Design and synthesis of water soluble (metallo)porphyrins with pendant arms: Studies of binding to xanthine oxidase

Article abstract of DOI:10.1039/b9nj00736a

The tethering of a substrate analogue to a covalently attached luminophore may give rise to a probe for reporting enzyme activity spectrofluorometrically. The overall design incorporates a water-soluble porphyrin luminophore, a substrate analogue and a pl

Full text of DOI:10.1039/b9nj00736a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/njc | New Journal of Chemistry
Design and synthesis of water soluble (metallo)porphyrins with pendant
arms: studies of binding to xanthine oxidasew
Elizabeth A. Gibson, Anne-K. Duhme-Klair* and Robin N. Perutz*
Received (in Montpellier, France) 6th December 2009, Accepted 3rd March 2010
First published as an Advance Article on the web 7th April 2010
DOI: 10.1039/b9nj00736a
The tethering of a substrate analogue to a covalently attached luminophore may give rise to a
probe for reporting enzyme activity spectrofluorometrically. The overall design incorporates a
water-soluble porphyrin luminophore, a substrate analogue and a planar conjugated bridge of the
appropriate length designed to bind in the substrate-access channel of xanthine oxidase (XO).
5,10,15-Tris(N-methyl-4-pyridiniumyl)-20-phenyl porphyrins was functionalised at the 4-position
of the phenyl group with amide-linked 2-methoxy-benzamide groups, [(2-methoxy-4-amino-
phenylcarbonyl)-amino] or [(2-methoxy-4-[(pyridine-4-carbonyl)-amino]-phenylcarbonyl)-amino].
The products were isolated as free base or zinc porphyrins with chloride or hexafluorophosphate
counterions and characterised by optical emission and absorption in addition to other
spectroscopic methods. Binding studies of bovine XO to the zinc porphyrin trichloride derivatives
with the above linkers were used to determine the IC50 of 81 and 113 mM and Ki values of
4.6 and 6.5 mM, respectively. Furthermore, addition of XO to an aqueous solution of the
products caused quenching of the porphyrin emission and a red shift in the absorption spectrum.
protein-induced emission enhancement on binding to trypto-
Introduction
phanase and probes for bovine serum albumin, estrogen
receptors and biotin conjugates which bind to avidin.3 Gray
Photochemical devices, such as ‘sensors’ or ‘probes’, are used
to transmit information about events occurring on
a
and co-workers have developed probes for heme active sites,
such as cytochrome P450cam and nitric oxide synthase.4 The
binding is sufficiently strong to allow detection of the enzyme
at submicromolar concentrations and is reported by a decrease
in emission intensity of the probe luminophore. Both cases
demonstrated that changing the length of the linker may
simultaneously alter the binding affinity and the rate of
energy/electron transfer processes.
microscopic or nanoscopic scale via light signalling. Ideally
they should exhibit high sensitivity, on–off switching, ease of
communication and a rapid response. They can be specifically
constructed to suit a range of applications and target
analytes.1 Luminescent probes may be designed to be bound
within the active site of enzymes by hydrophobic effects and
non-covalent interactions.2 The general principle is to attach a
signalling unit (luminophore) covalently to a binding unit via a
bridge, which will enable communication between the two.
Binding of the substrate to the enzyme active site can then be
communicated to the signaling unit by energy transfer or
electron transfer and the information is reported as a change
in the luminescent properties. Each of these distinct
components can be specifically designed according to the
intended application to maximise supramolecular interactions,
tune the spectroscopic properties and redox potentials of the
components and to achieve the required water solubility of the
molecule.
Since luminescent probes can be used to study substrate
binding and electron transfer reactions of enzymes, we have
begun to apply them to another class of clinically significant
redox-enzymes, the Xanthine Oxidoreductases (XORs).5
Xanthine Dehydrogenase (XDH) and Xanthine Oxidase
(XO) belong to the multifunctional, multisubstrate Xanthine
Oxidoreductase (XOR) family of enzymes.6,7 These enzymes
catalyse the hydroxylation of aromatic heterocycles such as
purines, pteridines and aromatic aldehydes (eqn (1)):5
RH + H2O - ROH + 2H+ + 2eÀ
(1)
Both Gray and Lo have combined simplified substrates with
covalently attached luminophores giving rise to novel probes
for reporting enzyme activity. Examples reported by Lo et al.
include probes for indole-binding proteins which show
The oxidation of xanthine to uric acid by the XORs is
particularly important in the catabolism of purines, such as
ATP.5,8,9 The reaction occurs at a molybdopterin active site
and electrons released are transferred to a second, flavin-
containing cofactor.10 In humans, the enzyme is synthesised
as the xanthine dehydrogenase (XDH) form, which uses
NAD+ as the terminal electron acceptor. In the tissues,
however, XDH is partially converted to xanthine oxidase
(XO), which uses dioxygen as electron acceptor. Con-
sequently, reactive oxygen species, such as hydrogen peroxide
as well as superoxide, are produced. XO is therefore a target
Department of Chemistry, The University of York, Heslington, York,
YO10 5DD, UK. E-mail: rnp1@york.ac.uk;
Tel: +44 (0)1904 432549
w Electronic supplementary information (ESI) available: Dihedral
angles and packing diagram of compound 2, examples of plots used
to determine Ki and 1H and 13C NMR assignments for porphyrins 3–8.
CCDC reference number 757513. For ESI and crystallographic data in
CIF or other electronic format see DOI: 10.1039/b9nj00736a
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1125–1134 | 1125

View Article Online
for drugs against oxygen radical-induced tissue damage and
oxidative stress, including myocardial reperfusion injury and
heart failure.11 Since a number of medical conditions are
associated with XO, there are potential clinical applications
for XOR inhibitors which can decrease the activity of the
enzyme. A significant number of compounds which bind near
the molybdenum cofactor active site have been reported.10,12
The most potent inhibitors are planar aromatic systems with
hydrogen bonding substituents, such as alloxanthine, which
are stabilised in the active site by complementary hydrogen
bonding residues and p–p-stacking. Drugs can bind tightly in
the long, narrow access channel leading to the Mo active site,
without directly binding to the molybdenum.12,13
Scheme 2 An example of a probe for XOR. The grey loop represents
the active site of XOR.
The purpose of our substrate analogue attached to the
probe is to bind within the access channel to the active site
and to communicate with the molybdenum cofactor. The
substrate should be bound in the channel by hydrogen bonds
and p–p stacking in a similar manner to the natural substrates
and so planar heteroaromatic groups are appropriate.10,14 In
addition, direct coordination to the Mo-centre, as observed
with the substrate xanthine, may also be possible if suitable
donor atoms are introduced, as was demonstrated for the
inhibitor Fyx-051 (Scheme 1).15 Fyx-051 interacts directly with
the Mo-centre through a bridging oxygen, which stems from
the catalytically active hydroxyl ligand of the Mo-centre. The
formation of this intermediate makes Fyx-051 a particularly
effective and specific drug candidate.
450 nm. Porphyrins fulfil these requirements and, in addition,
can be functionalized at the periphery to suit particular
requirements, especially water solubility in biological systems.
We developed our methods by synthesising triphenyl-
porphyrins functionalised by a probe arm consisting of
amide-linked 2-methoxy, 4-amino benzamide units.19 The
success of this approach enabled us to move on to water-
soluble analogues. Here we report the synthesis and testing
of analogous probe molecules based on a water-soluble
tris(methylpyridinium) metalloporphyrin with similar pendant
arms of different lengths designed to enter the access channel
(Scheme 2).
