Synthesis, Characterization and Metal Ion Detection of Novel Fluoroionophores Based on Heterocyclic Substituted Alanines

Abstract

The synthesis of new fluorescent probes containing the thiophene and benzoxazole moieties combined with an alanine residue is described. The resulting highly fluorescent heterocyclic alanine derivatives respond via a quenching effect, with paramagnetic Cu(II) and Ni(II) metal ions and with diamagnetic Hg(II), as shown by the absorption and steady-state fluorescence spectroscopy studies. The formation of mononuclear or dinuclear metal complexes was postulated based on the presence of the free carboxylic acid as binding site and also with the interaction with the donor atoms in the chromophore. Interaction with other important biological metal ions such as Zn(II), Ca(II) and Na(I) was also explored.

1. Introduction

In the last few years various 2-substituted benzoxazole derivatives were studied extensively for their antitumor, antiviral, antimicrobial activities as non-nucleoside topoisomerase 1 poison, HIV-1 reverse transcriptase and/or DNA gyrase inhibitors [1]. Other applications include in vivo probes for positron emission tomography (PET) and single-photon emission computed tomography for early detection of amyloid plaque formation and for visualization of brain dopamine D₃ receptors [2-4]. Derivatives containing the benzoxazole or the thiophene nucleus combined with amino acid moieties have also important biological applications such as potent inhibitors of human cystein proteases, as ligands of the NMDA receptor, in molecular recognition, as biomarkers or biosensors [5-8]. Owing to their excellent optical properties and high lipophilicity, it is possible for benzoxazole compounds to be used as sensing materials. Therefore, several authors described recently the application of benzoxazole derivatives as fluorescent and/or colorimetric sensors for metals, anionic species and pharmaceutical analysis [9-17].

The interesting biological activity and photophysical properties of thiophene and benzazole derivatives lead us to the synthesis of alanine derivatives containing these heterocyclic nuclei [18-24] The thienyl-benzoxazole amino acid derivatives can be incorporated into peptide chains and as such used as energy donors in conformational studies of peptides by means of fluorescence or be used as fluorescence markers. To our knowledge this is the first time that a fluorescent sensor containing the thiophene and benzoxazole fluorophores combined with the alanine residue selectively deprotected is described.

As a part of our ongoing research, in the present work we have combined the ability of amino acids for metal ion chelation with the strong fluorescence properties of the 2-substituted benzoxazole moiety. Thus, we were interested in studying the influence of both potential chelating units incorporated: i) the amino acid residue, bearing an amino and carboxylic acid terminal groups, partially or totally deprotected and ii) the aromatic chelating unit in the chromophore, with two possibilities for chelation: a NS or OS donor sets. The presence of two or more coordination centres in these compounds can lead to the design of interesting supramolecular structures for biological and environmental applications.

2. Results and Discussion

2.1. Synthesis

Starting from commercially available 3-nitro-l-tyrosine, reaction with thionyl chloride in methanol yielded the corresponding methyl ester 1a. Subsequent reaction with t-butyl pyrocarbonate resulted in the N-Boc protected derivative 1b. The nitro group in this compound was reduced to the amino group, by catalytic hydrogenation, yielding compound 1c [25], which was condensed with 2-formylthiophene, in order to obtain the imino derivative 1d. By reaction with lead tetraacetate (LTA) in DMSO [26], compound 1d was oxidised to the [2-(2′-thienyl)benzoxazol-5-yl] alanine derivative 2a.

Compound 2a was then selectively deprotected at its C- and N-terminus, yielding the corresponding N-Boc protected compound 2b and the methyl ester derivative 2c, as well as the fully deprotected alanine derivative 2d. By using similar procedures, compound 1c was reacted with 2,4,5-trimethoxybenzaldehyde, yielding imine 1e. Oxidation of this compound yielded [2-(2′,4′,5′-trimethoxyphenyl)benzoxazol-5-yl] alanine derivative 2e [27]. Selective removal of the protecting groups resulted in the corresponding N-Boc protected 2f and the fully deprotected 2g alanine derivatives (Scheme 1). Compounds 2e-g based on trimethoxyphenyl were prepared in order to confirm the involvement of the sulphur atom at the thiophene ring in the coordination process. All the compounds were obtained in good to excellent yields (

Table 1. Synthesis data of tyrosine derivatives 1d-e and alanine derivatives 2a-g.

Compd	Yield (%)	IR (cm−1)	1H NMR (δ, ppm)
1d	98	3375, 1754	8.76 (N=CH) 6.90-6.95 (H-2), 7.05 (H-5)
1e	97	3430, 1741	9.01 (N=CH) 6.88-6.94 (H-2), 6.99 (H-5)
2a	56	3358, 1734	7.48 (H-4), 7.45 (H-7)
2b	84	3357, 1752	7.63 (H-4), 7.48 (H-7)
2c	74	3310, 1741	7.56-7.59 (H-4), 7.49 (H-7)
2d	54	3413, 1620	7.93-7.96 (H-4), 7.62-7.67 (H-7)
2e	65	---	7.50 (H-4), 7.45 (H-7)
2f	82	3340, 1761	7.66 (H-4), 7.50 (H-7)
2g	44	3400, 1618	7.95(H-4), 7.60 (H-7)

) and were completely characterised by the usual techniques (MS, IR, UV-vis, ¹H and ¹³C NMR spectroscopy).

