Synthesis, characterization, and metal coordinating ability of multifunctional carbohydrate-containing compounds for Alzheimer's therapy

Article abstract of DOI:10.1021/ja068965r

Dysfunctional interactions of metal ions, especially Cu, Zn, and Fe, with the amyloid-β (Aβ) peptide are hypothesized to play an important role in the etiology of Alzheimer's disease (AD). In addition to direct effects on Aβ aggregation, both Cu and Fe ca

Full text of DOI:10.1021/ja068965r

Published on Web 05/19/2007
Synthesis, Characterization, and Metal Coordinating Ability of
Multifunctional Carbohydrate-Containing Compounds for
Alzheimer’s Therapy
Tim Storr,† Michael Merkel,† George X. Song-Zhao,† Lauren E. Scott,†
David E. Green,† Meryn L. Bowen,† Katherine H. Thompson,† Brian O. Patrick,†
Harvey J. Schugar,*,‡ and Chris Orvig*,†
Contribution from the Medicinal Inorganic Chemistry Group, Department of Chemistry,
UniVersity of British Columbia, 2036 Main Mall, VancouVer, British Columbia, V6T 1Z1,
Canada and Department of Chemistry and Chemical Biology, Rutgers, The State UniVersity of
New Jersey, 610 Taylor Road, Piscataway, New Jersey 08854-8087
Received December 31, 2006; E-mail: orvig@chem.ubc.ca, schugar@rutchem.rutgers.edu
Abstract: Dysfunctional interactions of metal ions, especially Cu, Zn, and Fe, with the amyloid-â (Aâ) peptide
are hypothesized to play an important role in the etiology of Alzheimer’s disease (AD). In addition to direct
effects on Aâ aggregation, both Cu and Fe catalyze the generation of reactive oxygen species (ROS) in
the brain further contributing to neurodegeneration. Disruption of these aberrant metal-peptide interactions
via chelation therapy holds considerable promise as a therapeutic strategy to combat this presently incurable
disease. To this end, we developed two multifunctional carbohydrate-containing compounds N,N′-bis[(5-
â-D-glucopyranosyloxy-2-hydroxy)benzyl]-N,N′-dimethyl-ethane-1,2-diamine (H2GL1) and N,N′-bis[(5-â-D-
glucopyranosyloxy-3-tert-butyl-2-hydroxy)benzyl]-N,N′-dimethyl-ethane-1,2-diamine (H2GL2) for brain-
directed metal chelation and redistribution. Acidity constants were determined by potentiometry aided by
UV-vis and 1H NMR measurements to identify the protonation sites of H2GL1,2. Intramolecular H bonding
between the amine nitrogen atoms and the H atoms of the hydroxyl groups was determined to have an
important stabilizing effect in solution for the H2GL1 and H2GL2 species. Both H2GL1 and H2GL2 were found
to have significant antioxidant capacity on the basis of an in vitro antioxidant assay. The neutral metal
complexes CuGL1, NiGL1, CuGL2, and NiGL2 were synthesized and fully characterized. A square-planar
arrangement of the tetradentate ligand around CuGL2 and NiGL2 was determined by X-ray crystallography
with the sugar moieties remaining pendant. The coordination properties of H2GL1,2 were also investigated
by potentiometry, and as expected, both ligands displayed a higher affinity for Cu2+ over Zn2+ with H2GL1
displaying better coordinating ability at physiological pH. Both H2GL1 and H2GL2 were found to reduce
Zn2+- and Cu2+- induced Aâ1-40 aggregation in vitro, further demonstrating the potential of these
multifunctional agents as AD therapeutics.
Introduction
the disease. More specifically, the amyloid hypothesis postulates
that plaque Aâ depositions, or partially aggregated soluble Aâ,
Alzheimer’s disease (AD) currently affects approximately 2%
of the population in North America with a predicted 3-fold
increase in incidence over the next 50 years.1 Neurodegenerative
diseases, such as AD, Parkinson’s disease (PD), Creutzfeldt
Jakob disease (CJD), and amyotrophic lateral sclerosis (ALS),
are all defined by the progressive loss of neuronal cell
populations, protein aggregation, and extensive evidence of
oxidative stress.2,3 The amyloid hypothesis has long been the
dominant theory to explain the etiology of AD,4 linking the 39-
42 residue amyloid-â (Aâ) peptide to the neurodegeneration in
trigger a neurotoxic cascade resulting in AD pathology.5
Amyloid plaques have been described as “metallic sinks”
because remarkably high concentrations of Cu, Fe, and Zn have
been found within these deposits in AD brains.6,7 Cu, Fe, and
Zn are increased from 3 to 5 times in AD brains as compared
to those in age-matched controls.7 Fenton-type processes involv-
ing redox-active metal ions,8 possibly in concert with impaired
(4) (a) Hardy, J.; Selkoe, D. J. Science 2002, 297, 353. (b) Hardy, J. A.; Higgins,
G. A. Science 1992, 256, 184.
(5) Selkoe, D. J. Physiol. ReV. 2001, 81, 741.
(6) Smith, M. A.; Harris, P. L. R.; Sayre, L. M.; Perry, G. Proc. Natl. Acad.
Sci. U.S.A. 1997, 94, 9866.
† University of British Columbia.
‡ Rutgers University.
(7) Lovell, M. A.; Robertson, J. D.; Teesdale, W. J.; Campbell, J. L.;
Markesbery, W. R. J. Neurol. Sci. 1998, 158, 47.
(1) www.alz.org, 10/13/06; www.alzheimer.ca, 10/13/06.
(2) Barnham, K. J.; Masters, C. L.; Bush, A. I. Nat. ReV. Drug DiscoV. 2004,
3, 205.
(8) (a) Hensley, K.; Floyd, R. A. Arch. Biochem. Biophys. 2002, 397, 377. (b)
Halliwell, B.; Gutteridge, J. M. Biochem. J. 1984, 219, 1. (c) Frederickson,
C. J.; Koh, J. Y.; Bush, A. I. Nat. ReV. Neurosci. 2005, 6, 449. (d) Gaggelli,
E.; Kozlowski, H.; Valensin, D.; Valensin, G. Chem. ReV. 2006, 106, 1995.
(e) Stadtman, E. R. Free Radical Biol. Med. 1990, 9, 315. (f) Bush, A. I.
Trends Neurosci. 2003, 26, 207.
(3) (a) Doraiswamy, P. M.; Finefrock, A. E. Lancet Neurol. 2004, 3, 431. (b)
Binolfi, A.; Rasia, R. M.; Bertoncini, C. W.; Ceolin, M.; Zweckstetter,
M.; Griesinger, C.; Jovin, T. M.; Fernandez, C. O. J. Am. Chem. Soc. 2006,
128, 9893.
9
10.1021/ja068965r CCC: $37.00 © 2007 American Chemical Society
J. AM. CHEM. SOC. 2007, 129, 7453-7463
7453

A R T I C L E S
Storr et al.
Chart 1. Structures of Metal-Ion Chelators Showing Promise as AD Therapeutics
cellular energy metabolism, lead to increased oxidative stress,
a major feature of age-related neurodegenerative diseases. The
interaction of Aâ with metal ions in vitro leads to aggregation
with Zn2+ accelerating the deposition process more rapidly
compared to Cu2+ and Fe3+ at pH 7.4.9 Cu2+-induced Aâ
aggregation is, however, exaggerated under slightly acidic
conditions (pH 6.8), and altered H+ homeostasis (cerebral
acidosis) is hypothesized to occur in AD.10,11 Metal ions have
also been shown to modify the Aâ peptide via cross-linking12
and oxidation.2,13 Further in vitro studies have demonstrated that
Cu2+ and Fe3+ potentiate the neurotoxicity of Aâ via redox
cycling and production of H2O2 in the presence of dioxygen.14,15
Zn and Cu have been found to co-purify with Aâ from AD
brain tissue, unlike Fe.16 Fe in plaques is most likely involved
in neuritic processes (for example, complexed to ferritin17) and
thus may not directly interact with Aâ. However, oxygen-
mediated production of ROS is often associated with increased
labile Fe stores,7 and Fe in human amyloid deposits has, in fact,
been shown to be redox active.6 Surprisingly, a recent in vitro
study highlighted the potential role of Fe in blocking Cu-
mediated neurotoxicity.18 The extensive evidence of disrupted
iron metabolism in AD still leaves many unanswered questions
about the role of this particular metal ion in the disease.19
The mounting evidence supporting the role(s) of metal ions
in the pathophysiology of AD has rendered metal-ion chelation
therapy a promising treatment strategy. Appropriate chelators
can be designed to potentially minimize oxidative stress as well
as disrupt Aâ pathology. In vitro studies have shown that
production of H2O2 via interaction of Cu2+ and Fe3+ with Aâ1-42
is significantly attenuated by the presence of the chelators
diethylenetriaminepentaacetic acid (DTPA), triethylenetet-
raamine (TETA), and desferrioxamine (DFO, Chart 1).14 Treat-
ment of the Tg2576 transgenic mouse (model for AD) with
clioquinol (Chart 1), an 8-hydroxyquinoline chelator, reduced
brain amyloid accumulation,20 and the effect was recently
reproduced with the lipophilic chelator DP-109 (Chart 1).21 An
early clinical study with the strong metal-ion chelator DFO
showed a significant decrease in the rate of decline of treated
subjects compared with the control group over a 24-month
period.22 Clioquinol has also shown indications of slowing
neurological decline in early-stage clinical trials.23 Although
metal chelating agents may represent a promising therapeutic
strategy, long-term use of strong chelators that are not tissue
specific, such as DFO, likely affects the homeostasis of
numerous biometals and the normal physiological functions of
essential metal-requiring biomolecules such as metalloen-
zymes.24 Specific, rather than systemic, chelation of excess
metals in the brain of AD patients is therefore highly preferred.
Ligands with intermediate affinity may be well suited to this
(9) Bush, A. I.; Pettingell, W. H.; Multhaup, G.; Paradis, M. d.; Vonsattel,
J.-P.; Gusella, J. F.; Beyreuther, K.; Masters, C. L.; Tanzi, R. E. Science
1994, 265, 1464.
(10) Atwood, C. S.; Moir, R. D.; Huang, X.; Scarpa, R. C.; Bacarra, N. M.;
Romano, D. M.; Hartshorn, M. A.; Tanzi, R. E.; Bush, A. I. J. Biol. Chem.
