Design of Novel, Water Soluble and Highly Luminescent Europium Labels with Potential to Enhance Immunoassay Sensitivities

To meet the continual demands of more-sensitive immunoassays, the synthesis of novel luminescent Eu(III) chelate labels having similar substituted 4-(phenylethynyl)pyridine chromophores in three different chelate structure classes are reported. Significantly enhanced luminescence intensities were obtained, evidently caused by the intra-ligand charge transfer (ILCT) mediated sensitization, but the alternative ligands triplet state process cannot be ruled out. Based on the present study, even quite small changes on the chelate structure, and, especially, on the substituents’ donor/acceptor strength on both ends of 4-(phenylethynyl)pyridine subunits have an unpredictable effect on the luminescence. The highest observed brightness was 16,400 M−1 cm−1 in solution and 69,500 M−1 cm−1 on dry surface, being 3.4 and 8.7 fold higher compared to the reference chelate. The new label chelates provide solutions for improved assay sensitivity up-to tenfold from the present concepts.


Introduction
Time-resolved fluorometry (TRF) employing long-lifetime emitting luminescent lanthanide chelates has been applied in many specific binding assays, such as immunoassays, DNA hybridization assays, receptor-binding assays, enzymatic assays, bio-imaging such as immunocytochemical, immunohisto-chemical assays or cell based assays to measure the wanted analyte at very low concentration [1][2][3]. Lanthanide chelates have special properties offering proper alternative markers in bio-affinity assays: (i) the difference between the excitation and emission wavelengths is large; (ii) normally, they have long emission life-time compared to the background, and, thus, the time-resolved fluorescence (TR-FIA) measurement techniques together with narrow emission lines eliminate the fluorescence background almost completely; (iii) the concentration quenching is small and enables using wide dynamic assay range without additional dilutions, and (iv) several different sample matrixes e.g., whole blood, serum and plasma can be used for the assays. Applications based on TRF have been commercialized in the fields of clinical diagnostics and drug discovery [1].
For TRF applications, an optimal label has to fulfill several strict requirements. First, the label has to be photo-chemically stable both in the ground and excited states, and it has to be kinetically and chemically stable. The excitation wavelength has to be as high as possible, preferable over 300 nm. It has to have efficient cation emission i.e., intense brightness (excitation coefficient x quantum yield, εΦ). The lanthanide chelate has to have a high absorptivity and efficient energy transfer from the excited level of the so-called antenna ligand (i.e., chromophore) to the proper emitting level of the lanthanide ion. Luminescence quenching by water molecules should be prevented to enable a long luminescence lanthanide ion. Luminescence quenching by water molecules should be prevented to enable a long luminescence decay time, but, at the same time, the chelate has to have good water solubility. For the purpose of labeling, it should have a reactive group to allow covalent attachment to a bio-specific binding reactant, and the affinity and nonspecific binding properties of the labeled biomolecules have to be retained. It is challenging to prepare a chelate-label molecule that fulfills all requirements, and, therefore, certain compromises are generally made in the development of suitable labels. As a consequence hereof, a number of attempts (see e.g., [4]) have been made to tune the photo-physical properties of the chelate labels suitable for TRF applications, and only a few viable chelate labels are in commercial use. Since the discovery in the mid-1980s and the later publication of the phenylethynylpyridine antenna for Eu(III) chelates [5,6], several new chelates utilizing the chromophore have been described in literature, inter alia three [6][7][8][9][10], seven [6][7][8][11][12][13] and nine-dentate acyclic [8,14,15] as well as macro-cyclic ligands [8,[16][17][18][19]. The nine-dentate Eu(III) label (1 in Scheme 1) has been applied already almost for 15 years in TRF immunoassays [14,[20][21][22][23][24] especially in near-patient testing such as critical care and emergency situations. The technology has offered quick, easy-to-use, high quality and quantitative point-of-care testing platform called as AQT90 FLEX (Radiometer Medical ApS). The Since the discovery in the mid-1980s and the later publication of the phenylethynylpyridine antenna for Eu(III) chelates [5,6], several new chelates utilizing the chromophore have been described in literature, inter alia three [6][7][8][9][10], seven [6][7][8][11][12][13] and nine-dentate acyclic [8,14,15] as well as macro-cyclic ligands [8,[16][17][18][19]. The nine-dentate Eu(III) label (1 in Scheme 1) has been applied already almost for 15 years in TRF immunoassays [14,[20][21][22][23][24] especially in near-patient testing such as critical care and emergency situations. The technology has offered quick, easy-to-use, high quality and quantitative point-of-care testing platform called as AQT90 FLEX (Radiometer Medical ApS). The aim of the present work is to enhance the brightness and to red-shift the excitation wavelength of the present label 1 to be able to use a UV LED based excitation [25] with significantly improved assay sensitivity. For that purpose, we prepared three different nine-dentate chelate label designs i.e., with two and three individual chromophore moieties in acyclic and macrocyclic ligand formats (chelates 2-8 in Scheme 1). The objective was to study the effect of electron-releasing substituents (OCH 2 COO − ) in mesomeric paraand orto-position of phenylethynylpyridine moiety together with electron-withdrawing (COO − ) as well as releasing (CH 2 ) groups at the other end of the chromophoric moiety in those three different ligand groups. The novel chelates were tested after coupled with taurine (2a-8a) and protein (2b-8b) being comparable to the derivative 1a,b of the used reference chelate 1.