2.2. Photophysical study

The absorption and emission spectra of benzoxazolyl-alanines 2a-d bearing a thiophene heterocycle and trimethoxy-phenyl alanine derivatives 2e-g in absolute ethanol solution at 298 K were performed. The absorption and emission bands of the 2-substituted benzoxazole derivatives were centered at 315-336 and 393-398 nm, respectively. The fluorescence quantum yields for compounds 2a-g were determined and the derivatives 2a-d with a thiophene moiety were found to be strongly emissive (0.66 < Φ_F < 0.80) while compounds 2e-g displayed much lower quantum yields (0.26 < Φ_F < 0.47) (

Table 2. UV-vis and fluorescence data for alanine derivatives 2a-g.

Compound	UV-vis λmax (nm)	λem (nm)	Fluorescence ΦF	Stokes' shift (nm)
2a	315	394	0.80	79
2b	316	393	0.66	77
2c	315	394	0.77	79
2d	315	393	0.76	78
2e	334	396	0.47	62
2f	334	395	0.44	61
2g	336	398	0.26	62

). As can be seen by the results, the selective deprotection at the amino and carboxylic groups of the amino acid residue had only a minor influence on the fluorescence quantum yields. No solvatochromic effect in the absorption spectra was observed in absolute ethanol/water (1:1), absolute ethanol, acetonitrile, dioxane and cyclohexane for all compounds studied.

Compounds 2a-g are amino acid derivatives, which in solution are present in zwitterionic form, and during our studies their zwitterionic nature was taken into account. All of our studies were performed in absolute ethanol and/or in water-absolute ethanol mixtures and the measured pH value was always between 5-6, at which the zwitterionic form predominates. At this pH value, the amine group in alanine derivatives is protonated and the carboxylic group is deprotonated, being available for coordination with the metal ions studied, Cu(II), Hg(II) and Ni(II), in these conditions.

2.3. Spectrofluorimetric titrations and metal sensing effect

In order to evaluate the sensor ability of systems 2a-g in solution towards Zn(II), Hg(II), Cu(II), Ni(II), Ca(II) and Na(I) in low acidic or neutral conditions, UV-vis and fluorescence studies were performed.

2.3.1. Protonation effects

In the presence of increasing amount of HBF₄, only compounds 2a and 2f, in acetonitrile or absolute ethanol were slightly affected. The maximum of absorption at 315 nm (2a) and 334 nm (2f), showed a 4 nm red shift upon acid addition. For compound 2f, a new band centered at 380 nm was observed when a larger amount of acid was added (30 to 190 equivalents). As previously reported for related quinoxalinyl derivatives [15], compounds 2b-d and 2e,g did not reveal any significant change in the absorption spectrum upon acid addition in absolute ethanol and absolute ethanol/water (1:1). With further addition of acid, a decrease in the intensity of the fluorescence emission was observed. This quenching may be a result of protonation on the benzoxazole nitrogen atom.

2.3.2. Deprotonation effects

Addition of bases such as potassium hydroxide, triethylamine and tetramethylammonium hydroxide solutions to ligands 2b, 2d, 2f and 2g in ethanolic solutions (with a free carboxylic acid group) were performed.

The maximum of absorption at 315 nm for 2a, and 334 nm for 2f, was slightly affected upon addition of 100 equivalents of tetrabutylammonium hydroxide solution. With further addition of base (more than 500 equivalents), a 50% quenching effect in fluorescence emission was observed. This effect can be attributed to a photoinduced electron transfer (PET) process from the NH₂ deprotonated group to chromophore. At this point, the pH value was 10-11 and addition of Hg(II) ions to a basic solution of ligand 2d led to the precipitation of a solid (hydroxide complex), thus preventing complexation studies in basic conditions.

2.3.3. Metal sensing effects

Compounds 2a and 2e, protected at amino and carboxylic terminals, did not show any change in the ground (absorption) and excited (emission) states after the addition of Na(I), Ca(II), Zn(II), Cu(II) and Ni(II) cations. In the titration of 2a with a Hg(CF₃SO₂)₂ ethanolic solution, the intensity of fluorescence was reduced in 30% after the addition of 30 equivalents of metal ion, while the spectrum of trimethoxy-phenyl alanine derivative 2e was not affected by Hg(II) addition. This result can be due to the involvement of the S atom in complexation.

Selective deprotection of the carboxylic terminal in 2a and 2e, gave compounds 2b and 2f. Both systems responded to the presence of Cu(II), Ni(II) and Hg(II), while addition of increasing amounts of Na(I) and Ca(II) did not change either the absorption and emission spectra. A rising CHEQ effect (Chelation Enhancement of the Quenching) [28-29] was observed in the cases of Cu(II), Ni(II) and Hg(II). This CHEQ effect is commonly observed in polyamine ligands containing aromatic fluorophores, and can be attributed to an energy transfer quenching of the π* emissive state through low-lying metal-centre unfilled d-orbitals for Cu(II), and to an intersystem crossing mechanism due to the heavy atom effect for Hg(II) [29]. While the quenching of fluorescence is predominantly connected with the nature of metal ion, the ligand fluorescence enhancement could be result from changes of the geometry between the steady and excited states, or be also induced by the ligand. A very small CHEF effect (Chelation enhancement of the Fluorescence Emission) was observed only when compound 2b was titrated with Zn(II) [30-31].

Although Cu(II) and Ni(II) affected only the excited state of the chromophore in both aforementioned ligands 2b and 2f, Hg(II) affected the ground and excited states when ligand 2b, bearing a thiophene unit, was used. This result suggests that the metal ion interacts with both chelating units present in the compound: the carboxylic acid group of the amino acid residue and the S atom of the thiophene heterocycle [32]. For ligand 2b, complexation constants using the SPECFIT/32 program [33] were obtained only for the interaction with Hg(II). The results suggested that two metal ions are coordinated to the ligand. In the case of ligand 2f all metal complexation studies showed values that agree with 2:1 metal to ligand molar ratio for Hg(II) and 1:1 metal to ligand ratio for Cu(II) and Ni(II) (

Table 3. Complexation constants for alanine derivatives 2b-d and 2f-g with Cu(II), Ni(II) and Hg(II) in absolute ethanol.