Experimental
Attachment and communication between the substrate
analogue and luminophore is to occur through a linker of
suitable length that positions the substrate close to the
molybdenum cofactor in the active site. Computer simulation
was used to estimate the length of the linker, although
computer simulation from crystal structures can be an
over-simplification.4 Substituents may introduce hydrogen
bonding within the molecule, increase rigidity and aid electron
transfer. Aromatic groups have been shown to be particularly
good bridges.16 Amide bonds provide a means of connection
whilst being able to encourage electron transfer via
superexchange.17
General methods
Chemicals came from the following suppliers: triethylamine,
ammonium chloride, nicotinic acid (BDH); 4-amino salicylic
acid, tetrabutyl ammonium chloride, methyl iodide, nicotinic
acid anhydride, isonicotinic acid anhydride, potassium
phosphate, Florisils, xanthine oxidase (bovine milk, grade
III, 22 mg mlÀ1, 1.30 U mlÀ1 as a suspension in 2.3 M
(NH4)2SO4, 10 mM sodium phosphate buffer, pH 7.8, containing
1 mM EDTA and 1 mM sodium salicylate), Sephadex G-10
(Sigma-Aldrich); allopurinol (Avocado); ammonium hexa-
fluorophosphate (Fluorochem); xanthine (Alfa Aesar);
ammonia solution (Fisons); silica gel 60, DMAP (Fluka);
NsCl (Lancaster); [TMPyP][OTs]4 kindly donated by Prof.
John Lindsay-Smith.
Metalloporphyrins are ideal electron-donating lumino-
phores, which have high extinction coefficients as well as
tunable fluorescence emission and redox potentials. Absorp-
tion and emission of the chromophore must avoid the region
in the spectrum dominated by the cofactors of XO itself
(lmax = 280 nm).5,18 The iron–sulfur clusters of XO contribute
absorption maxima at 420 nm, 470 nm and a shoulder at
550 nm and the absorption maxima of the flavin are at 360 and
All reactions were carried out under an argon atmosphere
unless otherwise stated. All other materials and methods have
been described elsewhere.19
Synthesis
Methyl 4-amino-2-methoxybenzoate was prepared according
to the procedure reported by Hewlett et al.20 Free base 5,10,15-
tris(4-pyridyl)-20-(4-aminophenyl)porphyrin was prepared
according to the method described by Li et al.21
Methyl 4-[(pyridine-3-carbonyl)-amino]-2-methoxy benzoate, 1.
Triethylamine (0.7 cm3, 5.0 mmol) was added to a stirred
solution of nicotinic acid (0.31 g, 2.5 mmol), DMAP (0.061 g,
0.50 mmol) and NsCl (0.55 g, 2.5 mmol) in acetonitrile
(30 cm3) under argon. The solution was stirred for 20 min at
which point a solution of methyl 4-amino-2-methoxybenzoate
Scheme 1 Proposed interactions between Fyx-051 and the molybdop-
terin cofactor in XDH according to ref. 15.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
1126 | New J. Chem., 2010, 34, 1125–1134

View Article Online
(0.50 g, 2.8 mmol) in acetonitrile (10 cm3) was added at 0 1C.
Stirring was continued for 24 h until the reaction was complete
by tlc (ethyl acetate). The reaction mixture was concentrated
and extracted into 2 M hydrochloric acid. The acidic aqueous
phase was washed with ethyl acetate, neutralised with sodium
hydrogen carbonate and the resulting precipitate re-extracted
into ethyl acetate. The organic fraction was washed with
sodium hydrogen carbonate, water and brine and dried over
magnesium sulfate. The solvent was removed in vacuo to give 1
as a pale yellow powder (0.27 g, 0.94 mmol, 34%); mp 144–146 1C
(from ethyl acetate); lmax(DMSO)/nm 284 (e/dm3 molÀ1 cmÀ1
14 000) and 308 (13 000); nmax (KBr disc)/cmÀ1 3401w, 3342s,
3199w, 3150w, 3127w, 3090w, 3056w, 3039w, 3012w, 2964w,
2935w, 2913w, 2863w, 2837w, 1687s, 1663s, 1593s, 1532s,
1479w, 1466m, 1448m, 1429m, 1406s, 1341w, 1298s, 1257s,
1202w, 1187w, 1173w, 1147m, 1117w, 1090m, 1026m, 976w,
958w, 886w, 860m, 839w, 827w, 780m, 734w, 704w, 682w,
640w, 620w and 583w; dH (300 MHz; CDCl3) 9.11 (1H, s,
pyridyl), 8.77 (1H, d, J = 6.0 Hz, pyridyl), 8.33 (1H, s, NH),
8.22 (1H, d, J = 7.3 Hz, pyridyl) 7.85 (1H, d, J = 8.6 Hz,
phenyl), 7.72 (1H, d, J = 1.8 Hz, phenyl), 7.44 (1H, dd, J =
4.6, 7.7 Hz, pyridyl), 7.04 (1H, dd, J = 1.7, 8.6, phenyl), 3.91
(3H, s, OCH3) and 3.87 (3H, s, OCH3); dC (75 MHz; CDCl3)
166.4, 164.5, 160.1, 152.6, 147.9, 143.0, 136.5, 133.4, 130.0,
124.4, 116.1, 111.4, 104.1, 56.5 and 52.4; m/z (ESIÀ) 285.0
(M À H+, 100%); m/z (HR ESI+) 287.1025 (M + H+.
C15H15N2O4 requires 287.1026).
ammonia solution and the resulting purple precipitate was
extracted with chloroform. The organic fraction was dried
over sodium sulfate and the solution was then concentrated
in vacuo and the crude product purified by column chromato-
graphy (Si-60, 10
: 1 chloroform : methanol v/v) to
give 3 (0.71 g, 0.88 mmol, 87%); lmax (DMSO)/nm 422
(e/dm3 molÀ1 cmÀ1 167 000), 516 (8210), 552 (4140), 590
(2710) and 647 (1790); nmax (KBr disc)/cmÀ1 3429br m,
3314m, 3077w, 3023w, 2925w, 2854w, 2756w, 2615w, 2536w,
2400w, 2349w, 1716w, 1662m, 1638s, 1589m, 1567w, 1522s,
1480w, 1459m, 1401m, 1383w, 1346m, 1321m, 1279w, 1251m,
1210w, 1182s, 1158w, 1123w, 1085m, 1018m, 999m, 981m,
968s, 897w, 885m, 856m, 801s, 729s, 664w and 639w;
dH (500 MHz; CDCl3) 10.02 (1H, s, NH), 9.05 (6H, d,
o-pyridyl), 8.97 (2H, d, b-pyrrole), 8.85 (6H, m, b-pyrrole),
8.58 (1H, d, J = 8.6 Hz, phenyl), 8.23 (2H, d, J = 8.5 Hz,
o-aminophenyl), 8.17 (6H, d, J = 8.7 Hz m-pyridyl), 8.11
(2H, d, J = 8.5 Hz, m-aminophenyl), 8.04 (1H, dd, J = 2.0,
8.5 Hz, phenyl), 7.98 (1H, d, J = 2.0 Hz, phenyl), 4.29
(3H, s, OCH3) and À2.88 (2H, br s, inner NH); dC
(176 MHz; CDCl3) 161.6, 157.4, 150.8, 150.2, 150.1, 148.3,
138.0, 137.9, 135.3, 134.0, 129.4, 127.3, 121.1, 119.1, 117.5,
117.1, 116.6, 107.1 and 57.3; m/z (ESI+) 812.3 (M + H+,
100%) and 406.6 (M + 2H+, 15%); m/z (HR ESI+) 812.2708
(M + H+. C49H34N9O4 requires 812.2728).
