Eu(III) complexes of functionalized octadentate 1-hydroxypyridin-2-ones: Stability, bioconjugation, and luminescence resonance energy transfer studies

Article abstract of DOI:10.1021/ic101133w

The synthesis, stability, and photophysical properties of several Eu(III) complexes featuring the 1-hydroxypyridin-2-one (1,2-HOPO) chelate group in tetradentate and octadentate ligands are reported. These complexes pair highly efficient emission with exc

Full text of DOI:10.1021/ic101133w

9928 Inorg. Chem. 2010, 49, 9928–9939
DOI: 10.1021/ic101133w
Eu(III) Complexes of Functionalized Octadentate 1-Hydroxypyridin-2-ones:
Stability, Bioconjugation, and Luminescence Resonance Energy Transfer Studies
Evan G. Moore, Jide Xu, Christoph J. Jocher, Todd M. Corneillie, and Kenneth N. Raymond*
Department of Chemistry, University of California, Berkeley, California 94720-1460, and Chemical Science
Division, Glenn T. Seaborg Center, Lawrence Berkeley National Laboratory, Berkeley, California 94720
Received June 4, 2010
The synthesis, stability, and photophysical properties of several Eu(III) complexes featuring the 1-hydroxypyridin-
2-one (1,2-HOPO) chelate group in tetradentate and octadentate ligands are reported. These complexes pair highly
efficient emission with exceptional stabilities (pEu ∼ 20.7-21.8) in aqueous solution at pH 7.4. Further analysis of
their solution behavior has shown the observed luminescence intensity is significantly diminished below about pH ∼ 6
because of an apparent quenching mechanism involving protonation of the amine backbones. Nonetheless, under
biologically relevant conditions, these complexes are promising candidates for applications in Homogeneous Time-
Resolved Fluorescence (HTRF) assays and synthetic methodology to prepare derivatives with either a terminal
amine or a carboxylate group suitable for bioconjugation has been developed. Lastly, we have demonstrated the use
of these compounds as the energy donor in a Luminescence Resonance Energy Transfer (LRET) biological assay
format.
Introduction
We recently reported6,7 on the exceptional luminescence
properties of 1-hydroxypyridin-2-one (1,2-HOPO) based
compounds as sensitizers for Eu(III), and our progress with
these compounds was also summarized in a recent review.8
Herein, we expand on our earlier communication9 regarding
the aqueous stability and photophysical properties of Eu(III)
complexes with the tetradentate 5LINMe-1,2-HOPO and
octadentate H(2,2)-1,2-HOPO ligands (Chart 1). We have
found that protonation of the apical amines is important in
determining the luminescence intensity and resulting photo-
physical properties for these complexes, and with the help of
TD-DFT calculations, suggest a possible rationale for this
effect. We also introduce two new functionalized octadentate
derivatives which bear either a pendant amine or a carbox-
ylate arm to enable conjugation of the complexes with
relevant biomolecules. We have measured the stability and
photophysical properties of these new compounds, and find
they are essentially unchanged when compared to the parent
complex. Lastly, we demonstrate the feasibility of using these
complexes as a long-lived luminescent tag in a bioassay
format, using a 1,2-HOPO labeled streptavidin bioconjugate
Recently, there have been several reports1-3 on the use of
luminescent Eu(III) and Tb(III) complexes for the detection
of important biomolecules such as citrate, lactate, or urate in
biological fluids. These assays utilize subtle changes in the
characteristic lanthanide emission upon analyte binding,
and offer both improved detection limits and allow for signi-
ficantly smaller sample volumes to be used. Concurrently,
Time-Resolved Luminescence Resonance Energy Transfer
(TR-LRET) assays have continued to be developed,4
enabling the convenient and sensitive high-throughput
study of molecular interactions such as those occurring
between proteins and their inhibitors.5 In particular, the
sharp emission spectra, large effective Stokes shifts, and
long-lived luminescent lifetimes makes trivalent lantha-
nide cations such as Eu(III) very attractive, since the use
of both spectral and time-resolved discrimination of the
luminescent signal from background autofluorescence
results in improved sensitivity.
*To whom correspondence should be addressed. E-mail: raymond@
socrates.berkeley.edu. Phone: þ1 510 642 7219. Fax: þ1 510 486 5283.
(1) Leonard, J. P.; dos Santos, C. M. G.; Plush, S. E.; McCabe, T.;
Gunnlaugsson, T. Chem. Commun. 2007, 2, 129–131.
(2) Pal, R.; Parker, D.; Costello, L. C. Org. Biomol. Chem. 2009, 7, 1525–
1528.
(3) Esplin, T. L., Cable, M. L., Gray, H. B., Ponce, A., Inorg. Chem. 2010,
49, 4643-4647.
(4) Vogel, K. W.; Vedvik, K. L. J. Biomol. Screening 2006, 11, 439–443.
(5) Bergendahl, V.; Heyduk, T.; Burgess, R. R. Appl. Environ. Microbiol.
2003, 69, 1492–1498.
(6) Moore, E. G.; Xu, J.; Jocher, C. J.; Werner, E. J.; Raymond, K. N.
J. Am. Chem. Soc. 2006, 128, 10648–10649.
(7) Moore, E. G.; Xu, J.; Jocher, C. J.; Castro-Rodriguez, I.; Raymond,
K. N. Inorg. Chem. 2008, 47, 3105–3118.
(8) Moore, E. G.; Samuel, A. P. S.; Raymond, K. N. Acc. Chem. Res.
2009, 42, 542–552.
(9) Moore, E. G.; Jocher, C. J.; Xu, J.; Werner, E. J.; Raymond, K. N.
Inorg. Chem. 2007, 46, 5468–5470.
r
pubs.acs.org/IC
Published on Web 09/27/2010
2010 American Chemical Society