Compd.	Metal complex	Log K	M.L.Ha
2b	2bHg	5.01 ± 4.0E-02 9.37 ± 7.0E-02	1.1.0 2.1.0
2d	2dHg	7.78 ± 2.0E-02	2.1.0
2d	2dCu	9.77 ± 2.0E-02	2.1.0
2f	2fHg	7.09 ± 8.0E-02	2.1.0
2f	2fCu	6.22 ± 2.0E-02	1.1.0
2f	2fNi	6.67 ± 6.0E-02	1.1.0
2g	2gCu	6.10 ± 6.0E-02	1.1.0
2g	2gNi	10.26 ± 8.0E-02	2.1.0

^aM.L.H means Metal. Ligand. proton ratio

).

In Figure 1 is represented the absorption (A) and emission (B) spectra of compound 2b in the presence of increasing amount of Hg(II). The increase of Hg(II) ion concentration in absolute ethanol solution caused the decrease in the absorbance in the range of 280-340 nm, and increase in absorption between 340-390 nm. A sharp isosbestic point at λ= 330 nm was detected (Fig. 1A). After the addition of small molar ratio equivalents of metal ions (0.25 to 1 equivalent) a small effect in absorption and emission was observed. By increasing the metal-ligand molar ratio, strong changes in absorption and emission were observed (inset Figure 1B).

The absorption and emission spectra of compound 2b with increasing amount of Cu(II) is represented in figure 2. It can be seen that while the emission was strongly affected by metal ion complexation, no major changes were seen in the absorption. These results did not allow us to calculate the stoichiometry of complexes with Cu(II) and Ni(II), formed in the ground state. Probably the formation of the 1:1 M:L metal complexes has relatively low values for the binding constants. Increasing the metal ion concentration forces the formation of dinuclear species. Due to the spectral response, complexation probably takes place firstly at the carboxylic acid group of the aminoacid residue, farther from the chromophore, and the second metal ion complexation occurs at the donor atoms present in the chromophore unit. As observed before, interaction of ligand 2b with Hg(II), Ni(II) and Cu(II) also caused a strong CHEQ effect, suggesting the interaction with these metal ions with the chromophore.

The complexation of ligand 2f with Hg(II), Cu(II) and Ni(II) in absolute ethanol can be seen in Figures 3 and 4, respectively. The amount of metal ion necessary to quench the fluorescence emission was higher than that used for the thiophenic ligand 2b. In the case of compound 2f, fifty equivalents were used, in contrast with the six equivalents of Hg(II) needed to quench ligand 2b. These results suggest lower complexation constants for compound 2f. In the cases of Ni(II) and Cu(II), 5 equivalents of metal ion were enough in all cases to reduce the intensity of emission in 80%.

Selective deprotection at the amino terminal of the amino acid gave compound 2c with a free NH₂ group. Due to the zwitterionic form of compound 2c studied (pH = 5-6) in which the amine group is protonated, NH₃⁺, a strong decrease in the coordination ability in absolute ethanol and absolute ethanol/water (1:1) was observed when compared with ligand 2b, with a free carboxylic acid group. After addition of Na(I), Ca(II), Zn(II), and Ni(II), only minor changes were observed in the absorption and emission spectra. In the case of Hg(II) and Cu(II) a decrease in the intensity of the fluorescence emission in 40% and 30%, respectively, was observed when a metal to ligand molar ratio of 30:1 was achieved.

Total deprotection of both amino and carboxylic acid terminals in alanines 2a and 2e gave compounds 2d and 2g in the form of free amino acid residues. In our conditions, the zwiterionic form was present in solution. Although higher solubility in water would be expected, ligands 2d and 2g were sparingly soluble in water probably due to the heteroaromatic moiety present at the ligands. Therefore, the complexation studies were carried out in acetonitrile, absolute ethanol and/or water-absolute ethanol mixtures, in order to compare the results for all ligands studied. Once again, no interaction was observed upon complexation with Na(I), Ca(II) and Zn(II) for both ligands, 2d and 2g, in the absorption and emission spectra.

In figure 5 is represented the Hg(II) titration with ligand 2d, showing that both, ground and excited states, are affected by the complexation and an isosbestic point at 325 nm is observed. In contrast, interaction of Hg(II) with compound 2g affected only the excited state, and in this case the intensity of the emission was reduced in 30%. This result is comparable with the results obtained for compounds 2a and 2b, suggesting the importance of the sulphur atom for the sensing of this metal ion, due to its role on the complexation. The values for the complexation constants were calculated and are presented in Table 3.

Cu(II) and Ni(II) complexation are represented in figure 6A and 6B for ligand 2d, and in figures 7A and 7B for ligand 2g. Stronger interaction with Cu(II) was observed in both ligands. Five and two Cu(II) equivalents were enough to totally quench the fluorescence emission of 2d and 2g, while 40 and 10 equivalents of Ni(II), respectively, were necessary to quench the fluorescence emission to 90%. A dinuclear complex could be postulated by the global log K calculated using the SPECFIT32 programme in the case of Ni(II), and a mononuclear species was postulated for Cu(II) (see table 3). A schematic drawing of the proposed coordination process is displayed in Scheme 2. Firstly, the metal ion binds through the carboxylic acid at the alanine residue with small change in the fluorescence emission, followed by another metal ion interaction at the coordinative site present in the chromophore, which strongly affects the emission spectrum.

2.4. Conclusions

The synthesis of new highly fluorescent probes containing the thiophene and benzoxazole moieties combined with an alanine residue was achieved by simple procedures in excellent yields.

