The identification of a potent, water soluble inhibitor of lipoprotein-associated phospholipase A2

Article abstract of DOI:10.1016/S0960-894X(01)00038-5

Modification of the pyrimidone 5-substituent in a series of 1-((amidolinked)-alkyl)-pyrimidones, lipophilic inhibitors of lipoprotein-associated phospholipase A₂, has given inhibitors of nanomolar potency and improved physicochemical properties

Full text of DOI:10.1016/S0960-894X(01)00038-5

Bioorganic & Medicinal Chemistry Letters 11 (2001) 701±704
The Identi®cation of a Potent, Water Soluble Inhibitor of
Lipoprotein-Associated Phospholipase A2
Helen F. Boyd, Beverley Hammond, Deirdre M. B. Hickey, Robert J. Ife,
Colin A. Leach, V. Ann Lewis, Colin H. Macphee, Kevin J. Milliner, Ivan L. Pinto,
Stephen A. Smith,* Ian G. Stans®eld, Colin J. Theobald and Caroline M. Whittaker
GlaxoSmithKline Pharmaceuticals, New Frontiers Science Park, Third Avenue, Harlow, Essex, CM19 5AW, UK
Received 28 November 2000; revised 8 January 2001; accepted 10 January 2001
AbstractÐModi®cation of the pyrimidone 5-substituent in a series of 1-((amidolinked)-alkyl)-pyrimidones, lipophilic inhibitors of
lipoprotein-associated phospholipase A2, has given inhibitors of nanomolar potency and improved physicochemical properties.
Compound 23 was identi®ed as a potent, highly water soluble, CNS penetrant inhibitor suitable for intravenous administration.
# 2001 Elsevier Science Ltd. All rights reserved.
Current therapy for atherosclerosis is broadly based on
the regulation of plasma lipid levels, particularly LDL
cholesterol. The statins, although only really e€ective in
around 30% of patients, have achieved medical and
commercial success in this role.1 Therapies which
directly in¯uence atherosclerotic plaque formation and
stability are less well precedented and represent an
exciting newopportunity to treat many more of the at
risk population. To this end, we have focused our atten-
tion on a novel serine dependant lipaseÐlipoprotein-
associated phospholipase A2 (Lp-PLA2)2Ðthat is able
to hydrolyse oxidatively modi®ed phosphatidylcholines
to release oxidised fatty acids and lysophos-
phatidylcholine (lyso-PC). Both of these hydrolysis
products are known to be pro-in¯ammatory and have
been implicated in atherosclerosis.3 Furthermore, a
recent study has shown a strong, positive correlation
between Lp-PLA2 levels and coronary events in asymp-
tomatic, hypercholesterolemic men and suggested that
Lp-PLA2 is a new, independent marker of coronary
heart disease risk.4 The identi®cation of inhibitors of
Lp-PLA2 would then aid our evaluation of the role of
this enzyme in atherosclerosis and additionally, could
also be of potential value in the treatment of other
in¯ammatory vascular diseases involving oxidative
stress (e.g., stroke).
inhibitors of Lp-PLA2 which showed activity in the
Watanabe hereditable hypolipodaemic rabbit (WHHL
rabbit). Those inhibitors were however rather lipophilic
and, as a result, poorly water soluble. In this commu-
nication we describe our studies towards the identi®ca-
tion of less lipophilic inhibitors via modi®cation of the
pyrimidone 5-substituent.
Compounds 3 were prepared via the acylisothiocyanate
2 in an analogous manner to that previously described
(Scheme 1).5 O-Demethylation of 3 was achieved with
B-bromocatecholborane6 to give compound 4 in high
yield. Other dealkylation conditions proved less e€ec-
tiveÐboron tribromide for example gave a less clean
product that was dicult to purify. N-Alkylation with
ethyl bromoacetate and hydrolysis gave acid 5 (Y=OH)
which was subsequently converted to the amide with
EDC/HOBT. Other N-alkylations of compound 4 were
performed similarly. All compounds in Tables 1 and 2
were evaluated using human Lp-PLA2 (hLp-PLA2).
Recently5 we described the identi®cation of a series of 1-
((amidolinked)-alkyl)-pyrimidones 1, as highly potent
*Corresponding author. Fax:+44-1279-627841; e-mail: stephen_1_
smith@sbphrd.com
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00038-5