Article
Inorganic Chemistry, Vol. 49, No. 21, 2010 9929
Chart 1. Chemical Structures of Tetradentate and Octadentate 1,2-
HOPO Based Ligands Discussed in This Paper
formate (1.5 equiv) in Et2O (150 mL) was added over a 2 h
period. The solution was stirred overnight, warmed to room
temperature, and the aqueous phase was extracted with CH2Cl2
(5 ꢀ 30 mL). The organic phases were combined and passed
through a flash silica gel pad, then concentrated to afford the
1
desired Cbz-aziridine (7.2 g, 81%) as a thick, colorless oil. H
NMR (500 MHz, CDCl3, 25 °C): δ = 2.22 (s, 4H, CH2), 5.14
(s, 2H, CH2), 7.34 (m, 5H, ArH). 13C NMR (125 MHz, CDCl3,
25 °C): δ = 20.51, 65.93, 127.3, 127.4, 128.7, 140.9, 159.4.
Cbz2-TREN. Benzyl Phenyl Carbonate (4.56 g, 20 mmol) was
added to a stirring solution of tris(ethylenediamine) (TREN,
1.46 g, 10 mmol) in absolute EtOH (50 mL) while cooling with
an ice bath. The reaction mixture was stirred overnight at room
temperature, and the volatiles were removed in vacuo. Distilled
water (50 mL) was added to the oily residue, and the pH was
adjusted to 3 by addition of aqueous HCl (2 M), again using an
external ice bath for cooling. The acidified solution was ex-
tracted with CH2Cl2 (4 ꢀ 50 mL). The cooled aqueous phase was
then made strongly alkaline by addition of aqueous NaOH (2 M)
and extracted with CH2Cl2 (3 ꢀ 80 mL). The organic phase was
dried over Na2SO4, filtered, and loaded onto a basic alumina
column. Gradient chromatography with 3-6% CH3OH in CH2-
Cl2 and then concentration afford pure Cbz2-TREN (3.10 g,
75%) as a thick, colorlessoil. 1H NMR (500 MHz, CDCl3, 25°C):
δ = 2.48 (t, 2H, J = 5.5 Hz, CH2), 2.56 (s,br, 4H, CH2), 2.67 (t,
2H, J = 5.5 Hz, CH2), 3.23 (s,br, 4H, CH2), 5.06 (s, 2H, CH2),
5.72 (s, 2H, CbzNH), 7.04 (m, 10H, ArH). 13C NMR (125 MHz,
CDCl3, 25 °C): δ = 38.9, 39.4, 53.8, 52.7, 66.3, 127.78, 127.82,
136.6, 156.7. MS (FAB, NBA) C22H30N4O4: [MþH]þ 415.0.
Cbz2-TREN-Cbz-Lys(Boc). Cbz2-TREN (4.14 g, 10 mmol)
and Cbz-Lys(Boc)-aldehyde (3.64 g, 10 mmol) were mixed in
THF (50 mL) at room temperature under N2. The mixture was
stirred for 3 h, then sodium triacetoxyborohydride (3.18 g, 15
mmol) was added and the mixture stirred at room temperature
under a N2 atmosphere for 24 h. Aqueous NaOH (1 M) was
added to quench the reaction mixture, and the mixture was
extracted with CH2Cl2 (3 ꢀ 50 mL). The CH2Cl2 extract was
loaded onto a flash silica gel column. The appropriate fractions
of a gradient elution with 3-10% MeOH in CH2Cl2 were
collected and evaporated to dryness to give the desired product
as the energy donor and a biotinylated organic dye, Alexa
Fluor 594, as the energy acceptor.
Experimental Section
General Procedures. All solvents for reactions were dried
using standard methodologies. Biocytin Alexa Fluor 594 was
purchased from Invitrogen as the sodium salt and was used as
supplied. Thin-layer chromotography (TLC) was performed
using precoated Kieselgel 60 F254 plates. Flash chromato-
graphy was performed using EM Science Silica Gel 60 (230-
400 mesh). NMR spectra were obtained using either Bruker
AM-300 or DRX-500 spectrometers operating at 300 (75) MHz
1
1
and 500 (125) MHz for H (or 13C) respectively. H (or 13C)
chemical shifts are reported in parts per million (ppm) relative to
the solvent resonances, taken as δ 7.26 (δ 77.0) and δ 2.49 (δ
39.5) for CDCl3 and (CD3)2SO, respectively, while coupling
constants (J) are reported in hertz (Hz). The following standard
abbreviations are used for characterization of 1H NMR signals:
s = singlet, d = doublet, t = triplet, q = quartet, quin = quintet,
m = multiplet, dd = doublet of doublets. Fast-atom bombard-
ment mass spectra (FABMS) were performed using 3-nitrobenzyl
alcohol (NBA) or thioglycerol/glycerol (TG/G) as the matrix.
Elemental analyses were performed bythe Microanalytical Labo-
ratory, University of California, Berkeley, CA.
1
(6.27 g, 82%) as a thick, pale beige oil. H NMR (500 MHz,
CDCl3, 25 °C): δ = 1.17 (s,br, 2H, Lys CH2), 1.24-1.89 (m, 4H,
Lys CH2), 1.41 (s, 9H, Boc CH3), 2.53 (s, br, 8H, Tren CH2), 3.03
(s, 2H, BocNHCH2), 3.14 (s, 2H, CbzNHCH2), 3.20 (s, 2H,
CbzNHCH2), 3.64 (s,br, 1H, Lysine chiral CH), 4.54 (s, 1H,
BocNH), 5.05(m, 6H, CbzCH2), 7.27 (m, 15H, ArH). 13C NMR
(125 MHz, CDCl3, 25 °C): δ, 22.8, 28.3, 29.5, 32.5, 39.4, 39.9,
47.4, 50.9, 53.8, 54.3, 66.5, 78.9, 79.3, 127.9, 128.1, 128.4, 136.6,
156.0, 156.7, 156.9. MS (FAB, NBA) C41H58N6O8: [MþH]þ
763.5.
Synthesis. The detailed syntheses of 5LINMe-1,2-HOPO and
H(2,2)-1,2-HOPO and their Eu(III) complexes have been pre-
viously reported,9 as well as many of the relevant precursors (see
Scheme 1) including the activated 1,2-HOPO-Bn-acid-chloride.10
Aziridine was prepared from 2-aminoethyl hydrogen sulfate
using literature methodologies.11 Similarly, the synthesis of Cbz-
Lys(Boc)-aldehyde and Benzyl Phenyl Carbonate have been
reported elsewhere.12,13
Cbz4-Lys-Boc-H(2,2)-amine. Cbz2-TREN-Cbz-Lys(Boc) (3.81
g, 5 mmol) and Cbz-aziridine (1.24 g, 7 mmol) were mixed in
t-butanol (50 mL) at room temperature under an N2 atmosphere.
The mixture was heated to 80 °C and stirred under N2 for 16 h,
when TLC showed the reaction was complete. The volatiles were
removed under vacuum, and the residue was dissolved in CH2Cl2.
The appropriate fractions of a gradient flash silica gel column
(1-7% MeOH in CH2Cl2) were collected and evaporated to
dryness to give the desired product as a thick, pale beige oil (3.98
g, 84.7%). 1HNMR(500 MHz, CDCl3, 25 °C): δ= 1.24-1.87 (m,
6H, Lys CH2), 1.42 (s, 9H, Boc CH3), 2.26 (s, br, 2H, CH2), 2.47
(m,br,, 6H, CH2), 2.58 (s,br, 2H, CH2), 2.93-3.35 (m,br, 8H,
CH2), 5.05 (m, 8H, CbzCH2), 7.27 (m, 20H, ArH). 13C NMR (125
MHz, CDCl3, 25 °C): δ 22.8, 28.4, 29.7, 31.2, 32.9, 38.5, 38.7, 40.1,
49.4, 52.1, 52.3, 59.4, 66.5, 69.1, 78.9, 127.9, 127.9, 128.0, 128.1,
128.36, 128.38, 136.57, 136.64, 156.0, 156.7. MS (FAB, NBA)
C51H69N7O10: [MþH]þ 940.5.
Cbz-aziridine. The concentrations of aziridine in the freshly
prepared distillate11 were calibrated by acid-base titration and
were found to range from 0.1 to 0.2 M, which was used directly
for preparation of Cbz-aziridine. Hence, to a stirred solution of
the distillate (500 mL, 0.1 mol) in a round-bottom flask with
external cooling via an ice bath, 3 equiv of potassium carbonate
were added. After all the solid had dissolved, benzyl choloro-
(10) Xu, J.; Churchill, D. G.; Botta, M.; Raymond, K. N. Inorg. Chem.
2004, 43, 5492–5494.
(11) Reeves, W. A.; Drake, G. L., Jr.; Hoffpauir, C. L. J. Am. Chem. Soc.
1951, 73, 3522.
(12) McConnell, R. M.; Barnes, G. E.; Hoyng, C. F.; Gunnl, J. M. J. Med.
Chem. 1990, 33, 86–93.
(13) Pittelkow, M.; Lewinsky, R.; Christensen, J. B. Synthesis 2002, 15,
2195–2202.
Boc-Lys-H(2,2)-amine. To a glass hydrogenation container
with 200 mg of wet 5% Pd/C catalyst, 5 mL of MeOH was

9930 Inorganic Chemistry, Vol. 49, No. 21, 2010
Moore et al.
Scheme 1. Synthetic Scheme for Preparation of Functionalized Octadentate 1,2-HOPO Ligands
carefully added along the glass wall to cover the catalyst (Caution!
Pd/C catalyst is pyrophoric). A solution of Cbz4BocLys-H(2,2)-
amine (0.94 g, 1 mmol) in MeOH (30 mL) was added to the
container, which was placed in a Parr bomb apparatus and
hydrogenated at 500 psi overnight. Subsequent TLC analysis
confirmed the absence of starting material, and the solvent was
removed in vacuo. The Boc-Lys-H(2,2)-amine was obtained as a
colorless thick oil (0.52 g, 90%), isolated as the tetracarbonate
form through reaction with atmospheric CO2. The desired free
amine form was prepared by passage of this product through a
strongly basic ion exchangecolumn. 1H NMR(500 MHz, CDCl3,
25 °C): δ = 1.24-1.87 (m, 6H, Lys CH2), 1.42 (s, 9H, Boc CH3),
2.41-2.82 (m, br, 10H, CH2), 2.82-3.23 (m, br, 9H, CH2), 3.29
(s, br, 1H, CH), 5.48(s, 1H, Boc NH), 8.52 (s, 3H, CarbonateNH).
13C NMR (125 MHz, CDCl3, 25 °C): δ 22.6, 28.4, 29.5, 31.1, 37.0,
39.9, 49.8, 51.9, 68.8, 78.6, 156.2, 169.0. MS (FAB, NBA)
C51H69N7O10: [MþH]þ 404.4.
1H), 6.60 (d, 3H), 6.94 (d, 1H), 7.0-7.4 (m, 19H), 7.4-7.5 (m,
8H), 7.71 (s, 1H, NH). 13C NMR (125 MHz, CDCl3, 25 °C): δ =
22.8, 28.2, 29.3, 31.9, 37.4, 37.5, 39.8, 48.7, 51.5, 51.9, 52.8, 53.3,
53.7, 59.0, 78.6, 79.0, 104.6, 105.0, 123/1, 128.3, 129.1, 129.7,
129.9, 130.0, 133.2, 133.3, 137.9, 138.1, 142.9, 143.1, 155.9, 158.1,
158.2160.5, 160.7. MS (FAB, NBA) C71H81N11O14: [MþH]þ
1312.0.
Lys-H(2,2)-1,2-HOPO. Boc-Lys-H(2,2)-1,2-HOPO-Bn (0.26
g, 0.2 mmol) was dissolved in a 1:1 (v/v) mixture of concentrated
HCl (12 M) and glacial CH3COOH (20 mL) and was stirred at
room temperature for 2 days. Filtration followed by removal of
the solvent gave the desired product as a beige solid (0.18 g,
81%). 1H NMR (300 MHz, DMSO-d6, 25 °C): δ = 1.2-1.7 (m,
br, 6H), 2.74 (s, br, 3H), 3.05 (s, br, 3H), 3.15 (s, br, 5H), 3. 55 (s,
br, 3H), 3.61 (s, br, 6H), 4.15 (s, br, 1H), 6.30-6.45 (m, 4H), 6.60
(d, 4H), 7.39 (m, 5H), 7.88 (s, br, 3H), 8.85 (s, br, 1H), 9.00 (s, br,
1H), 9.06 (s, br, 2H). MS (FAB, NBA): 851 [MþH]þ. Anal. for
C38H49N11O12 2HCl H2O 2CH3OH (1066.88 g mol-1), Calcd
Boc-Lys-H(2,2)-1,2-HOPO-Bn. To a freshly prepared solu-
tion of Boc-Lys-H(2,2)-amine (0.2 g, 0.5 mmol) in CH2Cl2 (20
mL), cooled with an ice bath and vigorously stirred, was added
2 mL of 40% K2CO3 solution. A solution of 1,2-HOPO-Bn-
acid-chloride (freshly prepared from 0.75 g 1,2-HOPO-Bn acid,
3 mmol) in dichloromethane (20 mL) was added slowly via a
Teflon tube equipped with a glass capillary tip over a period of
1 h. The reaction mixture was allowed to warm to room tem-
perature and was stirred overnight, and then the mixture was
washed with 1 M HCl (20 mL), then saline (20 mL), and was
loaded onto a flash silica column. Elution with 2-8% MeOH in
CH2Cl2 allows the separation of the benzyl-protected precursor
Boc-Lys-H(2,2)-1,2-HOPO-Bn as a beige foam (0.46 g, 70%).
1H NMR (300 MHz, DMSO-d6, 25 °C): δ = 1.15 (s,br, 4H),
1.41 (s, 9H, BocH), 1.9-2.8 (m, 14H), 2.8-3.3 (m, 8H), 3.79 (s,
br, 1H), 4.90-5.40 (m, 8H), 6.03 (s, 1H), 6.17 (d, 3H), 6.53 (d,
3
3
3
3
(Found): C, 47.72 (47.60); H, 6.11 (6.17); N, 15.30 (15.34).
Lys-EtGlutA-H(2,2)-1,2-HOPO-Bn. To a solution of Boc-
Lys-H(2,2)-1,2-HOPO-Bn (0.26 g, 0.2 mmol) in 2 mL of CH2Cl2
cooled with an ice bath was added 2 mL of CF3COOH. The
mixture was stirred for 4 h then evaporated to dryness at room
temperature, and TLC analysis confirmed removal of the Boc
protecting group. The residue was dissolved in dry THF (20
mL), and dry Et3N (0.5 mL) was added, cooling with an ice bath.
To this cold mixture, an excess of ethyl glutarate N-hydroxy-
succinimide ester (0.1 g, 0.4 mmol) was added under nitrogen.
The mixture was stirred for 4 h, and the volatiles were removed
in vacuo. The residue was dissolved in CH2Cl2 and loaded onto a
flash silica column. Elution with 2-8% methanol in CH2Cl2
allows the separation of the benzyl-protected precursor Lys-
EtGlutA-H(2,2)-1,2-HOPO-Bn as a beige foam (0.19 g, 70%).

