 30 lM, respectively, in assay buffer (1 mM
l
M
5.2.40. N-(2-Trimethylaminoethyl)-O-tetradecyl
phosphoramidate (11a)
in assay buffer. All assays were carried out in 96 well plates in
a BioTek Synergy-2 plate reader (BioTek, Winooski, VT, USA)
with excitation and emission wavelengths of 485 and 538 nm,
respectively. Fluorescent emission detection occurred every
5 min and data were shown as percent ATX inhibition, with re-
spect to vehicle control after subtraction of fluorescence with no
CCM, at the 1 h time point, where fluorescence detection with
respect to time is linear. All data were reported as mean stan-
dard deviation with at least three wells.
Rf = 0.16
(chloroform/methanol/ammonium
hydroxide
60:40:7). 1H NMR (chloroform-d/methanol-d4 2:1), d = 0.65 (t,
3H), 1.03 (br, 22 H), 1.34 (m, 2H), 2.96 (s, 9H), 2.98–3.23 (overlap-
ping multiplets, 4H), 3.53 (q, 2H). 31P NMR: d = 11.61. MS (ESI): m/
z = 379.38 [M]+. Elemental Anal. Calcd: C, 56.84; H, 10.79; N, 6.98;
P, 7.71. Found: C, 54.70; H, 10.10; N, 6.90; P, 7.50.
5.2.41. N-(2-Trimethylaminoethyl)-O-hexadecyl
Phosphoramidate (11b)
Acknowledgments
Rf = 0.16
(chloroform/methanol/ammonium
hydroxide
60:40:7). 1H NMR (chloroform-d/methanol-d4 2:1), d = 0.64 (t,
3H), 1.02 (br, 26 H), 1.37 (m, 2H), 2.95 (s, 9H), 2.98–3.24 (overlap-
ping multiplets, 4H), 3.52 (q, 2H), 3.87 (s, 1H). 31P NMR: d = 11.61.
MS (ESI): m/z = 407.40 [M]+. Elemental Anal. Calcd: C, 58.72; H,
11.03; N, 6.52; P, 7.21. Found: C, 55.67; H, 10.90; N, 6.47; P, 6.40.
The authors acknowledge the National Institutes of Health (NIH
R01 HL 084007) for their financial support and the National Sci-
ence Foundation (NSF CHE 0443627, NSF CHE 0619682) for their
financial aid in the acquisition of a Varian 500 MHz NMR and a
Thermoelectron LTQ-XL LC-MS.

3442
E. J. North et al. / Bioorg. Med. Chem. 17 (2009) 3433–3442
20. Zalatan, J. G.; Fenn, T. D.; Brunger, A. T.; Herschlag, D. Biochemistry 2006, 45,
References and notes
9788.
21. Saunders, L. P.; Ouellette, A.; Bandle, R.; Chang, W. C.; Zhou, H.; Misra, R. N.;
Cruz, E. M. D. L. Mol. Cancer Ther. 2008, 7, 3352.
22. Parrill, A. L.; Echols, U.; Nguyen, T.; Pham, T.-C. T.; Hoeglund, A.; Baker, D. L.
Bioorg. Med. Chem. 2008, 16, 1784.
23. Moulharat, N.; Fould, B.; Giganti, A.; Boutin, J. A.; Ferry, G. Chem. Biol. Interact.
2008, 172, 115.
24. Clair, T.; Koh, E.; Ptaszynska, M.; Bandle, R.; Liotta, L.; Schiffmann, E.; Stracke,
M. Lipids Health Dis. 2005, 4.
25. van Meeteren, L.; Ruurs, P.; Christodoulou, E.; Goding, J.; Takakusa, H.; Kikuchi,
K.; Perrakis, A.; Nagano, T.; Moolenaar, W. J. Biol. Chem. 2005, 280, 21155.
26. Cui, P.; Tomsig, J. L.; McCalmont, W. F.; Lee, S.; Becker, C. J.; Lynch, K. R.;
Macdonald, T. L. Bioorg. Med. Chem. Lett. 2007, 17, 1634.
27. Durgam, G.; Virag, T.; Walker, M.; Tsukahara, R.; Yasuda, S.; Liliom, K.; van
Meeteren, L.; Moolenaar, W.; Wilke, N.; Siess, W.; Tigyi, G.; Miller, D. J. Med.
Chem. 2005, 48, 4919.
28. Ferry, G.; Moulharat, N.; Pradere, J.-P.; Desos, P.; Try, A.; Genton, A.; Giganti, A.;
Beucher-Gaudin, M.; Lonchampt, M.; Bertrand, M.; Saulnier-Blache, J.-S.;
Tucker, G. C.; Cordi, A.; Boutin, J. A. J. Pharmacol. Exp. Ther. 2008, 327, 809.
29. Gududuru, V.; Zeng, K.; Tsukahara, R.; Makarova, N.; Fujiwara, Y.; Pigg, K. R.;
Baker, D. L.; Tigyi, G.; Miller, D. D. Bioorg. Med. Chem. Lett. 2006, 16, 451.
30. Baker, D. L.; Fujiwara, Y.; Pigg, K. R.; Tsukahara, R.; Kobayashi, S.; Murofushi, H.;
Uchiyama, A.; Murakami-Murofushi, K.; Koh, E.; Bandle, R. W.; Byun, H.-S.;
Bittman, R.; Fan, D.; Murph, M.; Mills, G. B.; Tigyi, G. J. Biol. Chem. 2006, 281,
22786.