Free base 5,10,15-tris(N-methyl-4-pyridiniumyl)-20-[4-[(2-
methoxy-4-nitro-phenylcarbonyl)-amino]phenyl]porphyrin tris-
(iodide), [4][I]3. Free base 3 (0.067 g, 0.082 mmol) was
dissolved in dimethylformide (5 cm3). Excess methyl iodide
(0.1 cm3) was added and the solution was stirred at room
temperature for 12 h. The solvent was removed in vacuo to give
[4][I]3 which was dried in a vacuum oven at 70 1C for 48 h
(0.11 g, 0.086 mmol, 96%); lmax (pH 7.5 phosphate buffer)/nm
431 (e/dm3 molÀ1 cmÀ1 142 000), 524 (12 300), 564 (7710), 592
(6350) and 649 (2630); nmax (KBr disc)/cmÀ1 3430br m, 3314w,
3077w, 3023br m, 2925br m, 2854w, 2756w, 2615w, 2536w,
2400w, 2349w, 1662m, 1638s, 1589m, 1567w, 1521s, 1479w,
1459m, 1401m, 1383w, 1346m, 1322m, 1279w, 1251m, 1210w,
1182s, 1158w, 1123w, 1083m, 1018m, 999m, 981m, 968s, 897w,
885m, 856m, 801s, 719s, 664m and 639m; dH (500 MHz;
d6-DMSO) 10.92 (1H, s, NH), 9.49 (6H, d, m-pyridyl), 9.01
(6H, m, o-pyridyl), 9.18 (4H, s, b-pyrrole), 9.10 (4H, m,
b-pyrrole), 8.25 (4H, m, o-, m-aminophenyl), 8.03 (2H, m,
phenyl), 7.94 (1H, m, phenyl), 4.73 (9H, s, NCH3), 4.12 (3H, s,
OCH3) and À2.88 (2H, br s, inner NH); dC (176 MHz; CDCl3)
164.5, 157.4, 157.1, 157.0, 150.0, 144.7, 139.7, 136.2, 135.4,
132.6, 130.6, 127.6, 123.2, 118.7, 116.2, 115.8, 115.0, 107.5,
57.3 and 48.5; m/z (ESI+) 856.3 (M3+ + 2eÀ, 25%) and 427.7
(M3+ À H+, 15%); m/z (HR ESI+) 427.6627 (M3+ À H+.
C52H41N9O4 requires 427.6635).
Methyl 4-[(pyridine-4-carbonyl)-amino]-2-methoxy benzoate, 2.
The procedure described above was carried out using isonico-
tinic acid to give 2 as a pale yellow powder (0.30 g, 1.0 mmol,
40%); mp 196–197 1C (from ethyl acetate); dH (300 MHz;
CDCl3) 8.82 (2H, dd, J = 4.5, 1.6 Hz, pyridyl), 8.04
(1H, s, NH), 7.86 (1H, d, J = 8.5 Hz, phenyl), 7.74 (1H, d,
J = 1.8 Hz, phenyl), 7.71 (2H, dd, J = 1.5, 4.4 Hz, pyridyl),
6.99 (1H, dd, J = 1.9, 8.5, phenyl), 3.94 (3H, s, OCH3) and
3.88 (3H, s, OCH3); dC (75 MHz; CDCl3): 166.3, 164.3, 161.0,
151.3, 142.7, 141.9, 133.4, 121.3, 116.3, 111.3, 104.0, 56.6 and
52.4; m/z (ESIÀ) 285.0 (M À H+, 100%); m/z (HR ESI+)
287.1023 (M + H+. C15H15N2O4 requires 287.1026).
Free base 5,10,15-tris(4-pyridyl)-20-[4-[(2-methoxy-4-nitro-
phenylcarbonyl)-amino]phenyl]porphyrin, 3. Dicyclohexyl-
carbodiimide (1.40 g, 6.8 mmol) was added to a stirred
solution of 2-methoxy-4-nitrobenzoic acid (1.20 g, 6.1 mmol)
in acetonitrile (25 cm3) at 0 1C. After 15 min the reaction was
brought to room temperature and stirred for B24 h until the
reaction was complete by tlc (10 : 1 chloroform : acetonitrile
v/v). The solution was filtered and the solid by-product was
washed thoroughly with acetonitrile and the washings added
to the filtrate. Removal of the solvent yielded a pale yellow
solid, which was used without any further purification. The
anhydride residues were taken up in acetonitrile (20 cm3) and
added dropwise to an ice cold solution of free base 5,10,15-
tris(4-pyridyl)-20-(4-aminophenyl)porphyrin (0.64 g, 1.01 mmol)
in chloroform with stirring. The reaction was heated at reflux
for 48 h. The solvent was removed and the purple residue was
extracted into 1 M hydrochloric acid and washed with chloro-
form (3 Â 100 cm3). The green solution was neutralised with
Free base 5,10,15-tris(N-methyl-4-pyridiniumyl)-20-[4-[(2-
methoxy-4-amino-phenylcarbonyl)-amino]phenyl]porphyrin tris-
(chloride) [5][Cl]3. Free base [4][I]3 (0.100 g, 0.082 mmol) was
dissolved in 25 cm3 ethanol. Saturated ammonium chloride
solution (0.1 cm3) and indium powder (100 mesh, 0.05 g) were
added and the solution stirred for 72 h in air, under reflux. The
reaction mixture was diluted with water, adjusted to pH 3 with
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1125–1134 | 1127

View Article Online
0.1 M hydrochloric acid and filtered through Celite. The
solvent was removed in vacuo to give a green/purple solid.
The crude porphyrin was taken up in water and ammonium
hexafluorophosphate was added to precipitate the product
as the hexafluorophosphate salt. This was filtered, washed
with cold water and dried to give [5][PF6]3 (0.070 g,
0.055 mmol, 64%).
(2H, d, J = 5.3 Hz, o-pyridyl), 8.29 (2H, d, J = 7.8 Hz,
m-aminophenyl), 8.23 (2H, d, J = 7.9 Hz, o-aminophenyl),
7.95 (1H, s, phenyl), 7.87 (1H, d, J = 8.6 Hz, phenyl), 7.72
(1H, d, J = 8.6, phenyl), 4.75 (9H, s, NCH3), 4.06 (3H, s,
OCH3) and À2.95 (2H, br s, inner NH); dC ([6][PF6]3; 150 MHz;
CD3CN) 164.0, 163.4, 159.4, 158.3, 149.6, 148.3, 144.0, 143.4,
138.7, 138.4, 135.2, 132.8, 132.7, 132.3, 131.6, 123.8, 121.3,
118.7, 118.3, 114.9, 114.0, 112.2, 103.4, 56.3 and 48.4;
To obtain the water soluble chloride salt, tetrabutyl ammo-
nium chloride solution in acetonitrile was added dropwise to
an acetonitrile solution of [5][PF6]3 to precipitate [5][Cl]3. This
salt was further purified by passing an aqueous solution
through Sephadex G10 size exclusion gel and precipitation
from aqueous solution by addition of acetone; lmax (DMSO)/nm
429 (e/dm3 molÀ1 cmÀ1 124 000), 522 (9780), 564 (8750), 595
(5840) and 653 (3510); nmax (KBr disc)/cmÀ1 3426br s, 2961w,
2926w, 2855m, 1638s, 1599s, 1590s, 1511s, 1463s, 1400m,
1384s, 1311s, 1271s, 1244s, 1210m, 1182m, 1135m, 1094m,
1052w, 1023s, 1014s, 999m, 981w, 968s, 885w, 821m, 797s,
763m, 729m, 710s and 672m; nmax (ATR)/cmÀ1 1636s, 1597s,
1588s, 1559m, 1509s, 1468s, 1460s, 1400m, 1359w, 1349w,
1329w, 1312s, 1271s, 1245s, 1211m, 1182s, 1155m, 1136m,
1094m, 1054w, 1024s, 1014s, 998m, 968s, 945w, 927w, 883m,
853s, 793s, 764s, 710s and 663s; dH (500 MHz; d6-DMSO)
10.21 (1H, s, NH), 9.53 (6H, d, m-pyridyl), 9.13 (8H, m,
b-pyrrole), 9.00 (6H, m, o-pyridyl), 8.21 (2H, d, J = 8.2 Hz,
m-aminophenyl), 8.19 (2H, d, J = 8.2 Hz, o-aminophenyl),
7.74 (1H, d, J = 9.0 Hz, phenyl), 6.40 (1H, d, J = 1.9 Hz,
phenyl), 6.33 (1H, dd, J = 9.0, 1.9 Hz, phenyl), 4.73 (9H, s,
NCH3), 4.01 (3H, s, OCH3) and À2.99 (2H, br s, inner NH);
dC (176 MHz; d6-DMSO) 164.7, 159.6, 157.1, 157.0, 154.6,
144.7, 140.3, 135.7, 135.4, 133.2, 132.9, 132.6, 123.6, 118.8,
115.8, 114.9, 109.7, 107.0, 96.6, 56.3 and 48.3; m/z (ESI+)
412.7 (M3+ À H+, 15%), 275.5 (M3+, 100%) and 206.8
m/z (ESI+) 465.2 (M3+ À H+, 10%) and 310.5 (M3+
,
100%); m/z (HR ESI+) 310.4596 (M3+. C58H47N10O3 requires
310.4605).