The resulting heterocyclic alanine derivatives respond with a quenching effect to the presence of Cu(II), Ni(II) and Hg(II), as shown by the absorption and steady-state fluorescence spectroscopy studies. Selective deprotection of the carboxylic acid group increased the coordination ability, while that deprotection of the amino group reduced notably this effect. The sensor effect takes place by both coordinative sites, for systems 2b and 2d, involving first the terminal carboxylic acid group of the amino acid, and finally the donor atoms present in the chromophores. The synthetic, photophysical and metal ion sensing properties displayed by these compounds showed that they are promising candidates as fluorescent probes and also for sensory applications.

3. Experimental Section

3.1. Synthesis general

All melting points were measured on a Gallenkamp melting point apparatus and are uncorrected. TLC analyses were carried out on 0.25 mm thick precoated silica plates (Merck Fertigplatten Kieselgel 60F₂₅₄) and spots were visualised under UV light. Chromatography on silica gel was carried out on Merck Kieselgel (230-240 mesh). IR spectra were determined on a Perkin Elmer FTIR-1600 using KBr discs. UV/visible spectra were run on a Shimadzu UV/2501PC or a Perkin Elmer lamda-35 spectrophotometers. ¹H NMR spectra were recorded on a Varian 300 spectrometer in CDCl₃ or DMSO-d₆ at 300 MHz at 25 °C. All chemical shifts are given in ppm using δ_H Me₄Si = 0 ppm as reference and J values are given in Hz. ¹³C NMR spectra were run in the same instrument at 75.4 MHz using the solvent peak as internal reference. Assignments were made by comparison of chemical shifts, peak multiplicities and J values and were supported by spin decoupling-double resonance and bidimensional heteronuclear HMBC and HMQC correlation techniques. Mass spectrometry analyses were performed at the “C.A.C.T.I. -Unidad de Espectrometria de Masas”, University of Vigo, Spain. Elemental analyses were carried out on a Leco CHNS 932 instrument. Fluorescence spectra were collected using a Perkin Elmer LS45.

3-Nitro-l-tyrosine methyl ester hydrochloride (1a)

Thionyl chloride (0.65 mL, 8.85 × 10⁻³ mol) was added drop wise with stirring to methanol (10 mL), cooled in an ice bath, followed by the addition of 3-nitrotyrosine (2.00 g, 8.85 × 10⁻³ mol). The solution was heated at 40 °C for 5 h. The solvent was evaporated under reduced pressure, yielding a yellowish green solid (2.44 g, quant.). M.p. > 300 °C [lit. [25] data not quoted]. ¹H NMR (DMSO-d₆): δ = 3.11 (d, J = 6.3 Hz, 2H, β CH₂), 3.71 (s, 3H, CH₃), 4.30 (t, J = 6.6 Hz, 1H, αH), 7.15 (d, J = 8.4 Hz, 1H, H-5), 7.38 (dd, J = 2.1 and 8.4 Hz, 1H, H-6), 7.78 (d, J = 2.1 Hz, 1H, H-2), 8.55 (br s, 3H, ⁺NH₃), 11.08 (br s, 1H, OH) ppm. ¹³C NMR (DMSO-d₆): δ = 34.26 (β CH₂), 52.75 (CH₃), 52.94 (αC), 119.36 (C5), 125.46 (C1), 126.15 (C2), 136.30 (C3), 136.59 (C6), 151.51 (C4), 169.24 (C=O) ppm. IR (KBr): ν = 3388 (br), 3204, 2856, 1746, 1635, 1589, 1544, 1493, 1445, 1390, 1318, 1241, 1064, 990 cm⁻¹.

N-tert-Butyloxycarbonyl-3-nitro-l-tyrosine methyl ester (1b)

Compound 1a (1.52 g, 5.5 × 10⁻³ mol) was added to 1,4-dioxane (11 mL), water (5.5 mL) and sodium hydroxide 1 m aqueous solution (5.5 mL), with stirring in an ice bath. t-Butyl pyrocarbonate (1.36 g, 1.1 eq, 6.05 × 10⁻³ mol) was added and the mixture stirred at room temperature for 2 days. The volume was reduced by half in a rotary evaporator under reduced pressure and ethyl acetate was added to cover the aqueous layer. The mixture was acidified by addition of KHSO₄ 1 m aqueous solution, until acidic pH with congo red paper, and extracted with ethyl acetate (3 × 10 mL), followed by washing of the organic layer with water (3 × 10 mL). The organic layer was dried with anhydrous magnesium sulphate and evaporated under reduced pressure, resulting a yellow solid (1.68 g, 90%). M.p. 97.3-98.4 °C [lit. [25] 90-91°C]. ¹H NMR (CDCl₃): δ = 1.43 (s, 9H, C(CH₃)₃), 2.95-3.04 (m, 1H, β CH₂), 3.12-3.22 (m, 1H, β CH₂), 3.76 (s, 3H, CH₃), 4.54-4.60 (m, 1H, αH), 5.07 (d, J = 6.6 Hz, 1H, NH), 7.10 (d, J = 8.4 Hz, 1H, H-5), 7.37 (dd, J = 1.8 and 8.4 Hz, 1H, H-6), 7.87 (d, J = 1.8 Hz, 1H, H-2), 10.50 (s, 1H, OH) ppm. ¹³C NMR (CDCl₃) δ = 28.21 (C(CH₃)₃), 37.23 (β CH₂), 52.51 (CH₃), 54.18 (αC), 80.27 (C(CH₃)₃), 120.09 (C5), 125.25 (C2), 128.62 (C1), 133.28 (C3), 138.65 (C6), 154.13 (C4 and C=O Boc), 171.69 (C=O ester) ppm. IR (KBr): ν = 3335, 3023, 3007, 2984, 2955, 2936, 1732, 1685, 1629, 1578, 1536, 1439, 1324, 1221, 1162, 1061, 994, 838 cm⁻¹.