Syntheses
The syntheses of two different terminal acetylenes 14 and 15 used in the study started from 4-bromobenzene-1,3-diol and benzene-1,3,5-triol, which were transformed to the corresponding intermediates 11 and 9 with the reaction of ethyl bromoacetate in dry MeCN and using K 2 CO 3 as the needed base (Scheme 2). For the synthesis of compound 9, careful control of reaction conditions is required to achieve a reasonable yield; otherwise, a dramatic drop in yield together with purification problems will be the result. Direct iodination of the compound 9 with I 2 and NaHCO 3 in H 2 O/CHCl 3 mixture afforded poor results, whereas the reaction with N-chlorosuccimide and NaI in acetic acid described by Yamamoto et al. [26] produced the iodo derivative 10 in quantitative yield. As an alternative route to the compound 10, a reaction of 2-iodo-benzene-1,3,5-triol [27] with ethyl bromoacetate was entirely unsuccessful. Both halides 10 and 11 were conjugated with trimethysilylacetylene using the Sonogashira reaction in the presence of a catalytic amount of Pd(II) catalyst and CuI under argon. Microwave heating at 100 • C for 30-45 min in diethyl amine (DEA) and dimethylformamide (DMF) gave appropriate yields of products. In the reaction with 11, an improved yield was obtained by using PPh 3 in the mixture. We also tested using a corresponding bromo derivative to the compound 10-nicely obtained by the reaction of 9 with N-bromosuccimide, but, unfortunately, without success. Finally, the deprotection with tetrabutylammonium fluoride gave the wanted phenylacetylenes 14 and 15 with quantitative yields.
The second Sonogashira reaction between the acetylenes 14 and 15 and the dibromo intermediate 16 [14] gave ligand esters 17 and 18, respectively. The esters 17 and 18 were hydrolyzed with KOH in EtOH-H 2 O followed by the forming of chelates 19 and 20 with EuCl 3 in slightly acidic conditions and precipitating the excess europium as Eu(OH) 3 by adjusting the pH to 8.5. The chelates 19 and 20 were purified by the semi-preparative HPLC.
The activation of the amino group of the chelates by transformation to the corresponding isothiocyanato group in the final labeling reagents 2 and 3 was performed with thiophosgene in H 2 O/CHCl 3 .
The synthesis of the acyclic labeling regents (4 and 5 in Scheme 3) with three chromophores started from the protection of aminophenyl acetylene with trifluoroacetic acid anhydride to get compound 21. After the Sonagashira coupling reaction with the 6-bromo-2,6-di-hydroxy-methylpy-ridine 22 [28], diol 23 was transformed with PBr 3 into bis(bromomethyl) derivative 24. This product was conjugated with two pyridine subunits 26 in the dry MeCN in the presence of K 2 CO 3 to get the key intermediate 27.
