Amides of xanthurenic acid as zinc-dependent inhibitors of Lp-PLA 2

Article abstract of DOI:10.1016/j.bmcl.2011.12.045

AX10185, the phenyl amide of xanthurenic acid, was found to be a sub-100 nM inhibitor of Lp-PLA₂. However, in the presence of EDTA the inhibitory activity of AX10185 was extinguished while the enzymatic activity of Lp-PLA₂ did not change. Subsequent metal screening experiments determined the inhibition to be Zn²⁺ dependent. Structure-activity relationship studies indicated the presence of the 4-hydroxy group to be critical and selected substituted phenyl, polycyclic, and cycloaliphatic amides of xanthurenic acid to be well tolerated.

Full text of DOI:10.1016/j.bmcl.2011.12.045

Bioorganic & Medicinal Chemistry Letters 22 (2011) 868–871
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Amides of xanthurenic acid as zinc-dependent inhibitors of Lp-PLA2
Emme C.K. Lin a, , Yi Hu a, , Christopher M. Amantea a, Lan M. Pham a, Julia Cajica a, Eric Okerberg a,
Heidi E. Brown a, Allister Fraser a, Lingling Du a, Yasushi Kohno b, Junichi Ishiyama b,
John W. Kozarich a, Kevin R. Shreder a,
⇑
a ActivX Biosciences, Inc., 11025 N. Torrey Pines Road, La Jolla, CA 92037, USA
b Discovery Research Laboratories, Kyorin Pharmaceutical Co. Ltd, 2399-1, Nogi, Nogi-machi, Shimotsuga-gun, Tochigi, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
AX10185, the phenyl amide of xanthurenic acid, was found to be a sub-100 nM inhibitor of Lp-PLA2. How-
ever, in the presence of EDTA the inhibitory activity of AX10185 was extinguished while the enzymatic
activity of Lp-PLA2 did not change. Subsequent metal screening experiments determined the inhibition
to be Zn2+ dependent. Structure–activity relationship studies indicated the presence of the 4-hydroxy
group to be critical and selected substituted phenyl, polycyclic, and cycloaliphatic amides of xanthurenic
acid to be well tolerated.
Received 12 October 2011
Revised 5 December 2011
Accepted 8 December 2011
Available online 13 December 2011
Keywords:
AX10185
Ó 2011 Elsevier Ltd. All rights reserved.
Xanthurenic acid
Lp-PLA2
Zn2+ dependent inhibitor
Lipoprotein-associated Phospholipase A2 (Lp-PLA2) is a vascu-
lar-specific, inflammatory enzyme that circulates primarily bound
to low density lipoproteins (LDL) and is responsible for >95% of the
phospholipase activity associated with LDL.1 One important func-
tion of this Ca+2-independent phospholipase A2 superfamily mem-
ber is to hydrolyze oxidized phospholipids in LDL particles, which
in turn contributes to the process of atherosclerotic plaque pro-
gression. Multiple epidemiological studies have established a cor-
relation between elevated Lp-PLA2 levels and increased risk for
cardiovascular events such as myocardial infarction and stroke,
even after multivariable adjustment for traditional risk factors.2
There are several reports in the literature citing the potential of
Lp-PLA2 as a therapeutic target for the treatment of coronary artery
disease and atherosclerosis.3 Along those lines, a recent phase II
clinical trial demonstrated that inhibiting circulating Lp-PLA2
activity (by 59%) prevented expansion of the necrotic core in
coronary plaque.4 Current phase III trials are in place to assess
the efficacy of direct Lp-PLA2 inhibition in reducing adverse events
in patients with atherosclerosis. There is therefore a need for effec-
tive Lp-PLA2 inhibitors as therapeutics for treatment of Lp-PLA2-
mediated diseases.
plasma (php-Lp-PLA2),6 AX10185 was identified as a sub-100 nM
inhibitor (see Table 1). In a subsequent modification of this assay
to include EGTA or EDTA, we found that the addition of these anti-
coagulants, well-known chelators of a variety of metals, dimin-
ished or extinguished the Lp-PLA2 inhibitory activity of AX10185
(e.g., see Table 1) without changing the enzymatic activity of
Lp-PLA2. These experiments coupled with the observation that
the structure of 8-hydroxy-quinoline (compound 2, another
chelate of various metals including zinc, cadmium, and iron)7
was present as a substructure in AX10185, led us to hypothesize
that the inhibition of Lp-PLA2 by AX10185 was metal dependent.