N-tert-Butyloxycarbonyl-3-amino-l-tyrosine methyl ester (1c)

Compound 1b (1.00 g, 2.94 × 10⁻³ mol), 1,4-cyclohexadiene (0.5 mL, 1.8 eq, 5.29 × 10⁻³ mol) and Pd/C (147 mg) were dissolved in methanol (15 mL) and refluxed for 24h. The solvent was evaporated and the residue submitted to column chromatography with silica gel (eluent: CHCl₃/MeOH, 98:2). The fractions were combined and the product was obtained as a light brown solid (0.92 g, 98%). M.p. 52.0-52.9 °C [lit. [25] data not quoted]. ¹H NMR (CDCl₃): δ = 1.43 (s, 9H, C(CH₃)₃), 2.92 (d, J = 4.8 Hz, 2H, β CH₂), 3.74 (s, 3H, CH₃), 4.47-4.54 (m, 1H, αH), 5.05 (d, J = 8.4 Hz, 1H, NH), 6.37 (dd, J = 2.1 and 8.1 Hz, 1H, H-6), 6.53 (d, J = 2.1 Hz, 1H, H-2), 6.64 (d, J = 8.1 Hz, 1H, H-5) ppm. ¹³C NMR (CDCl₃) δ = 28.27 (C(CH₃)₃), 37.64 (β CH₂), 52.21 (CH₃), 54.57 (αC), 80.11 (C(CH₃)₃), 115.30 (C5), 117.33 (C2), 119.89 (C6), 128.32 (C1), 134.65 (C3), 143.33 (C4), 155.36 (C=O Boc), 172.78 (C=O ester) ppm. IR (KBr): ν = 3376, 3005, 2955, 1738, 1694, 1615, 1520, 1505, 1454, 1441, 1393, 1368, 1288, 1251, 1217, 1164, 1059, 1022, 912, 861, 799, 758 cm⁻¹. MS (FAB): m/z (%) = 311 (M⁺ + H, 41), 310 (M⁺, 100), 307 (20), 255 (59), 254 (47), 211 (57), 193 (27) 155 (24), 154 (83). HRMS (FAB): calcd. for C₁₅H₂₃N₂O₅ 311.1607, found 311.1568.

N-tert-Butyloxycarbonyl-3-[(thien-2′-ylmethylene)amino]-l-tyrosine methyl ester (1d)

Compound 1c (0.54 g, 1.74 × 10⁻³ mol) was stirred with 2-formylthiophene (0.16 mL, 1.76 × 10⁻³ mol) in absolute ethanol (5 mL) for 5 days. The solvent was evaporated and the crude residue was used in the next step without further purification, as light yellow oil (0.68 g, 98%). ¹H NMR (CDCl₃): δ = 1.42 (s, 9H, C(CH₃)₃), 2.98-3.10 (m, 2H, β CH₂), 3.72 (s, 3H, CH₃), 4.52-4.60 (m, 1H, αH), 5.02 (d, J = 7.80 Hz, 1H, NH), 6.90-6.95 (m, 2H, H-2 + H-6), 7.05 (br s, 1H, H-5), 7.14-7.17 (m, 1H, H-4′), 7.51-7.55 (m, 2H, H-3′ + H-5′), 8.76 (s, 1H, N=CH) ppm. ¹³C NMR (CDCl₃) δ = 28.28 (C(CH₃)₃), 37.84 (β CH₂), 52.23 (CH₃), 54.55 (αC), 79.99 (C(CH₃)₃), 114.99 (C2), 116.54 (C5), 127.56 (C1), 127.99 (C4′), 129.46 (C6), 130.82 (C3′ or C5′), 132.59 (C3′ or C5′), 134.98 (C3), 142.73 (C2′), 149.88 (CH), 151.15 (C4), 155.09 (C=O Boc), 172.41 (C=O ester) ppm. IR (KBr): ν = 3375, 2981, 2930, 1754, 1682, 1617, 1602, 1513, 1500, 1430, 1392, 1367, 1295, 1253, 1217, 1165, 1047, 1025, 855, 814 cm⁻¹. MS (FAB): m/z (%) = 405 (M⁺ + H, 40), 350 (22), 349 (100), 216 (33). HRMS (FAB): calcd. for C₂₀H₂₅N₂O₅S 405.1484, found 405.1488.

N-tert-Butyloxycarbonyl-3-[(2′,4′,5′-trimethoxyphenylmethylene)amino]-l-tyrosine methyl ester (1e)