702
H. F. Boyd et al. / Bioorg. Med. Chem. Lett. 11 (2001) 701±704
Assays were performed in duplicate.7 In order to factor
in any non-speci®c binding e€ects in plasma, com-
pounds were also assessed against the plasma enzyme in
both whole human and WHHL rabbit plasma at a single
concentration of inhibitor.5 Good activity in rabbit and
human plasma was required before compounds were
evaluated in vivo in either WHHL rabbits5 or in the rat.8,9
philic, but often more potent, (N-dodecyl-N-methyl-
carbonyl)methyl group. Activity versus hLp-PLA2 was
retained or enhanced on substitution of the (pyrimidin-
5-yl)methyl group at the pyrimidine 2-position. It is
interesting to note that both lipophilic and more
hydrophilic substituents are well tolerated (cf. 7 vs 10±
15 and 9 vs 16±18).
We quickly focused our attention (Table 1) on tertiary
acetamides, as their secondary counterparts (e.g., 6
and 8) proved very insoluble and as a result, dicult
to evaluate fully. Two N-1 substituents were chosen:
(N-methyl-N-octylcarbonyl)methyl and the more lipo-
As a result of the encouraging activity in whole plasma
for the more polar 2-oxopyrimidines 13, 15, 17 and 18,
we decided to investigate substitution of the 2-oxopyr-
imidine ring with more polar substituents (Table 2).
Very encouragingly, high potency versus hLp-PLA2 was
Scheme 1. Reagents: (i) (a) CH2ˆCHCO2Et, Pd(OAc)2, P(o-Tol)3, Et3N, (b) H2, Pd/C, EtOH/Et3N, (c) HCO2Et, NaH, DME, (d) Me2SO4, K2CO3,
DMF, (e) NaOH, H2O; (ii) (a) (COCl)2, ClCH2CH2Cl, (b) KSCN, CH3CN; (iii) (a) R1NH2, DMF then NaOEt, (b) R2Cl/Br, iPr2NEt, CH2Cl2;
(iv) B-bromocatecholborane, CH2Cl2; (v) (a) BrCH2CO2Et, K2CO3, DMF, (b) NaOH, H2O/dioxan, (c) R3R4NH, EDC, HOBT, CH2Cl2.
Table 1. Pyrimidone 5-substituent variation
Inhibition in plasma (%) @ 100 nM
No.a
Ra
Rb
Ar
IC50 nM
Human
Rabbit
6
7
8
9
10
11
12
13
14
15
16
17
18
n-C8H17
n-C8H17
n-C12H25
n-C12H25
n-C8H17
n-C8H17
n-C8H17
n-C8H17
n-C8H17
n-C8H17
n-C12H25
n-C12H25
n-C12H25
H
Me
H
Pyrimidin-5-yl
Pyrimidin-5-yl
Pyrimidin-5-yl
7
15
0.3
1
4
3
43
40
86
79
39
23
10
66
55
48
81
83
75
38
30
54
39
13
7
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Pyrimidin-5-yl
2-MeO-pyrimidin-5-yl
2-EtO-pyrimidin-5-yl
2-PhCH2O-pyrimidin-5-yl
2-Oxo-pyrimidin-5-yl
1-Me-2-oxo-pyrimidin-5-yl
1-Et-2-oxo-pyrimidin-5-yl
2-MeO-pyrimidin-5-yl
2-Oxo-pyrimidin-5-yl
1-Me-2-oxo-pyrimidin-5-yl
7
4
8
34
19
5
29
39
32
27
20
1
0.9
5
aAll newcompounds gave satisfactory analytical/spectral data. 10