Article
Inorganic Chemistry, Vol. 49, No. 21, 2010 9931
1H NMR (300 MHz, DMSO-d6, 25 °C): δ = 1.09 (s, br, 4H),
1.20 (t, 3H), 1.25 (t, 4H), 1.84 (t, 2H), 2.02 (m, 2H), 2.17 (t, 6H),
2.28 (t, 6H), 2.98-3.1 (m, 8H), 3.78 (s, br, 1H), 4.05 (q, 2H),
5.10-5.40 (m, 8H), 6.03 (s, 1H), 6.16 (d, 3H), 6.45 (s, 1H), 6.51
(d, 1H), 6.58 (d, 3H), 7.15 (dd, 1H), 7.18-7.40 (m, 18H),
7.40-7.51 (m, 8H), 8.00 (s, 1H, NH). 13C NMR (125 MHz,
CDCl3, 25 °C): δ = 14.4, 21.2, 23.2, 25.5, 28.9, 31.9, 33.7, 35.5,
37.7, 38.2, 39.0, 46.0, 49.1, 50.4, 52.1, 53.2, 54.5, 59.9, 60.6, 79.4,
105.1, 105.6, 123.3, 123.5, 128.8, 129.6, 130.3, 130.4, 130.6, 133.7,
133.8, 138.7, 143.4, 143.7, 143.8, 158.7, 158.8, 161.0, 161.2, 161.5,
173.0, 173.5. MS (FAB, NBA) 1354 [MþH]þ.
mined experimentally to be 0.486. The A341/280 ratio and a molar
extinction coefficient of 18,200 M-1 cm-1 for the parent Eu-
H(2,2)-1,2-HOPO complex were used to estimate a labeling ratio
of about 2.3 [Eu(Lys-GlutA-H(2,2)-1,2-HOPO)] molecules per
streptavidin molecule.
Physical Methods. Solution Thermodynamics. The general
procedure used to compare aqueous stabilities of the Eu(III)
complexes herein was that of competition batch titration as
described previously,15,16 using diethylenetriaminepentaacetic
acid (DTPA) as the known competitor. Briefly, varying volumes
of standardized DTPA stock solution were added to an aqueous
0.1 M HEPES buffer solution (pH 7.4) containing 0.1 M KCl
electrolyte, and a fixed amount of the appropriate ligand and
Eu(III) cation in 1:1 stoichiometry was delivered by Eppendorf
pipet. If necessary, the pH of the solutions was readjusted to
7.4 with HCl and/or KOH, and the solutions were diluted to
identical volumes with additional 0.1 M HEPES buffer (pH 7.4)
containing 0.1 M KCl. The initial concentrations of DTPA
relative to the Eu(III) complex ranged from about 1:10 to
3000:1. After standing for 24 h to ensure thermodynamic
equilibrium, the concentration of the remaining complexed
ligand was evaluated by monitoring the Eu(III) emission spec-
trum after 330 nm excitation, using the integrated area of the
luminescence spectra in the wavelength range between 570 and
710 nm and the spectra of the fully formed complexes in the
absence of added DTPA competitor as a reference.
Lys-GlutA-H(2,2)-1,2-HOPO. The deprotection of Lys-
EtGlutA-H(2,2)-1,2-HOPO-Bn was performed in two steps.
The first step was saponification of the pendant ethyl ester,
followed by deprotection of the four benzyl groups under
strongly acidic condition as described for Lys-H(2,2)-1,2-
HOPO-Bn. To a solution of Lys-EtGlutA-H(2,2)-1,2-HOPO-Bn
(0.27 g, 0.2 mmol) in 5 mL of methanol, cooled with an ice bath,
was added 2 mL of KOH solution (1 M). The mixture was stirred
for 4 h when TLC confirmed the hydrolysis of the ethyl ester was
complete. The mixture was evaporated to dryness at room tem-
perature, and the residue was dissolved in water (10 mL). The
hydrolyzed Lys-GlutA-H(2,2)-1,2-HOPO-Bn was precipitated
upon acidification with HCl (1 M). This was collected, rinsed with
cold water, and further deprotected by dissolving in a 1:1 (v/v)
mixture of concentrated HCl (12 M) and glacial CH3COOH
(20 mL). After stirring at room temperature for 2 days, filtration
followed by removal of the solvent gave the desired product as a
beige solid (0.15 g, 71%). 1H NMR (300 MHz, DMSO-d6, 25 °C):
δ = 1.37 (s, br, 4H), 1.67 (t, 7H), 1.84 (t, 2H), 2.03 (m, 2H), 2.17 (t,
6H), 2.28 (t, 6H), 2.98-3.1 (m, 8H), 3.78 (s, br, 1H), 4.05 (q, 2H),
6.41 (m, 4H), 6.60 (m, 4H), 7.39 (m, 4H), 7.78 (t, 1H), 8.83 (m, 1H),
8.80 (m, 1H), 9.05 (t, 2H). MS (FAB, NBA): 966.4 [MþH]þ. Anal.
for C43H55N11O15 HCl 4H2O (1074.48 g mol-1), Calcd (Found):
Photophysics. Typical sample concentrations for absorption
and fluorescence measurements were about 10-5 to 10-6 M and
1.0 cm cells in quartz suprasil or equivalent were used. UV-
visible absorption spectra were recorded on a Cary 300 double
beam absorption spectrometer. Emission spectra were acquired
on a HORIBA Jobin Yvon IBH FluoroLog-3 spectrofluori-
meter. Spectra were reference corrected for both the excitation
light source variation (lamp and grating) and the emission
spectral response (detector and grating). Quantum yields were
determined by the optically dilute method using the following
equation;
3
3
3
C, 48.07 (48.34); H, 6.00 (6.09); N, 14.34 (14.00).
[Eu(Lys-H(2,2)-1,2-HOPO)]. To a solution of Lys-H(2,2)-
1,2-HOPO (21.0 mg, 19.7 μmol) in MeOH (5 mL) was added 100
μL of pyridine, followed by 1.03 molar equivs of EuCl3 6H2O
3
"
#
ꢀ
ꢁ ꢀ
ꢁ
ꢀ
ꢁ
(7.85 mg, 20.3 μmol) in MeOH (1 mL). The resulting mixture
was refluxed for 4 h, then allowed to cool to room temperature.
Addition of diethyl ether (ca. 10 mL) induced precipitation of a
white solid, which was collected by vacuum filtration, washed
with diethyl ether, and air-dried to yield the desired product
(19.4 mg, 76.6%). Anal. for EuC38H45N11O12 3HCl 2CH3OH
Φx
Φr
ArðλrÞ IðλrÞ η2x Dx
AxðλxÞ IðλxÞ η2r Dr
¼
where A is the absorbance at the excitation wavelength (λ), I is
the intensity of the excitation light at the same wavelength, η is
the refractive index, and D is the integrated luminescence inten-
sity. The subscripts “x” and “r” refer to the sample and reference,
respectively. Quinine sulfate in 1.0 N sulfuric acid was used as the
reference (Φr = 0.546).17
3
3
3
6H2O (1282.37 g mol-1), Calcd (Found): C, 37.46 (37.24); H,
5.42 (5.28); N, 12.01 (11.83).
3
[Eu(Lys-GlutA-H(2,2)-1,2-HOPO)]. This complex was pre-
pared using the same methodology as the [Eu(Lys-H(2,2)-1,2-
HOPO)] complex, substituting the Lys-GlutA-H(2,2)-1,2-HOPO
ligand where appropriate (17.0 mg, 72.0%). Anal. for EuC43H52-
N11O15 HCl CH3OH 5H2O (1273.48 g mol-1), Calcd (Found):
Luminescence lifetimes were determined with a HORIBA
Jobin Yvon IBH FluoroLog-3 spectrofluorimeter, adapted for
time-resolved measurements. A submicrosecond Xenon flash
lamp (Jobin Yvon, 5000XeF) was used as the light source,
coupled to a double grating excitation monochromator for
spectral selection. The input pulse energy (100 nF discharge
capacitance) was about 50 mJ, yielding an optical pulse duration
of less than 300 ns at fwhm. A thermoelectrically cooled single
photon detection module (HORIBA Jobin Yvon IBH, TBX-04-
D) incorporating fast rise time PMT, wide bandwidth pream-
plifier and picosecond constant fraction discriminator was used
as the detector. Signals were acquired using an IBH DataStation
Hub photon counting module, and data analysis was performed
using the commercially available DAS 6 decay analysis software
package from HORIBA Jobin Yvon IBH. Goodness of fit was
assessed by minimizing the reduced chi squared function, χ2, and
3
3
3
3
C, 41.50 (41.39); H, 5.30 (5.28); N, 12.10 (12.01).
SAv-[Eu(Lys-GlutA-H(2,2)-1,2-HOPO)]. Lys-GlutA-H(2,2)-
1,2-HOPO (5.5 mg, 5.1 μmol) was dissolved with N-hydroxysul-
fosuccinimide (5.54 mg, 25.5 μmol) and 1-ethyl-3-(3-dimethyl-
aminopropyl)carbodiimide (4.75 mg, 30.6 μmol) in 127 μL of
dry DMF, and then stirred for 2 h. The activated Lys-GlutA-
H(2,2)-1,2-HOPO-sNHS ester was then directly added dropwise to
a solution of streptavidin (160 μM, 0.13 μmol) in 100 mM Na2CO3
buffer at pH 9 and allowed to stir overnight at room temperature.
The SAv-Lys-GlutA-H(2,2)-1,2-HOPO was separated from any
unreacted Lys-GlutA-H(2,2)-1,2-HOPO and buffer-exchanged
into 0.1 M TRIS at pH 7.0 using a Penefsky gel filtration column.14
EuCl3.6H2O (2.8 mg, 7.7 μmol) dissolved in sodium citrate buffer
was then added to the SAv-Lys-GlutA-H(2,2)-1,2-HOPO conju-
gate while stirring to form the metal complex. The A341/280
absorbance ratio for Eu-Lys-GlutA-H(2,2)-1,2-HOPO was deter-
(15) Pierre, V. C.; Botta, M.; Aime, S.; Raymond, K. N. Inorg. Chem.
2006, 45, 8355–8364.
(16) Doble, D. M. J.; Melchior, M.; O’Sullivan, B.; Siering, C.; Xu, J.;
Pierre, V. C.; Raymond, K. N. Inorg. Chem. 2003, 42, 4930–4937.
(17) Crosby, G. A. D., J. N. J. Phys. Chem. 1971, 75, 991–1024.
(14) Penefsky, H. S. Methods Enzymol. 1979, 56, 527–530.