Compound 1c (0.365 g, 1.18 × 10⁻³ mol) was stirred with 2,4,5-trimethoxybenzaldehyde (0.231 g, 1.18 × 10⁻³ mol) in absolute ethanol (5 mL) for 3 days. The solvent was evaporated and the residue was used in the next step without further purification as brown oil (0.56 g, 97%). ¹H NMR (CDCl₃): δ = 1.52 (s, 9H, C(CH₃)₃), 2.97-3.05 (m, 2H, β CH₂), 3.88 (s, 3H, OCH₃), 3.93 (s, 3H, OCH₃), 3.97 (s, 3H, OCH₃), 4.50-4.58 (m, 1H, α-H), 5.00 (d, J = 8.4 Hz, 1H, NH), 6.53 (s, 1H, H-3′), 6.88-6.94 (m, 2H, H-2 + H-6), 6.99 (br s, 1H, H-5), 7.63 (s, 1H, H-6′), 9.01 (s, 1H, N=CH) ppm. ¹³C NMR (CDCl₃) δ = 28.40 (C(CH₃)₃), 38.02 (β CH₂), 52.27 (CH₃), 54.65 (αC), 80.06 (C(CH₃)₃), 100.12 (C3′), 106.33 (C1′), 115.11 (C2), 115.71 (C6′), 116.33 (C5), 127.71 (C1), 129.76 (C6), 135.12 (C3), 135.79 (C5′), 144.76 (C4′), 149.79 (CH), 151.18 (C4), 152.06 (C2′), 155.01 (C=O Boc), 172.23 (C=O ester) ppm. IR (KBr): ν = 3430 (br), 2956, 2927, 2854, 1741, 1708, 1611, 1596, 1510, 1466, 1439, 1412, 1366, 1340, 1280, 1210, 1165, 029, 811 cm⁻¹. MS (FAB): m/z (%) = 489 (M⁺ + H, 53), 434 (31), 349 (100). HRMS (FAB): calcd. for C₂₅H₃₃N₂O₈ 489.2238, found 489.2252.

N-tert-Butyloxycarbonyl [2-(thien-2′-yl)benzoxazol-5-yl]-l-alanine methyl ester (2a)

The crude imine 1d (0.70 g 1.74 × 10⁻³ mol) and lead tetraacetate (1.17 g, 2.64 × 10⁻³ mol) were stirred at room temperature in DMSO (5 mL) for 3 days. The mixture was poured over water and extracted with ethyl acetate (3 × 10 mL). After drying with anhydrous magnesium sulphate and evaporation of the solvent, the resulting brown oil was submitted to column chromatography with silica gel (eluent: dichloromethane). The product was isolated as a light yellow solid (0.39 g, 56%). M.p. 106.6-107.8 °C. ¹H NMR (CDCl₃): δ = 1.42 (s, 9H, C(CH₃)₃), 3.15-3.28 (m, 2H, β CH₂), 3.73 (s, 3H, CH₃), 4.61-4.67 (m, 1H, αH), 5.04 (d, J = 7.8 Hz, 1H, NH), 7.10 (dd, J = 1.8 and 8.4 Hz, 1H, H-6), 7.17-7.20 (m, 1H, H-4′), 7.45 (d, J = 8.4 Hz, 1H, H-7), 7.48 (d, J = 1.8 Hz, 1H, H-4), 7.56 (dd, J = 1.2 and 5.1 Hz, 1H, H-5′), 7.90 (dd, J = 1.2 and 3.9 Hz, 1H, H-3′) ppm. ¹³C NMR (CDCl₃) δ = 28.24 (C(CH₃)₃), 38.27 (β CH₂), 52.29 (CH₃), 54.57 (αC), 80.00 (C(CH₃)₃), 110.26 (C7), 120.31 (C4), 126.32 (C6), 128.24 (C4′), 129.50 (C2), 129.97 (C3′), 130.31 (C5′), 132.79 (C5), 142.28 (C3a), 149.59 (C7a), 155.04 (C=O Boc), 159.41 (C2′), 172.08 (C=O ester) ppm. IR (KBr): ν = 3358, 2982, 2928, 1734, 1693, 1573, 1526, 1475, 1438, 1369, 1289, 1244, 1172, 1052, 1021, 857 cm⁻¹. UV/Vis (ethanol, nm): λ_max (ε) = 315 (2.13 × 10⁴). MS (FAB): m/z (%) = 403 (M⁺ + H, 73), 348 (23), 347 (100), 243 (29), 215 (27), 185 (21). HRMS (FAB): calcd. for C₂₀H₂₃N₂O₅S 403.1328, found 403.1324. C₂₀H₂₂N₂O₅S: calcd. C 59.69, H 5.51, N 6.96, S 7.97; found C 59.44, H 5.63, N 6.69, S 7.66.

N-tert-Butyloxycarbonyl [2-(thien-2′-yl)benzoxazol-5-yl]-l-alanine (2b)

Compound 2a (0.062 g, 1.54 × 10⁻⁴ mol) was dissolved in 1,4-dioxane (1 mL), in an ice bath, and sodium hydroxide 1 m aqueous solution (0.23 mL, 2.3 × 10⁻⁴ mol, 1.5 eq) was added drop wise. The mixture was stirred at room temperature for 3h. The pH was adjusted to 2-3 by addition of KHSO₄ 1 m aqueous solution and extracted with ethyl acetate (3 × 10 mL). After drying with anhydrous magnesium sulphate and evaporation of the solvent, the residue was triturated with diethyl ether and a off-white solid was obtained (0.050 g, 84%). M.p. 153.5-154.0 °C. ¹H NMR (CDCl₃): δ = 1.46 (s, 9H, C(CH₃)₃), 3.30-3.36 (m, 2H, β CH₂), 4.68-4.73 (m, 1H, αH), 5.17 (d, J = 7.5 Hz, 1H, NH), 7.16-7.22 (m, 2H, H-6 + H-4′), 7.48 (d, J = 8.4 Hz, 1H, H-7), 7.57 (dd, J = 1.2 and 5.1 Hz, 1H, H-5′), 7.63 (d, J = 2.1 Hz, 1H, H-4), 7.93 (dd, J = 1.2 and 3.9 Hz, 1H, H-3′) ppm. ¹³C NMR (CDCl₃) δ = 28.31 (C(CH₃)₃), 37.63 (β CH₂), 54.46 (αC), 80.14 (C(CH₃)₃), 110.34 (C7), 120.03 (C4), 127.05 (C6), 128.43 (C4′), 128.67 (C2), 130.73 (C3′), 130.84 (C5′), 133.24 (C5), 141.26 (C3a), 149.35 (C7a), 155.31 (C=O Boc), 159.49 (C2′), 174.66 (COOH) ppm. IR (KBr): ν = 3357, 3113, 2998, 2929, 1752, 1691, 1569, 1527, 1479, 1418, 1371, 1327, 1281, 1252, 1213, 1168, 1051, 856 cm⁻¹. UV/Vis (ethanol, nm): λ_max (ε) = 316 (2.73 × 10⁴). MS (FAB): m/z (%) = 389 (M⁺ + H, 57), 333 (45), 185 (100). HRMS (FAB): calcd. for C₁₉H₂₁N₂O₅S 389.1171, found 389.1173.