H. F. Boyd et al. / Bioorg. Med. Chem. Lett. 11 (2001) 701±704
Table 2. E€ect of substitution in C-5 pyrimidone ring
703
Inhibition in plasma (%) @ 100 nM
No.a
Ra
Rc
IC50 (nM)
Human
Rabbit
13
19
20
21
17
22
23
24
25
26
27
28
29
30
31
32
33
34
35
n-C8H17
n-C8H17
n-C8H17
n-C8H17
n-C12H25
n-C12H25
n-C12H25
n-C12H25
n-C8H17
n-C8H17
n-C8H17
n-C8H17
n-C8H17
n-C8H17
n-C8H17
n-C12H25
n-C12H25
n-C12H25
n-C12H25
H
CH2CO2Et
CH2CO2H
CH2CH2OH
H
CH2CO2Et
CH2CO2H
CH2CH2OH
CH2CONHMe
CH2CONHPr
4
75
14
5
0.9
4
1
1
6
11
8
66
NT
56
59
83
55
74
77
63
42
68
64
68
80
80
80
79
80
86
34
NT
27
36
39
15
28
36
42
16
40
36
40
55
49
28
37
30
49
CH2CONHCH2CH2OH
CH2CO(morpholin-4-yl)
CH2CO(1-Me-piperazin-4-yl)
CH2CO(2-oxo-piperazin-4-yl)
CH2CO(1-acetyl-piperazin-4-yl)
CH2CONHMe
CH2CONHCH2CH2OH
CH2CO(morpholin-4-yl)
CH2CO(2-oxo-piperazin-4-yl)
10
3
2
5
0.4
0.6
0.7
0.4
aAll newcompounds gave satisfactory analytical/spectral data. 10
maintained on the introduction of a wide range of sub-
stituents. Indeed, acidic, polar neutral and basic groups
were all well tolerated, suggesting that the substituent
may be directed away from the enzyme and towards the
surrounding aqueous environment. As has been pre-
viously observed, potency is higher in human than rab-
bit plasma suggesting that compounds may be more
potent in man than in the rabbit.
ꢀ85% liver blood ¯ow)8 and as a result little oral
activity/systemic exposure. We postulated that this pro-
®le, alongside the ease of formulating 20 and 23 for
intravenous dosing, would give these acids the correct
characteristics as drugs to be administered by infusion
dosingÐgood control of enzyme inhibition should be
achieved as levels of inhibitor would fall rapidly after
infusion is terminated. As a result, we proceeded to an
infusion study that also included measurement of CNS
penetration in the rat.9 Compounds 9, 13, 25 and 34,
although more dicult to formulate, were included in
this study for comparison. We were very pleased to
showthat, in contrast to the less soluble analogues,
both 20 and particularly 23 proved to be CNS penetrant
(Table 3). Furthermore, sampling during the 6±8 h time
period of the infusion, indicated that good blood levels
of 23 had been achieved (steady-state attained) along-
side a high level of inhibition of rat Lp-PLA2Ða dose of
2 mmol/kg/h gave a blood concentration of 533Æ50 nM
and 83% inhibition of rat Lp-PLA2.
Selected compounds from Tables 1 and 2 (16, 20, 23 and
34) were also screened against the most closely related
phospholipase A2 (human serine dependent-PLA2).11
Encouragingly, all compounds tested showed >1000-
fold selectivity for Lp-PLA2 over this lipase.
Whilst a number of compounds in Table 2 showed some
improvement in solubility over the parent pyrimidines
(7 and 9), our attention quickly focused on the acetic
acid derivatives 20 and 23 whose potency was matched
by high solubility (>15 mg/mL) in normal saline at
pH 7.4. Based on these data, it was decided to evaluate
compounds 20 and 23 in vivo in order to assess oral
availability, their use as intravenous agents and, with a
Table 3. CNS penetration results
No.
CNS penetration (%) (n)
viewto assessing the role of Lp-PLA
in centrally
2
mediated in¯ammatory vascular disease (e.g., stroke),
CNS penetration.
9
<2 (3)
7 (1)
10Æ1 (3)
37Æ8 (3)
<7 (1)
<7 (1)
13
20
23
25
34
Initial results in both the rat and WHHL rabbit indi-
cated a rather rapid clearance following intravenous
administration (e.g., 23: rat CLb=76 mL/min/kg,