9932 Inorganic Chemistry, Vol. 49, No. 21, 2010
Moore et al.
a visual inspection of the weighted residuals. Each trace con-
tained at least 10,000 points, and the reported lifetime values
result from at least three independent measurements.
Computational Studies. Ground state density functional theory
(DFT) and time-dependent DFT (TD-DFT) calculations were
performed at the Molecular Graphics and Computational Facility,
College of Chemistry, University of California, Berkeley. In both
cases, the B3LYP/6-311Gþþ (d,p) basis set provided in
Gaussian’0318 was used, with simplified input structures derived
from a previously reported7 crystal structure. All calculations were
done in the gas phase and geometry optimizations were performed
with no symmetry restraints.
Results and Discussion
Synthesis. The synthesis of the 5LINMe-1,2-HOPO and
H(2,2)-1,2-HOPO ligands were reported earlier,9 and rely
on an amide coupling reaction between primary amines of
the ligand backbone and benzyl protected 1,2-HOPO
carboxylic acid, activated by conversion to the acid
chloride. Adopting similar methodologies, the synthetic
route toward the new pivotal building block, Lys-H(2,2)-
amine, is illustrated in Scheme 1 and has also been
detailed in a recent patent.19 The selective Cbz protection
of TREN to yield Cbz2-TREN, and subsequent Schiff
base condensation product with Cbz-Lys(Boc)-aldehyde,
obtained using literature procedures,12 leads to Cbz2-
TREN-Cbz-Lys(Boc) after in situ reduction with NaBH-
(OAc)3. This intermediate can be coupled in good yield
with the highly reactive Cbz-aziridine to give the Cbz4-
Lys-Boc-H(2,2)-amine, which after catalytic hydrogena-
tion and purification by cation exchange chromatogra-
phy, yields the monoprotected Boc-Lys-H(2,2)-amine.
Standard coupling procedures with the aforementioned
benzyl protected 1,2-HOPO acid chloride yields Boc-Lys-
H(2,2)-1,2-HOPO-Bn, which after strongly acidic depro-
tection using 1:1 (v/v) conc. HCl/AcOH yields the desired
amine functionalized Lys-H(2,2)-1,2-HOPO chelate. Alter-
nately, selective deprotection of the Boc group using TFA
affords Lys-H(2,2)-1,2-HOPO-Bn which can be coupled
with activated ethyl glutarate N-hydroxysuccinimide ester
to give Lys-EtGlutA-H(2,2)-1,2-HOPO-Bn. Subsequent
stepwise saponification of the ethyl ester under basic con-
ditions and then removal of the benzyl protecting groups
under strong acidic conditions gives the carboxylate
functionalized Lys-GlutA-H(2,2)-1,2-HOPO chelate.
The complexation of either the 5LINMe-1,2-HOPO or
Figure 1. 1H NMR spectra at various pH values (left axis intercept) for
(a) the 5LI-1,2-HOPO(top) and(b) 5LINMe-1,2-HOPO(bottom) ligands.
achieved using the appropriate 1:2 or 1:1 metal/ligand
stoichiometries respectively in the presence of pyridine
acting as a base, and using MeOH as the solvent. After
a short reflux period of about 4-6 h, the desired com-
plexes were isolated upon cooling by filtration in excel-
lent yield and in analytically pure form. Similarly, the
complexation reactions of the new octadentate H(2,2)-
ligands bearing amine or carboxylate functional groups
were straightforward, and yielded the desired complexes
as the protonated trihydrochloride and monohydro-
chloride salts, respectively, because of the differing pro-
tonation state of the terminal functional group (i.e.,
-NH3þ versus -COO-, respectively).
H(2,2)-1,2-HOPO ligands with EuCl3 6H2O is readily
3
Solution Thermodynamics. The protonation constants
for the tetradentate 5LINMe-1,2-HOPO and octadentate
H(2,2)-1,2-HOPO ligands, as determined by potentio-
metric titration, were detailed earlier, together with the
experimental pEu values and calculated formation con-
stants with Eu(III).9 However, to gain further insight into
(18) Frisch, M. J., Trucks, G. W., Schlegel, H. B., Scuseria, G. E., Robb;
M. A., C., J. R., Montgomery, J. A., Jr., Vreven, T.; Kudin; K. N., B., J. C.,
Millam, J. M., Iyengar, S. S., Tomasi, J., Barone; V., M., B., Cossi, M.,
Scalmani, G., Rega, N., Petersson, G. A.; Nakatsuji, H., Hada, M., Ehara,
M., Toyota, K., Fukuda, R., Hasegawa; J., I., M., Nakajima, T., Honda, Y.,
Kitao, O., Nakai, H., Klene; M., L., X., Knox, J. E., Hratchian, H. P., Cross,
J. B., Bakken, V.; Adamo, C., Jaramillo, J., Gomperts, R., Stratmann, R. E.,
Yazyev; O., A., A. J., Cammi, R., Pomelli, C., Ochterski, J. W., Ayala; P. Y.,
M., K., Voth, G. A., Salvador, P., Dannenberg, J. J.; Zakrzewski, V. G.,
Dapprich, S., Daniels, A. D., Strain, M. C., Farkas; O., M., D. K., Rabuck,
A. D., Raghavachari, K., Foresman, J. B.; Ortiz, J. V., Cui, Q., Baboul,
A. G., Clifford, S., Cioslowski, J.; Stefanov, B. B., Liu, G., Liashenko, A.,
Piskorz, P., Komaromi, I.; Martin, R. L., Fox, D. J., Keith, T., Al-Laham,
M. A., Peng, C. Y.; Nanayakkara, A., Challacombe, M., Gill, P. M. W.,
Johnson, B., Chen; W., W., M. W., Gonzalez, C., Pople, J. A. Gaussian 03,
revision C.02; Gaussian, Inc.: Wallingford, CT. 2004.
1
the solution behavior, we have conducted additional H
NMR titrations for the ligand, in comparison to the
structurally analogous tetradentate 5LI-1,2-HOPO li-
gand, which lacks an alkyl amine in the backbone. As a
result, the previously assigned pKa at 8.13, which we
attributed to the amine, can be unequivocally confirmed
by the observed shift in the N-methyl group resonance
from about 2.2 ppm (basic) to about 3.0 ppm (acidic)
between pH 10 to pH 6 (Figure 1b), consistent with an
increase in deshielding of this resonance by the adjacent
(19) Raymond, K. N.; Corneillie, T. M.; Xu, J. Luminescent hydroxyi-
sophthalamide macrocyclic lanthanide complexes and derivatives. U.S.
Patent 2007-US76047 2008063721, 20070815, 2008.