To address which metal may be involved, a screening experi-
ment was conducted using php-Lp-PLA2 in the presence of 50 lM
added metal chloride salts (Fig. 1), none of which changed the enzy-
matic activity of Lp-PLA2. In the absence of added metal, the php-
Lp-PLA2 IC50 value for AX10185 was determined to be 100 nM.8
This value changed depending on the metal added in three different
ways: (1) The addition of CdCl2 and FeCl3 resulted in a significant
increase in the php-Lp-PLA2 IC50 value of AX10185 to 270 and
190 nM, respectively; (2) The addition of CoCl2, MnCl2, CaCl2, BaCl2,
and NiCl2 resulted in no significant change in the php-Lp-PLA2
inhibitory activity of AX10185; (3) The addition of ZnCl2 lowered
the IC50 value to 20 nM. When these metal-dependent IC50 experi-
ments were repeated with recombinant human Lp-PLA2 (rh-Lp-
PLA2),9 a more highly purified source of Lp-PLA2, analogous trends
were observed (Fig. 1).
As part of an effort to find novel inhibitors of Lp-PLA2, AX10185
was synthesized from the HATU catalyzed condensation of xanth-
urenic acid and aniline as shown in Scheme 1.5 When examined in
an Lp-PLA2 IC50 assay utilizing Lp-PLA2 purified from human
Because of the positive impact of zinc on the Lp-PLA2 inhibitory
activity, this specific metal was examined in greater detail. The
rh-Lp-PLA2 IC50 value of AX10185 was determined in the presence
⇑
Corresponding author. Tel.: +1 858 526 2517; fax: +1 858 587 4878.
E-mail address: kevins@activx.com (K.R. Shreder).
These authors contributed equally to this work.
0960-894X/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2011.12.045

E. C. K. Lin et al. / Bioorg. Med. Chem. Lett. 22 (2011) 868–871
100
869
80
60
40
20
Scheme 1. Synthesis of AX10185 from xanthurenic acid.
0.1
1
10
100
Table 1
[ZnCl2 added] (µM)
Lp-PLA2 IC50 values for AX10185 and compounds 2–5 where php-Lp-PLA2 = Lp-PLA2
purified from human plasma and rh-Lp-PLA2 = recombinant human Lp-PLA2
Figure 2. Change in the rh-Lp-PLA2 IC50 value for AX10185 (compound 1) as a
function of added ZnCl2.
ꢀ20
lM added ZnCl2. Moving forward in our studies, we chose to
include this concentration of ZnCl2 to our rh-Lp-PLA2 IC50 assay
to more closely mimic physiological plasma zinc concentrations.10
In contrast, no additional ZnCl2 was added the php-Lp-PLA2 assay.
Compounds containing some of the structural elements present
in AX10185 were examined in both Lp-PLA2 IC50 assays to investi-
gate which functional groups of AX10185 contributed to the
observed Lp-PLA2 inhibition (Table 1). Neither 8-hydroxy-quino-
line (2) nor its zinc complex 3 was found to inhibit Lp-PLA2.
Likewise, xanthurenic acid (4), also a zinc chelate,11 and the
4-dehydroxy version of AX10185 (5) lacked any Lp-PLA2 inhibitory
activity. This initial SAR foray indicated that the observed Zn2+
dependent inhibition of Lp-PLA2 by AX10185 could not be ascribed
to the simple presence of the 8-hydroxy-quinoline substructure.