1998, 273, 12817.
(11) Yates, C. M.; Butterworth, J.; Tennant, M. C.; Gordon, A. J. Neurochem.
1990, 55, 1624.
(12) (a) Barnham, K. J.; Haeffner, F.; Ciccotosto, G. D.; Curtain, C. C.; Tew,
D.; Mavros, C.; Beyreuther, K.; Carrington, D.; Masters, C. L.; Cherny,
R. A.; Cappai, R.; Bush, A. I. FASEB J. 2004, 18. (b) Atwood, C. S.;
Perry, G.; Zeng, H.; Kato, Y.; Jones, W. D.; Ling, K. Q.; Huang, X. D.;
Moir, R. D.; Wang, D. D.; Sayre, L. M.; Smith, M. A.; Chen, S. G.; Bush,
A. I. Biochemistry 2004, 43, 560.
(13) Barnham, K. J. et al. J. Biol. Chem. 2003, 278, 42959.
(14) Huang, X.; Atwood, C. S.; Hartshorn, M. A.; Multhaup, G.; Goldstein, L.
E.; Scarpa, R. C.; Cuajungco, M. P.; Gray, D. N.; Lim, J.; Moir, R. D.;
Tanzi, R. E.; Bush, A. I. Biochemistry 1999, 38, 7609.
(15) (a) da Silva, G. F. Z.; Tay, W. M.; Ming, L. J. J. Biol. Chem. 2005, 280,
16601. (b) Huang, X. et al. A. I. J. Biol. Chem. 1999, 274, 37111.
(16) Opazo, C.; Huang, X. D.; Cherny, R. A.; Moir, R. D.; Roher, A. E.; White,
A. R.; Cappai, R.; Masters, C. L.; Tanzi, R. E.; Inestrosa, N. C.; Bush, A.
I. J. Biol. Chem. 2002, 277, 40302.
(17) Grundkeiqbal, I.; Fleming, J.; Tung, Y. C.; Lassmann, H.; Iqbal, K.; Joshi,
J. G. Acta Neuropathol. 1990, 81, 105.
(19) Bishop, G. M.; Robinson, S. R.; Liu, Q.; Perry, G.; Atwood, C. S.; Smith,
M. A. DeV. Neurosci. 2002, 24, 184.
(20) Cherny, R. A. et al. Neuron 2001, 30, 665.
(21) Lee, J. Y.; Friedman, J. E.; Angel, I.; Kozak, A.; Koh, J. Y. Neurobiol.
Aging 2004, 25, 1315.
(22) Crapper McLachlan, D. R.; Dalton, A. J.; Kruck, T. P.; Bell, M. Y.; Smith,
W. L.; Kalow, W.; Andrews, D. F. Lancet 1991, 337, 1304.
(23) (a) Regland, B.; Lehmann, W.; Abedini, I.; Blennow, K.; Jonsson, M.;
Karlsson, I.; Sjogren, M.; Wallin, A.; Xilinas, M.; Gottfries, C. G. Dement.
Geriatr. Cogn. Disord. 2001, 12, 408. (b) Ritchie, C. W. et al. Arch. Neurol.
2003, 60, 1685.
(24) (a) Singh, S.; Khodr, H.; Taylor, M. I.; Hider, R. C. In Free Radicals and
OxidatiVe Stress: EnVironment, Drugs and Food AdditiVes; Rice-Evans,
C., Halliwell, B., Lunt, G. G., Eds.; Portland Press: London, 1995; pp
127. (b) Porter, J. B.; Huens, E. R. Bailliere’s Clin. Haematol. 1989, 2,
459.
(18) White, A. R.; Barnham, K. J.; Huang, X.; Voltakis, I.; Beyreuther, K.;
Masters, C. L.; Cherny, R. A.; Bush, A. I.; Cappai, R. J. Biol. Inorg. Chem.
2004, 9, 269.
9
7454 J. AM. CHEM. SOC. VOL. 129, NO. 23, 2007

Multifunctional Ligands for Alzheimer’s Therapy
A R T I C L E S
role; they would be capable of disrupting low affinity but
pathologically relevant metal-peptide interactions.
hydroxypyridinone glycoconjugate32 highlights the potential of
facilitated carbohydrate transport across the BBB.
Tetrahydrosalen compounds were investigated as chelators
in this work as the tetradentate N2O2 donor set is known to have
a strong affinity for metal ions,38,39 stronger than for the
hydroxypyridinone class of chelators.32,40 In addition, the
phenolic moieties of H2GL1,2 have the potential to act as
antioxidants.41-43 Antioxidant molecules such as vitamin E and
nonsteriodal anti-inflammatory drugs (NSAIDs) have been found
to slow neurological decline in AD trials.44 The development
of multifunctional metal-ion chelating antioxidant molecules
with attached carbohydrate moieties offers an attractive strategy
for Alzheimer’s treatment.
There is considerable promise in enhancing the targeting and
efficacy of metal chelators through ligand design.25 Reports of
chelators with additional antioxidant properties,26 Aâ peptide
intercalation ability,27 and amyloid binding properties28 are
interesting developments in the field of AD therapeutics. The
design of a pro-chelator, triggered by H2O2, was recently shown
to inhibit Fe-promoted OH• radical formation, a significant
advance in the design of selective metal passivating agents.29
In this work we report the synthesis and preliminary evaluation
of two carbohydrate-containing metal-ion chelators H2GL1,2
(Chart 1).
Linking of carbohydrates to drug molecules, forming new
derivatives and/or pro-drugs, offers the potential to increase
water solubility, minimize toxicity, and improve targeting. The
brain requires a significant amount of glucose to maintain normal
bodily functions (up to 30% of total body glucose consump-
tion30), and this demand is met by the high density of hexose
transporters (GLUTs)31 at the blood brain barrier (BBB). The
GLUTs, including GLUT1, at the BBB also offer the potential
for transporter-facilitated drug delivery31 for increased brain
access to drug molecules. The substrate specificity of GLUT-1
may limit the brain uptake of H2GL1,2 by this approach;
however, less selective glucose transporters, including GLUT-3
and GLUT-4, have been identified on BBB endothelial cells.
A recent report of a series of hydroxypyridinone glycoconjugate
(Chart 1) pro-ligands that, once enzymatically deprotected, act
as selective, tissue-dependent metal binders, as well as ROS
scavengers, is an intriguing development in AD therapy.32 This
glycoconjugate approach has been used in an effort to enhance
the central nervous system (CNS) targeting of anticonvulsants,33
analgesics,34 dopamine derivatives,35 anti-cancer agents,36 and
HIV therapies.37 Positive brain uptake by a radiolabeled
Experimental Section
General Methods. All solvents and chemicals (Fisher, Aldrich) were
reagent grade and used without further purification unless otherwise
specified. Water was deionized, purified (Barnstead D9802 and D9804
cartridges), and distilled with a Corning MP-1 Mega-Pure Still. DCl
and NaOD were purchased from Cambridge Isotope Laboratories.
Atomic absorption standards Cu(NO3)2 and ZnCl2 were purchased from
Aldrich and used directly in the potentiometry experiments. The Aâ1-40
peptide was purchased from Advanced ChemTech (Louisville, KY).
1H and 13C{1H} NMR spectra were recorded on a Bruker AV-300 or
AV-400 instrument at 300.13 (75.48 for 13C NMR) or 400.13 (100.62
for 13C NMR) MHz, respectively. Mass spectra (positive ion) were
obtained on a Kratos Concept II H32Q instrument (Cs+, LSIMS), a
Macromass LCT (electrospray ionization) instrument, or a Bruker Biflex
IV (MALDI) instrument. C, H, and N analyses were performed at UBC
by M. Lakha (Carlo Erba analytical instrumentation). Room-temperature
(293 K) magnetic susceptibilities were measured on a Johnson Matthey
balance using Hg[Co(NCS)4] as the susceptibility standard; diamagnetic
corrections were estimated using Pascal’s constants.45 Frozen solution
X-band EPR spectra were recorded on a Bruker ECS-106 EPR
spectrometer in 4-mm diameter quartz tubes. The temperature (∼130
K) was maintained by liquid nitrogen flowing through a cryostat in
conjunction with a Eurotherm B-VT-2000 variable-temperature control-
ler. The microwave frequency and magnetic field were calibrated with
an EIP 625A microwave frequency counter and a Varian E500
gaussmeter, respectively. Computer simulations of the EPR spectra were
performed using the XSophe-Sophe-XeprView simulation software
suite.46 UV-vis spectra were recorded using a Hewlett-Packard 8543
diode array spectrometer.
(25) Storr, T.; Thompson, K. H.; Orvig, C. Chem. Soc. ReV. 2006, 35, 534.
(26) (a) Bebbington, D.; Monck, N. J. T.; Gaur, S.; Palmer, A. M.; Benwell,
K.; Harvey, V.; Malcolm, C. S.; Porter, R. H. P. J. Med. Chem. 2000, 43,
2779. (b) Ji, H. F.; Zhang, H. Y. Bioorg. Med. Chem. Lett. 2005, 15, 1257.
(27) Cherny, R. A.; Barnham, K. J.; Lynch, T.; Volitakis, I.; Li, Q. X.; McLean,
C. A.; Multhaup, G.; Beyreuher, K.; Tanzi, R. E.; Masters, C. L.; Bush, A.
I. J. Struct. Biol. 2000, 130, 209.
(28) Dedeoglu, A.; Cormier, K.; Payton, S.; Tseitlin, K. A.; Kremsky, J. N.;
Lai, L.; Li, X. H.; Moir, R. D.; Tanzi, R. E.; Bush, A. I.; Kowall, N. W.;
Rogers, J. T.; Huang, X. D. Exp. Gerontol. 2004, 39, 1641.
(29) Charkoudian, L. K.; Pham, D. M.; Franz, K. J. J. Am. Chem. Soc. 2006,
128, 12424.
(30) La Manna, J. C.; Harik, S. I. Brain Res. 1985, 326, 299.
(31) Qutub, A. A.; Hunt, C. A. Brain Res. ReV. 2005, 49, 595.
(32) Schugar, H.; Green, D. E.; Bowen, M. L.; Scott, L. E.; Storr, T.; Bohmerle,
K.; Thomas, F.; Allen, D. D.; Lockman, P. R.; Merkel, M.; Thompson, K.