Zinc 5,10,15-tris(N-methyl-4-pyridiniumyl)-20-[4-[(2-methoxy-
4-amino-phenylcarbonyl)-amino]phenyl]porphyrin tris(chloride),
[7][Cl]3. Free base [5][Cl]3 (0.050 g, 0.054 mmol) and zinc
acetate dihydrate (0.11 g, 0.50 mmol) were dissolved in ethanol
(25 cm3) and heated in air, at reflux for 2 h. The reaction
mixture was cooled to room temperature and diluted with
water (50 cm3). Ammonium hexafluorophosphate was added
to precipitate the product as the hexafluorophosphate salt,
[7][PF6]3 (0.058 g, 0.044 mmol, 82%). This was filtered, washed
with cold water and dried. To obtain the water soluble
chloride salt, tetrabutyl ammonium chloride solution in aceto-
nitrile was added dropwise to an acetonitrile solution of
[7][PF6]3 to precipitate [7][Cl]3. This product was further
purified by passing a methanolic solution of [7][Cl]3 through
Sephadex LH20 size exclusion gel and precipitation from
aqueous solution by addition of acetone to give [7][Cl]3
(0.040 g, 0.041 mmol, 76%); lmax (pH 7.5 phosphate buffer)/
nm 436 (e/dm3 molÀ1 cmÀ1 107 000), 565 (11 200) and 615
(6890); nmax (KBr disc)/cmÀ1 3428br m, 3110w, 3044w, 2963w,
2921m, 2851m, 1636s, 1602s, 1559w, 1518s, 1508s, 1457m,
1400w, 1385m, 1359w, 1341w, 1311w, 1264m, 1241m, 1202w,
1182m, 1129m, 1108w, 1093m, 1016m, 992s, 855w, 797s,
759m, 713m and 673m; dH (600 MHz; d6-DMSO) 10.18
(1H, s, NH), 9.41 (6H, m, m-pyridyl), 8.98 (8H, m, b-pyrrole),
8.88 (6H, m, o-pyridyl), 8.17 (2H, d, J = 6.0 Hz, m-amino-
phenyl), 8.10 (2H, d, J = 6.0 Hz, o-aminophenyl), 7.76 (1H, d,
J = 7.5 Hz, phenyl), 6.39 (1H, s, phenyl), 6.32 (1H, d, J =
7.5 Hz, phenyl), 5.95 (2H, br s, NH2), 4.70 (9H, s, N+CH3)
and 4.02 (3H, s, OCH3); dC (150 MHz; d6-DMSO) 164.6,
159.0, 154.6, 150.1, 148.8, 148.6, 148.3, 144.1, 135.0, 133.8,
133.2, 132.6, 132.4, 131.9, 123.8, 115.9, 118.5, 113.9, 109.7,
(M3+ + H+, 10%); m/z (HR ESI+) 275.4526 (M3+
.
C52H44N6O2 requires 275.4534).
Free base 5,10,15-tris(N-methyl-4-pyridiniumyl)-20-[4-{(2-
methoxy-4-[(pyridine-4-carbonyl)-amino]-phenylcarbonyl)-amino}-
phenyl]porphyrin tris(chloride), [6][Cl]3. The isonicotinic acid
anhydride (0.050 g, 0.18 mmol) was taken up in acetonitrile
(50 cm3) and added dropwise to an ice cold solution of free
base [5][PF6]3 (0.068 g, 0.054 mmol) in acetonitrile with
stirring. The reaction was heated at reflux for 48 h. Tetra-
butylammonium chloride solution in acetonitrile was added
dropwise to the cool solution until a precipitate of [6][Cl]3
formed. This was filtered, washed with acetonitrile and dried.
The product was further purified by passing an aqueous
solution through Sephadex G10 size exclusion gel and
precipitating from aqueous solution by addition of acetone
to give [6][Cl]3 (0.043 g, 0.041 mmol, 76%); lmax (pH 7.5
phosphate buffer)/nm 434 (e/dm3 molÀ1 cmÀ1 117 000), 524
(11 600), 566 (11 800), 598 (8120) and 620 (6520); nmax (KBr
disc)/cmÀ1 3421s, 3109w, 3018w, 2958w, 2917w, 2837w,
1636br s, 1559w, 1539w, 1521m, 1507m, 1475w, 1457w,
1437w, 1401s, 1384s, 1323w, 1260w, 1232w, 1213w, 1182w,
1145w, 1090m, 1072m, 1026s, 964w, 882w, 856w, 794m, 763m,
698m, 668s, 659w and 617w; dH (600 MHz; d6-DMSO) 11.35
(1H, s, NH), 10.58 (1H, s, NH), 9.55 (6H, m, m-pyridyl), 9.10
(8H, m, b-pyrrole), 9.01 (8H, m, o-pyridyl, m-pyridyl), 8.35
106.9, 96.6, 56.3 and 48.2; m/z (HR ESI+) 296.0891 (M3+
.
C52H42N9O2Zn requires 296.0912).
Zinc 5,10,15-tris(N-methyl-4-pyridiniumyl)-20-[4-{(2-methoxy-
4-[(pyridine-4-carbonyl)-amino]-phenylcarbonyl)-amino}phenyl]-
porphyrin tris(chloride), [8][Cl]3. Free base [6][Cl]3 (0.043 g,
0.041 mmol) and zinc acetate dihydrate (0.075 g, 0.41 mmol)
were reacted and converted first to the hexafluorophosphate
salt and then the chloride salt as for [7][Cl]3, above, to give
[8][Cl]3 (0.030 g, 0.028 mmol, 67%); lmax (DMSO)/nm 442
(e/dm3 molÀ1 cmÀ1 111 000), 569 (12 500) and 614 (7520); nmax
(KBr disc)/cmÀ1 3443br w, 3116w, 3043w, 2963w, 2920w,
2849w, 1637s, 1594m, 1526m, 1514w, 1461m, 1453m, 1402m,
1384s, 1342w, 1314m, 1288w, 1262w, 1247w, 1185m, 1129w,
1092m, 1019w, 994s, 853w, 797s, 750w, 715w, 673w and
616w; dH (600 MHz; d6-DMSO) 11.18 (1H, s, NH), 10.98
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
1128 | New J. Chem., 2010, 34, 1125–1134

View Article Online
(1H, s, NH), 9.47 (6H, m, m-pyridyl), 9.00 (8H, m, b-pyrrole),
8.92 (6H, m, o-pyridyl), 8.81 (2H, d, J = 6.0 Hz, m-pyridyl),
8.23 (2H, d, J = 8.2 Hz, m-aminophenyl), 8.14 (2H, d, J = 8.2 Hz,
o-aminophenyl), 7.99 (2H, d, J = 6.0 Hz, o-pyridyl), 7.91 (1H,
s, phenyl), 7.82 (1H, d, J = 8.2 Hz, phenyl), 7.69 (1H, d,
J = 8.2, phenyl), 4.75 (9H, s, NCH3) and 4.05 (3H, s, OCH3);
dC (150 MHz; d6-DMSO) 164.8, 164.7, 159.0, 157.6, 150.9, 150.8,
148.6, 148.3, 144.1, 142.0, 137.5, 135.0, 133.8, 132.8, 131.8, 131.2,
126.6, 123.7, 122.2, 118.5, 115.9, 115.0, 112.8, 104.4, 55.6 and
48.1; m/z (ESI+) 496 (M2+, 100%) and 331 (M3+, 80%); m/z
(HR ESI+) 331.0971 (M3+. C58H45N1O3Zn requires 331.0984).
a
SMART CCD camera. Diffractometer control, data
collection, and initial unit-cell determination were performed
using ‘‘SMART’’ (v5.625 Bruker-AXS). Frame integration
and unit-cell refinement software was carried out with
‘‘SAINT+’’ (v6.22, Bruker AXS). Absorption corrections
were applied using SADABS (v2.03, Sheldrick). The structure
was solved by direct methods using SHELXS-97 and refined
by full-matrix least squares using SHELXL-97.23
Crystal data. C15H14N2O4, M = 286.28, monoclinic, a =
13.6031(9) A, b = 8.0726(5) A, c = 13.6339(9), b =
118.0610(10)1, T = 110(2) K, P2(1)/n, Z = 4, reflections
measured, 13 200, independent reflections 3270, Rint = 0.0194,
final R indices [I > 2s(I)] R1 = 0.0364, wR2 = 0.0979,
R indices (all data) R1 = 0.0403, wR2 = 0.1011.
