Structure-activity relationship on human serum paraoxonase (PON1) using substrate analogues and inhibitors

Article abstract of DOI:10.1016/S0960-894X(03)00290-7

Substrate analogues based on the parent compounds paraoxon and phenyl acetate were tested on human serum paraoxonase (PON1) to explore the active site of the enzyme. Replacement of the nitro group of paraoxon with an amine or hydrogen, as well as electronic changes to the parent compound, converted these analogues into inhibitors. Introduction of either electron-withdrawing or donating groups onto phenyl acetate resulted in reduction in their rate of hydrolysis by PON1.

Full text of DOI:10.1016/S0960-894X(03)00290-7

Bioorganic & Medicinal Chemistry Letters 13 (2003) 1623–1626
Structure–Activity Relationship on Human Serum Paraoxonase
(PON1) Using Substrate Analogues andInhibitors
Rakesh S. Bargota,a Mahmoud Akhtar,a Keith Biggadike,b David Gania
and Rudolf K. Allemanna,*
aSchool of Chemical Sciences, The University of Birmingham, Edgbaston, Birmingham B15 2TT, UK
bGlaxoSmithKline, Medicines Research Centre, Gunnels Wood Road, Stevenage, Hertfordshire SG1 2NY, UK
Received 7 February 2003; revised 17 March 2003; accepted 19 March 2003
Abstract—Substrate analogues based on the parent compounds paraoxon and phenyl acetate were tested on human serum para-
oxonase (PON1) to explore the active site of the enzyme. Replacement of the nitro group of paraoxon with an amine or hydrogen,
as well as electronic changes to the parent compound, converted these analogues into inhibitors. Introduction of either electron-
withdrawing or donating groups onto phenyl acetate resulted in reduction in their rate of hydrolysis by PON1.
# 2003 Elsevier Science Ltd. All rights reserved.
Human serum paraoxonase (PON1, EC 3.1.8.1) is a
calcium-dependent glycoprotein with hydrolase activ-
ity.1 The enzyme derives its name from its ability to
hydrolyse paraoxon 1a into p-nitrophenol and diethyl
phosphate, a reaction that was first demonstrated by
Aldridge in 1953 (Scheme 1).2 Paraoxon, a potent acetyl-
cholinesterase inhibitor, is metabolically generated in
vivo from the insecticide parathion by mitochondrial
oxidation involving the cytochrome-P450 pathway.3
PON1 is synthesised in the liver and then secreted into
the bloodstream where it becomes tightly bound to
high-density lipoprotein (HDL).4 The enzyme catalyses
the hydrolysis of a broad range of substrates including
arylesters 2a,5 carbamates,6 as well as cyclic carbonates
and lactones 3 (Scheme 1).7 It has been shown that the
incorporation of the latter two groups as hydrolysable
moieties onto a glucocorticosteroid nucleus generates an
ideal antedrug strategy that could offer safer treatment
of asthma.7 This antedrug strategy is possible because of
the tissue-specific distribution of PON1. Its presence in the
blood and liver renders the drug inactive. However, PON1
activity is absent in the lungs, where stability of the com-
pound is required for maximal efficacy. PON1 can also
activate an unsaturated cyclic carbonate prodrug of qui-
nolone antibiotic in the plasma8 and the enzyme can
detoxify nerve agents, such as sarin 49 (Scheme 1).