The needed building block 26 was prepared from 4-bromo-6-bromomethyl-2-carboxyethylpyridine 25 [14] using five-fold excess of glycine ethyl ester hydrochloride in dry MeCN and di-isopropylethylamine. After the triplet bond coupling reaction of the two bromo groups in the compound 27 with the acetylenes 14 and 15, the target ester ligands 28 and 29 were hydrolyzed, transformed to the corresponding Eu(III) chelates 30 and 31 including the HPLC purifications, and, finally, activated to the target labeling chelates 4 and 5 as described above with the chelates 2 and 3.  The synthesis of the unsymmetrical acyclic analogue to the labeling reagent 5 i.e., the Eu(III) chelate 6 in Scheme 4 was more complicated. First, the 6-bromo-2,6-dibromomethylpyridine 32 [28] was transformed to diester 33 by the reaction with two ethyl glysinates, and, after conjugation to one equivalent of the compound 25, the triester 34 was obtained with a reasonable yield. It is worth mentioning that, although in the last step only one main product spot is seen on TLC, besides the wanted product 34, the TLC spot actually also contains a product in which two equivalents of the compound 25 have reacted with the compound 33. With a careful column chromatography, the two products can be separated from each other. The secondary amino in 33 was coupled to the The synthesis of the unsymmetrical acyclic analogue to the labeling reagent 5 i.e., the Eu(III) chelate 6 in Scheme 4 was more complicated. First, the 6-bromo-2,6-dibromomethylpyridine 32 [28] was transformed to diester 33 by the reaction with two ethyl glysinates, and, after conjugation to one equivalent of the compound 25, the triester 34 was obtained with a reasonable yield. It is worth mentioning that, although in the last step only one main product spot is seen on TLC, besides the wanted product 34, the TLC spot actually also contains a product in which two equivalents of the compound 25 have reacted with the compound 33. from 37 and di-tert-butyl 1,4,7-triazacyclononane-1,4-dicarboxylate (diBocTACN) was deprotected with CF 3 COOH to get the compound 46. The two phenylethynylpyridine subunits 43 and 44 were coupled with 46 in dry MeCN and DIPEA with high yields. It is notable that, according to mass spectra analysis of the crude reaction mixture, a mass corresponding of an over conjugated product was found i.e., the reactivity of bromide in 43 and 44 is high enough to be able to form ternary substituted amino function, although steric hindrance could be assumed significant enough to prevent such a reaction. The synthesis strategy to prepare the chromophores before conjugation to the 1,4,7-triazacyclononane (TACN) was chosen as the organometallic coupling reaction was not efficient when TACN was present, probably due to possible complex formations between Cu(I)/Pd(II) and the azacrown, which lead to incomplete reaction.
The ethyl ester groups of 47 and 48 were saponified with KOH in EtOH-H 2 O mixture; however, the conventional Eu(III) chelate formation in slightly acidic solution at room temperature lead to incomplete chelate formation according to HPLC and MS analyses. After multiple tests, the reasonable complex formation of 49 and 50 was achieved within several days incubation at 95 • C and pH 9.5 in aqueous citric solution to prevent Eu(OH) 3 precipitation. The possible reason for the slow complexation could be the high negative charge of the two chromophores causing repulsion and slow rotation of the chromophores to reach the final complex i.e., high rigidity of the ligand and/or a possible strong encapsulation of potassium ion in the TACN cavity during the earlier phase, as has been seen with e.g., crown ethers and cryptands. After the HPLC purification of chelates 49 and 50, the amino groups were activated with thiophosgene to receive the labeling reagents 7 and 8.