[2-(Thien-2′-yl)benzoxazol-5-yl]-l-alanine methyl ester (2c)

Compound 2a (0.103 g, 2.56 × 10⁻⁴ mol) was stirred in a trifluoroacetic acid/dichloromethane solution (1:1, 1 mL) at room temperature for 2h. The solvent was evaporated, the solid residue dissolved in pH 8 aqueous buffer solution and extracted with ethyl acetate (3 × 10 mL). After drying with anhydrous magnesium sulphate and evaporation of the solvent, the product was isolated as a light yellow solid (0.054 g, 74%). M.p. 80.7-81.4 °C. ¹H NMR (CDCl₃): δ = 1.46 (s, 9H, C(CH₃)₃), 2.98-3.06 (m, 1H, β CH₂), 3.17-3.24 (m, 1H, β CH₂), 3.74 (s, 3H, CH₃), 3.77-3.82 (m, 1H, αH), 7.17-7.22 (m, 2H, H-6 + H-4′), 7.49 (d, J = 8.4 Hz, 1H, H-7), 7.56-7.59 (m, 2H, H-5′ + H-4), 7.91 (dd, J = 1.5 and 3.9 Hz, 1H, H-3′) ppm. ¹³C NMR (CDCl₃) δ = 40.77 (β CH₂), 52.07 (CH₃), 55.92 (αC), 110.25 (C7), 120.18 (C4), 126.33 (C6), 128.24 (C4′), 129.51 (C2), 129.97 (C3′), 130.31 (C5′), 133.85 (C5), 142.33 (C3a), 149.49 (C7a), 159.41 (C2′), 175.12 (C=O ester) ppm. IR (KBr): ν = 3310, 2925, 1741, 1590, 1478, 1415, 1325, 1260, 1063, 855 cm⁻¹. UV/Vis (ethanol, nm): λ_max (ε) = 315 (3.07 × 10⁴). MS (FAB): m/z (%) = 303 (M⁺ + H, 100), 215 (20). HRMS (FAB): calcd. for C₁₅H₁₅N₂O₃S 303.0803, found 303.0803.

[2-(Thien-2′-yl)benzoxazol-5-yl]-l-alanine (2d)

Compound 2b (0.100 g, 2.58 × 10⁻⁴ mol) was stirred in a trifluoroacetic acid/dichloromethane solution (1:1, 1 mL) at room temperature for 2h. The solvent was evaporated, the solid residue dissolved in pH 8 aqueous buffer solution and extracted with ethyl acetate (3 × 10 mL). After drying with anhydrous magnesium sulphate and evaporation of the solvent, the product was isolated as a light brown solid (0.040 g, 54%). M.p. 238.9-239.9 °C. ¹H NMR (DMSO-d₆): δ = 2.94-3.02 (m, 1H, β CH₂), 3.23-3.30 (m, 1H, β CH₂), 3.46-3.51 (m, 1H, αH), 7.28-7.31 (m, 2H, H-6 + H-4′), 7.62-7.67 (m, 2H, H-7 + H-5′), 7.93-7.96 (m, 2H, H-3′ + H-4) ppm. ¹³C NMR (DMSO-d₆) δ = 30.40 (β CH₂), 53.01 (αC), 109.92 (C5′), 119.40 (C7), 126.57 (C6), 128.50 (C4′), 129.72 (C2), 130.03 (C4), 131.38 (C3′), 132.34 (C5), 141.36 (C3a), 151.70 (C7a), 159.50 (C2′), 180.83 (C=O) ppm. IR (KBr): ν = 3413, 3072, 2940, 1620, 1572, 1548, 1479, 1407, 1346, 1312, 1287, 1269, 1217, 1195, 1160, 1049, 1024, 1005, 872, 853 cm⁻¹. UV/Vis (ethanol, nm): λ_max (ε) = 315 (2.17 × 10⁴). MS (FAB): m/z (%) = 289 (M⁺ + H, 30), 278 (22), 277 (21), 186 (71). HRMS (FAB): calcd. for C₁₄H₁₃N₂O₃S 289.0646, found 289.0647.

N-tert-Butyloxycarbonyl-3-[2-(2′,4′,5′-trimethoxyphenyl)benzoxazol-5-yl]-l-alanine methyl ester (2e)