704
H. F. Boyd et al. / Bioorg. Med. Chem. Lett. 11 (2001) 701±704
In conclusion, we have shown that high potency is
retained on modi®cation of the (pyrimidin-5-yl)methyl
group present in our previously described inhibitors.
Two of these new inhibitors, acetic acid derivatives 20
and 23, exhibit excellent water solubility at physio-
logical pH. Compound 23 shows an admirable pro®le
following infusion dosing, including CNS penetration,
and will be of great value in evaluating the role of Lp-
PLA2 in atherosclerosis and other in¯ammatory vas-
cular diseases involving oxidative stress (e.g., stroke) in
situations that require administration by this route.
6. King, P. F.; Stroud, S. G. Tetrahedron Lett. 1985, 26,
1415.
7. Boyd, H. F.; Flynn, S. F.; Hickey, D. M. B.; Ife, R. J.; Jones,
M.; Leach, C. A.; Macphee, C. H.; Milliner, K. J.; Rawlings,
D. A.; Slingsby, B. P.; Smith, S. A.; Stans®eld, I. G.; Tew, D. G.;
Theobald, C. J. Bioorg. Med. Chem. Lett. 2000, 10, 395.
8. Oral bioavailability and clearance parameters were deter-
mined (e.g., for the sodium salt of 23) by non-compartmental
pharmacokinetic analysis following iv infusion over 1 h (in
saline 0.9% w/v) at a target dose of 1.6 mmol/kg/h and oral
gavage administration (3 mmol/kg in distilled water) in the
conscious, cannulated male rat. Serial blood samples were
collected over 10 h post dose and analysed by LC/MS/MS.
9. CNS penetration at steady-state was investigated in the rat.
Compounds were dissolved in 2% (v/v) DMSO, 2% ethanol
and 10% EncapsinTM in saline and administered at a constant
rate infusion over 8 h at a target dose of 2 mmol/kg/h. Blood
samples were removed over the last 2 h of the infusion to
con®rm steady-state concentrations. Blood and brain samples
were analysed by LC/MS/MS. For compound 23, inhibition of
Lp-PLA2 was determined by a method similar to that descri-
bed in ref 5.
10. Representative examples: Compound 10 1H NMR
(CDCl3) d 0.8±0.95 (3H, m), 1.1±1.7 (12H, m), 2.95 and 2.99
(3H, 2Âs), 3.21 and 3.36 (2H, 2Ât), 3.66 (2H, s), 3.99 (3H, s),
4.48 (2H, s), 4.51 and 4.55 (2H, d), 6.80 (1H, s), 6.9±7.1
(2H, m), 7.3±7.45 (2H, m), 8.45 (2H, s); MS (APCI+) found
(M+1)=542; C28H36FN5O3S requires 541. Compound 23
(250 MHz) 1H NMR (DMSO-d6) d 0.85 (3H, t), 1.22 (18H,
m), 1.35±1.61 (2H, m), 2.78, 2.95 (3H, 2Âs), 3.20±3.35 (2H,
m), 3.58 (2H, s), 4.40 (2H, s), 4.55 (2H, s), 4.83 (2H, m), 7.10
(2H, m), 7.44 (2H, m), 7.54 and 7.57 (1H, 2Âs), 8.03 (1H, m),
8.54 (1H, m) 13.10 (1H, br. s); MS (APCI+) found
(M+1)=642, C33H44FN5O5S requires 641.
References
1. Farnier, M.; Davignon, J. Am. J. Cardiol. 1998, 82(4B), 3J
and references therein.
2. Also known as PAF acetyl hydrolase, this enzyme has
much broader substrate speci®city than this name implies. See
Tew, D. G.; Southan, C.; Rice, S. Q. J.; Lawrence, M. P.;
Haodong, L.; Boyd, H. F.; Moores, K.; Gloger, I. S.;
Macphee, C. H. Atheroscler. Thromb. Vasc. Biol. 1996, 16(4),
591.
3. Macphee, C. H.; Moores, K. E.; Boyd, H. F.; Dhanak, D.;
Ife, R. J.; Leach, C. A.; Leakes, D. S.; Milliner, K. J.; Patter-
son, R. A.; Suckling, K. E.; Tew, D. G.; Hickey, D. M. B.
Biochem. J. 1999, 338, 479 and references therein.
4. Packard, C. J.; O'Reilly, D. St. J.; Caslake, M. J.; McMa-
hon, A. D.; Ford, I.; Cooney, J.; Macphee, C. H.; Suckling,
K. E.; Krishna, M.; Wilkinson, F. E.; Rumley, A.; Lowe,
G. D. O. N. Engl. J. Med. 2000, 434, 1148.
5. Boyd, H. F.; Fell, S. C. M.; Flynn, S. T.; Hickey, D. M. B.;
Ife, R. J.; Leach, C. A.; Macphee, C. H.; Milliner, K. J.;
Moores, K. E.; Pinto, I. L.; Porter, R. A.; Rawlings, D. A.;
Smith, S. A.; Stans®eld, I. G.; Tew, D. G.; Theobald, C. J.;
Whittaker, C. M. Bioorg. Med. Chem. Lett. 2000, 10, 2557.
11. Rice, S. Q. J.; Southan, C.; Boyd, H. F.; Terrett, J. A.;
Macphee, C. H.; Moores, K.; Tew, D. G. Biochem. J. 1998,
330, 1309.