Photophysical Properties
The reference chelate (1) and prepared labeling reagents (2)(3)(4)(5)(6)(7)(8) were conjugated to taurine (→1a-8a) and protein (→1b-8b) followed by purifications. The photochemical properties i.e., the absorption maxima (λ abs ), the excitation maxima (λ exc ), decay-times (τ), the corresponding molar absorption coefficients (ε), brightness (εΦ) and quantum yield (Φ) in tris-saline-azide (TRIS) buffer (TRIS 50 mM, NaCl 0.9%, pH 7.75) are presented in Table 1. The table includes also the obtained enhanced signals from dried troponin I immunoassay wells, compared with the corresponding signal of the reference chelate with a known effect of drying [20]. Table 1. Absorption maxima (λ abs ) and excitation maxima (λ exc ), luminescence decay times (τ), molar absorption coefficients (ε), brightness (εΦ) and quantum yields (Φ) in TRIS buffer (pH 7.75) as well as the enhanced brightness on dried surface. The excitation wavelengths of the new chelates are ca. 15-20 nm red-shifted compared to the reference chelate, and are between 340-350 nm due to the additional electron-releasing ether functions at paraand meta-positions of the chromophores. The excitation maxima fit perfectly to a recently published new optical system based on UV light emitting diode excitation at 340 nm [25]. The third ether substituent (e.g., chelate 5 compared to 4) does not have any notable effect on the excitation wavelength, except with the chelate 2 (ca. 345 nm) compared to the corresponding complex 3 (ca. 350 nm). Generally, the increase of the chromophores decreases the H 2 O solubility as well as increases the formation of aggregates during the bio-molecule labeling process and non-specific binding properties for the labelled proteins. Aggregations will produce purification problems and reduce the yield of labelled material. Moreover, increased non-specific binding of the labelled biomolecule can enhance the assay background, and, thus, reduces the assay sensitivity. In the present study, the additional COO − Na + groups, together with the ether substituents offer high solubility in H 2 O, reduce the nonspecific binding of the labelled biomolecule, and the affinity properties of the used biomolecules are almost unchanged despite the increased negative charge of the labels. However, such groups close to each other might cause nonspecific binging to positive groups present in the assay, and, thus, interference with assay components might be obtained.
The absorptivities between the two chromophore chelates 1a-3a are close to each other from 55,000 to 58,000 M −1 cm −1 . As it could be assumed, the third chromophore increases the absorptivities. The measured values from 64,000 (5a) to 89,000 M −1 cm −1 (7a) express the higher value for the 2,4-substituted taurine chelate designs compared to the corresponding 2,4,6-substitued ones. The coupling of chelate to the protein changes the polarity of the chelates environment and seems to enhance absorptivity with all new chelates except 4b and 6b, and the change is strongest with the TACN based label 8b (104,000 vs. 77,000 M −1 cm −1 ).
The luminescence decay-times of the new chelates with two chromophores 2 and 3 are slightly shorter compared to 1, and are more reduced with the acyclic chelates with three chromophores 4-6 being lowest with TACN derivatives 7 and 8. Moreover, the life-times with 2,4,6-trisubstituted chelate designs are shorter compared to the corresponding 2,4-disubstitued analogues (3 vs. 2, 5 vs. 4 and 8 vs. 7). Moreover, the chromophore with an electron donating NH(C=S)NH group can have its own influence on the luminescence lifetime. This phenomenon of luminescence lifetime reduction is rather surprising as all chelate structures are nine-dentate and the chromophores are similar. Generally, the sensitization of lanthanide luminescence occurs through the ligands' triplet excited state. However, lately, an alternative route through an intra-ligand charge transfer (ILCT) state has been demonstrated [10,29]. Ligands, which contain a conjugated π-electron skeleton, such 4-(phenylethynyl)pyridine moiety, substituted by an electron-donating substituent together with an electron accepting substituent at the other end of the antenna, can be excited through the ligands's CT-state. Such antenna chromophore can present a dipolar (push-pull) geometry. The lanthanide ion acting as a Lewis acid increases the acceptor strength of this push-pull feature and lowers the energy of the relaxed CT excited state. When the CT state is low enough, the excited energy can transfer back from the excited state of lanthanide ion, which is observed as a reduced luminescence life-time. Thus, the shortened luminescence life-time can be assumed to reflect the energy transfer through the ILCT together with the original triplet state mediating route. It is known that the relaxation in the CT excitation path is strongly correlated with the donor/acceptor pair, which supports the shorter luminescence life-time of the chelates having chromophores with only one CH 2 group and one electron accepting COO − group in the pyridines (7 and 8) over the other chelates having chromophores also with two CH 2 groups. Thus, the chelates 7 and 8 have chromophores with lower CT states, whereas the others have subunits with CT state(s) at a higher energy level, which minimized both the energy transfer trough ILCT and the energy backflow. Similarly, the complexes with two chromophores 1-3 have only electron donating CH 2 groups in pyridine and show the longest lifetimes. Moreover, reference chelate 1 has only one electron donating substituent in the phenyl ring compared to others, and, thus, the longest decay time is observed. As a consequence, the increase of luminescence lifetime in order 1 > 2 > 3 > 4 > 5 = 6 > 7 > 8 can be attributed to the push-pull feature of the complex chromophores. Further evidence for the energy back-transfer phenomenon due to the low lying ILCT state indicates the luminescence lifetime measurements of chelates 6b (0.55 ms) and 8b (0.42 ms) in D 2 O, which rule out the possible lifetime reduction caused by water molecules.