Article
Inorganic Chemistry, Vol. 49, No. 21, 2010 9933
diamagnetic. Consistent with our initial reports, the in-
itial two protonation equilibria for the Y(III) complex
can be readily assigned to protonation of the two amine
backbones in the [EuL2]- complex to form [EuL2H2]þ,
since the apparent pKa’s of 7.5 and 6.5 determined by
potentiometry correlate well with the observed shift of the
N-methyl proton resonance from about 2.0 to 2.8 ppm
between pH 8.1 and 4.7 (i.e., within (1.5 pKa units;
Figure 2). The remaining pKa at 4.5 has been assigned
to protonation of one of the 1,2-HOPO chelate groups,
most likely at an exposed keto oxygen atom, and is
accompanied by a general broadening of the observed
1H NMR spectra and a slight upfield shift of the reso-
nances in the aromatic region, presumably because of the
weaker electrostatic interaction and hence altered ligand
exchange kinetics in going from an overall monocationic
[EuL2H2]þ complex involving two anionic ligands to
the dicationic [EuL2H3]2þ complex involving one anionic
and one neutral ligand. Unfortunately, because of only
limited solubility in aqueous solution (<0.1 mM), the pH
dependent 1H NMR titration data could not be obtained
for the H(2,2)-1,2-HOPO ligand, nor its Eu(III) or Y(III)
complexes.
Figure 2. 1H NMR spectra at various pH values (left axis intercept) for
[YIII(5LINMe-1,2-HOPO)2].
Chart 2. Structural Changes Associated with Protonation of the Tertiary
Amine Group in 5LINMe-1,2-HOPO
Having confirmed the protonation equilibria of com-
plexes with 5LI-1,2-HOPO and 5LINMe-1,2-HOPO, we
were interested in the pH dependent behavior of the
observed Eu(III) luminescence, and as such have under-
taken fluorescence titrations at variable pH, with result-
ing data for 10 μM solutions of the bis(tetradentate)
complexes as shown in Figure 3. When viewed together
with the calculated speciation from the thermodynamic
parameters, some striking features are seen. For the
[Eu(5LI-1,2-HOPO)2] complex, the luminescence inten-
sity is essentially constant across a wide pH range from
10 to 6, consistent with the expected speciation for the
EuL2 complex. Below pH 6, the concomitant formation
of EuL2H and, more importantly, hydrolysis of the EuL2
complex to form EuL(H2O)x contribute significantly to
the speciation. Since the latter will have much weaker
luminescence because of the well-known quenching effect
of inner sphere water molecules,22 this results in a de-
crease in the overall luminescence intensity. In stark con-
positive charge upon protonation. As previously assig-
ned20 using HMBC and HMQC 2D NMR techniques, we
attribute the peaks at 6.62, 7.24, and 6.99 ppm (pH ∼11)
to the H3, H4, and H5 aromatic protons, respectively
(Chart 1). Notably, concomitant with the spectral
changes observed in the alkyl region was a slight upfield
shift of these aromatic signals by 0.05, 0.1, and 0.18 ppm,
respectively, (see dotted lines, Figure 1b) upon proton-
ation of the amine backbone going from pH 10 to pH 6,
which were not evident in the case of 5LI-1,2-HOPO.
Since adjacent protonation should induce a downfield
shift, this cannot be due to an inductive effect. As was the
case for a related hexadentate ligand (Tren-1,2-HOPO),20
we assign these shiftstoa conformational changeinvolving
single bond rotation within the ligand backbone, facilitat-
ing intramolecular H-bonding interactions between the
protonated amine and the amide carbonyl group(s), as
showninChart 2. Further acidification beyond about pH 6
results in formation of the fully protonated 5LINMe-1,2-
HOPO ligand, and a corresponding downfield shift for the
aromatic H3 and H4 signals. The observation of the
sequential protonation equilibria for the equivalent N-
trast, the overall luminescence intensity of [Eu(5LINMe
-
1,2-HOPO)2] rapidly diminishes below pH 10 to about
50% of its maximum intensity at pH 7, which correlates
to maximum formation of the [EuL2H] complex. Upon
further acidification, the formation of [EuL2H2]þ be-
comes dominant at about pH 5.8, and with the loss of
the [EuL2]- and [EuL2H] species, the luminescence is
effectively switched off. Evidently, protonation of the
amine backbones has a dramatic influence on the ensuing
photophysical properties for the [Eu(5LINMe-1,2-
HOPO)2] complex.
1
hydroxyl groups was not possible by H NMR, presum-
ably because of their fast exchange on the NMR time scale.
1
The measured H NMR spectra of the corresponding
1
Although H NMR data could not be obtained, data
[Eu(5LINMe-1,2-HOPO)2] complex was very broad be-
cause of the paramagnetic nature of the metal ion and
could not be interpreted. To further elucidate the solution
behavior of this complex, we have substituted the lantha-
nide cation with Y(III), since this metal has an identical
charge and very similar ionic radius21 to Eu(III), but is
from fluorescence titrations that utilize significantly low-
er concentrations could be obtained for a 1 μM solution
of [Eu(H(2,2)-1,2-HOPO)], and the corresponding lumi-
nescence intensities at variable pH for this complex are
shown in Figure 3, together with the calculated speciation
(22) Beeby, A.; Clarkson, I. M.; Dickins, R. S.; Faulkner, S.; Parker, D.;
Royle, L.; de Sousa, A. S.; Williams, J. A. G.; Woods, M. J. Chem. Soc.,
Perkin Trans. 1999, 493–503.
(20) Jocher, C. J.; Moore, E. G.; Xu, J.; Avedano, S.; Botta, M.; Aime, S.;
Raymond, K. N. Inorg. Chem. 2007, 46, 9182–9191.
(21) Shannon, R. D. Acta Crystallogr. 1976, A32, 751–767.