1. Stracke, M.; Krutzch, H.; Unsworth, E.; Arestad, A.; Cioce, V.; Schiffmann, E.;
Liotta, L. J. Biol. Chem. 1992, 267, 2524.
2. van Meeteren, L. A.; Moolenaar, W. H. Prog. Lipid Res. 2007, 46, 145.
3. Umezu-Goto, M.; Kishi, Y.; Taira, A.; Hama, K.; Dohmae, N.; Takio, K.; Yamori,
T.; Mills, G.; Inoue, K.; Aoki, J.; Arai, H. J. Cell Biol. 2002, 158, 227.
4. Tokumura, A.; Majima, E.; Kariya, Y.; Tominaga, K.; Kogure, K.; Yasuda, K.;
Fukuzawa, K. J. Biol. Chem. 2002, 277, 39436.
5. Tabata, K.-I.; Baba, K.; Shiraishi, A.; Ito, M.; Fujita, N. Biochem. Biophys. Res.
Commun. 2007, 363, 861.
6. Mills, G. B.; Moolenaar, W. H. Nat. Rev. Cancer 2003, 3, 582.
7. Lee, C.-W.; Rivera, R.; Gardell, S.; Dubin, A. E.; Chun, J. J. Biol. Chem. 2006, 281,
23589.
8. Pasternack, S. M.; von Kugelgen, I.; Aboud, K. A.; Lee, Y.-A.; Ruschendorf, F.;
Voss, K.; Hillmer, A. M.; Molderings, G. J.; Franz, T.; Ramirez, A.; Nurnberg, P.;
Nothen, M. M.; Betz, R. C. Nat. Genet. 2008, 40, 329.
9. Murakami, M.; Shiraishi, A.; Tabata, K.; Fujita, N. Biochem. Biophys. Res.
Commun. 2008. doi:10.1016/j.bbrc.2008.04.145.
10. Parrill, A. L. Biochim. Biophys. Acta 2008, 1781, 540.
11. McIntyre, T. M.; Pontsler, A. V.; Silva, A. R.; Hilaire, A. S.; Xu, Y.; Hinshaw, J. C.;
Zimmerman, G. A.; Hama, K.; Aoki, J.; Arai, H.; Prestwich, G. D. Proc. Natl. Acad.
Sci. U.S.A. 2003, 100, 131.
12. Hama, K.; Aoki, J.; Fukaya, M.; Kishi, Y.; Sakai, T.; Suzuki, R.; Ohta, H.; Yamuri,
T.; Watanabe, M.; Chun, J.; Arai, H. J. Biol. Chem. 2004, 279, 17634.
13. Boucher, J.; Quilliot, D.; Pradere, J.-P.; Simon, M.-F.; Gres, S.; Guigne, C.; Prevot,
D.; Ferry, G.; Boutin, J. A.; Carpene, C.; Valet, P.; Saulnier-Blache, J. S.
Diabetologia 2005, 48, 569.
14. Zhao, C.; Fernandes, M. J.; Prestwich, G. D.; Turgeon, M.; Battista, J. D.; Clair, T.;
Poubelle, P. E.; Bourgoin, S. G. Mol. Pharmacol. 2008, 73, 587.
15. Inoue, M.; Ma, L.; Aoki, J.; Chun, J.; Ueda, H. Mol. Pain 2008, 4.
16. Rother, E.; Brandl, R.; Baker, D. L.; Goyal, P.; Gebhard, H.; Tigyi, G.; Siess, W.
Circulation 2003, 108, 741.
31. Jiang, G.; Xu, Y.; Fujiwara, Y.; Tsukahara, T.; Tsukahara, R.; Gajewiak, J.; Tigyi,
G.; Prestwich, G. D. ChemMedChem 2007, 2, 679.
32. van Meeteren, L. A.; Brinkmann, V.; Saulnier-Blache, J.-S.; Lynch, K. R.;
Moolenaar, W. H. Cancer Lett. 2008, 266, 203.
33. Ferguson, C. G.; Bigman, C. S.; Richardson, R. D.; van Meeteran, L. A.;
Moolenaar, W. H.; Prestwich, G. D. Org. Lett. 2006, 8, 2023.
34. Erukulla, R. K.; Byun, H.-S.; Locke, D. C.; Bittman, R. J. Chem. Soc., Perkin Trans. 1
1995, 18, 2199.
35. Kley, J.; Unger, C.; Massing, U. Monatsh. Chem. 1998, 129, 173.
36. Garrido-Hernandez, H.; Moon, K. D.; Geahlen, R. L.; Borch, R. F. J. Med. Chem.
2006, 49, 3368.
17. Gijsbers, R.; Ceulemans, H.; Stalmans, W.; Bollen, M. J. Biol. Chem. 2001, 276,
1361.
18. Jansen, S.; Callewaert, N.; Dewerte, I.; Andries, M.; Ceulemans, H.; Bollen, M. J.
Biol. Chem. 2007, 282, 11084.
37. Chen, L.; Wu, L.; Otaka, A.; Smyth, M. S.; Roller, P. P.; Burke, T. R., Jr.; Hertog, J.
D.; Zhang, Z.-Y. Biochem. Biophys. Res. Commun. 1995, 216, 976.
19. Galperin, M. Y.; Jedrzejas, M. J. Prot. Struct. Funct. Genet. 2001, 45, 318.