H.; Orvig, C. Angew. Chem., Int. Ed. 2007, 46, 1716-1718; Angew. Chem.
2007, 119, 1746-1748.
(33) Battaglia, G.; La Russa, M.; Bruno, V.; Arenare, L.; Ippolito, R.; Copani,
A.; Bonina, F.; Nicoletti, F. Brain Res. 2000, 860, 149.
(34) Bilsky, E. J.; Egleton, R. D.; Mitchell, S. A.; Palian, M. M.; Davis, P.;
Huber, J. D.; Jones, H.; Yamamura, H. I.; Janders, J.; Davis, T. P.; Porreca,
F.; Hruby, V. J.; Polt, R. J. Med. Chem. 2000, 43, 2586.
(35) (a) Fernandez, C.; Nieto, O.; Rivas, E.; Montenegro, G.; Fontenla, J. A.;
Fernandez-Mayoralas, A. Carbohydr. Res. 2000, 327, 353. (b) Bonina, F.;
Puglia, C.; Rimoli, M. G.; Melisi, D.; Boatto, G.; Nieddu, M.; Calignano,
A.; La Rana, G.; De Caprariis, P. J. Drug Targeting 2003, 11, 25.
(36) (a) Halmos, T.; Santarromana, M.; Antonakis, K.; Scherman, D. Eur. J.
Pharm. 1996, 318, 477. (b) Dunsaed, C. B.; Dornish, J. M.; Aastveit, T.
E.; Nesland, J. M.; Pettersen, E. O. Anti-Cancer Drugs 1995, 6, 456. (c)
Zhang, M.; Zhang, Z. H.; Blessington, D.; Li, H.; Busch, T. M.; Madrak,
V.; Miles, J.; Chance, B.; Glickson, J. D.; Zheng, G. Bioconjugate Chem.
2003, 14, 709. (d) Luciani, A.; Olivier, J. C.; Clement, O.; Siauve, N.;
Brillet, P. Y.; Bessoud, B.; Gazeau, F.; Uchegbu, I. F.; Kahn, E.; Frija, G.;
Cuenod, C. A. Radiology 2004, 231, 135.
(38) (a) Baran, P.; Bottcher, A.; Elias, H. Z. Naturforsch. 1992, 47b, 1681. (b)
Garcia-Zarracino, R.; Ramos-Quinones, J.; Hopfl, H. J. Organomet. Chem.
2002, 664, 188. (c) Bottcher, A.; Elias, H.; Eisenmann, B.; Hilms, E.; Huber,
A.; Kniep, R.; Rohr, C.; Zehnder, M.; Neuburger, M.; Springborg, J. Z.
Naturforsch. B 1994, 49, 1089. (d) Bottcher, A.; Elias, H.; Glerup, J.;
Neuburger, M.; Olsen, C. E.; Paulus, H.; Springborg, J.; Zehnder, M. Acta
Chem. Scand. 1994, 48, 967. (e) Hinshaw, C. J.; Peng, G.; Singh, R.;
Spence, J. T.; Enemark, J. H.; Bruck, M.; Kristofzski, J.; Merbs, S. L.;
Ortega, R. B.; Wexler, P. A. Inorg. Chem. 1989, 28, 4483. (f) Hormnirun,
P.; Marshall, E. L.; Gibson, V. C.; White, A. J. P.; Williams, D. J. J. Am.
Chem. Soc. 2004, 126, 2688. (g) Klement, R.; Stock, F.; Elias, H.; Paulus,
H.; Pelikan, P.; Valko, M.; Mazur, M. Polyhedron 1999, 18, 3617. (h)
Correia, I.; Dornyei, A.; Jakusch, T.; Avecilla, F.; Kiss, T.; Pessoa, J. C.
Eur. J. Inorg. Chem. 2006, 2819.
(39) Gruenwedel, W. Inorg. Chem. 1968, 7, 495.
(40) Martell, A. E.; Smith, R. M. Critical Stability Constants; Plenum: New
York, 1974-1989; Vols. 1-6.
(41) Burton, G. W.; Ingold, K. U. J. Am. Chem. Soc. 1981, 103, 6472.
(42) Ponticelli, F.; Trendafilova, A.; Valoti, M.; Saponara, S.; Sgaragli, G. P.
Carbohydr. Res. 2001, 330, 459.
(43) Valoti, M.; Sipe, H. J.; Sgaragli, G.; Mason, R. P. Arch. Biochem. Biophys.
1989, 269, 423.
(44) (a) Sano, M.; Ernesto, C.; Thomas, R. G.; Klauber, M. R.; Schafer, K.;
Grundman, M.; Woodbury, P.; Growdon, J.; Cotman, D. W.; Pfeiffer, E.;
Schneider, L. S.; Thal, L. J. N. Engl. J. Med. 1997, 336, 1216. (b) Veld, B.
A.; Ruitenberg, A.; Hofman, A.; Launer, L. J.; van Duijn, C. M.; Stijnen,
T.; Breteler, M. M. B.; Stricker, B. H. C. N. Engl. J. Med. 2001, 345,
1515.
(37) (a) Bonina, F.; Puglia, C.; Rimoli, M. G.; Avallone, L.; Abignente, E.;
Boatto, G.; Nieddu, M.; Meli, R.; Amorena, M.; de Caprariis, P. Eur. J.
Pharm. Sci. 2002, 16, 167. (b) Rouquayrol, M.; Gaucher, W.; Greiner, J.;
Aubertin, A. M.; Vierling, P.; Guedj, R. Carbohydr. Res. 2001, 336, 161.
(45) Mabbs, F. E.; Machin, D. J. Magnetism and Transition Metal Complexes;
Chapman and Hall: London, 1961.
9
J. AM. CHEM. SOC. VOL. 129, NO. 23, 2007 7455

A R T I C L E S
Storr et al.
Synthesis. N,N′-Bis[(5-â-D-glucopyranosyloxy-2-hydroxy)benzyl]-
N,N′-dimethyl-ethane-1,2-diamine (H2GL1). A solution of 4-hydrox-
yphenyl-â-D-glucopyranoside (arbutin) 1 (10.01 g, 36.8 mmol), paraform-
aldehyde (1.20 g, 40.00 mmol), and N,N′-dimethyl-1,2-ethane-diamine
2 (1.470 g, 16.68 mmol) in EtOH (150 mL) was heated to reflux for
24 h, and then the solvent was removed in vacuo. The residue was
purified by silica-gel chromatography (1:1 MeOH:EtOAc eluent) to
N,N′-Bis[(5-â-D-glucopyranosyloxy-3-tert-butyl-2-hydroxy)benzyl]-
N,N′-dimethyl-ethane-1,2-diamine (H2GL2). Compound 5 (5.90 g,
7.68 mmol) was dissolved in dry MeOH (100 mL), and NaOMe (1.15
g, 21.3 mmol) was added with stirring. Rexyn (H+ form) was added
after 2 h, and then the mixture was filtered and the solvent removed
under reduced pressure. The crude material was purified by silica gel
chromatography (4:1 EtOAc:MeOH eluent) to afford the product H2GL2
as a white solid (3.81 g, 93%).1H NMR (MeOH-d4, 300 MHz): δ 7.02
1
afford the product H2GL1 as a white solid (4.122 g, 39%). H NMR
(MeOH-d4, 300 MHz): δ 6.96 (dd, 4Jb,d ) 2.8 Hz, 3Ja,b ) 8.7 Hz, 2H;
(d, Jb,d ) 2.8 Hz, 2H; ArH), 6.73 (d, Jb,d ) 2.8 Hz, 2H; ArH), 4.86
4
4
4
3
3
3
2
ArH), 6.90 (d, Jb,d ) 2.8 Hz, 2H; ArH), 6.70 (d, Ja,b ) 8.7 Hz, 2H;
(d, J1,2 ) 7.5 Hz, 2H; H-1), 3.83 (dd, J5,6b ) 1.7 Hz, J6a,6b ) 12.1
Hz, 2H; H-6b), 3.74 (dd, 3J5,6a ) 3.1 Hz, 2J6a,6b ) 12.1 Hz, 2H; H-6a),
3.68 (s, 4H; ArCH2), 3.44 (m, 8H; H-2, H-3, H-4, H-5), 2.66 (s, 4H,
NCH2CH2N), 2.27 (s, 6H, NCH3), 1.39 (s, 18H; C(CH3)3). 13C NMR
(MeOH-d4, 75.48 MHz): δ 153.56, 151.84 (ArC), 139.02 (ArC), 124.52
(ArC), 117.11, 116.93 (ArCH), 104.2 (C1), 78.46 (C3/C5), 75.45 (C2),
71.96 (C4), 63.57, 63.08 (C6/ArCH2), 54.80 (NCH2CH2N), 41.95
(NCH3), 36.16 (C(CH3)3), 30.52 (C(CH3)3). MS (+ES-MS) m/z (relative
intensity) ) 791 ([L + Na]+, 10), 769 ([L + H]+, 100). UV-vis
(MeOH): λmax (ꢀmax) ) 231 nm (1.2 × 104 L mol-1 cm-1), 288 nm
(6.2 × 103 L mol-1 cm-1). Anal. Calcd (found) for C38H60N2O14‚H2O:
C, 58.00 (58.40); H, 7.94 (7.92); N, 3.56 (3.57).
3
3
ArH), 4.77 (d, J1,2 ) 7.5 Hz, 2H; H-1), 3.89 (dd, J5,6b ) 1.7 Hz,
2J6a,6b ) 12.0 Hz, 2H; H-6b), 3.73 (dd, J5,6a ) 5.4 Hz, J6a,6b ) 12.0
Hz, 2H; H-6a), 3.68 (s, 2H; ArCH2), 3.30-3.45 (m, 8H; H-2, H-3,
H-4, H-5), 2.69 (s, 4H, NCH2CH2N), 2.28 (s, 6H, NCH3). 13C NMR
(MeOH-d4, 75.48 MHz): δ 153.91, 152.26 (ArC), 124.50 (ArC),
119.47, 118,70, 117.26 (ArCH), 103.68 (C1), 78.12 (C3/C5), 75.14
(C2), 71.62 (C4), 62.76, 61.36 (C6/ArCH2), 55.03 (NCH2CH2N), 42.06
(NCH3). MS (+ES-MS) m/z (relative intensity) ) 657 ([L + H]+, 100),
373.6 ([L - C13H15O7]+, 40). UV-vis (MeOH): λmax (ꢀmax) ) 226 nm
(1.1 × 104 L mol-1 cm-1), 289 nm (5.4 × 103 L mol-1 cm-1). Anal.