PON1 is a 354 amino acid polymorphic protein of 43–
45k Da with amino acid substitutions at positions 54 (L
or M) and 191 (Q or R) giving rise to several isozymes
in the human blood which display marked differences in
activities towards some substrates.10 For example, the
differential rates of detoxification of warfare agents by
Q, R and QR phenotypes has been linked to suscept-
ibility of soldiers to developing the Gulf War Syn-
drome.11
Over the last decade it has become apparent that PON1
also performs an important physiological function by
retarding the oxidative modification of low-density
lipoprotein (LDL) in the arterial wall. This oxidation is
thought to be the initial step in the cascade of events
which eventually lead to atherosclerosis.12 The R/L
phenotype has been associated with a risk of developing
cardiovascular disease, while in populations with the Q/
M phenotype, a lower risk of atherosclerosis has been
observed,13 although this is still controversial.14
As yet, the natural substrate of PON1 remains in doubt,
but a recent report suggests that it might be platelet-
activating factor (PAF).15 This phospholipid is a good
substrate for PON1, with a catalytic efficiency 750-fold
greater than that for paraoxon. Interestingly, PAF
seems to be hydrolysed in the absence of Ca2+ ions in
vitro, possibly suggesting the presence of two different
active sites on the enzyme. At present, little is known
about how PON1 processes any of its substrates.
*Corresponding author. Tel.:+44-121-4144359; fax:+44-121-
4144446; e-mail: r.k.allemann@bham.ac.uk
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00290-7

1624
R. S. Bargota et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1623–1626
limited to substrates and products exhibiting chromo-
phoric groups with different spectral properties. We
have now developed a novel PON1 assay system that
employs NMR spectroscopy to monitor continuous
enzyme-catalysed reactions.20 This system has none of
the limitations of spectrophotometric techniques and it
can be used to generate qualitative and quantitative
data on novel compounds.
Using this assay protocol, we tested the hydrolysis of
our compounds with serum PON1 by measuring the
relative changes of reactants and products over time.
From this data, the time taken for 50% of the test
compound to be consumed by serum PON1 was deter-
mined, which is reported as t1/2 (Table 1). A wide range
of activities was observed. Substrate 2a was hydrolysed
most rapidly (t1/2 1.5 min), while the rate of hydrolysis
of the phosphate ester 1a was very low (t1/2 1460 min).
No hydrolysis was detected with compounds 1b, 1c, 5a
and 5b even after 24 h incubation with serum. Control
experiments indicated that diluted serum [10% (v/v)]
was stable at 30 ꢀC for up to 21 h without loss of PON1
activity. Introduction of the electron-withdrawing nitro-
group to the para-position of the parent ester 2a slowed
down the hydrolysis of the resultant molecule 2b more
than 2-fold. Conversely, the presence of NO2 group was
essential for the hydrolysis of 1a because its removal
from the phenyl ring rendered the analogue 1c stable.
Attachment of electron-donating groups (NH2/OMe) to
Scheme 1. Some known enzymatic reactions of PON1 with synthetic
substrates.
Reports concerning the relationship between structure
and activity of PON1 using synthetic substrates are
scarce, but recently lactone substrate specificity and
inhibition by lactams on PON1 has been investigated.16
Previously, plasma arylesterase activity was quantita-
tively related to structural effects of various aryl and
aliphatic esters.17 Studies that offer a mechanistic and
catalytic insight into the action of PON1 could provide
important clues to the in vivo function of the enzyme,
thereby leading the way to new antedrug and prodrug
strategies.
We report here results from the use of substrate analo-
gues based on the parent compounds paraoxon 1a and
phenyl acetate 2a (Scheme 2) to define the relation
between the structure and the function of human serum
PON1. Analogues 1b, 1c, 2c, and 2d were synthesised in
our laboratory (Scheme 3) whilst others were commer-
cially available.