The luminescence effectiveness of the chelates is affected by the molar absorption coefficient (ε) and the quantum yield (Φ), and it can be illustrated by the product εΦ, called brightness. From the chelates with two chromophores 2a gives almost two fold higher signal, whereas 3 is about at the same level compared to reference 1. The acyclic chelates 4-6 show notable enhanced luminescence intensities (9700-16,400 M −1 cm −1 ), being 2.0-3.1 times higher compared to the chelate 1. In addition, the quantum yields (13-26%) are significantly increased. A remarkable drop of brightness and quantum yield is seen with the chelates 7 and 8. The reason behind the obtained luminescence intensities cannot be explained by differences between triplet state energy levels as those can be assumed to be near to each other. Earlier triplet state energy level measurements from several different chelate designs having the same (phenylethynyl)pyridine subunit including three, seven and nine-dentate chelates have given the triplet state energy values in a range of 21,600-21,830 cm −1 [8]. Therefore, the observed effect on the luminescence intensity should be attributed to the excited energy transfer through the low lying ILCT states, although the triplet mediated process cannot be entirely ruled out. If the ILCT state is too low, the signal is reduced in aqueous solutions. Further evidence of the energy transfer through the ILCT state is seen from the brightness measurements on the surface dried assay wells after an immunoreaction (Table 1) compared to the corresponding wet measurements. The signal increase of 60% (i.e., 1.7 fold) for the chelate 1 has been demonstrated in literature [20]. It is known that with lanthanide chelates, which have low-lying CT state, the signal is highly dependent on solvent or environment polarity [29]. The enhanced luminescence on the dried surface are without doubt partly caused by the loss of H 2 O molecules causing the signal reduction due to the IR overtones of OH bonds in aqueous solution but also by the polarity change. The signal enhancement between wet and dry formats is ca. 2.8-4.1, 1.5-3.8 and 69-80 folds for the two chromophores containing chelates 2 and 3, the acyclic chelates 4-6 and the TACN chelates 7 and 8, respectively (see also Figure 1). When compared to the dry signal level of chelate 1, brightness improvement from 2.0 to 8.7 fold is obtained. More evidence of the CT route and its significance on the luminescence is clearly seen from Figure 2. A quite clear correlation is seen between the decay constants (k chel = 1/τ) in TRIS buffer and the measured enhanced brightness of the different chelates (1b-8b) after the drying. The luminescence effectiveness of the chelates is affected by the molar absorption coefficient (ε) and the quantum yield (Φ), and it can be illustrated by the product εΦ, called brightness. From the chelates with two chromophores 2a gives almost two fold higher signal, whereas 3 is about at the same level compared to reference 1. The acyclic chelates 4-6 show notable enhanced luminescence intensities (9700-16,400 M −1 cm −1 ), being 2.0-3.1 times higher compared to the chelate 1. In addition, the quantum yields (13-26%) are significantly increased. A remarkable drop of brightness and quantum yield is seen with the chelates 7 and 8. The reason behind the obtained luminescence intensities cannot be explained by differences between triplet state energy levels as those can be assumed to be near to each other. Earlier triplet state energy level measurements from several different chelate designs having the same (phenylethynyl)pyridine subunit including three, seven and nine-dentate chelates have given the triplet state energy values in a range of 21,600-21,830 cm −1 [8]. Therefore, the observed effect on the luminescence intensity should be attributed to the excited energy transfer through the low lying ILCT states, although the triplet mediated process cannot be entirely ruled out. If the ILCT state is too low, the signal is reduced in aqueous solutions. Further evidence of the energy transfer through the ILCT state is seen from the brightness measurements on the surface dried assay wells after an immunoreaction (Table 1) compared to the corresponding wet measurements. The signal increase of 60% (i.e., 1.7 fold) for the chelate 