Calcd (found) for C30H44N2O14‚H2O: C, 53.41 (53.65); H, 6.87 (7.12);
N, 4.15 (4.03).
3
2
CuGL1‚2H2O. H2GL1 (0.096 g, 0.15 mmol) and Cu(ClO4)2‚6H2O
(0.054 g, 0.15 mmol) were dissolved in MeOH (5 mL), and aqueous
NaOH (0.5 mL, 1 M) was added with stirring. The resulting green
suspension was stirred for 2 h, and then the solvent was removed in
vacuo. The residue was dissolved in a minimum amount of H2O and
purified by size-exclusion chromatography on Sephadex G-10 (H2O
eluent) to afford the product CuGL1‚2H2O as a dark green solid (0.047
g, 47%). MS (+ES-MS) m/z (relative intensity) ) 720/718 ([M + H]+,
100), 657 ([L + H]+, 20), 557/555 ([M - C6H11O5]+, 50). EPR (130
K, MeOH): A ) 20 × 10-4 cm-1, g ) 2.037, A| ) 179 × 10-4
cm-1, g| ) 2.215. UV-vis (MeOH): λmax (ꢀmax) ) 243 nm (2.8 × 104
L mol-1 cm-1), 290 nm (1.5 × 104 L mol-1 cm-1), 420 nm (2.0 × 103
L mol-1 cm-1), 601 nm (7.0 × 102 L mol-1 cm-1). The room-
temperature solid-state magnetic moment µeff ) 1.74 µB. Anal. Calcd
(found) for C30H42CuN2O14‚2H2O: C, 47.77 (47.82); H, 6.15 (5.90);
N, 3.71 (3.63).
NiGL1‚3H2O. A stirred solution of H2GL1 (0.101 g, 0.15 mmol)
and Ni(ClO4)2‚6H2O (0.055 g, 0.15 mmol) in MeOH (4 mL) turned
red upon addition of NEt3 (0.031 g, 0.31 mmol). After 12 h, addition
of CH3CN (2 mL) caused the product to separate as an oily residue.
The residue was recovered by decantation and redissolved in 5 mL of
H2O The product NiGL1‚3H2O was precipitated as a red-brown solid
following addition of CH3CN (2 mL) and dried (0.027 g, 25%). MS
(+ES-MS) m/z (relative intensity) ) 713 ([M + H]+, 100). Anal. Calcd
(found) for C30H42N2NiO14‚3H2O: C, 46.95 (47.03); H, 6.30 (6.21);
N, 3.65 (3.42).
CuGL2‚2MeOH. Aqueous NaOH (0.5 mL, 1 M) was added to a
green solution of H2GL2 (0.082 g, 0.11 mmol) and Cu(ClO4)2‚6H2O
(0.040 g, 0.11 mmol) in MeOH (4 mL). The solution was stirred for 2
h, and then the solvent was removed in vacuo. The residue was
dissolved in a minimum amount of MeOH and purified by size-
exclusion chromatography on Sephadex G-10 (MeOH eluent) to afford
the product CuGL2‚2MeOH as a dark green solid (0.050 g, 56%). MS
(+ES-MS) m/z (relative intensity) ) 854/852 ([M + Na]+, 100), 830
([M + H]+, 10). EPR (130 K, MeOH): Ax ) 24 × 10-4 cm-1, gx )
2.040, Ay ) 30 × 10-4 cm-1, gy ) 2.020, Az ) 178 × 10-4 cm-1, gz
) 2.215. UV-vis (MeOH): λmax (ꢀmax) ) 247 nm (1.7 × 104 L mol-1
cm-1), 297 nm (1.2 × 104 L mol-1 cm-1), 445 nm (2.7 × 103 L mol-1
cm-1), 620 nm (1.4 × 103 L mol-1 cm-1). The room-temperature solid-
state magnetic moment µeff ) 1.79 µB. Anal. Calcd (found) for C38H58-
CuN2O14‚2MeOH: C, 53.71 (53.38); H, 7.44 (7.30); N, 3.13 (3.23).
NiGL2‚H2O. Triethylamine NEt3 (0.036 g, 0.36 mmol) was added
to a stirred solution of H2GL2 (0.051 g, 0.07 mmol) and Ni(ClO4)2‚
6H2O (0.024 g, 0.07 mmol) in MeOH (2 mL). The color of the reaction
mixture changed from light pink to dark red over a period of 24 h.
3-tert-Butyl-4-hydroxyphenyl-2,3,4,6-tetra-O-acetyl-â-D-glucopy-
ranoside (4). The title compound was synthesized by a method different
from that available in the literature.42 tert-Butyl-hydroquinone 3
(0.535 g, 3.22 mmol) and pentaacetylglucose (1.05 g, 2.69 mmol)
were dissolved in dry CH2Cl2 (30 mL), and with stirring, BF3·OEt2
(1.90 g, 13.41 mmol) was added dropwise. A saturated NaHCO3 (30
mL) solution was added after 4 h, and the resulting mixture was
extracted with CH2Cl2 (2 × 30 mL). The combined organic extracts
were dried over Na2SO4 and filtered, and then the solvent was removed
in vacuo. The crude product was purified by silica-gel chromatography
(1:1 then 2:1 Et2O:pet. ether eluent) to afford the product 4 as a white
solid (0.931 g, 70%). 1H NMR (CHCl3-d1, 300 MHz): δ 6.94 (d, 4Jb,d
4
3
) 2.9 Hz, 1H; ArH), 6.71 (dd, Jb,d ) 2.9 Hz, Ja,b ) 8.5 Hz, 1H;
ArH), 6.57 (d, 3Ja,b ) 8.5 Hz, 1H; ArH), 5.19 (m, 3H; H-2, H-3, H-4),
4.96 (d, 3J1,2 ) 7.5 Hz, 1H; H-1), 4.73 (br. s, 1H; OH), 4.22 (dd, 3J5,6a
) 5.2 Hz, 2J6a,6b ) 12.6 Hz, 1H; H-6a), 4.15 (dd, 3J5,6b ) 2.4 Hz, 2J6a,6b
3
3
3
) 12.6 Hz, 1H; H-6b), (ddd, J4,5 ) 9.5 Hz, J5,6a ) 5.2 Hz, J5,6b
)
2.4 Hz, 2H; H-5), 2.08 (s, 6H; 2× COCH3), 2.04 (s, 3H; COCH3),
2.03 (s, 3H; COCH3), 1.38 (s, 18H; C(CH3)3). MS (+CI-MS) m/z
(relative intensity) ) 514 ([L + NH4]+, 90), 331 ([L - C10H13O2]+,
100). Anal. Calcd (found) for C24H32O11: C, 58.06 (57.95); H, 6.50
(6.47).
N,N′-Bis-(3-(5-tert-butyl-4-hydroxyphenyl-2,3,4,6-tetra-O-acetyl-
â-D-glucopyranoside)benzyl)-N,N′-dimethyl-ethane-1,2-diamine (5).
A solution of 4 (3.90 g, 7.86 mmol), paraformaldehyde (0.26 g, 8.57
mmol), and N,N′-dimethyl-ethane-1,2-diamine 2 (0.315 g, 3.57 mmol)
in dry benzene (40 mL) was stirred and heated to reflux for 22 h. The
solvent was then removed in vacuo, and the residue was purified by
silica-gel chromatography (3:1 Et2O:pet. ether eluent) to afford the
1
product 5 as a white solid (2.49 g, 63%). H NMR (CHCl3-d1, 300
4
MHz): δ 10.60 (s, 2H; OH), 6.87 (d, Jb,d ) 2.7 Hz, 2H; ArH), 6.50
4
(d, Jb,d ) 2.7 Hz, 2H; ArH), 5.25 (m, 6H; H-2, H-3, H-4), 4.94 (d,
3
2
3J1,2 ) 7.5 Hz, 2H; H-1), 4.27 (dd, J5,6a ) 5.2 Hz, J6a,6b ) 12.3 Hz,
2H; H-6a), 4.15 (dd, 3J5,6b ) 2.2 Hz, 3J6a,6b ) 12.3 Hz, 1H; H-6b), 3.81
(ddd, 3J5,6a ) 5.2 Hz, 3J5,6b ) 2.2 Hz, 3J4,5 ) 9.6 Hz, 2H; H-5), 3.61 (s,
4H; ArCH2), 2.59 (s, 4H; NCH2CH2N), 2.25 (s, 6H; NCH3), 2.08 (s,
6H; COCH3), 2.07 (s, 6H; COCH3), 2.04 (s, 6H; COCH3), 2.0 (s, 6H;
COCH3), 1.36 (s, 18H; C(CH3)3). MS (+LSIMS) m/z (relative intensity)
) 1105 ([L + H]+, 30), 552 ([L - C27H38NO11]+, 100). Anal. Calcd
(found) for C54H76N2O22: C, 58.69 (59.11); H, 6.93 (7.07); N, 2.53
(2.69).