Enzyme assay of XO
50 mM potassium phosphate buffer, pH 7.5. Potassium
phosphate, monobasic, anhydrous (1.36 g, 10 mmol) was
dissolved in deionized water (190 cm3). A solution of EDTA
(0.1 M, 0.2 cm3) was added and the pH of the solution was
adjusted to 7.5 at 25 1C with 1 M potassium hydroxide
solution. The volume of the solution was increased to
200 cm3 with deionized water.
Results
Preparation and binding studies of the substrate–linker
conjugates
A number of substrate–linker conjugates were prepared so
that their affinity for the XO active site could be assessed and
the best combination identified for the design of the probe.
Commercially available 2-methoxy-4-amino salicylic acid was
chosen as the ‘linker’. As mentioned above, the enzyme active
site accommodates planar aromatic compounds, such as
salicylic acid.10 The acid and amine functionalities provide a
means of attachment to the luminophore and substrate by
amide coupling. The methoxy group introduces a potential
hydrogen bond acceptor which could bind either intra-
molecularly, to improve rigidity, or intermolecularly to
hydrogen bond donor residues lining the access channel of
the protein.
0.15 mM solution of xanthine. Xanthine (2.3 mg, 0.015 mmol)
was dissolved in the minimum volume of aqueous potassium
hydroxide (1 M). The solution was diluted to 90 cm3 with
deionised water and the pH was adjusted to 7.5 with dilute
hydrochloric acid. The solution was made up to 100 cm3 with
deionised water.
Solution of xanthine oxidase. 17 ml of a stock solution
of xanthine oxidase (bovine milk, grade III, 22 mg mlÀ1
,
1.30 U mlÀ1) was diluted to 5 cm3 with 50 mM potassium
phosphate buffer and stored over ice.
Assay procedure
Since pyridyl compounds have been shown to be suitable
inhibitors, nicotinic and isonicotinic acid were coupled to the
linker.12 The DMAP-catalysed reaction shown in Scheme 3
gave 1 and 2 in 34% and 40% yield, respectively, after
recrystallisation from ethyl acetate.24 The successful amide
coupling was indicated by the downfield chemical shift of
The rate of formation of uric acid was measured spectro-
photometrically at 295 nm using the method of Sweeney et al.
with modifications.22 All solutions were freshly prepared and
stored in an ice bath before use. Enzyme solution (100 ml,
3 mU mlÀ1) in phosphate buffer (50 mM, pH 7.5) was added to
a 2900 ml solution of xanthine and inhibitor in phosphate
buffer (50 mM, pH 7.5, 25 1C) giving a final assay concentra-
tion of 0.050 mM xanthine (saturated) and 0.1 mU mlÀ1
enzyme. The reaction mixture was stirred and maintained at
25 1C for the duration of the reaction (10 min). A blank was
run in the absence of the inhibitor and a control reaction was
carried out in the absence of XO. Three replicates were
recorded for each inhibitor concentration. The stock solutions
of inhibitors that were not water soluble were made up in
dimethyl sulfoxide. The final concentration of dimethyl
sulfoxide (1%) was not sufficient to affect the assay. In those
cases the control reactions were carried out with 1% dimethyl
sulfoxide in the final assay.
1
the NH proton in the H NMR spectrum at d 8.77 for 1 and
d 8.04 for 2.
The IC50 was calculated using the sigmoidal fit tool
in Origin 6.1 (OriginLab Corporation, Northampton, MA).
An example is shown in Fig. S4 in the ESI.w
X-Ray crystallography
Diffraction data were collected on a Bruker Smart Apex
diffractometer with MoKa radiation (l = 0.710 73 A) using
Scheme 3 Preparation of substrate–linker conjugates 1 and 2.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1125–1134 | 1129

View Article Online
initial rate without the inhibitor. The %inhibition was found
to vary by Æ5% between the three replicate runs. A plot of
%inhibition vs. log [I], where [I] is the inhibitor concentration,
is fitted to a sigmoidal function to yield the IC50 (Table 2,
estimated experimental error Æ10%). Some of the inhibitors
were not sufficiently soluble at concentrations that gave close
to 100% inhibition. In those cases, the upper limit was fixed at
100%. The UV/vis spectrum of the inhibitors was recorded
over time to ensure no degradation reaction was taking place.
In order to provide a means of comparison with literature
data, the IC50 of allopurinol was measured as a control
experiment. Since the IC50 is dependent on the conditions of
the assay, of which there is a great variation across published
data, the binding constant, Ki, was calculated from eqn (2).
Fig. 1 ORTEP view of 2 with 50% thermal ellipsoids; hydrogen
atoms not shown.
Table 1 Selected bond lengths and angles for 2
Bond
Length/A
Bond
Angle/1
%inhibition = 100(1 À n/n0)
IC50 = (1 + [S]/Km)Ki
(1)
(2)
C(1)–O(2)
C(1)–O(1)
C(1)–C(2)
C(3)–O(3)
C(5)–N(1)
C(8)–O(4)
C(8)–N(1)
C(14)–O(1)
C(15)–O(3)
1.2153(13)
1.3315(12)
1.4817(14)
1.3583(12)
1.4042(13)
1.2194(13)
1.3627(13)
1.4450(12)
1.4311(12)
O(2)–C(1)–C(2)
O(1)–C(1)–C(2)
O(3)–C(3)–C(4)
O(3)–C(3)–C(2)
O(4)–C(8)–N(1)
O(4)–C(8)–C(9)
N(1)–C(8)–C(9)
C(8)–N(1)–C(5)
123.21(9)
115.48(9)
121.89(9)
117.78(9)
124.21(9)
119.97(9)
115.82(9)
127.50(9)
where IC50 is the concentration of inhibitor required to reduce
the rate of the reaction by 50%, [S] is the concentration of the
substrate. Km was estimated using the double-reciprocal plot
of 1/v vs. 1/[S] which gives a straight line of gradient Km/Vmax
and intercept on the 1/v axis of 1/Vmax and on the 1/[S] axis of
À1/Km (see ESIw for figures). A value of Km = 3.05 mM was
obtained which is consistent with that reported.25
Compound 1 was a powder that showed little tendency to
form crystals from organic solvents, while 2 readily formed
large, transparent blocks on slow evaporation of solvent from
a concentrated solution in ethyl acetate. The molecular structure
was determined by X-ray diffraction (Fig. 1, Table 1). The two
aromatic rings in 2 and the amide and ester groups deviate
from co-planarity as indicated by the dihedral angles
N(1)–C(8)–C(9)–C(13) of 26.721 and O(2)–C(1)–C(2)–C(7) of
12.451, respectively. The angle between the fitted planes of the
two aromatic rings is 29.35 (0.04)1 (see ESIw for dihedral
angles, Table S1). A possible reason for this twist can be
inferred from the packing diagram. A weak hydrogen bond
between N(1) of one molecule and the ester carbonyl oxygen,
O(2), on another molecule creates a supramolecular array of
species (Fig. S1 (ESIw), N(1)–H(1)Á Á ÁO(2)#1 d(NÁ Á ÁO) =
2.062(15) A). The pyridyl nitrogen does not act as a H-bond
acceptor, nor is there any p–p-stacking between the aromatic
groups.