1 has been demonstrated in literature [20]. It is known that with lanthanide chelates, which have low-lying CT state, the signal is highly dependent on solvent or environment polarity [29]. The enhanced luminescence on the dried surface are without doubt partly caused by the loss of H2O molecules causing the signal reduction due to the IR overtones of OH bonds in aqueous solution but also by the polarity change. The signal enhancement between wet and dry formats is ca. 2.8-4.1, 1.5-3.8 and 69-80 folds for the two chromophores containing chelates 2 and 3, the acyclic chelates 4-6 and the TACN chelates 7 and 8, respectively (see also Figure 1). When compared to the dry signal level of chelate 1, brightness improvement from 2.0 to 8.7 fold is obtained. More evidence of the CT route and its significance on the luminescence is clearly seen from Figure 2. A quite clear correlation is seen between the decay constants (kchel = 1/τ) in TRIS buffer and the measured enhanced brightness of the different chelates (1b-8b) after the drying.  Finally, the dependency of luminescence on the measurement temperature between three chelates 1b, 6b and 8b in Figure 3 shows increasing luminescence quenching in order 8b > 6b > 1b being accordance with above observations, and is in agreement with the behavior of ILCT transition i.e., hypsochromic shift upon decreasing the temperature obtained earlier with phenylethynylpyri-dine Eu(III) chelates [29].  Finally, the dependency of luminescence on the measurement temperature between three chelates 1b, 6b and 8b in Figure 3 shows increasing luminescence quenching in order 8b > 6b > 1b being accordance with above observations, and is in agreement with the behavior of ILCT transition i.e., hypsochromic shift upon decreasing the temperature obtained earlier with phenylethynylpyri-dine Eu(III) chelates [29]. Finally, the dependency of luminescence on the measurement temperature between three chelates 1b, 6b and 8b in Figure 3 shows increasing luminescence quenching in order 8b > 6b > 1b being accordance with above observations, and is in agreement with the behavior of ILCT transition i.e., hypsochromic shift upon decreasing the temperature obtained earlier with phenylethynylpyri-dine Eu(III) chelates [29].  The luminescence intensities after drying are really remarkable e.g., when compared to the brightness of the well-known dissociation-enhanced lanthanide fluorescent immunoassays (DELFIA ® ) enhancement solution (26,320 M −1 cm −1 in [30]) used in many sensitive bio-affinity assays. Therefore, it should be possible to design highly sensitive TR-FIA assays by using the novel labeling chelates. Actually, preliminary troponin I assays performed have given three, five and seven folds improved assay sensitivities by using e.g., chelate labels 3, 6 and 8, respectively.

Materials
All commercially available solvents and reagents were used without further purifications. MeCN, THF and TEA were dried with molecular sieves and K 2 CO 3 was dried overnight at 140 • C before use. The organic intermediates were purified by silica gel 60 (Merck, Darmstadt, Germany) column chromatography. The low fluorescence MaxiSorp single wells made of irradiated polystyrene were purchased from Nunc (Thermo Fisher Scientific, Waltham, MA, USA). The monoclonal anti-troponin I detection antibodies were manufactured by International Point of Care Inc., Toronto, ON, Canada. The troponin I capture antibody and antigen were purchased from HyTest Ltd. (Turku, Finland).

Syntheses
The general synthetic pathways used for the Eu(III) Chelates 2-8 are shown in Schemes 2-5. The more detailed synthetic procedures and the results of the spectroscopic product characterization are presented in the Supporting information. Microwave assisted syntheses were performed with a Initiator+ microwave synthesizer from Biotage (Uppsala, Sweden).
The Eu(III) chelates were analysed and purified by using a reversed phase HPLC (2996 Photodiode Array Detector, 600 Controller, Delta 600 Fraction Collector III; Waters, Milford, MA, USA) with a RP-18 column. The solvents were A: triethylammonium acetate buffer (20 mM, pH 7) and B: 50% acetonitrile in triethylammonium acetate buffer (20 mM, pH 7). The gradient was started from 5% of solvent B and the amount of solvent B was linearly raised to 100% within 30 min.