(46) (a) Wang, D. M.; Hanson, G. R. J. Mag. Reson., Ser. A 1995, 117, 1. (b)
Hanson, G. R.; Gates, K. E.; Noble, C. J.; Griffin, M.; Mitchell, A.; Benson,
S. J. Inorg. Biochem. 2004, 98, 903.
9
7456 J. AM. CHEM. SOC. VOL. 129, NO. 23, 2007

Multifunctional Ligands for Alzheimer’s Therapy
A R T I C L E S
Table 1. Crystal Data for H2GL2, CuGL2, and NiGL2
Table 2. Selected Bond Lengths (Å) in H2GL2
cryst data
H2GL2
CuGL2
NiGL2
C(19)-C-(38)
N(1)-C(17)
N(1)-C(19)
N(2)-C(36)
N(2)-C(38)
C(10)-O(7)
C(29)-O(14)
1.53(1)
C(7)-O(1)
C(1)-O(1)
C(20)-O(8)
C(26)-O(8)
1.407(8)
1.380(9)
1.398(7)
1.404(7)
formula
C38H68N2O18.25
844.95
triclinic, P1 (#1) triclinic, P1 (#1) monoclinic, P21 (#4)
C38H58CuN2O14 C42H74N2NiO18
830.40 953.74
1.471(8)
1.469(9)
1.468(8)
1.462(9)
1.387(8)
1.374(7)
fw
cryst syst, space
group
a (Å)
b (Å)
10.498 (1)
10.748 (1)
39.414 (3)
87.66 (1)
86.31 (1)
89.89 (2)
4434.3 (8)
4, 1.266
10.691 (2)
8.7559 (9)
13.602 (2)
95.700 (1)
90.080 (1)
92.050 (7)
2349.4 (7)
2, 1.174
14.993 (3)
9.2197 (18)
16.607 (3)
90.00
94.823 (7)
90.00
2287 (7)
2, 1.385
50.0
H bonding
O(7)....N(1)
O(14)....N(2)
2.66(1)
2.68(1)
c (Å)
R (deg)
â (deg)
γ (deg)
V [Å3]
hydrogen concentration probe by titrating known amounts of HCl with
CO2-free NaOH solutions and determining the equivalence point by
Gran’s method.57 Acidity constants for the compounds H2GL1,2 were
determined by titrating 50 mL of aqueous 2.5 mM HCl in the presence
of 0.6 mM H2GL1 or H2GL2 with 0.1 M NaOH. The constants were
calculated with data in the range 4.5 e pH e 10.5 (for H2GL1) and
3.7 e pH e 10.0 (for H2GL2) using the computer program HYPER-
QUAD;58,59 an average of eight independent titrations were used in
determining the final values. The stability constants of Cu2+ and Zn2+
with H2GL1,2 were determined by titrating a solution of 0.6 mM (Cu2+
or Zn2+) and 0.6 mM (H2GL1 or H2GL2) with 0.1 M NaOH. The
stability constants were calculated using HYPERQUAD58,59 with data
in the range 4.5 e pH e 10.0. The acidity constants for H2GL1,2 as
well as the respective metal hydrolysis constants60 were used in the
stability constant calculations. The final values for the stability constants
were an average of at least five independent titrations. Speciation
diagrams for H2GL1,2 and the M2+:H2GL1,2 (M ) Cu2+, Zn2+) systems
were calculated using the program HYSS.59 pM values were calculated
with a locally written Basic program for Cu2+ and Zn2+ (1 mM) and
selected ligands for a ratio of 1:1 M2+:ligand.
Z, Dcalcd [g/cm3]
µ(Mo KR) [cm-1
F000
]
1.00
1824
52.3
882
1024
173 (2)
temp. (K)
min-max, deg
reflns collcd/
unique
173 (2)
173 (2)
θ
1.55, 22.49
53 561/18 652
(Rint ) 0.066)
wR2 ) 0.179
2.16, 27.87
20 637/15 770
(Rint ) 0.044)
wR2 ) 0.213
2.46, 21.96
12 661/5383
(Rint ) 0.089)
wR2 ) 0.189
residuals
(F2, all data)
residuals
R1 ) 0.063
R1 ) 0.082
R1 ) 0.084
(F, I > 2σ(I))
The reaction mixture was filtered, and the filtrate was left to stand for
3 days over which time red crystals of the product NiGL2‚H2O formed.
X-ray-quality crystals were isolated by decanting the supernatant liquid
and washing with a minimum amount of cold MeOH to afford the
crystalline product NiGL2‚H2O (0.012 g, 20%). MS (+ES-MS) m/z
(relative intensity) ) 847 ([M + Na]+, 60), 826 ([M + H]+, 90), 792
([L + Na]+, 10), 769 ([L + H]+, 100). Anal. Calcd (found) for C38H58N2
NiO14‚H2O: C, 54.10 (53.86); H, 7.17 (7.18); N, 3.32 (3.42).
X-ray Crystallography. Crystals of H2GL2 and CuGL2 were grown
from concentrated MeOH/H2O solutions of the respective compounds,
whereas crystals of NiGL2 were obtained via slow evaporation of a
concentrated MeOH solution. The crystals were mounted on a glass
fiber, and measurements were made on a Bruker X8 APEX diffracto-
meter for H2GL2 and a Rigaku/ADSC CCD area detector for CuGL2
and NiGL2 with graphite-monochromated Mo KR radiation. All data
were processed and corrected for Lorentz and polarization effects and
absorption using the SADABS47 program for H2GL2 and the d*TREK48
program for CuGL2 and NiGL2. The structures were solved by direct
methods49 and expanded using Fourier techniques.50 A large number
of disordered water and MeOH molecules were removed from the unit
cell of CuGL2, and the program PLATON/SQUEEZE51 was used to
account for the residual electron density. Final refinements were carried
out using SHELXL-9752 for H2GL2 and teXsan53 for CuGL2 and
NiGL2. The relevant parameters for crystal data, structure solution, and
refinement are summarized in Table 1. CIF files are found in the
Supporting Information.
UV-vis Determination of Acidity Constants. Solutions of either
H2GL1 or H2GL2 (0.125 mM, I ) 0.16 M NaCl, T ) 298 K) were
prepared, the pH of each solution was adjusted, and the UV-vis
spectrum was then recorded. At least 20 readings were taken in the pH
range 8-13.5. Four separate experiments were carried out for each
ligand, and each experiment was monitored at two different wavelengths
(237 and 308 nm for H2GL1; 248 and 310 nm for H2GL2). pKa values
were calculated by fitting the absorption values and corresponding pH
values using a locally written Basic program employing a Newton-
Gauss nonlinear-least-squares curve-fit procedure.
Antioxidant Studies. Compounds H2GL1,2 were tested for their
ability to scavenge free radicals using the trolox equivalent antioxidant
capacity (TEAC) assay.61 An 2,2′-azinobis(3-ethylbenzothiazoline-6-
sulfonic acid (ABTS•+) radical cation decolorization assay61 was used
to determine relative TEAC values. ABTS was dissolved in water (7
mM) and reacted with potassium persulfate (2.45 mM) in the dark for
16 h to form the colored ABTS•+ radical cation. The ABTS•+ solution
was diluted with MeOH to an absorbance value of 0.70 ((0.02) at
740 nm after equilibrating to 30 °C. Solutions of H2GL1,2 in MeOH
(20 µL, 2.5-7.5 µM) were added to 2 mL of ABTS•+ solution to initiate
the reaction. After initial mixing, A740 nm was measured at 30 °C after
Potentiometric Measurements. Equilibrium constants for proto-
nation and complexation reactions with H2GL1,2 were determined by
pH-metric measurements (pH ) -log [H+]) in 0.16 M NaCl at 298 (
0.1 K using a system that has been described previously.54,55 The value
of pKw used was 13.76.56 The glass electrode was calibrated as a
(54) (a) Caravan, P.; Gelmini, L.; Glover, N.; Herring, F. G.; Li, H.; McNeill,
J. H.; Rettig, S. J.; Setyawati, I. A.; Shuter, E.; Sun, Y.; Tracey, A. S.;
Yuen, V. G.; Orvig, C. J. Am. Chem. Soc. 1995, 117, 12759. (b) Song, B.;
Kurokawa, G. S.; Liu, S.; Orvig, C. Can. J. Chem. 2001, 79, 1058. (c)
Song, B.; Storr, T.; Liu, S.; Orvig, C. Inorg. Chem. 2002, 41, 685.
(55) Song, B.; Mehrkhodavandi, P.; Buglyo, P.; Mikata, Y.; Shinohara, Y.;
Yoneda, K.; Yano, S.; Orvig, C. J. Chem. Soc., Dalton Trans. 2000, 8,
1325.
(47) SADABS, Version 2.05; Bruker AXS Inc.: Madison, WI, 1999.
(48) d*TREK: Area Detector Software, Version 7.11; Molecular Structure Corp.,
The Woodlands, TX, 2001.
(49) SIR97: Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G. L.;
Giacovazzo, C.; Guagliardi, A.; Moliterni, A. G. G.; Polidori, G.; Spagna,
R. J. Appl. Crystallogr. 1999, 32, 115.
(50) DIRDIF94: Beurskens, P. T.; Admiraal, G.; Beurskens, G.; Bosman, W.
P.; de Gelder, R.; Israel, R.; Smits, J. M. M. The DIRDIF-94 program
system; Technical Report of the Crystallography Laboratory, University
of Nijmegen: The Netherlands, 1994.
(56) Martell, A. E.; Motekaitis, R. J. Determination and Use of Stability
Constants; VCH: New York, 1988.
(57) Gran, G. Analyst 1952, 77, 661.
(51) PLATON/SQUEEZE. (a) Spek, A. L. Acta Crystallogr. 1990, A46, C34.
(b) Sluis, P. V. D.; Spek, A. L. Acta Crystallogr. 1990, A46, 194-201. (c)
Spek, A. L. PLATON, A Multipurpose Crystallographic Tool; Utrecht
University: Utrecht, The Netherlands, 1998.
(52) SHELXL, Version 5.1; Bruker AXS Inc.: Madison, WI, 1997.
(53) TeXsan: Crystal Structure Analysis Package; Molecular Structure Corp.,
The Woodlands, TX, 1992.
(58) Gans, P.; Sabatini, A.; Vacca, A. Talanta 1996, 43, 1739.
(59) Alderighi, L.; Gans, P.; Ienco, A.; Peters, D.; Sabatini, A.; Vacca, A. Coord.
Chem. ReV. 1999, 184, 311.
(60) Baes, C. F., Jr.; Mesmer, R. E. The Hydrolysis of Cations; R. E. Krieger
Publishing Co.: Malabar, FL, 1986.
(61) Re, R.; Pellegrini, N.; Proteggente, A.; Pannala, A.; Yang, M.; Rice-Evans,
C. Free Radical Biol. Med. 1999, 26, 1231.
9
J. AM. CHEM. SOC. VOL. 129, NO. 23, 2007 7457

A R T I C L E S
Storr et al.
Scheme 1. Synthesis of H2GL1 a
a (a) Paraformaldehyde, EtOH, 39%.
Scheme 2. Synthesis of H2GL2 a
a (a) Pentaacetylglucose, BF3‚OEt2, CH2Cl2, 70%. (b) Paraformaldehyde, benzene, 63%. (c) NaOMe, MeOH, 93%.