9934 Inorganic Chemistry, Vol. 49, No. 21, 2010
Moore et al.
Figure 4. Double log plots of equilibrium Eu(III) complex concentra-
tions determined from luminescence spectra versus equilibrium concen-
trations for (a) the Lys-H(2,2)-1,2-HOPO (left) and (b) Lys-GlutA-
H(2,2)-1,2-HOPO (right) ligands. Solid lines are corresponding fits used
to evaluate ΔpM versus Eu(DTPA), corresponding to pEu’s of 20.7(1)
and 21.8(4), respectively.
β110 when these constants are included, we still obtain a
theoretical dilution limit with a similar order of magni-
tude of 1.6 ꢀ 10-16 M at pH 7.4 under the same tolerance
condition of 1% free Eu(III), compared to 7 ꢀ 10-9
M
for the [Eu(5LINMe-1,2-HOPO)2] complex, with the
exceptional resistance to hydrolysis being driven by the
entropic chelate effect of the octadentate versus bis-
(tetradentate) complex topologies.
As was the case for the tetradentate ligands, we note the
luminescence at variable pH has a maximum intensity at
about pH 10, which correlates well with the calculated
speciation of the EuL complex. Upon acidification, the
luminescence is again diminished, although the decrease
is not as significant as was the case with the 5LINMe-1,2-
HOPO ligand, amounting to only a about 13% decrease
at pH 7.4 where the calculated [EuLH] speciation is at a
maximum. Below about pH 6, where the formation of the
[EuLH2]þ complex becomes dominant, we note the same
dramatic decrease in overall Eu(III) based luminescence,
again suggesting that protonation of the amine back-
bones effectively switches off luminescence from the
metal center. Although the mechanism of this quenching
is not readily apparent, it is most likely due to disruption
of the overall sensitization process, as evidenced by our
computational results (vide infra).
Despite this decrease in luminescence intensity at low
pH, we note the H(2,2)-1,2-HOPO complex with Eu(III)
still retains about 87% of its maximum luminescence at a
biologically relevant pH of 7.4. As such, we carried on
with the preparation of the Eu(III) complexes of the Lys-
H(2,2)-1,2-HOPO and Lys-GlutA-H(2,2)-1,2-HOPO li-
gands, bearing pendant amine and carboxyl groups,
respectively, and have assessed their aqueous stability at
pH 7.4 by competition batch titration using DTPA as a
known competitor (pEu ∼ 19.04), with results as shown in
Figure 4. The addition of increasing equivalents of DTPA
results in the steady decrease of the characteristic Eu(III)
luminescence for both the [Eu(Lys-H(2,2)-1,2-HOPO)]
and the [Eu(Lys-GlutA-H(2,2)-1,2-HOPO)] complexes.
From the accompanying double log plots, the point at
which speciation between the ligand and DTPA is equiva-
lent can be easily determined, and from the x intercept,
the calculated pEu’s for each complex at pH 7.4 were
evaluated to be 20.7(1) and 21.8(4) with the Lys-H(2,2)-
1,2-HOPO and Lys-GlutA-H(2,2)-1,2-HOPO ligands,
Figure 3. Combined pH dependent total integrated luminescence in-
tensity profiles (right axis) (λex = 330 nm, λem = 520-720 nm) in aqueous
solution with expected speciation (left axis) from solution thermodynamic
data for about 10 μM solutions of Eu(III) complexes with (a) 5LI-1,2-
HOPO (top), (b) 5LINMe-1,2-HOPO (middle), and for a about 1 μM
solution of the Eu(III) complex with (c) H(2,2)-1,2-HOPO (bottom).
obtained from the solution thermodynamic model. For
this complex, given a tolerance of 1% free Eu(III), we
previously reported9 a calculated theoretical limit of
stability of 5 ꢀ 10-17 M at pH 7.4. Upon re-examining
this calculation, we noted that hydrolysis constants for
the lanthanide cation were omitted, and the reported
stability constants do not accurately reproduce the ex-
perimentally determined pEu values when these values
are included. As such, the corrected values are reported
here to be β110 = 22.3(1), β111 = 31.2(1), and β112 =
37.3(1). Nonetheless, because of a slight improvement in

Article
Inorganic Chemistry, Vol. 49, No. 21, 2010 9935
Figure 5. Calculatedoutputgeometries obtainedfromstatic B3LYP/6-311Gþ2 (d,p) geometry optimizationand relevantmolecular orbital diagrams from
TD-DFT electronic structure calculations for (a) deprotonated [Na(Me2N-Et-NHC(O)-1,2-HOPO)] (top), (b) protonated [Na(Me2NþH-Et-NHC(O)-1,2-
HOPO)] (middle), and (c) protonated [Na(Me2NþH-Et-NHC(O)-1,2-HOPO)] (bottom) with “fixed” input geometry (see text).
respectively. These values are very similar to the previously
reported pEu of 21.2 (1) at pH 7.4 for the unsubstituted
[Eu(H(2,2)-1,2-HOPO)], and are also in accord with
our earlier studies23 of hexadentate HOPO chelates with
Gd(III), where we noted that the overall complex charge
can influence the ensuing aqueous stability.
Electronic Structure Calculations. In an effort to ratio-
nalize the mechanism for the quenching of Eu(III) cen-
tered luminescence after protonation of the adjacent
amine backbones, TD-DFT calculations were applied
to several simplified model systems. In accordance with
strategies originally adopted24 by Picard et al. and our
own more recent reports,25,26 we have replaced the triva-
lent lanthanide with monovalent sodium to avoid com-
putationally expensive calculations involving the 4f
metal. Although this is a crude model, we have found
this approach to be useful and generally applicable in
several cases as a tool for understanding the electronic
structure of the lowest excited singlet and triplet states of
the ligand.27
Hence, the model system as shown in Figure 5a was
considered, which features a tertiary amine, separated by
an ethyl chain from the 1,2-HOPO amide linkage, and
this molecular fragment was first geometry optimized
with no symmetry constraints, with the resulting output
geometry as shown. Upon inspection, we note that the
1,2-HOPO motif including the amide group attachment is
essentially planar, and the short C-O bond length of
˚
about 1.26 A for the coordinated oxygen atom sup-
ports the keto resonance form. A symmetrical bidentate
coordination mode with the monovalent sodium cation is
apparent, with almost identical bond lengths of about
˚
2.16 A, mirroring that expected upon complexation with
a trivalent lanthanide cation. The protons of the pendent
ethyl chain between the amide nitrogen atom and the
amine are staggered, although the amine and the amide
nitrogen atoms have adopted a gauche conformation,
(23) Pierre, V. C.; Botta, M.; Aime, S.; Raymond, K. N. Inorg. Chem.
2006, 45, 8355–8364.
(24) Gutierrez, F.; Tedeschi, C.; Maron, L.; Daudey, J.-P.; Azema, J.;
Tisnes, P.; Picard, C.; Poteau, R. THEOCHEM 2005, 756, 151–162.
(25) Moore, E. G.; Xu, J.; Jocher, C. J.; Castro Rodriguez, I.; Raymond,
K. N. Inorg. Chem. 2008, 47, 3105–3118.
(26) Moore, E. G.; Xu, J.; Dodani, S. C.; Jocher, C. J.; D’Aleo, A.; Seitz,
M.; Raymond, K. N. Inorg. Chem. 2010, 49, 4156–4166.
(27) Samuel, A. P. S.; Xu, J.; Raymond, K. N. Inorg. Chem. 2009, 48, 687–
698.

9936 Inorganic Chemistry, Vol. 49, No. 21, 2010
Moore et al.
Table 1. Summary of Electronic Structure Calculations for Deprotonated [Na(Me2N-Et-NHC(O)-1,2-HOPO)] and Protonated [Na(Me2NþH-Et-NHC(O)-1,2-HOPO)]
Model Complexes
excited state
multiplicity
energy (eV)
wavelength (nm)
oscillator strength
orbital(s) involved n f n0
transition coefficient
[Na(Me2N-Et-NHC(O)-1,2-HOPO)] (Relaxed Geometry)
1
2
3
4
T1
T2
S1
T3
2.620
3.175
3.192
3.216
473.3 (∼ 21,130 cm-1
)
0.0000
0.0000
0.0055
0.0000
65 f 67 (HOMOfLUMOþ1)
0.7628
0.7028
0.7035
0.1182
0.6656
0.3681
0.1523
0.6357
390.6
388.4
385.5
65 f 66
65 f 66
62 f 67
63 f 67
65 f 71
63 f 71
65 f 67
5
S2
3.719
333.4
0.1167
[Na(Me2NþH-Et-NHC(O)-1,2-HOPO)] (Relaxed Geometry)
1
2
T1
T2
2.172
3.147
570.9 (∼ 17,516 cm-1
)
0.0000
0.0000
65 f 66 (HOMO f LUMO)
64 f 66
64 f 69
65 f 69
64 f 69
65 f 66
64 f 66
65 f 68
65 f 69
65 f 67
0.7758
0.7071
0.1056
0.2730
0.1244
0.6383
-0.2431
0.1403
0.6529
0.6939
394.0
3
4
S1
3.261
3.749
380.2
330.7
0.0872
0.0000
T3
5
T4
3.845
322.4
0.0000
[Na(Me2NþH-Et-NHC(O)-1,2-HOPO (Fixed Geometry)
1
2
T1
T2
2.462
2.618
503.6 (∼ 19,860 cm-1
)
0.0000
65 f 66 (HOMO f LUMO)
65 f 67
65 f 68
65 f 67
65 f 66
65 f 67
64 f 66
64 f 68
65 f 74
0.7428
0.1184
0.1146
0.6890
0.1327
0.6993
0.6695
-0.1349
0.3240
-0.1256
0.6400
473.7
0.0000
3
4
S1
T3
2.640
3.272
469.6
379.0
0.0029
0.0000
5
S2
3.568
347.5
0.1102
64 f 74
65 f 66
rather than the more stable anti configuration. This is
stabilized by a weak intramolecular H-bonding inter-
action between the amine lone pair and the amide proton,
adopted a gauche rather than anti conformation, and the
proton of the amine group is directed toward the amide
carbonyl, forming a very strong intramolecular H-bonding
˚
˚
with a separation of about 2.66 A. The amide proton is
similarly involved in a much stronger intramolecular
H-bonding interaction with the enolate oxygen atom,
contact of about 1.55 A. This change in conformation is
reminiscent of that observed for the 5LINMe-1,2-HOPO
ligand at low pH, as monitored by 1H NMR spectroscopy.
Furthermore, since the amide NH proton is no longer
involved in a bifurcated interaction with the tertiary amine
lone pair, the strength of the H-bond between this group and
the enolate oxygen atom is increased, as evidenced by a
decrease in the intramolecular separation from about 1.78
˚
with a separation of about 1.78 A.
The corresponding electronic structure calculation as
determined by TD-DFT techniques using the same basis
set is summarized in Table 1, and relevant molecular
orbitals involved in these transitions are also depicted in
Figure 5a. Most importantly, the lowest energy triplet
state is predicted to occur at about 21,130 cm-1, which is
almost identical to the value of about 21,365 cm-1 pre-
viously calculated25 for the Naþ complex of a simpler
6-MeAmide-1,2-HOPO model system using the same
TD-DFT techniques. Evidently, the presence of the ethyl
amine chain has little influence over the electronic struc-
ture of the model systems when deprotonated.
Using the ground state optimized geometry for the
deprotonated system as a starting point, but including an
additional proton at the tertiary amine, we have performed
a second geometry relaxation with the resulting output
geometry as shown in Figure 5b. While the 1,2-HOPO motif
and attached amide group are once again essentially planar
as expected, the most significant difference, readily apparent
on inspection, is a change in the -C(O)-NH-CH2-CH2-
bond torsion from -153° to þ66°. As a consequence of the
rotation of this bond, the ethyl amine substituent has
˚
to 1.65 A, and the corresponding distance from the enolate
oxygen to chelated metal cation is slightly elongated from
˚
about 2.15 to 2.22 A.
The consequences of these changes in geometry on the
electronic structure of the 1,2-HOPO chromophore sub-
sequent to amine protonation were similarly assessed by
TD-DFT techniques, with resulting lowest energy transi-
tions as summarized in Table 1, and relevant molecular
orbitals as depicted in Figure 5b. Clearly, the lowest
energy triplet state is largely affected, with a significant
decrease in the predicted energy of this transition from
about 21,130 cm-1 to 17,516 cm-1. Also apparent is a
change in the nature of the HOMO-1 orbital from essen-
tially an amine based lone pair in the former case to a π type
orbital in the latter. As a result of the about 3600 cm-1
decrease in energy for the lowest energy triplet state
(typically considered as the donor in the so-called “anten-
nae” effect), the position of this level is about 1500 cm-1