Our initial approach was to examine the time course of
hydrolysis of compounds 1a–1c, 2a–2d, 5a and 5b using
whole serum, which had been pre-treated with phe-
nylmethylsulfonyl fluoride (PMSF) to inactivate other
esterases within it, particularly acetylcholinesterase
which is known to have arylesterase activity.1 We have
demonstrated that 4 mM PMSFcauses a 20% reduc-
tion in arylesterase activity, using 2a as substrate, which
is primarily due to inhibition of acetylcholinesterase
(unpublished data). We further confirmed this by
showing that in the presence of serum treated with
PMSF, no significant hydrolysis of acetylcholine occur-
red (data not shown). All substrates and analogues were
tested under identical assay conditions with fixed
amounts of compound and serum. Although crude
serum has previously been used to measure PON1
activities spectrophotometrically,18 this technique is
Scheme 3. Reagents and conditions for the synthesis of substrate
analogues of PON1. All final compounds exhibited satisfactory an-
alytical data in accordance with the structures proposed.19
Table 1. Effect of substrates and their analogues on PON1
Compd
t1/2 (min)
IC50 (mM)
2a
2b
2c
2d
1a
1b
1c
5b
5a
1.5
3.5
20
20
1460
nh
nh
nh
nh
na
na
na
na
na
1.28
3.27
3.25
18.5
IC50 values were determined from plots of enzyme activity versus
inhibitor concentration using GraFit (Erithacus Software, Staines,
UK). na=not active, nh=not hydrolysed.
Scheme 2. Substrate analogues of PON1.

R. S. Bargota et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1623–1626
1625
the para position of 2a caused over 13-fold reduction in
the rate of hydrolysis (2c and 2d, respectively). Sub-
stitution of the NO2-group of paraoxon 1a with a NH2-
group rendered the substrate analogue 1b non-hydro-
lysable by serum PON1. Because compounds 1b, 1c, 5a
and 5b did not serve as substrates for serum PON1, we
tested whether these acted as inhibitors of the enzyme.21
All four derivatives caused varying degrees of inhibition
with 1b, 1c, and 5b displaying IC50 values between 1.28
and 3.27 mM (Table 1), whilst 5a was the weakest inhi-
bitor in the series (IC50 of 18.5 mM).
mode of inhibition of the paraoxonase derivatives also
needs to be evaluated. Data from these concerted efforts
will further our knowledge of the active site of PON1
and pave the way towards understanding the catalytic
mechanism of this physiologically important enzyme.
Acknowledgements
The authors thank GlaxoSmithKline for financial sup-
port, Dr. Serge Zeler for technical assistance, Dr. David
Miller for helpful discussions and critical reading of the
manuscript, and Dr. Gabra at the Birmingham Blood
Centre for donating surplus serum.
These studies provide curious observations associated
with substituent effects for carboxylic ester substrates of
PON1. Coupled with this study, we have previously
demonstrated that phenyl acetate 2a is a better substrate
than the 4-nitro analogue 2b for PON1 in that 2a shows
a 50-fold greater Vmax/Km value, despite both exhibiting
similar Km values of around 0.7 mM (unpublished
results). These findings suggest that the Lewis basicity of
the carbonyl oxygen is more important in stabilizing the
rate-limiting transition state than a low pKa for the
leaving group of structures 2a–2d. The enzyme could
stabilize the negative charge on oxygen through inter-
action with a metal ion such as Ca2+ or by protonation.
References andNotes
1. Gan, K. N.; Eckerson, H. W.; La Du, B. N. Drug. Metab.
Dispos. 1991, 19, 100.
2. Aldridge, W. N. Biochem. J. 1953, 53, 117.
3. La Du, B. N. In Pharmacogenetics of Drug Metabolism;
Kalow, W, Ed.; Pergamon: New York, 1992; p 51.
4. Mackness, B.; Durrington, P. N.; Mackness, M. I. Gen.
Pharmac. 1998, 31, 329.
5. Mackness, M. I.; Mackness, B.; Durrington, P. N.; Con-
nelly, P. W.; Hegele, R. A. Curr. Opin. Lipidol. 1996, 7, 69.
6. Sogorb, M. A.; Vilanova, E. Toxicol. Lett. 2002, 128, 215.
7. Biggadike, K.; Angell, R. M.; Burgess, C. M.; Farrel, R. M.;
Hancock, A. P.; Harker, A. J.; Irving, W. R.; Ioannou, C.;
Procopiou, P. A.; Shaw, R. E.; Solanke, Y. E.; Singh, O. M. P.;
Snowden, M. A.; Stubbs, R. J.; Walton, S.; Weston, H. E. J.