Table 3. Deprotonation Constants (pKa’s) of H2GL1,2 Determined
by Potentiometric Measurement in 0.16 M NaCl at 298 K (errors
are for the last digit)
1, 3, and 6 min, with each experiment performed in triplicate. The
percentage change of the absorbance at 740 nm was calculated and
plotted as a function of H2GL1,2 concentration. The slopes were then
compared to the standard, trolox (6-hydroxy-2,5,7,8-tetramethylchro-
mane-2-carboxylic acid), with its TEAC value normalized to 1.
Aâ Turbidity Measurements. Lyophilized synthetic human Aâ1-40
amyloid peptide was prepared as a ∼200 µM solution in distilled,
deionized water. To achieve full peptide dissolution while avoiding
frothing, intermittent sonication (1 min on, 30 s off) was repeated 3
times; the solution was then filtered through a 0.2-µm syringe filter
(Whatman) to remove any microparticulate protein matter. A bicin-
choninic acid (BCA) assay was performed to quantify (relatively, versus
bovine serum albumin, BSA) the concentration of the Aâ peptide.62
Two Chelex-treated 4-(2-hydroxyethyl)-1-piperazineethanesulphonic
acid (HEPES, 20 µM) buffer solutions containing 150 µM NaCl were
prepared with distilled, deionized water to pH values of 6.6 and 7.4.
These solutions were filtered through 0.22-µm acetate filters (Millipore)
to remove any particulate material and used in subsequent preparations
of metal, ligand, and reaction mixture solutions. Stock solutions of Cu2+
and Zn2+ were prepared from standards to final concentrations of 200
µM using pH 6.6 and 7.4 buffers, respectively. Solutions of the free
ligands (DTPA, H2GL1, and H2GL2) were prepared in HEPES buffer
to final concentrations of 200 µM. Turbidity assays were performed in
flat-bottomed 200 µL 96-well microtiter plates (Falcon). To evaluate
the effects of the test ligands, the ligand of interest (50 µM) was added
after a short delay (∼2 min) to a solution of the Aâ1-40 peptide (∼ 25
µM) and either Zn2+ (25 µM) or Cu2+ (25 µM). After a 45-min
incubation, the 405-nm absorbances of all test solutions were measured
using a Labsystems iEMS (Zn2+) or Molecular Devices Thermomax
(Cu2+) microplate reader programmed to agitate the plate for 30 s to
reaction
H2GL1
H2GL2
[HGL1,2]-/[GL1,2
H2GL1,2/[HGL1,2
[H3GL1,2]+/H2GL1,2 + H+ (pKa2)
[H4GL1,2 2+/[H3GL1,2 + + H+ (pKa1)
]
]
2- + H+ (pKa4)
11.3(3)a
10.30(5)b
8.03(6)
13.7(5)a
13.1(5)a
6.47(4)
4.02(9)
- + H+ (pKa3)
]
]
5.01(4)
a Determined spectrophotometrically. b Determined by both spectropho-
tometric (10.2(2)) and potentiometric (10.30(5)) methods.
evenly suspend any aggregates before all readings. In all cases,
appropriate blank trials were run and accounted for; wells containing
Aâ peptide and the respective test ligand in HEPES buffer were used
as blanks for spectrophotometric analysis. MALDI spectroscopic
analysis of selected wells performed after turbidity assay exhibited peaks
consistent with intact Aâ1-40 peptide, confirming that no degradation
of the peptide occurred during the assay.
Results and Discussion
Synthesis and Characterization of the Carbohydrate-
Appended Ligands H2GL1,2. The pro-ligands H2GL1 and
H2GL2 were designed to chelate metal ions in their deprotonated
forms while carrying carbohydrate moieties to enhance solubility
and improve targeting ability. H2GL1 was synthesized in
moderate yield by the coupling of commercially available
arbutin 1 with N,N′-dimethylethylenediamine via a Mannich
condensation63 (Scheme 1).
The synthesis of H2GL2 is shown in Scheme 2. In the first
step tert-butylhydroquinone 3 was glycosylated with pen-
(62) Walker, J. M., Ed. The Protein Protocols Handbook; 2nd ed.; Humana
Press: New Jersey, 2002.
(63) Tramonti.M. Synthesis 1973, 703.
9
7458 J. AM. CHEM. SOC. VOL. 129, NO. 23, 2007

Multifunctional Ligands for Alzheimer’s Therapy
A R T I C L E S
Figure 1. Ellipsoid plot (50% probability, H atoms omitted for clarity) of H2GL2.
spectrophotometry was used to determine pKa values outside
the usable pH range of potentiometry. In addition, as potenti-
ometry does not provide microscopic information involving
identification of protonation sites on the ligands, spectroscopic
methods (UV-vis/1H NMR) were used to assign pKa values to
ionizable sites on the ligands. There are four ionizable groups
of interest on H2GL1 or H2GL2: two tertiary amines and two
phenol moieties as it is unlikely that deprotonation of the OH
groups of glucose occurs in the pH range examined in this
work.64,65
For H2GL1, three acidity constants (pKa1, pKa2, and pKa3)
could be accurately determined by potentiometry while two
acidity constants (pKa3 and pKa4) could be evaluated spectro-
photometrically. The values for pKa3 determined by the two
methods (pKa3 ) 10.2(2) via UV-vis; pKa3 ) 10.30(5) via
potentiometry) were found to agree very well. The first two
acidity constants (pKa1, pKa2) for H2GL1 were assigned to the
Figure 2. Variable pH (pH 9.7-13.5) UV spectra of H2GL1 ([H2GL1] )
0.13 mM, 25 °C, I ) 0.16 M NaCl).
1
tertiary amine moieties on the basis of variable pD H NMR
(data not shown) and pH UV-vis studies (Figure 2). The pH
UV-vis experiment exhibited a bathochromic shift of the
aromatic π f π* transition (and an increase in ꢀmax) due to
deprotonation of the two phenol moieties in the basic pH
region.66
taacetylglucose to afford compound 4. BF3‚OEt2 was used as
the Lewis acid promoter, and the yield of 4 was almost identical
to a literature report of the same coupling using p-toluene-
sulfonic acid under Helferich conditions.42 Glycosylation oc-
curred exclusively at the less hindered alcohol in 72% overall
yield (97% â and 3% R). The â anomer was then coupled with
N,N′-dimethylethylenediamine via a Mannich condensation, and
the acetyl groups were removed to afford H2GL2.
An ellipsoid plot of H2GL2 is shown in Figure 1 with selected
bond lengths and intramolecular H-bonding interactions pre-
sented in Table 2. The carbohydrate moieties participate in a
network of intermolecular hydrogen-bonding interactions in the
solid state, accounting for the observed packing arrangement
in the crystal. Intramolecular H-bonding interactions between
the amine nitrogen atoms and the H atoms of the hydroxyl
groups are evident (shown by dashed lines in Figure 1). These
H bonds were determined to have an important stabilizing effect
in solution for the neutral H2GL1 and H2GL2 species (vide
infra).
The order of deprotonation for the ammonium and phenol
moieties in H2GL1 parallels that for related aminophenol
compounds.39,67,68 The weak basicity of the ammonium moieties
is most probably due to the stability afforded by formation of
intramolecular hydrogen bonds (6-membered rings) between the
amine nitrogen atoms and the H atoms of the hydroxyl groups
for the neutral species H2GL1. H bonding thus weakens the
proton affinity of the amino groups, increasing the stability of
the phenol O-H bond.39,68,69 This same H-bonding pattern is
evident in the X-ray structure (Figure 1) of H2GL2.
The acidity constants for H2GL2 were determined and
assigned in much the same way as for H2GL1. Due to the
increased basicity of the phenolic moieties of H2GL2, both pKa3
and pKa4 were determined spectrophotometrically (see Support-
ing Information). Simulation of the spectrophotometric titration
Both H2GL1 and H2GL2 were investigated by potentiometry
to determine the ability of these compounds to compete with
biological ligands (notably Aâ) for excess Cu and Zn. Poten-
tiometric titrations of these two ligands were first completed to
determine the acidity constants (Ka1-Ka4) of these compounds
(Table 3); the results were then used as constants in the Zn and
Cu metal binding studies.
(64) Nielsen, H.; Sorensen, P. E. Acta Chem. Scand. A 1983, 37, 105. (b)
Ballinger, P.; Long, F. A. J. Am. Chem. Soc. 1960, 82, 795.
(65) Kapinos, L. E.; Song, B.; Sigel, H. Chem. Eur. J. 1999, 5, 1794.
(66) Silverstein, R. M.; Bassler, G. C.; Morrill, T. C. Spectrometric Identification
of Organic Compounds, 4th ed.; Wiley: New York, 1981.
(67) Wong, E.; Caravan, P.; Liu, S.; Rettig, S. J.; Orvig, C. Inorg. Chem. 1996,
35, 715.
(68) L’Eplattenier, R.; Murase, I.; Martell, A. E. J. Am. Chem. Soc. 1967, 89,
837.
(69) Freedman, H. H. J. Am. Chem. Soc. 1961, 83, 2900.
Due to the limitations of potentiometry (inaccurate due to
ionic strength changes when pH < 2.5 or > 11), UV-vis
9
J. AM. CHEM. SOC. VOL. 129, NO. 23, 2007 7459

A R T I C L E S
Storr et al.
Figure 3. Solution speciation diagrams for H2GL1 (0.6 M) and H2GL2 (0.6 M); FL ) fraction of compound with given protonation.
pherol, and the results are shown in Figure 4. Both H2GL1,2
exhibited TEAC values that were enhanced in comparison to
(()-R-tocopherol and butylated hydroxytoluene (BHT); how-
ever, it should be noted that each pro-ligand contains two
phenolic moieties capable of quenching the ABTS•+ radical
cation. H2GL1,2 are essentially protected hydroquinones, and
oxidation to the associated quinones could be the reason for
the potent antioxidant properties. The increased activity of
H2GL2, as compared to the hydrogen analog H2GL1, is most
likely due to the increased stabilization of the phenoxy radical
through inductive and/or steric effects from the tert-butyl group
ortho to the ring hydroxyl.43 It is clear from this study that
H2GL1,2 have the capability to act as antioxidant compounds
contributing to the multifunctional nature of these potential AD
therapeutics. Generation of ROS from reaction of the Cu
complexes of H2GL1,2 (vide infra) with dioxygen in vivo is a
possibility which has not been explored in this work.