After stirring overnight at RT, the pH was adjusted to 8.5 with 1 M NaOH. The precipitate was removed by centrifugation and the supernatant evaporated to dryness. The product was purified by HPLC. After stirring for 45 min at RT, the two phases were separated and the aqueous phase was washed with CHCl 3 (3 × 1.3 mL). The product was precipitated with acetone (ca. 45 mL), isolated by centrifugation, washed with acetone (2 × 10 mL) and dried overnight in vacuum desiccator. The products were used as such for next phase or for labeling the antibodies.  1 (1H, s). 13 (24). PBr 3 (0.35 mL, 3.67 mmol) was added in a suspension of the compound 23 (0.86 g, 2.45 mmol) in CHCl 3 (100 mL). After stirring for 20 h at 60 • C, the mixture was neutralized with 5% NaHCO 3 (50 mL). The aqueous phase was extracted with CHCl 3 (50 mL) and the combined organic phases were dried with Na 2 SO 4 , filtered and evaporated to dryness. The product 24 (1.09 g, 93%) was used for the next step without further purifications. 1 (26). A mixture of 4-bromo-6bromomethyl-2-carboxyethylpyridine (25; [14]) (1.66 g, 5.15 mmol), glycine ethyl ester hydrochloride (3.60 g, 25.8 mmol) and di-isopropylethylamine (9.1 mL) in dry MeCN (70 mL) was stirred overnight at RT. The mixture was evaporated to dryness and the product was purified by silica gel column chromatography using MeOH/CH 2 Cl 2 (3:97) as an eluent. Yield 1.55 g (87%). 1 13  General Procedure for the Synthesis of Ligand Esters 28 and 29. A mixture of compound 27 (0.120 g, 0.148 mmol), compound 14 or 15 (0.355 mmol) in dry TEA (1 mL) and THF (2 mL) was de-aerated with argon. After an addition of bis(triphenylphosphine)palladium(II) dichloride (10 mg, 14 µmol) and CuI (6 mg, 28 µmol), the mixture was stirred overnight at 55 • C and evaporated to dryness. The residue was dissolved in CH 2 Cl 2 (30 mL), washed with H 2 O (2 × 15 mL), dried with Na 2 SO 4 and purified by silica gel column chromatography.
Ligand Ester 28. The product purified by silica gel column chromatography using 10% EtOH/CH 2 Cl 2 as an eluent. Yield: 69%. 1 13 13  General Procedure for the Synthesis of Eu(III) Chelates 30 and 31. Compound 28 or 29 (90 µmol) was stirred in 0.5 M KOH in EtOH (13 mL) for 30 min at RT. H 2 O (10 mL) was added and the mixture was further stirred at RT for 3 h. After evaporation of EtOH and an additional overnight stirring at RT, the pH was adjusted to 6.5 by addition of 6 M HCl. EuCl 3 (33 mg, 90 µmol) in water (0.5 mL) was added within 5 min and the pH was maintained at 6.0-6.5 with suitable additions of solid NaHCO 3 .
After stirring overnight at RT, the pH was adjusted to 8.5 with 1 M NaOH. The precipitate was removed by centrifugation and the supernatant evaporated to dryness. The product was purified by HPLC. General Procedure for the Synthesis of Eu(III) Chelates 4 and 5. An aqueous solution (2.7 mL) of the Eu(III) chelate 30 or 31 (0.10 mmol) was added within 5 min to a mixture of CSCl 2 (53 µL, 0.70 mmol) and NaHCO 3 (67 mg, 0.80 mmol) in CHCl 3 (2.7 mL). After stirring for 30 min at RT, the two phases were separated and the aqueous phase was washed with CHCl 3 (3 × 3 mL). The product was precipitated with acetone (ca. 45 mL), isolated by centrifugation, washed with acetone (2 × 10 mL) and dried overnight in vacuum desiccator. General Procedure for the Synthesis of Eu(III) Chelates 4a and 5a. A mixture of Eu(III) chelate 4 or 5 (6.3 µmol) and taurine (8 mg, 63 µmol) was stirred in aqueous 50 mM Na 2 CO 3 buffer (0.64 mL, pH 9.8) and DMF (0.64 mL) for overnight at RT. The product was purified by HPLC.