Med. Chem. 2000, 43, 19.
The reduction in the rate of hydrolysis caused by the
presence of amine or methoxy groups on phenyl acetate
was most likely due to steric hindrance. Substrate ana-
logues of paraoxon in the phosphate ester series (1b, 1c,
5a and 5b) were all inhibitors suggesting that they bind
to PON1, but the enzyme does not catalyze their
hydrolysis at a significant rate. These inhibitors either
had an overall negative charge or were hydrophobic,
both of which characteristics were detrimental to
hydrolysis by PON1. There is a free cysteine residue at
position 283 in PON1 which is not involved in aryl-
esteraser/paraoxon catalysis. Evidence for this came
from site-directed mutagenesis (SDM) experiments in
which cysteine 283 was replaced with either alanine or
serine and the resulting mutants retained both para-
oxonase and arylesterase activities.22 Other SDM stud-
ies have revealed that several histidine and tryptophan
residues are essential for PON1 arylesterase and para-
oxonase activities.23 The aromatic nature of these amino
acids suggests that the active site of the enzyme is
hydrophobic, which could also explain why PON1
favours lipophilic substrates, such as 2a and PAF. But,
paraoxon 1a, which itself has a polar NO2 group, is also
hydrolyzed by PON1, albeit at a much reduced rate,
suggesting that there may be a different mode of inter-
action with phosphate and carboxylic acid esters.
8. Tougou, K.; Nakamura, A.; Okuyama, Y.; Morino, A.
Drug. Metab. Dispos. 1998, 26, 355.
9. (a) For examples, see: Davies, H. G.; Richter, R. J.; Keifer,
M.; Broomfield, C. A.; Sowalla, J.; Furlong, C. E. Nat. Gen.
1996, 14, 334. (b) Costa, L. G.; Li, W. F.; Richter, R. J.; Shih,
D. M.; Lusis, A.; Furlong, C. E. Chem. Biol. Interact. 1999,
120, 429.
10. (a) For examples, see: Smolen, A.; Eckerson, H. W.; Gan,
K. N.; Hailat, N.; La Du, B. N. Drug. Metab. Dispos. 1991,
19, 107. (b) Mackness, B.; Mackness, M. I.; Arrol, S.; Turkie,
W.; Durrington, P. N. Br. J. Pharmacol. 1997, 112, 265.
11. Haley, R. W.; Billecke, S.; La Du, B. N. Toxicol. Appl.
Pharmacol. 1999, 227.
12. (a) For reviews and examples, see: Watson, A. D.; Berli-
ner, J. A.; Hama, S. Y.; La Du, B. N.; Faull, K. F.; Fogelman,
A. M.; Navab, M. J. Clin. Invest. 1995, 96, 2882. (b) Aviram,
M. Mol. Med. Today 1999, 5, 381. (c) Aviram, M.; Hardak, E.;
Vaya, J.; Mahmood, S.; Milo, S.; Hoffman, A.; Billicke, S.;
Draganov, D.; Rosenblat, M. Circulation 2000, 101, 2510. (d)
Durrington, P. N.; Mackness, B.; Mackness, M. I. Arter-
ioscler. Thromb. Vasc. Biol. 2001, 21, 473.
13. (a) Ruiz, J.; Blanche, H.; James, R. W.; Garin, M.-C.B.;
Vaisse, C.; Charpentier, G.; Cohen, N.; Morabia, A.; Passa,
P.; Froguel, P. Lancet 1995, 346, 869. (b) Mackness, B.;
Mackness, M. I.; Arrol, S.; Turkie, W.; Durrington, P. N.