Table 2 lists the IC50 and Ki values determined for each of
the compounds that were shown to inhibit XO. Both the linker
motif and the proposed substrate bound to XO with Ki values
in the micromolar range, suggesting that they are suitably
accommodated in the XO active site, competing with the
binding of xanthine. The compounds with free carboxylic acid
groups bound less tightly than the ester equivalents. Since the
Ki for the isonicotinic acid-based compound 2 was slightly
lower than that of the nicotinic acid isomer, 1, and structural
information was available for 2, isonicotinic acid was chosen
as the substrate analogue for the porphyrin-based probe.
Preparation of the porphyrin derivatives
Our approach to the synthesis of the porphyrin derivatives
involved coupling of 5,10,15-tris(4-pyridyl)-20-(4-aminophenyl)-
porphyrin, prepared using the method described by Li et al., to
2-methoxy-4-nitrobenzoic acid using the DCC-mediated
method described for TPP derivatives (Scheme 4).21
In order to compare the binding affinity of the compounds
prepared, enzyme inhibitor studies were carried out.22 The
percent inhibition is calculated from eqn (1), where v is
the initial rate of formation of uric acid monitored at the
absorption maximum, 295 nm with inhibitor and v0 is the
Compound 3 was obtained, in 79% yield, using the
anhydride procedure previously described.19 Nitro compound
3 was methylated with methyl iodide as described by Li et al.21
Methylation in all three positions was shown to be complete
by 1H NMR spectroscopy and ESI-MS. The iodide salt, [4][I]3,
was reasonably soluble in water and ethanol. [4][I]3 was
reduced with indium and ammonium chloride in 64% yield
and precipitated as the hexafluorophosphate salt to yield
[5][PF6]3.26 The 1H NMR spectrum and ESI-MS indicated
that the free base compound, not the indium porphyrin, was
present at this stage.19 Water solubility of the compound could
be achieved by conversion to the chloride salt [5][Cl]3.
Table 2 IC50 and Ki values for each of the compounds with XO.
(Errors in IC50 Æ 10%)
Compound
IC50/mM
Ki/mM
Allopurinol
Methyl 4-amino-2-methoxybenzoate
4-Amino-2-methoxybenzoic acid
Nicotinic acid
Isonicotinic acid
1
2
17
163
416
946
1270
179
130
81
0.98
9.4
24
54
73
10
Isonicotinic acid itself was not sufficiently soluble in the
solvents appropriate for the DCC-mediated coupling proce-
dure to attach the linker. Therefore commercially available
7.5
4.6
6.5
[7][Cl]3
[8][Cl]3
113
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
1130 | New J. Chem., 2010, 34, 1125–1134

View Article Online
Miyatani and Amao.27 The progress of the reaction could be
monitored by UV-Vis spectroscopy. The metalloporphyrins
were purified by addition of ammonium hexafluorophosphate
solution to precipitate [7][PF6]3 and [8][PF6]3. The water
soluble derivatives could be obtained once more by adding
tetrabutylammonium chloride to a solution of [7][PF6]3 or
[8][PF6]3 in acetonitrile yielding the porphyrin–linker conjugate,
[7][Cl]3, and probe compound, [8][Cl]3, respectively.
Characterisation
1
Porphyrins 3–8 were characterized by 1D H and 13C NMR
spectroscopy and 2D 1H–1H COSY and 1H–13C HMQC
and HMBC experiments. The spectra of the water-soluble
porphyrins were recorded in d6-DMSO, since protic solvents
such as D2O and CD3OD exchange with the NH protons.
Despite broadening of the peaks, the spectra were of higher
quality for the chloride salts in d6-DMSO than the PF6À salts
in CD3CN. The assignments are provided in the ESI.w
For free base meso-5,10,15-tris(4-pyridyl)-20-(4-acetamido-
phenyl)porphyrin the amide NH resonance was located at
d 7.74 whereas for compound 3 the resonance corresponding
to the amide NH protons was shifted further downfield
(d E 10) suggestive of intermolecular hydrogen bonding. In
general the signals for compound [4]3+ were broader than for
compound 3 due to the more viscous solvent, d6-DMSO. The
amide NH proton was shifted considerably downfield and the
inner protons were shifted considerably upfield relative to
those of 3, possibly due to hydrogen bonding with the solvent.
The 1H NMR spectra of [6]3+ included additional doublets
at d 8.35 and at d 9.01 corresponding to the pyridyl protons
that were absent in [5]3+. Two amide signals are located at
d 11.35 and 10.58. The resonance of the NH nearest to the
phenyl linker is shifted further downfield due to hydrogen
bonding to the neighbouring methoxy compared to that of the
NH nearest to the pyridyl substrate for which there is no
neighbouring hydrogen bond acceptor. The resonance of the
latter is shifted downfield compared to that recorded for
compound 2 (d 8.04) possibly due to hydrogen bonding to
the solvent (2 was recorded in CDCl3).
In the ESI-MS, the M + H+ adduct was observed for 1 and
2 but a different ionisation pathway was observed for 4.
A peak at m/z = 856 corresponding to reduction of the
porphyrin, possibly by iodide (M3+ + 2eÀ), was observed.
The rest of the porphyrins were already charged and therefore
the corresponding M3+ ions were observed.
Table 3 Porphyrin Q and Soret (S) absorption maxima (nm) and
absorption coefficients e (103 dm3 molÀ1 cmÀ1) in 0.5 mM potassium
phosphate buffer, pH 7.5
S
Q1
lmax (e)
Q2
lmax (e)
Q3
lmax (e)
Q4
lmax (e)
lmax (e)
Scheme 4 Synthesis of porphyrin–linker conjugates [7][Cl]3 and
[8][Cl]3.
3a
TMPyP 422 (235) 518 (16.3) 554 (6.43) 585 (7.52) 640 (2.00)
422 (167) 516 (8.21) 552 (4.14) 590 (2.71) 647 (1.79)
[4]I3
431 (142) 524 (12.3) 564 (7.71) 592 (6.35) 649 (2.63)
429 (124) 522 (9.78) 564 (8.75) 595 (5.84) 653 (3.51)
434 (117) 524 (11.6) 566 (11.8) 598 (8.12) 620 (6.52)
436 (108) 565 (10.3) 615 (6.04)
[5]Cl3
[6]Cl3
[7]Cl3
[8]Cl3
isonicotinic acid anhydride was used directly and compound
[6][Cl]3 was prepared in 76% yield after purification.
442 (110) 569 (11.5) 614 (6.52)
The porphyrins [5][Cl]3 and [6][Cl]3 were metallated
with Zn(OAc)2Á2H2O following the procedure reported by
a
3 in chloroform.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1125–1134 | 1131

View Article Online
Fig. 3 Absorption spectrum for XO and emission spectrum for XO
(lex = 325 nm).
the absorption and emission spectra for XO in phosphate
buffer (50 mM, pH 7.5). The porphyrin absorption bands
occur at longer wavelength than the absorption maximum for
the enzyme. Similarly, the emission bands are at longer
wavelength for porphyrins 5–8 than that of XO. These
properties are essential for an efficient probe.
A solution of XO in phosphate buffer (bovine milk, grade III,
22 mg cmÀ3, 1.30 U cmÀ3, 10 ml) was added to each of the
water-soluble porphyrins, 5–8. In each case the absorption
decreased by B30% and a small red-shift (1–3 nm) was
observed. Fig. 4a illustrates the spectrum of 5 in the presence
of XO. This was matched by a 30% decrease in emission
Fig. 2 Corrected and optically matched emission spectra in 50 mM
potassium phosphate buffer, pH 7.5: (a) [4][I]3, [6][Cl]3, lex = 420 nm,
[5][Cl]3; (b) [7][Cl]3 and [8][Cl]3, lex = 435 nm.
Each porphyrin displayed characteristic absorption bands
in the visible region with significant variation with the
functional groups i.e. nitro, amine vs. pyridine (Table 3).