Figure 4. Trolox equivalent antioxidant capacity (TEAC) values at 1, 3,
and 6 min for H2GL1,2, (()-R-tocopherol, and BHT. Error bars represent
(SD above and below the average TEAC value (determined in triplicate).
Ni2+ and Cu2+ Metal Complexes of [GL
]
1,2 2-. The neutral
data for H2GL2 afforded only one pKa with a value of 13.4(3).
The calculated value most likely corresponds to an average of
pKa3 and pKa4 as the two phenols are equivalent, and thus, the
acidity constants should be similar. Statistical effects can be
used to separate pKa3 and pKa4 from the single measured value.55
For deprotonation of the first phenol there are two positions
that can be deprotonated but only one position that can be
protonated. This requires the measured acidity constant to be
twice as acidic. For the second deprotonation step the reverse
situation exists. Considering this statistical effect, the two acidity
constants should differ by a factor of 4, and the values for pKa3
and pKa4 can be calculated according to eqs 1 and 2 (Table 3)
Ni2+ and Cu2+ complexes of [GL1,2 2-
]
were synthesized in
MeOH with the addition of base. The Cu complexes were
characterized by MS, EA, UV-vis, EPR, and µeff, while the
Ni complexes were characterized by MS and EA. The Cu and
Ni complexes of [GL ]
1,2 2- were further characterized by X-ray
crystallography. The green Cu complexes exhibited character-
istic [Cu + Na]+ and/or [Cu + H]+ ion peaks with the correct
isotope patterns by +ES-MS. Elemental analysis of the bulk
samples correlated well for CuGL1‚2H2O and CuGL2‚2MeOH,
respectively. The different associated solvents (H2O vs MeOH)
are a result of the purification procedure used, and these could
not be removed upon prolonged drying. It is assumed that due
to Jahn-Teller distortion, there is a weak axial interaction
between the residual solvents and the square-planar Cu centers.
The red/brown Ni complexes exhibited characteristic [Ni + Na]+
and/or [Ni + H]+ ion peaks by +ES-MS. Elemental analysis
of the bulk samples correlated well for NiGL1‚3H2O and
NiGL2‚H2O, respectively. Ni complexes with N-glycoside
ligands have been shown to exhibit antifungal activity.70 The
experimental magnetic moments for CuGL1 (1.74 BM) and
CuGL2 (1.79 BM) were found to be very close to the spin-
only value (1.73 BM) at room temperature. Frozen solution (T
) 130 K) X-band EPR spectra for the two Cu complexes were
recorded, and appropriate spin Hamiltonian parameters were
derived from simulation of the experimental spectra. As
expected, both CuGL1 and CuGL2 exhibited patterns typical
for tetragonal Cu(II)-N2O2 centers (see Supporting Informa-
tion).71
pKa3 ) (pKa3 + pKa4)/2 - 0.3
pKa4 ) (pKa3 + pKa4)/2 + 0.3
(1)
(2)
Using the acidity constants determined for H2GL1,2, solution
speciation diagrams were calculated and are shown in Figure
3. Due to the differences in the basicities of the phenolic
moieties the neutral species H2GL2 exists over a much larger
pH range as compared to H2GL1. Indeed, while H2GL2 is the
predominant species above pH 7, [H3GL1]+, H2GL1, [HGL1]-,
and [GL1]2- exist in the solution speciation diagram for H2GL1.
Antioxidant Activity of H2GL1,2. H2GL1,2 were monitored
for antioxidant activity by the TEAC assay, which has been
used to quantify the antioxidant activity of biological fluids,
extracts, and pure compounds by measuring the disappearance
of the ABTS•+ radical cation via UV-vis spectroscopy.61 The
ability of H2GL1,2 to quench the ABTS•+ radical cation was
compared to Trolox, a more water-soluble analog of R-toco-
(70) Yano, S.; Inoue, S.; Nouchi, R.; Mogami, K.; Shinohara, Y.; Yasuda, Y.;
Kato, M.; Tanase, T.; Kakuchi, T.; Mikata, Y.; Suzuki, T.; Yamamoto, Y.
J. Inorg. Biochem. 1998, 69, 15.
9
7460 J. AM. CHEM. SOC. VOL. 129, NO. 23, 2007

Multifunctional Ligands for Alzheimer’s Therapy
A R T I C L E S
Figure 5. Ellipsoid plot (50% probability, H atoms omitted for clarity) of CuGL2.
Table 4. Selected Bond Lengths (Å) and Angles (deg) in CuGL2
and NiGL2
Table 5. Stability Constants of the Cu2+ and Zn2+ Complexes with
H2GL1,2 Determined by Potentiometric Measurement in 0.16 M
NaCl at 298 K
M
)
Cu
M
)
Ni
M
)
Cu
M ) Ni
log K
M(1)-N(1)
M(1)-N(2)
M(1)-O(7)
2.002(6) 1.90(1)
2.042(7) 1.942(9) N(1)-M(1)-O(7)
1.893(5) 1.907(8) N(2)-M(1)-O(14)
N(1)-M(1)-N(2)
86.4(3)
92.7(2)
93.2(2)
92.2(2)
87.8(4)
94.2(4)
92.6(4)
87.0(3)
H2GL1
H2GL2
+
+
Cu2
+
+
Zn2
Zn2
Cu2
M(1)-O(14) 1.906(5) 1.838(8) O(7)-M(1)-O(14)
2+
M2+ + H2GL1,2/[MH2GL1,2
]
7.23(5)
14.05(6)
5.75(8)
C(7)-O(6)
C(1)-O(6)
C(20)-O(13) 1.381(8) 1.42(1)
C(1)-O(1)
C(10)-O(7)
C(29)-O(14) 1.343(9) 1.32(1)
1.430(8) 1.42(1)
1.415(8) 1.40(1)
M(1)-N(1)-C(17) 112.1(4) 113.9(8)
M(1)-N(2)-C(36) 109.4(5) 111.2(8)
O(6)-C(1)-O(1)
O(13)-C(20)-O(8) 106.5(5) 105.9(9)
M2+ + [HGL1,2]-/[MHGL1,2
]
7.8(2)
9.3(2)
14.83(5)
+
M2+ + [GL1,2
]
2-/MGL1,2
12.90(8) 20.46(4) 16.83(5) 24.18(2)
106.7(6)
105(1)
MGL1,2 + H2O/[MGL1,2(OH)]- -11.0(6) -11.1(4) -11.1(4) -10.9(3)
+ H+
1.449(8) 1.42(1)
1.307(9) 1.38(1)
C(17)-N(1)
C(36)-N(2)
1.470(9) 1.51(1)
1.499(9) 1.47(2)
scopic scale. Evidence for complexation via +ES-MS ([Zn +
Na]+ and/or [Zn + H]+) was encouraging, yet the complexes
could not be purified satisfactorily.
C(19)-C(38) 1.53(1)
1.51(2)
There are a significant number of Cu and Ni complexes
containing carbohydrate-derived ligands reported in the litera-
ture;72-74 in almost all cases, however, the carbohydrate moiety
is bound to the metal center. Only two X-ray structures of Ni
and Cu complexes containing pendant carbohydrate ligands
exist: a Ni complex of a 1,3-diaminosugar73 and a Cu complex
with xylose-appended serine ligands.74 Importantly, the carbo-
hydrate moieties remain pendant in both the Ni and Cu com-
plexes of H2GL2. The structure of CuGL2 is shown in Figure
5 with selected bond lengths and angles for the Cu and Ni
derivatives presented in Table 4. CuGL2 possesses a distorted
square-planar geometry; the dihedral angle between the least-
squares planes through Cu(1)N(1)O(7) and Cu(1)N(2)O(14) is
22.1°.
CuGL2 crystallized with two molecules in the asymmetric
unit, one of each diastereomer. The carbohydrate moieties
participate in a network of hydrogen-bonding interactions, both
between and along the sheets, accounting for the packing
arrangement in the crystal. Removal of the solvent molecules
in the structure solution precluded the observation of H-bonding
interactions mediated by the solvent molecules. The structure
of the Ni complex was very similar to that for Cu and is included
in the Supporting Information. Complexation of the pro-ligands
H2GL1,2 with Zn was only partially successful on the macro-
The solution speciation of Cu2+ and Zn2+ with H2GL1,2 was
studied by potentiometric titrations. The pKa values for the
ligands as well as the hydrolysis reactions of free Cu2+ and
Zn2+ were included as constants in the calculations.60 The values
in Table 5 show that H2GL2 exhibits larger binding constants
2+
(log KMGL ) with Zn and Cu2+ than does H2GL1.
1,2
This result is expected as binding constants for a particular
metal ion should increase with increasing ligand basicity for a
series of similar ligands.65,75,76 On the basis of the calculated
constants, solution speciation diagrams were calculated for Zn2+
and Cu2+ with H2GL1 and H2GL2 (Figure 6). Except for the
presence of [CuH2GL1,2 2+ in the Cu2+ systems, the species
]
present in the pH range examined are identical for both Zn2+
and Cu2+. From the speciation diagrams it is obvious that no
free Cu2+ exists above pH 5 with either ligand, while free Zn2+
exists up to pH 7. Due to the higher stability of the CuGL1,2
species as compared to ZnGL1,2, the former species are more
predominant at lower pH and exist over a larger pH range.
Indeed, at pH 7, CuGL1,2 is the sole species present for Cu2+
,
while ZnGL1,2, Zn2+, and [ZnHGL1,2 + exist for the Zn2+
]
systems. No metal hydrolysis species of significance were
present in any of the systems studied.
Comparison of the calculated Cu2+ and Zn2+ stability
constants of H2GL1,2 with relevant chelators is of significant
interest. However, while stability constants are a measure of
the overall stability of a metal-ligand species, they do not reflect
the metal-binding affinity of ligands at a specific pH value, and
thus a direct comparison can be misleading. For a given complex
equilibrium system and pH the concentration of unchelated metal
ion (referred to as pM ) -log[Munchelated]) represents a direct
(71) Peisach, J.; Blumberg, W. E. Arch. Biochem. Biophys. 1974, 165, 691.