FEBS Lett. 1998, 423, 57. (c) Mackness, B.; Mackness, M. I.;
Durrington, P. N.; Arrol, S.; Evans, A. E.; McMaster, D.;
Ferrieres, J.; Ruidavets, J.-B.; Williams, N. R.; Howard, A. N.
Eur. J. Clin. Invest. 2000, 30, 4. (d) Malin, R.; Jarvinen, O.;
Sisto, T.; Koivula, T.; Lehtimaki, T. Atherosclerosis 2001, 157,
301. (e) Gnasso, A.; Motti, C.; Irace, C.; Gennaro, D. I.; Pujia,
A.; Leto, E.; Ciamei, M.; Crivaro, A.; Bernardini, S.; Federici,
G.; Cortese, C. Atherosclerosis 2002, 164, 289.
In summary, we have demonstrated that both electron-
withdrawing and -donating groups in the para-position
of the parent phenyl acetate esters lead to significant
reductions in the rate of hydrolysis. Using the NMR
assay described here, it should be possible to determine
Km and Vmax for these and other substrate analogues of
PON1 using purified single isomorphs. The structure–
activity data generated could then be used to extra-
polate whether the analogues have an effect on the
enzyme through binding or catalysis. In addition, the
14. (a) Antikainen, M.; Murtomaki, S.; Syvanne, M.; Pahlman,

1626
R. S. Bargota et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1623–1626
R.; Tahvanainen, E.; Jauhiainen, M. J. Clin. Invest 1996, 883.
(b) Hasselwander, O.; Savage, D. A.; McMaster, D.; Lough-
rey, C. M.; McNamee, P.; Middleton, D.; Nicholls, D. P.;
Maxwell, A. P.; Young, I. S. Kidney Int. 1999, 289. (c) Ferre,
N.; Tous, M.; Paul, A.; Zamora, A.; Vendrell, J. J.; Bardaji,
A.; Camps, J.; Richart, C.; Joven, J. Clin. Biochem. 2002, 35,
197.
(m, 2H). 13C NMR (75.5 Hz, CDCl3): d 170.3 (C¼O), 144.2,
142.9, 122.2 and 115.6 (ArC), 21.0 (CH3). IR (KBr, cmꢂ1):
3361, 3338, 3016, 1751, 1616, 1508, 1219, 1186, 1014, 752. 2d
HRMS calcd for C9H10O3Na (M + Na+) 189.0528, found
189.0522. 1H NMR (300 MHz, CDCl3): d 2.28 (s, 3H), 3.79 (s,
3H), 6.88 (m, 2H), 7.00 (m, 2H). 13C NMR (75.5 Hz, CDCl3):
d 169.9 (C¼O), 157.2, 144.2, 122.3 and 114.4 (ArC), 55.6
(OCH3), 21.0 (CH3). IR (KBr, cmꢂ1): 3015, 2940, 2907, 2832,
1753, 1594, 1509, 1462, 1298, 1097, 909.
15. Rodrigo, L.; Mackness, B.; Durrington, P. N.; Hernandez,
A.; Mackness, M. I. Biochem. J. 2001, 354, 1.
16. Billecke, S.; Draganov, D.; Counsell, R.; Stetson, P.;
Watson, C.; Hsu, C.; La Du, B. N. Drug. Metab. Dispos. 2000,
28, 1335.
17. Augustinsson, K.-B.; Ekedahl, G. Acta Chem. Scand.
1962, 16, 240.
20. PON1 enzyme activity assays were conducted in a total
volume of 2.5 mL containing 20 mM Tris–HCl buffer in D2O
(pH 7.4), 2 mM CaCl2 and 5 mm of the test compound. The
reaction was initiated by adding 100 mL crude serum, which
had been pre-treated for 1 h with PMSF, and an aliquot of the
reaction mixture was quickly transferred to an NMR tube for
18. (a) Nishio, E.; Watanabe, Y. Biochem. Biophys. Res. Com.