Furthermore, both the 4-hydroxy group (compare AX10185 to
compound 5) and the 2-phenylamide of AX10185 are critical
pharmacophores.
Compound
php-Lp-PLA2 IC50 (nM)
rh-Lp-PLA2 IC50 (nM)
AX10185
74
27
AX10185a
>100,000a
>100,000
>100,000
>100,000
>100,000
>100,000a
>100,000
>100,000
>100,000
>100,000
2
3
4
5
To further explore and optimize the amide group of AX10185,
both phenyl-substituted derivatives (Table 2, compounds 6–14)
and amide analogs (Table 3, compounds 15–21) of AX10185 were
evaluated in the php-Lp-PLA2 IC50 assay. Various para-substituted
derivatives of AX10185 were synthesized and examined. Incorpora-
tion of a para-methoxy (6) or para-chloro group (7) resulted in a 55-
fold and 2.4-fold decrease in potency, respectively, relative to
AX10185. By contrast, incorporation of a para-fluoro group (8)
yielded an inhibitor more potent than the parent compound. Consis-
tent with the observed intolerance in the para-position for steric
bulk, the para-methyl derivative 9 was significantly less potent than
AX10185. However, meta-methyl substitution (10) was tolerated
and ortho-methyl substitution (11) yielded a 2.3-fold improvement
in potency relative to AX10185. Various difluoro-phenylamides
(12–14) were also tested. Of these, the 3,4-difluoro-phenyl amide
(14) was the most potent with a php-Lp-PLA2 IC50 value of 23 nM.
Selected amide analogs of AX10185 were found to be reasonable
substitutes of the phenyl amide present in AX10185 (Table 3).
Substitution of the phenyl group of AX10185 with cyclohexyl (15)
was found to yield an inhibitor of similar potency as the parent
compound. Ring contraction (16) resulted in a loss of potency while
ring expansion (17) resulted in an equipotent inhibitor. Substitu-
tion of the AX10185 phenylamide group with a 1-adamantyl amide
(18) was also well tolerated. When compared to the cyclopentyl
amide 16, the linear n-pentyl amide 19 lost potency to yield an
inhibitor with a micromolar IC50 value. Homologation of the amide
aryl group of compound 8 also resulted in a loss of potency (com-
pare compounds 8 and 20). Finally, the piperidinyl amide 21 was
found to be the poorest inhibitor in this study. Based on this analy-
sis, a secondary amide derived from a cyclic aryl or saturated amine
was identified as a preferred pharmacophore.
a
Determined in the presence of 10 mM EDTA.
no added metal
Cd(II)Cl2
Fe(III)Cl3
Co(II)Cl2
Mn(II)Cl2
Ca(II)Cl2
Mg(II)Cl2
Ba(II)Cl2
Ni(II)Cl2
php-Lp-PLA2
rh-Lp-PLA2
Zn(II)Cl2
0
0
0
100
200
3
Lp-PLA2 IC50 (nM)
compound 1
Figure 1. Influence of 50 lM added metal chloride salt on the php-Lp-PLA2 and
rhLp-PLA2 IC50 values of AX10185 (compound 1).
of increasing amounts of added ZnCl2 (Fig. 2). In the absence of
added ZnCl2, the rh-Lp-PLA2 IC50 was determined to be 96 nM.
With added ZnCl2 this value dropped as low as 33 nM, with the
SAR gleaned from the rh-Lp-PLA2 assay (Tables 2 and 3) was in
line with that seen in the php-Lp-PLA2 IC50 assay. While there was
maximal inhibitory activity observed at
a concentration of

870
E. C. K. Lin et al. / Bioorg. Med. Chem. Lett. 22 (2011) 868–871
Table 2
Lp-PLA2 IC50 values for AX10185 and 6–14 where php-Lp-PLA2 = Lp-PLA2 purified from human plasma and rh-Lp-PLA2 = recombinant human Lp-PLA2
Compound
R2
R3
R4
php-Lp-PLA2 IC50 (nM)
rh-Lp-PLA2 IC50 (nM)
ABP Lp-PLA2 IC50 (nM)
AX10185
6
7
8
H
H
H
H
H
H
Me
F
H
H
H
H
H
Me
H
H
F
H
74
27
270
61
n.d.