(72) (a) Tanase, T.; Kurihara, K.; Yano, S.; Kobayashi, K.; Sakurai, T.;
Yoshikawa, S.; Hidai, M. Inorg. Chem. 1987, 26, 3134. (b) Tanase, T.;
Yasuda, Y.; Onaka, T.; Yano, S. Dalton Trans. 1998, 345. (c) Tanase, T.;
Inukai, H.; Onaka, T.; Kato, M.; Yano, S.; Lippard, S. J. Inorg. Chem.
2001, 40, 3943. (d) Tanase, T.; Doi, M.; Nouchi, R.; Kato, M.; Sato, Y.;
Ishida, K.; Kobayashi, K.; Sakurai, T.; Yamamoto, Y.; Yano, S. Inorg.
Chem. 1996, 35, 4848. (e) Mikata, Y.; Sugai, Y.; Yano, S. Inorg. Chem.
2004, 43, 4778.
(73) Yano, S. et al. Chem. Lett. 1999, 255.
(74) Delbaere, L. T. J.; Kamenar, B.; Prout, K. Acta. Crystallogr., Sect. B 1975,
31, 862.
(75) Chiu, Y.; Canary, J. W. Inorg. Chem. 2003, 42, 5107.
(76) Martin, R. B.; Sigel, H. Comments Inorg. Chem. 1988, 6, 285.
9
J. AM. CHEM. SOC. VOL. 129, NO. 23, 2007 7461

A R T I C L E S
Storr et al.
Figure 6. Solution speciation diagrams for (A) H2GL1 and Zn2+, (B) H2GL1 and Cu2+, (C) H2GL2 and Zn2+, and (D) H2GL2 and Cu2+ ([H2GL1,2]:[M2+],
1.1:1; [M2+] ) 0.6 mM).
Table 6. Calculated pM (-log[M]unchelated; M ) Zn2+, Cu2+) for a
et al.80 is most likely a result of rapid peptide aggregation (under
kinetic control) during the competitive metal capture assay. The
pZn values for H2GL1,2 correspond to approximate Kd values
in the micromolar range, comparable to reported values (2-
300 µM) for Zn binding to the Aâ peptide.78,81 Thus, the
affinities of H2GL1,2 for Zn and Cu at physiological pH suggest
that the chelators will compete with all soluble forms of Aâ
for these metal ions (vide infra).
Solution Containing a 1:1 Mixture of Indicated Compound and
Metal Ion ([M2+
]
) [chelator]tot ) 0.001 M)
tot
pZn
pCu
compound
pH 7.4
pH 6.6
pH 7.4
H2GL1
H2GL2
DTPA
4.2
3.9
8.4a
6.8
6.6
9.3a
7.9
7.5
10.1a
Aâ Aggregation Inhibition Studies. The ability of H2GL1,2
to inhibit metal-ion-induced Aâ-peptide aggregation was as-
sessed via a turbidometry assay. It has been shown previously
that metal ions such as Zn2+ and Cu2+ promote the rapid
aggregation of Aâ and that metal chelators can attenuate these
effects.9,10,28,82 Similarly, in this study both H2GL1,2 were shown
to reduce Zn2+-induced Aâ1-40 aggregation by approximately
50% at physiological pH (Figure 7). The strong metal-ion
chelator DTPA was, as expected, even more effective at
reducing Zn2+-induced aggregation of the Aâ peptide. The
slightly enhanced activity of H2GL1 over H2GL2 can be
attributed to the increased metal-binding affinity of the former
at physiological pH. The effect of these ligands on Cu2+-induced
Aâ1-40 aggregation was also assessed but at a lower pH (pH
6.6). The interaction of the Aâ peptide with Cu2+ is known to
be pH dependent10 with no aggregation apparent at physiological
pH (data not shown). At this lower pH Cu2+-induced Aâ1-40
a Values calculated from ref 40.
estimate of the ligand-metal affinity taking into account all
relevant equilibria.75,77 pM depends on ligand and metal
concentration, temperature, ionic strength, and pH of the
solution. Table 6 shows the calculated pM values (pH 7.4) for
solutions containing 1 mM of metal (M ) Zn2+ or Cu2+) and
1 mM of either H2GL1, H2GL2, or DTPA. In addition, pCu
values were calculated at lower pH (pH 6.6) where Cu-induced
aggregation of Aâ is accelerated.10 Interestingly, although
H2GL2 has larger zinc and copper binding constants compared
to those of H2GL1, the affinity of H2GL2 for these metals is
lower at pH 6.6 and 7.4 due to the greater proton affinity of
H2GL2. At physiological pH (pH ) 7.4) the pZn values for
H2GL1 (pZn ) 4.2) and H2GL2 (pZn ) 3.9) are quite similar
yet substantially lower than that for DTPA (pZn ) 8.4). Again,
the pZn values can be attributed to the difference in proton
affinity of H2GL1,2 compared with that of DTPA. Much the
same trend exists for copper; the related pCu values (at either
pH 6.6 or 7.4) are lower for H2GL1,2 compared with DTPA
(Table 6). The pCu values for H2GL1,2 correspond to ap-
proximate dissociation constants (Kd) in the nanomolar range,
slightly lower than those reported (0.1-5 µM) in Cu-Aâ
peptide binding studies.10,78,79 The substantially lower Kd value
(attomolar) for the Cu-Aâ1-42 complex described by Atwood
(78) Garzon-Rodriguez, W.; Yatsimirsky, A. K.; Glabe, C. G. Bioorg. Med.
Chem. Lett. 1999, 9, 2243. (a) Danielsson, J.; Pierattelli, R.; Banci, L.;
Graslund, A. FEBS Lett. 2007, 274, 46.
(79) (a) Hou, L. M.; Zagorski, M. G. J. Am. Chem. Soc. 2006, 128, 9260. (b)
Syme, C. D.; Nadal, R. C.; Rigby, S. E. J.; Viles, J. H. J. Biol. Chem.
2004, 279, 18169. (c) Guilloreau, L.; Damian, L.; Coppel, Y.; Mazarguil,
H.; Winterhalter, M.; Faller, P. J. Biol. Inorg. Chem. 2006, 11, 1024. (d)
Kowalik-Jankowska, T.; Ruta, M.; Wisniewska, K.; Lankiewicz, L. J. Inorg.
Biochem. 2003, 95, 270.
(80) Atwood, C. S.; Scarpa, R. C.; Huang, X.; Moir, R. D.; Jones, W. D.; Fairlie,
D. P.; Tanzi, R. E.; Bush, A. I. J. Neurochem. 2000, 75, 1219.
(81) (a) Clements, A.; Allsop, D.; Walsh, D. M.; Williams, C. H. J. Neurochem.
1996, 66, 740. (b) Bush, A. I.; Pettingell, W. H., Jr.; Paradis, M. d.; Tanzi,
R. E. J. Biol. Chem. 1994, 269, 12152.
(82) Huang, X. D.; Atwood, C. S.; Moir, R. D.; Hartshorn, M. A.; Vonsattel, J.
P.; Tanzi, R. E.; Bush, A. I. J. Biol. Chem. 1997, 272, 26464.
(77) (a) Fahrni, C. J.; O’Halloran, T. V. J. Am. Chem. Soc. 1999, 121, 11448.
(b) Martell, A. E.; Hancock, R. D. Metal Complexes in Aqueous Solutions;
Plenum Press: New York, 1996.
9
7462 J. AM. CHEM. SOC. VOL. 129, NO. 23, 2007

Multifunctional Ligands for Alzheimer’s Therapy
A R T I C L E S
moderate affinity for Cu2+ and Zn2+ and decrease Aâ1-40
aggregation induced by these metal ions. H2GL1,2 may be ideally
suited for disruption of the abnormal metal-protein interactions
in the brain possibly responsible for the observed toxicity in
AD. The moderate affinity of H2GL1,2 for metal ions at
physiological pH may obviate the toxicity commonly associated
with chelation therapy, especially when coupled with the
potential for increased tissue specificity due to the attached
carbohydrate functions. In addition, the phenolic moieties of
the apo-chelators can act as antioxidants, contributing to the
multifunctional nature of these potential Alzheimer’s disease
therapeutics.
Acknowledgment. The authors gratefully acknowledge the
Canadian Institutes of Health Research (CIHR) for a Proof of
Principle grant. T.S. acknowledges NSERC for a postgraduate
scholarship, L.E.S. recognizes support from the Alzheimer’s
Society of Canada for a training award, and M.M. acknowledges
the Alexander von Humboldt Foundation for a Feodor Lynen
postdoctoral fellowship. Dr. Bin Song is thanked for help with
the potentiometric analysis. Dr. Wilf Jeffries is acknowledged
for generous donation of a sample of Aâ1-40 peptide.
Figure 7. Aggregation inhibition assay of Aâ1-40. Data indicate the mean
absorbance (at least three trials) at 405 nm of Aâ solutions in buffer: in
the absence of metal ion (M), in the presence of M (Zn2+ at pH 7.4 or
Cu2+ at pH 6.6), and in the presence of M and test ligand (DTPA, H2GL1,
or H2GL2). Error bars represent (SD above (and below, not shown) the
average absorbance value.
aggregation was significantly attenuated with both H2GL1,2 as
well as DTPA (Figure 7). We cannot rule out that both H2GL1,2,
or their respective Zn or Cu complexes, inhibit peptide ag-
gregation at least partially through intercalation.27,83
Summary. In this work we have shown that the water-
soluble, carbohydrate-containing compounds H2GL1,2 have
significant antioxidant capacity on the basis of an in vitro
antioxidant assay. In addition, the two compounds have a
Supporting Information Available: Complete refs 13, 15b,
20, 23b, and 73, variable pH UV-vis titration data for H2GL2,
the ellipsoid plot of NiGL2, as well as EPR spectra and
simulations of CuGL1 and CuGL2; X-ray crystallographic files
in CIF format of H2GL2, CuGL2, and NiGL2. This material is
available free of charge via the Internet at http://pubs.acs.org.
(83) Raman, B.; Ban, T.; Yamaguchi, K.; Sakai, M.; Kawai, T.; Naiki, H.; Goto,
Y.; J. Biol. Chem. 2005, 280, 16157.
JA068965R
9
J. AM. CHEM. SOC. VOL. 129, NO. 23, 2007 7463