1997, 236, 289. (b) Hasselwander, O.; McMaster, D.; Fogarty,
D. G.; Maxwell, A. P.; Nicholls, D. P.; Young, I. S. Clin.
Chem. 1998, 44, 179. (c) Sams, C.; Mason, H. J. Hum. Exp.
Toxicol. 1999, 18, 653. (d) Aviram, M.; Dornfeld, L.; Rosen-
blat, M.; Volkova, N.; Kaplan, M.; Coleman, R.; Hayek, T.;
Presser, D.; Fuhrman, B. Am. J. Clin. Nutr. 2000, 71, 1062 ref
13(b).
time course analysis, which was done at 30 ꢀC. From the H
1
NMR spectra, integration assignment was performed to
determine the amount of substrate remaining over time, which
was plotted to obtain t1/2 of the compound. A control tube
was also prepared for each compound which did not contain
serum.
21. Inhibition of serum PON1 activity was tested with 1 mM
paraoxon in the presence of various concentrations of inhib-
itor with crude serum, which had been pre-treated for 1 h with
PMSF. This mixture was incubated at 37 ꢀC for 10 min prior
to assay. Assays were preformed as described,1 except that 20
mM Tris–HCl (pH 7.4) buffer containing 2 mM CaCl2 was
used and that 200 mL treated serum were assayed in a total
volume of 3 mL at 30 ꢀC.
22. Sorenson, R. C.; Primo-Parmo, S. L.; Kuo, C.-L.; Adkins,
S.; Lockridge, O.; La Du, B. Proc. Natl. Acad. Sci. 1995, 92,
7187.
23. (a) Josse, D.; Xie, W.; Renault, F.; Rochu, D.; Schopfer,
L. M.; Masson, P.; Lockridge, O. Biochemistry 1999, 38, 2816.
(b) Doorn, J. A.; Sorenson, R. C.; Billecke, S. S.; Hsu, C.;
La Du, B. N. Chem. Bio. Interact. 1999, 120, 235. (c) Josse,
D.; Lockridge, O.; Xie, W.; Bartels, C. F.; Schopfer, L. M.;
Masson, P. J. Appl. Toxicol. 2001, 21, S7.
19. 1b HRMS calcd for C10H16NO4NaP (M
+
Na+)
268.0715, found 268.0725. 1H NMR (300 MHz, CDCl3): d
1.32 (t, J=6.99 Hz, 6H), 4.17 (m, 4H), 6.61 (d, J=8.81 Hz,
2H), 6.99 (d, J=8.81 Hz, 2H). 13C NMR (75.5 Hz, CDCl3): d
143.5, 143.2, 120.8 and 115.8 (ArC), 64.4 and 64.4 (2ꢁCH2),
16.1 and 16.0 (2ꢁCH3). 31P NMR (121.5 Hz, CDCl3): d ꢂ5.68.
IR (KBr, cmꢂ1): 3380 br, 3025, 1504, 1210, 1028, 967, 757. 1c
HRMS calcd for C10H15O4NaP (M + Na+) 253.0606, found
253.0599. 1H NMR (300 MHz, CD3OD): d 1.34 (t, J=7.06
Hz, 6H), 4.22 (m, 4H), 7.22 (m, 3H), 7.38 (m, 2H). 13C NMR
(75.5 Hz, CDCl3): d 148.9, 129.6, 124.8, 112.0 and 119.8 (ArC),
64.5 and 64.4 (2ꢁCH2), 16.0 and 15.9 (2ꢁCH3). 31P NMR
(121.5 Hz, CDCl3): dP ꢂ5.62. IR (KBr, cmꢂ1): 3013, 1601,
1488, 1273, 1221, 1041, 970, 749. 2c HRMS calcd for
C8H9NO2Na (M + Na+) 174.0531, found 174.0532. 1H
NMR (300 MHz, CDCl3): d 2.26 (s, 3H), 6.66 (m, 2H), 6.86