The extinction coefficients for the zinc porphyrins were lower
than those of the free base porphyrins, possibly as a result of
coordination to water. The absorption spectra for compounds
[4]3+, [5]3+ and [6]3+ were all substantially red-shifted
compared to those of 3 and TMPyP. The Soret bands of
probes [7]3+ and [8]3+ are both red-shifted relative to the
corresponding free base compounds, causing their green colour.
Compounds 4–8 dissolved in phosphate buffer each gave
broad emission bands between 600 and 800 nm upon excitation
at 420 nm. The quantum yields were lower than that of TMPyP
but were sufficiently high to compete with the absorption from
XO under assay conditions. Fig. 2a shows the emission spectra
of free base porphyrins [4]3+, [5]3+ and [6]3+. The emission
spectrum of compound [4]3+ was similar to that of TMPyP in
shape and emission maximum; the quantum yield was ca. 50%
of that of TMPyP. The quantum yields of compounds 5 and 6
were much lower and the spectra contained a more conspicuous
shoulder at shorter wavelength. The emission spectra for the
zinc porphyrins 7 and 8 (Fig. 2b) are similar with respect to
intensity and lmax. The spectra are blue-shifted and more
intense relative to those of the free base derivatives.
Effects of XO on the probe absorption and emission spectra
Enzyme affinity bioassays were carried out for probes 7 and 8.
The IC50 and Ki for each are listed in Table 2 and Fig. 3 shows
Fig. 4 Effect of adding XO to [5][Cl]3 in phosphate buffer pH 7.5; (a)
absorption spectra; (b) corrected emission spectra (lex = 420 nm).
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
1132 | New J. Chem., 2010, 34, 1125–1134

View Article Online
compounds 3 and 4 and extended porphyrin–linker–substrate
compounds 5–8.
4-Amino-2-methoxybenzoic acid was chosen as the linker
since it contains aromatic rings and possible hydrogen bond
donor and acceptor groups. It can therefore interact with
residues lining the access channel leading to the active site of
the enzyme. The neutral ester derivative, methyl 4-amino-2-
methoxybenzoate, was shown to bind to XO with Ki
=
9.4 mM. Since the enzyme catalyses the oxidation of aromatic,
nitrogen containing heterocycles, pyridine derivatives were
also studied. Isonicotinic acid was shown to bind to the
enzyme with Ki = 73 mM. However, when coupled to the
linker (2) the binding was improved to Ki = 7.5 mM, making it
a suitable choice as the substrate analogue.
The porphyrin luminophore absorbed strongly in the visible
region of the electromagnetic spectrum where there is little
competition with absorption of the chromophore of XO. The
chloride salts of the pyridinium derivatives were highly water-
soluble, which is required for the XO affinity bioassay. There is
also a possibility for interaction of the cationic groups with the
negatively charged surface areas of the protein.10 Zinc was
incorporated to give an electron donating metalloporphyrin. It
was anticipated that photoinduced electron transfer from the
metalloporphyrin to the molybdenum cofactor active site on
excitation into the probe bands would result in quenching of
the luminescence and would therefore signal binding of
the probe.
Fig. 5 Effect of adding XO to [8][Cl]3 in phosphate buffer pH 7.5;
(a) absorption spectra; (b) corrected emission spectra (lex = 570 nm).
Enzyme affinity bioassays were carried out on the porphyrin
compounds to investigate their ability to bind to XO. The IC50
for the substrate-analogues and substrate–linker conjugates in
xanthine oxidase affinity assays were measured by spectro-
photometric rate determinations. The values obtained were
used to aid design of [8][Cl]3. The Ki for each of the
compounds was found to be in the micromolar range
(Table 2). The binding affinity for the substrate–linker
conjugates was much higher than that for the linker and
substrate analogues themselves. Ki for the reference compound,
allopurinol, was found to be 0.98 mM, in agreement with the
value reported by Ho and Clifford, Ki = 0.91 mM.28
The IC50 of [7][Cl]3 (81.2 mM) was approximately five times
that of allopurinol (IC50 17 mM), suggesting that the porphyrin–
linker conjugate already has a reasonable affinity for xanthine
oxidase. The value was half that of the linker alone, possibly
due to improved water-solubility and additional ionic inter-
actions of the cationic porphyrin periphery with the anionic
residues around the channel on the enzyme surface.10 Since
compound [7][Cl]3 was not displaced by allopurinol, the
channel may be blocked by the porphyrin, preventing access.
ZnTMPyP did not inhibit XO and, therefore, inhibition by
compound [7][Cl]3 is likely to be associated with interactions
between the enzyme and the probe ‘arm’, linker and substrate
moieties.
intensity (Fig. 4b). The absorption spectrum for [8]3+ was also
red-shifted on addition of XO (441–446 nm) (Fig. 5a). The
relative quantum yields were compared by examining the ratio
of the integrated emission intensity (Fig. 5b) per unit of
absorbance; the quantum yield was reduced by 25%. These
observations show that the porphyrin does report the presence
of XO by a change in luminescence and absorption. Addition
of allopurinol did not result in restoration of the luminescence.
Therefore it has not been conclusively proven that the
quenching is due to binding to the active site and not just to
the enzyme surface. However, neither of these effects was
observed for TMPyP, suggesting that the change is not due
to interaction with the enzyme surface alone.
Discussion
Water-soluble zinc porphyrin derived probes, designed to
report binding to xanthine oxidase (XO), have been prepared
according to Scheme 4. The probes are composed of a
porphyrin luminophore, a linker and a substrate analogue.
A straightforward strategy was used to couple the aromatic
units sequentially, using amide coupling chemistry based
on the experiments with triphenyl porphyrins reported
previously.19 The need for amine protecting groups was
avoided by using the corresponding nitro compound and
selectively reducing the coupled species with indium and
ammonium chloride to give the amine. This procedure was
suitable for preparing the simple substrate–linker compounds
1 and 2 in addition to neutral and cationic porphyrin–linker
For [7][Cl]3, the Ki = 4.6 mM is lower than that of [8][Cl]3
(Ki = 6.5 mM). The longer probe ‘arm’ of 8 may affect the
binding in the channel in two ways: matching to the depth of
the channel and the potential for direct interaction with the
molybdenum active site. Nonetheless an IC50 in the 100 mM
scale is suitable for an application as a probe for XO since
it is sufficient to be measured but low enough to allow
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010 New J. Chem., 2010, 34, 1125–1134 | 1133

View Article Online
displacement by strong inhibitors (IC50o1 mM). The absorp-
tion spectrum for the probe, [8][Cl]3, was red-shifted in the
presence of XO and the quantum yield was 25% lower than in
the absence of the enzyme.
D. C. M. Ng, Coord. Chem. Rev., 2006, 250, 1724; K. K. W. Lo
and J. S. Y. Lau, Inorg. Chem., 2007, 46, 700.
4 S. M. Contakes, Y. H. L. Nguyen, H. B. Gray, E. C. Glazer,
A.-M. Hays and D. B. Goodin, Struct. Bonding, 2007, 123, 177;
Y. H. L. Nguyen, J. R. Winkler and H. B. Gray, J. Phys. Chem. B,
2007, 111, 6628; E. C. Glazer, Y. H. L. Nguyen, H. B. Gray and
D. B. Goodin, Angew. Chem., Int. Ed., 2008, 47, 898; C. A. Whited,
W. Belliston-Bittner, A. R. Dunn, J. R. Winkler and H. B. Gray,
J. Inorg. Biochem., 2009, 103, 906; C. Shih, A. K. Museth,
M. Abrahamsson, A. M. Blanco-Rodriguez, A. J. Di Bilio,
J. Sudhamsu, B. R. Crane, K. L. Ronayne, M. Towrie, A. Vlcek
Jr., J. H. Richards, J. R. Winkler and H. B. Gray, Science, 2008,
320, 1760.