130
53
16
30
154
24
59
MeO
Cl
F
Me
H
H
F
H
F
4100
180
31
560
97
32
96
55
23
n.d.
n.d.
82
n.d.
64
9
10
11
12
13
14
44
177
n.d.
48
F
H
F
n.d. = not determined.
Table 3
Lp-PLA2 IC50 values for compounds 15–21 where php-Lp-PLA2 = Lp-PLA2 purified from human plasma and rh-Lp-PLA2 = recombinant human Lp-PLA2
Compound
X
php-Lp-PLA2 IC50 (nM)
98
rh-Lp-PLA2 IC50 (nM)
34
ABP Lp-PLA2 IC50 (nM)
48
15
16
17
170
92
n.d.
n.d.
n.d.
n.d.
18
19
88
56
n.d.
n.d.
11,000
360
20
1200
91
n.d.
n.d.
21
>200,000
2900
n.d. = not determined.
a modest correlation between the IC50 values in both assays
(R2 = 0.71), the general trends were the same. For example, the
ortho-methyl and 3,4-difluoro-phenyl amide derivatives 11 and
14 were identified as the most potent Lp-PLA2 inhibitors in both
assays. Conversely, the linear n-pentyl and piperidinyl amides 19
and 21 were identified as the least potent Lp-PLA2 inhibitors in
both assays. One difference to be noted between the assays was
that consistently lower IC50 values were seen in the recombinant
Lp-PLA2 version. This difference as well as any subtle SAR differ-
ences between the assays could be attributed to the addition of
zinc to the recombinant Lp-PLA2 assay and/or differences between
the native and recombinant enzymes.
To rule out the participation of any thiol-derived small molecule
components used in the substrate-based assay (e.g., 2-thio-PAF or
DNTB) on the observed Zn2+ dependent inhibition, we developed
a substrate-free assay12 that relied on the inhibition of rh-Lp-PLA2
labeling by AX4870, a TAMRA tagged, fluorophosphonate activity-
based probe (ABP). Such probes have been shown to covalently
and irreversibly label the active site of serine hydrolases, including
lipases,13 and have previously been used to determine IC50 values

E. C. K. Lin et al. / Bioorg. Med. Chem. Lett. 22 (2011) 868–871
871
for small molecule enzyme inhibitors.14 For selected potent Lp-PLA2
inhibitors, ABP IC50 values (see Tables 2 and 3) were calculated by
plotting the % inhibition of AX4870 labeling of rh-Lp-PLA2 versus
inhibitor concentration. These experiments were conducted in a
buffer identical to that which was used in the rh-Lp-PLA2 IC50 assay
catalysts for peptide bond formation, see: Bienert, M.; Henklein, P.; Beyermann,
M.; Carpino, L. A. In Synthesis of Peptides and Peptidomimetics; Goodman, M.,
Ed.; Thieme Stuttgart: New York, 2004; Vol. 1, pp 555–581.
6. php-Lp-PLA2 IC50 assay: Lp-PLA2 purified from human plasma (Cayman
Chemical, Ann Arbor, MI) was diluted 1:25 into an assay buffer of phosphate-
buffered saline (PBS) containing 0.5 mM 5,50-dithio-bis(2-nitrobenzoic acid)
(DTNB). Sixteen serial dilutions of inhibitor were prepared at
a
50ꢁ
(20 lM ZnCl2/PBS). All inhibitors tested in the ABP assay exhibited
concentration in DMSO and added to the Lp-PLA2 to a final 1ꢁ concentration
for a preincubation period of 20 min at room temperature. The substrate 2-
thio-PAF (Cayman Chemical) was prepared by evaporation of the ethanol stock
solution under nitrogen which was then reconstituted by adding PBS and
vortexing to make a 4 mM stock solution. The substrate was added to the assay
to a final concentration of 0.4 mM and incubated for 1 h at room temperature
before reading the absorbance at 405 nM. Absorbance versus inhibitor
concentration was plotted and fit to the Hill equation to determine IC50
values using the curve-fitting software Prism (GraphPad Software, Inc.). No
zinc was added to the assay.