Article
Inorganic Chemistry, Vol. 49, No. 21, 2010 9937
the new octadentate derivatives reported herein featuring
either pendant amine or carboxylate arms are essentially
identical when compared with the parent H(2,2)-1,2-HOPO
complex. The UV-visible spectrum of [Eu(Lys-H(2,2)-1,2-
HOPO)] is shown as an example in Figure 6. A broad
electronic envelope assigned to n-π* transitions of the 1,2-
HOPO chromophore is evident, with peak maxima at about
342 nm, and at higher energy, a more intense band assigned to
π-π* transitions of the chromophore is observed at about
233 nm.
Theemissionspectrumofthe[Eu(Lys-H(2,2)-1,2-HOPO)]
complex is also shown in Figure 6, and is characteristic of
metal centered Eu(III) luminescence, with sharp bands attri-
butable to the 5D0 f 7FJ (J = 0, 1, 2, 3, and 4) transitions
readily apparent at about 577, 585-597, 612, 650, and
680-700 nm. Notably, the splitting of the J = 1 transi-
tion into 3 peaks (see inset, Figure 6) is typical of that for
the Eu(III) cation in a low-symmetry environment, and is
consistent with our previous assignment9 of local C2
symmetry for the parent complex. The spectrum of the
corresponding [Eu(Lys-GlutA-H(2,2)-1,2-HOPO)] was
essentially identical. The overall luminescence quantum
yields, Φtot, were determined using quinine sulfate in
0.5 M H2SO4 as a reference, and were evaluated to be
4.3% and 4.2% respectively for the [Eu(Lys-H(2,2)-1,2-
HOPO)] and the [Eu(Lys-GlutA-H(2,2)-1,2-HOPO)]
complexes in 0.1 M TRIS buffered aqueous solution at
pH 7.4. These are identical within experimental error, and
are slightly higher than the parent form of the complex.
The observed increase may be due to a slightly more
shielded environment for the Eu(III) cation, because of
the additional steric bulk of the alkyl linked pendant
amine or carboxylate arms, which would contribute
toward reducing the number of nearby aqueous solvent
molecules in the second solvation sphere of the complex.
The luminescence lifetimes of the Eu(III) centered
emission at 612 nm for each of the new pendant armed
octadentate complexes were evaluated in 0.1 M TRIS
buffered aqueous solution at pH 7.4, and these results are
summarized in Table 2. Notably, the luminescence life-
times of these two new compounds are almost identical,
and remain unchanged when compared to the parent
[Eu(H(2,2)-1,2-HOPO)(H2O)] complex. Using the corre-
sponding luminescence lifetimes measured in deuterated
solvent allowed us to estimate “q”, the number of bound
inner sphere solvent molecules, using the improved em-
pirical Horrock’s equation.30 As was the case for the
parent [Eu(H(2,2)-1,2-HOPO)(H2O)] complex, this anal-
ysis reveals the presence of a single inner sphere water
molecule, also contributing toward the lower overall
Figure 6. Absorption spectrum (black, left axis) and emission spectra at
λex = 330 nm (red, right axis) for [Eu(Lys-H(2,2)-1,2-HOPO)] in 0.1 M
TRIS buffered aqueous solution at pH 7.4. Inset: Expansion of [Eu(Lys-
H(2,2)-1,2-HOPO)] emission showing 5D0 f7F0 and 5D0f7F1 transitions
(red), and corresponding fit (blue) to four Voigt functions (black).
below the 5D1 excited state, and only 239 cm-1 above the
emitting 5D0 level of the Eu(III) cation. In this region, the
effects of thermal back energy transfer will be very signifi-
cant, since the optimum energy separation for efficient
energy transfer has been previously suggested to be at least
about 1850 cm-1 above the emitting level of the Ln(III)
cation.28 This important result may explain the loss of
luminescence intensity upon protonation of the amine back-
bones as observed by pH dependent fluorescence titrations.
Lastly, although molecular modeling at the MM2 level
of theory for both the [Eu(5LINMe-1,2-HOPO)2] and the
[Eu(H(2,2)-1,2-HOPO)(H2O)] complexes have shown
that rotation of the ligand backbones (at least in part) is
feasible (see Supporting Information, Figure S1), it could
be argued that the extent of “protonation induced” amide
rotation in the Eu(III) complexes is more conformation-
ally restricted. To address this possibility, we have under-
taken a third TD-DFT electronic structure calculation
utilizing the fixed output geometry derived from the
deprotonated amine case, but with the addition of a single
proton to the tertiary amine nitrogen. As summarized in
Table 1, the predicted energy of the lowest energy triplet
state is once again reduced, from about 21,130 cm-1 to
19,860 cm-1, although the decrease of 1270 cm-1 is not as
dramatic. Nonetheless, even in this case, the location of
this critical sensitizing level is only 830 cm-1 above the
5D1 excited state of the Eu(III) cation, which is suffi-
ciently close to facilitate efficient back energy transfer.
Indeed, it has been shown previously that, based on
selection rules for the direct energy transfer process, direct
quantum yield values, Φtot, when compared to 5LINMe
-
5
energy transfer to the D0 level is not allowed, and it is
likely the intermediate D1 state which is the primary
accepting level for energy transfer into the Eu(III) excited
5
1,2-HOPO complexes of Eu(III), because of well-known
non-radiative quenching by bound inner sphere solvent
water molecules. This increase in non-radiative quench-
ing of Eu(III) luminescence is reflected in a higher value
for knr, the non-radiative deactivation rate constant,
which was determined together with the radiative decay
rate constant, kr, following the work of Verhoeven et al.31
state manifold.29
Photophysics. The absorption spectra of the [Eu(5LINMe
1,2-HOPO)2] and [Eu(H(2,2)-1,2-HOPO)] complexes were
-
detailed in the preceding communication,9 and the spectra of
(28) Latva, M. T. H.; Mukkala, V.-M.; Matachescu, C.; Rodriguez-Ubis,
J. C.; Kankare, J. J. Lumin. 1997, 75, 149–169.
(30) Supkowski, R. M.; Horrocks, W. D., Jr. Inorg. Chim. Acta 2002, 340,
(29) de Sa, G. F.; Malta, O. L.; de Mello Donega, C.; Simas, A. M.;
Longo, R. L.; Santa-Cruz, P. A.; da Silva, E. F., Jr. Coord. Chem. Rev. 2000,
196, 165–195.
44–48.
(31) Werts, M. H. V.; Jukes, R. T. F.; Verhoeven, J. W. Phys. Chem.
Chem. Phys. 2002, 4, 1542.