Conclusions
The compounds described here have been shown to interact
with XO. The porphyrins report interaction with the protein
by a change in the absorption and luminescence properties, as
anticipated. Compound 8 had a lower affinity for XO (IC50
=
5 R. Hille, Chem. Rev., 1996, 96, 2757.
6 J. C. Enroth, B. T. Eger, K. Okamoto, T. Nishino, T. Nishino and
E. F. Pai, Proc. Natl. Acad. Sci. U. S. A., 2000, 97, 10723.
7 T. Nishino, K. Okamoto, Y. Kawaguchi, H. Hori, T. Matsumura,
B. T. Eger, E. F. Pai and T. Nishino, J. Biol. Chem., 2005, 280, 24888.
8 R. Hille and R. F. Anderson, J. Biol. Chem., 2001, 276, 31193.
113, Ki = 6.5 mM) than compounds 7 and methyl 4-amino-2-
methoxybenzoate. Despite its weaker binding to XO, the
longer probe ‘arm’ should approach more closely or bind
directly to the active site molybdenum. As a result there is
more opportunity for quenching of the luminescence by
electron transfer. Indeed, addition of XO to an aqueous
solution of [8][Cl]3 did cause quenching of the porphyrin
emission and a red shift in the absorption spectrum. The Ki
is in the low micromolar region which may be suitable for use
in screening for XO inhibitors, where the displacement of the
probe by active analytes (such as febuxostat, Ki = 20 nM)29
would restore the porphyrin luminescence provided that the
access is not prevented by the porphyrin unit.
9 E.-Y. Choi, A. L. Stockert, S. Leimkuhler and R. Hille, J. Inorg.
Biochem., 2004, 98, 841.
¨
10 J. Truglio, K. Theis, S. Leimkuhler, R. Rappa, K. V. Rajagopalan
¨
and C. Kisker, Structure, 2002, 10, 115; R. Hille, Arch. Biochem.
Biophys., 2005, 433, 107; J. Doonan, A. Stockert, R. Hille and
G. N. George, J. Am. Chem. Soc., 2005, 127, 4518.
11 M. M. Kittleson and J. M. Hare, Eur. Heart J., 2005, 26, 1458.
12 F. Borges, E. Fernandes and F. Roleira, Curr. Med. Chem., 2002,
9, 195; Okamoto, B. T. Egers, T. Nishino, S. Kondo, E. F. Pai and
T. Nishino, J. Biol. Chem., 2003, 278, 1848; K. Okamoto,
K. Matsumoto, R. Hille, B. T. Eger, E. F. Pai and T. Nishino,
Proc. Natl. Acad. Sci. U. S. A., 2004, 101, 7931; K. Okamoto,
A. Fukunari, T. Nishino, B. T. Eger, E. F. Pai, M. Kamezawa,
I. Yamada and N. Kato, J. Pharmacol. Exp. Ther., 2004, 311, 519;
T. A. Krenitsky, W. W. Hall, P. de Miranda, L. M. Beauchamp,
H. J. Schaeffer and P. D. Whiteman, Proc. Natl. Acad. Sci.
U. S. A., 1984, 81, 3209; T. Sato, N. Ashizawa, T. Iwanaga,
H. Nakamura, K. Matsumoto, T. Inoue and O. Nagata, Bioorg.
Med. Chem. Lett., 2009, 19, 184.
13 P. Keith and W. R. Gilliland, Am. J. Med., 2007, 120, 221.
14 G. Biagi, A. Costantini, L. Costantino, I. Giorgi, O. Livi,
P. Pecorari, M. Rinaldi and V. Scartoni, J. Med. Chem., 1996,
39, 2529.
15 K. Okamoto, K. Matsumoto, R. Hille, B. T. Eger, E. F. Pai and
T. Nishino, Proc. Natl. Acad. Sci. U. S. A., 2004, 101, 7931.
16 K. Kilsa, J. Kajanus, A. N. Macpherson, J. Martensson and
B. Albinsson, J. Am. Chem. Soc., 2001, 123, 3069; M. U. Winters,
K. Pettersson, J. Martensson and B. Albinsson, Chem.–Eur. J., 2005,
11, 562.
The synthetic procedure has been shown to be highly
versatile, enabling synthesis of triphenyl porphyrin derivatives
as well as water-soluble trimethylpyridiniumyl porphyrins.
Further modification of the linker and substrate-analogue,
such as substitution of different N-heterocycles for isonicotinic
acid, should be reasonably straightforward. Likewise,
derivatisation around the porphyrin periphery is not restricted
to methylation of the pyridyl nitrogens.
At this stage we have successfully designed and synthesized
probe molecules that bind to XO. It is not yet proved
conclusively that they bind in the substrate access channel.
Possible future work includes co-crystallisation of the probes
with XO and porphyrin excitation in the presence of indicators
that would show the formation of reactive oxygen species such
as peroxides and superoxides formed during the reaction
where oxygen is the final electron acceptor.
17 N. Paddon-Row, Aust. J. Chem., 2003, 56, 729.
18 V. Massey, P. E. Brumby and H. Komai, J. Biol. Chem., 1969, 244,
1682.
19 E. A. Gibson, PhD thesis, University of York, 2007.
20 W. A. Hewlett, T. de Paulis, N. S. Mason, D. E. Schmidt,
B. L. Trivedi, Z.-J. Zhang and M. H. Ebert, Chem. Pharm. Bull.,
1997, 45, 2079.
Acknowledgements
EAG is grateful to EPSRC for a studentship. We thank
Dr Adrian Whitwood for assistance with the X-ray structure.
21 H. Li, O. S. Fedorova, A. N. Grachev, W. R. Trumble,
G. A. Bohach and L. Czuchajowski, Biochim. Biophys. Acta,
1997, 1354, 252.
22 A. P. Sweeney, S. G. Wyllie, R. A. Shalliker and J. L. Markham,
J. Ethnopharmacol., 2001, 75, 273.
23 G. M. Sheldrick, SHELXS-97, Program for Structure Solution,
Notes and references
1 L. Fabbrizzi and A. Poggi, Chem. Soc. Rev., 1995, 24, 197;
J. F. Callan, A. P. de Silva and D. C. Magria, Tetrahedron,
2005, 61, 8551.
University of Gottingen, Gottingen, Germany, 1997; G. M. Sheldrick,
¨ ¨
SHELXL-97: Program for the Refinement of Crystal Structures,
University of Gottingen, Gottingen, Germany, 1997.
¨
¨
2 E. Meggers, Chem. Commun., 2009, 1001.
24 J. C. Lee, Y. H. Cho, H. K. Lee and S. H. Cho, Synth. Commun.,
1995, 25, 2877.
25 J. Escribano, F. Garcia-Canovas and F. Garcia-Carmona,
Biochem. J., 1988, 254, 829.
26 M. R. Pitts, J. R. Harrison and C. J. Moody, J. Chem. Soc., Perkin
Trans. 1, 2001, 955.
27 R. Miyatani and Y. Amao, Photochem. Photobiol. Sci., 2004, 3, 681.
28 C. Y. Ho and A. J. Clifford, J. Nutr., 1976, 106, 1600.
29 N. L. Edwards, Rheumatology, 2009, 48, ii15–ii19.
3 J. S.-Y. Lau, P.-K. Lee, K. H.-K. Tsang, C. H.-C. Ng, Y.-W. Lam,
S.-H. Cheng and K. K. W. Lo, Inorg. Chem., 2009, 48, 708;
K. K. W. Lo, K. H.-K. Tsang, K.-S. Sze, C.-K. Chung,
T. K.-M. Lee, K. Y. Zhang, W.-K. Hui, C.-K. Li, J. S.-Y. Lau,
D. C.-M. Ng and N. Zhu, Coord. Chem. Rev., 2007, 251, 2292;
K. K. W. Lo, M.-W. Louie, K.-S. Sze and J. S.-Y. Lau, Inorg.
Chem., 2008, 47, 602; K. K. W. Lo, W. K. Hui, C. K. Chung,
K. H. K. Tsang, T. K. M. Lee, C. K. Li, J. S. Y. Lau and
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
1134 | New J. Chem., 2010, 34, 1125–1134