inhibitory activity that paralleled that seen in the substrate-based
rh-Lp-PLA2 IC50 assay. The lack of thiol-containing components in
the ABP assay format thus rules out any significant contribution
of thiol binding to the metal/inhibitor/enzyme complex.
In conclusion, the phenylamide of xanthurenic acid, AX10185,
was identified as a Zn2+-dependent inhibitor of Lp-PLA2 with a
php-Lp-PLA2 IC50 value of 74 nM. Efforts to optimize the amide
group of AX10185 yielded Lp-PLA2 inhibitors with improved po-
tency and demonstrated a defined structure–activity relationship
bearing the hallmarks of small-molecule enzyme inhibition. In
addition to opening the door to the development of novel metal-
dependent Lp-PLA2 inhibitors, this work also demonstrates the
importance of considering the influence of metals on the en-
zyme/receptor binding of small molecules. Indeed, we had already
initiated lead optimization before the revelation that our inhibitors
were functioning in a metal-dependent manner. Without this dis-
covery, we would have moved forward unaware of the role that
zinc played in the mechanism of binding. Such Zn2+-mediated inhi-
bition of serine hydrolases has been described before. For example,
bis(5-amidino-2-benzimidazolyl)methane (BABIM) was identified
as an inhibitor of human trypsin with Ki values of 90 nM and
7. Stary, J.; Zolotov, A.; Petrukhin, O. M. In Critical Evaluation of Equilibrium
Constants Involving 8-Hydroxyquinoline and its Metal Chelates; Pergamon Press:
New York, 1979. First ed..
8. The concentration of background zinc in the php-Lp-PLA2 IC50 assay (i.e., php-
Lp-PLA2 + DNTB + PBS) was determined to be 1.2
spectroscopy (Columbia Analytical Services).
lM using atomic absorption
9. rh-Lp-PLA2 IC50 assay: rh-Lp-PLA2 (recombinant N-terminal hexahistidine-
tagged human protein purified from E. coli, Cayman Chemical, purity P95%)
was diluted to 0.25
with or without 20
l
g/ml in an assay buffer of PBS containing 0.5 mM DTNB
Zn(II)Cl2 (Sigma, St. Louis, MO). IC50 values for
l
M
inhibitors were assayed as described above for php-Lp-PLA2.
10. (a) Bloxam, D. L.; Tan, J. C.; Parkinson, C. E. Clin. Chim. Acta 1984, 144, 81; (b
Katsuhiko, Y.; Egger, N. G.; Ramanujam, V. M. S.; Alcock, N. W.; Dayal, H. H.;
Penland, J. G.; Sandstead, H. H. Am. J. Physiol. Endocrinol. Metab. 2003, 285,
E1010; (c) Masuoka, J.; Hegenauer, J.; Van Dyke, B. R.; Saltman, P. J. Biol. Chem.
1993, 268, 21533.