9938 Inorganic Chemistry, Vol. 49, No. 21, 2010
Moore et al.
Table 2. Summary of Experimental and Calculated Photophysical Parameters for Various 1,2-HOPO Based Eu(III) Complexes in 0.1 M TRIS Buffered Aqueous Solution,
pH 7.4
εmax
Φtot (H2O)
(%)
τ
obs H2O
q
calcd
kr (ms-1
calcd
)
knr (ms-1
calcd
)
ΦEu
calcd
ηsens
calcd
complex
λmax (nm)
(M-1 cm-1
)
{D2O} (msec)
[Eu(5LINMe-1,2-HOPO)2]
[Eu(H(2,2)-1,2-HOPO)]
[Eu(Lys-H(2,2)-1,2-HOPO)]
[Eu(Lys-GlutA-H(2,2)-1,2-HOPO)]
332
341
342
342
18,750
18,200
18,400
18,350
17.3
3.6
4.3
4.2
0.73 {1.00}
0.48 {1.22}
0.47 {0.99}
0.46 {0.98}
0.1
1.0
0.9
0.9
0.61
0.77
1.75
1.79
1.83
0.442
0.178
0.158
0.158
0.391
0.202
0.272
0.266
0.375
0.335
0.343
and Beeby et al.,32 and a complete summary of these data
are also reported in Table 2.
Turning once again to the parent H(2,2)-1,2-HOPO
ligand, we have prepared the Gd(III) complex to investi-
gate the energetic position of the lowest energy ligand-
centered triplet state. This metal cation has a similar size
and atomic weight when compared to Eu(III), but lacks
an appropriately positioned electronic acceptor level,
making it amenable to the observation of phosphores-
cence from the ligand’s lowest energy T1 state at low
temperature. Measurements were performed at 77 K in a
solid matrix of 1:1 (v/v) MeOH:EtOH which also con-
tained about 5% (v/v) of 1.0 M TRIS buffered aqueous
solution at pH 7.4 to enable pH control of the otherwise
nonaqueous solvent. The resulting spectrum for [Gd(H(2,2)-
1,2-HOPO)(H2O)] (Supporting Information, Figure S2)
shows an intense unstructured emission band apparent
from 450 to 650 nm, which we attribute to phosphore-
scence from the ligand. The lowest energy zero phonon
Figure 7. LRET assay plots from a 500 nM solution of SAv-[Eu(Lys-
GlutA-H(2,2)-1,2-HOPO)] in 0.1 M TRIS buffered aqueous solution at
pH 7.4 (a) Eu(III) centered emission intensity and (b) Eu(III) lumines-
cence lifetime (λex = 335 nm) upon incremental addition of 0, 0.5, 2.0,
3.75, and 7.5 mol equiv of a biotinylated Alexa Fluor 594 organic dye.
was determined to be about 2.3 from analysis of the
UV-visible absorption spectrum at 280 and 342 nm,
corresponding to the λmax of the protein and complex,
respectively. The steady state Eu(III) centered emission
spectrum of the SAv-[Eu(Lys-GlutA-H(2,2)-1,2-HOPO)-
(H2O)x] bioconjugate upon excitation of the 1,2-HOPO
chromophore at about 340 nm was unchanged when
compared to that of the [Eu(Lys-GlutA-H(2,2)-1,2-
HOPO)(H2O)x] complex. The luminescence lifetime of
0
(v0-0 ) band of the T1 state was estimated by spectral de-
convolution of the 77 K luminescence signal into several
overlapping Gaussian functions, and the lowest energy
peak was evaluated to be about 20,385 cm-1. Impor-
tantly, the position of this state is about 875 cm-1 lower
in energy, when compared to the values obtained via
identical analysis25 of a bis(tetradentate) Gd(III) com-
plex which lacks an alkyl amine in the ligand backbone.
In the latter cases, the energy gap between the lowest
the Eu(III) bioconjugate was also determined at λem
=
612 nm, and yielded a satisfactory fit to a monoexponen-
tial decay, with an observed lifetime of τobs = 0.60 ms. In
comparison to the unlabeled complex (τobs = 0.46 ms),
we attribute the slight increase in lifetime as a result of the
more shielded protein environment, which serves to
protect the complex from deactivation by non-radiative
interactions with the solvent.
Using a commercially available biotinylated Alexa
Fluor 594 fluorescent dye, we have examined the photo-
physical changes evident upon interaction of these mole-
cules in buffered aqueous solution at pH = 7.4 using both
steady state and time-resolved luminescence. This organic
chromophore was chosen as an acceptor specifically
5
energy T1 triplet state and the D1 accepting state is
optimal at about 2230 cm-1. The decrease of this energy
gap can be expected to significantly decrease the sensi-
tization efficiency for the H(2,2)-1,2-HOPO complex
(and pendant armed derivatives), which is corroborated
by lower values of ηsens as shown in Table 2, and is also
in accordance with our TD-DFT calculations, strongly
suggesting that protonation of the adjacent amines
influences the energetic position of the triplet state,
and hence the overall luminescence performance of
Eu(III) in 1,2-HOPO complexes containing alkyl amine
backbones.
€
because of its good Forster overlap with the Eu(III)
Bioconjugation Studies. Lastly, in a proof of principle
experiment, a bioconjugate of the carboxy functionalized
[Eu(Lys-GlutA-H(2,2)-1,2-HOPO)(H2O)] complex was
prepared with Streptavidin (SAv), a 52.8 kDa tetrameric
protein commonly utilized in molecular biology because
of its extraordinarily strong non-covalent interaction
with biotin. This SAv-[Eu(Lys-GlutA-H(2,2)-1,2-HOPO)-
(H2O)q] bioconjugate was prepared using standard NHS
activated coupling techniques, and was purified from
unreacted starting materials using gel filtration. The
degree of labeling (DOL) of the purified bioconjugate
centered emission of the complex. The resulting plot of
observed overall luminescence upon the addition of in-
creasing equivalents of the biotinylated Alexa Fluor 594
dye (Supporting Information, Figure S3), and the inten-
sity of the Eu(III) emission, obtained by subtracting the
known emission of the organic dye, versus added equiva-
lents is shown in Figure 7. A sigmoidal decrease in Eu(III)
centered emission was observed which clearly plateaus
after about 4 equiv of Alexa Fluor 594, consistent with the
tetrameric nature of the SAv-biotin interaction. Further-
more, by monitoring the luminescence lifetimes, we ob-
served an almost identical decrease in lifetime from about
0.6 to 0.3 ms upon addition of up to about 4 equiv of
(32) Beeby, A.; Bushby, L. M.; Maffeo, D.; Williams, J. A. G. J. Chem.
Soc., Dalton Trans. 2002, 48.

Article
Inorganic Chemistry, Vol. 49, No. 21, 2010 9939
biotinylated Alexa Fluor 594, consistent with quenching
of the Eu(III) centered luminescence by intramolecular
energy transfer. The long-lived Eu(III) signal was com-
pletely quenched upon further addition of the biotiny-
lated Alexa Fluor 594 organic dye. The intramolecular
nature of the LRET quenching was verified by a titration
in the presence of an initially added excess of unlabeled
biotin. In this case, no significant changes were observed
in the lifetime of the Eu(III) signal because of presatura-
tion of the available Streptavidin binding sites.
Through a thorough understanding of their solution
behavior, we have also identified that the use of basic alkyl
amine backbones to link adjacent 1,2-HOPO chelates can
influence the resulting triplet energies of the chromophore,
resulting in less efficient sensitization of the bound Eu(III)
metal center. As a result, we are now exploring methodo-
logies to exclude the remaining inner sphere water molecule
(by introducing additional steric bulk at the “open” face of
the tetrapodal complexes), and to replace the alkyl amine
backbone groups with either branched carbon linkages or
by utilizing less basic amide groups,33 with the hope of
preparing new octadentate 1,2-HOPO derivatives, whose
complexes with Eu(III) have overall quantum yields more
similar to Φtot ∼ 20% obtained for related bis(tetradentate)
complex topologies.
Conclusions
The 1-hydroxypyridin-2-one chelate group has proven to
be a remarkably efficient sensitizer for long-lived Eu(III)
luminescence. By using an exclusively oxygen containing
donor set, the resulting compounds also form extremely stable
complexes in aqueous solution with oxophilic lanthanides,
essential for their practical usage in bioassay applications,
where complexes must be highly resistant toward decomplexa-
tion, even in the presence of strong competitors such as EDTA
commonly used in biotechnology. Despite the presence of a
single inner sphere water molecule, these compounds exhibit
useful quantum yields in aqueous solution of Φtot ∼ 4%.
Furthermore, we have now developed synthetic methodo-
logies toward the preparation of pendant armed versions of
these ligands, containing either terminal amine or carboxylate
functional groups, suitable for the attachment to relevant
biomolecules, and have demonstrated the use of the resulting
bioconjugate in a simple LRET bioassay platform.
Acknowledgment. This work was partially supported by
the NIH (Grant HL69832) and supported by the Direc-
tor, Office of Science, Office of Basic Energy Sciences,
and the Division of Chemical Sciences, Geosciences, and
Biosciences of the U.S. Department of Energy at LBNL
under Contract No. DE-AC02-05CH11231. This tech-
nology is licensed to Lumiphore, Inc. in which some of the
authors have a financial interest. Financial support was
provided to C.J.J. by the German Research Foundation
(DFG).
Supporting Information Available: MM2 optimized mole-
cular models of protonated complexes, low temperature lumine-
scence spectra of Gd(III) complex, and emission spectra from
LRET assay. This material is available free of charge via the
Internet at http://pubs.acs.org.
(33) Moore, E. G.; D’Aleo, A.; Xu, J.; Raymond, K. N. Aust. J. Chem.
2009, 62, 1300–1307.