11. Bag, S. P.; Fernando, Q.; Freiser, H. Inorg. Chem. 1964, 3, 93.
12. rh-Lp-PLA2 activity-based probe IC50 assay: rh-Lp-PLA2 (Cayman Chemical)
was diluted to 10 lg/ml in an assay buffer of PBS with 20 lM Zn(II)Cl2. Eight
19 l
M in the presence of Zn2+ and EDTA, respectively.15 The benzi-
serial dilutions of inhibitor were made at a 10ꢁ concentration in DMSO, and
added to the Lp-PLA2 to a final 1ꢁ concentration for a pre-incubation period of
20 min at room temperature. The fluorescent, active-site-directed probe
AX4870 was added to a final concentration of 500 nM and incubated for
2 min at room temperature prior to quenching the probe labeling reaction with
Laemmli sample buffer. The samples were run on a 12.5% SDS–PAGE gel and
fluorescent signal corresponding to rh-Lp-PLA2 was detected on an FMBIO II
(Hitachi) flat bed scanner. The fluorescent signal was used to calculate
inhibitor dose–response curves using Prism.
midazole groups of BABIM were shown via X-ray crystallography
to coordinate to Zn2+ which in turn was complexed to two active
site residues of the catalytic triad, Ser195 and His57. Future work
on the continued lead optimization of AX10185 and additional me-
tal binding studies will be reported in due course.
References and notes
1. Caslake, M. J.; Packard, C. J.; Suckling, K. E.; Holmes, S. D.; Chamberlain, P.;
Macphee, C. H. Atherosclerosis 2000, 150, 413.
2. (a) Anderson, J. L. Am. J. Cardiol. 2008, 101, 23F; (b) Corson, M. A.; Jones, P. H.;
Davidson, M. H. Am. J. Cardiol. 2008, 101, 41F.
3. (a) Racherla, S.; Arora, R. Am. J. Ther. 2010 epub ahead of print.; (b) Wilensky, R.
L.; Shi, Y.; Mohler, E. R., III; Hamamdzic, D.; Burgert, M. E.; Li, J.; Postle, A.;
Fenning, R. S.; Bollinger, J. G.; Hoffman, B. E.; Pelchovitz, D. J.; Yang, J.; Mirabile,
R. C.; Webb, C. L.; Zhang, L.; Zhang, P.; Gelb, M. H.; Walker, M. C.; Zalewski, A.;
Macphee, C. H. Nat. Med. 2008, 14, 1059; (c) Macphee, C. H. Curr. Opin.
Pharmacol. 2001, 1, 121; (d) Leach, C. A.; Hickey, D. M.; Ife, R. J.; Macphee, C. H.;
Smith, S. A.; Tew, D. G. Farmaco 2001, 56, 45.
4. Serruys, P. W.; García-García, H. M.; Buszman, P.; Erne, P.; Verheye, S.;
Aschermann, M.; Duckers, H.; Bleie, O.; Dudek, D.; Bøtker, H. E.; von Birgelen,
C.; D’Amico, D.; Hutchinson, T.; Zambanini, A.; Mastik, F.; van Es, G. A.; Van der
Steen, A. F.; Vince, D. G.; Ganz, P.; Hamm, C. W.; Wijns, W.; Zalewski, A.
Circulation 2008, 118, 1172.
13. (a) Liu, Y.; Patricelli, M. P.; Cravatt, B. F. Proc. Natl. Acad. Sci. USA 1999, 96,
14694; (b) Simon, G. M.; Cravatt, B. F. J. Biol. Chem. 2010, 285, 11051.
14. Ahn, K.; Johnson, D. S.; Fitzgerald, L. R.; Liimatta, M.; Arendse, A.; Stevenson, T.;
Lund, E. T.; Nugent, R. A.; Nomanbhoy, T. K.; Alexander, J. P.; Cravatt, B. F.
Biochemistry 2007, 46, 13019.
15. (a) Katz, B. A.; Luong, C. J. Mol. Biol. 1999, 292, 669; (b) Katz, B. A.; Clark, J. M.;
Finer-Moore, J. S.; Jenkins, T. E.; Johnson, C. R.; Ross, M. J.; Luong, C.; Moore, W.
R.; Stroud, R. M. Nature 1998, 391, 608.
5. For an overview on the use of uronium salts, including HATU (O-(7-
azabenzotriazol-1-yl)-N,N,N0,N0-tetramethyluronium hexafluorophosphate), as