The crude imine 1e (0.470 g, 9.63 × 10⁻⁴ mol) and lead tetraacetate (0.640 g, 1.44 × 10⁻³ mol, 1.5 eq.) were stirred at room temperature in DMSO (5 mL) for 3 days. The mixture was poured over water and extracted with ethyl acetate (3 × 10 mL). After drying with anhydrous magnesium sulphate and evaporation of the solvent, the resulting brown oil was submitted to column chromatography with silica gel (eluent: chloroform). The product was isolated as light brown oil (0.304 g, 65%) [lit. [27] M.p. 137-140 °C]. ¹H NMR (CDCl₃): δ = 1.43 (s, 9H, C(CH₃)₃), 3.22 (d, J = 4.2 Hz, 2H, β CH₂), 3.74 (s, 3H, CH₃), 3.97 (s, 3H, OCH₃), 4.00 (s, 3H, OCH₃), 4.02 (s, 3H, OCH₃), 4.62-4.66 (m, 1H, α-H), 5.01 (d, J = 8.1 Hz, 1H, NH), 6.66 (s, 1H, H-3′), 7.08 (dd, J = 1.8 and 8.4 Hz, 1H, H-6), 7.45 (d, J = 8.4 Hz, 1H, H-7), 7.50 (d, J = 1.8 Hz, 1H, H-4) 7.65 (s, 1H, H-6′) ppm. UV/Vis (ethanol, nm): λ_max (ε) = 334 (1.41 × 10⁴). MS (FAB): m/z (%) = 489 (M⁺ + H, 100), 488 (21), 433 (14), 300 (38), 154 (14). HRMS (FAB): calcd. for C₂₅H₃₃N₂O₈ 489.2237, found 489.2238.

N-tert-Butyloxycarbonyl-3-[2-(2′,4′,5′-trimethoxyphenyl)benzoxazol-5-yl]-l-alanine (2f)

Compound 2e (0.060 g, 1.23 × 10⁻⁴ mol) was dissolved in 1,4-dioxane (1 mL), in an ice bath, and sodium hydroxide 1 m aqueous solution (0.19 mL, 1.9 × 10⁻⁴ mol, 1.5 eq) was added drop wise. The mixture was stirred at room temperature for 3h. The pH was adjusted to 2-3 by addition of KHSO₄ 1 m aqueous solution and extracted with ethyl acetate (3 × 10 mL). After drying with anhydrous magnesium sulphate and evaporation of the solvent, the residue was triturated with diethyl ether and a brown solid was obtained (0.048 g, 82%). M.p. 135.5-137.0 °C. ¹H NMR (CDCl₃): δ = 1.45 (s, 9H, C(CH₃)₃), 3.30-3.34 (m, 2H, β CH₂), 3.94 (s, 3H, OCH₃), 3.97 (s, 3H, OCH₃), 3.98 (s, 3H, OCH₃), 4.64-4.68 (m, 1H, α-H), 5.12 (d, J = 8.4 Hz, 1H, NH), 6.62 (s, 1H, H-3′), 7.18 (dd, J = 1.5 and 8.4 Hz, 1H, H-6), 7.50 (d, J = 8.4 Hz, 1H, H-7), 7.60 (s, 1H, H-6′), 7.66 (s, 1H, H-4) ppm. ¹³C NMR (CDCl₃) δ = 28.36 (C(CH₃)₃), 37.67 (β CH₂), 54.46 (αC), 80.20 (C(CH₃)₃), 100.35 (C3′), 106.43 (C1′), 110.35 (C7), 114.95 (C6′), 120.15 (C4), 126.96 (C6), 128.90 (C2), 133.34 (C5), 135.93 (C5′), 141.18 (C3a), 145.02 (C4′), 149.54 (C7a), 152.26 (C2′), 155.40 (C=O Boc), 174.48 (COOH) ppm. IR (KBr): ν = 3340, 3108, 2988, 2930, 1761, 1684, 1569, 1527, 1420, 1377, 1343, 1265, 1250, 1178, 1049, 850 cm⁻¹. UV/Vis (ethanol, nm): λ_max (ε) = 334 (1.75 × 10⁴). MS (FAB): m/z (%) = 473 (M⁺ + H, 34), 472 (M⁺, 16), 417 (16), 307 (17), 195 (21), 155 (27), 154 (100). HRMS (FAB): calcd. for C₂₄H₂₉N₂O₈ 473.1924, found 473.1935.

3-[2-(2′,4′,5′-Trimethoxyphenyl)benzoxazol-5-yl]-l-alanine (2g)

Compound 2f (0.035 g, 7.41 × 10⁻⁵ mol) was stirred in a trifluoroacetic acid/dichloromethane solution (1:1, 1 mL) at room temperature for 2h. The solvent was evaporated, the solid residue dissolved in pH 8 aqueous buffer solution and extracted with ethyl acetate (3 × 10 mL). After drying with anhydrous magnesium sulphate and evaporation of the solvent, the product was isolated as an off-white solid (0.012 g, 44%). M.p. 251.6-253.3 °C. ¹H NMR (DMSO-d₆): δ = 2.94-3.02 (m, 1H, β CH₂), 3.23-3.30 (m, 1H, β CH₂), 3.46-3.51 (m, 1H, αH), 4.05 (s, 3H, OCH₃), 4.11 (s, 3H, OCH₃), 4.14 (s, 3H, OCH₃), 7.28-7.31 (m, 2H, H-6 + H-3′), 7.60 (d, J = 8.4 Hz, 1H, H-7), 7.65 (s, 1H, H-6′), 7.95 (br s, 1H, H-4) ppm. ¹³C NMR (DMSO-d₆) δ = 30.28 (β CH₂), 53.11 (αC), 101.30 (C3′), 105.99 (C1′), 115.34 (C6′), 119.43 (C7), 126.35 (C6), 129.82 (C2), 130.06 (C4), 132.55 (C5), 135.41 (C5′), 141.34 (C3a), 145.13 (C4′), 151.57 (C7a), 152.25 (C2′), 180.88 (C=O) ppm. IR (KBr): ν = 3400, 3052, 2954, 1618, 1570, 1490, 1350, 1325, 1307, 1250, 1198, 1163, 1057, 1030, 1001, 853 cm⁻¹. UV/Vis (ethanol, nm): λ_max (ε) = 336 (2.29 × 10⁴). MS (FAB): m/z (%) = 373 (M⁺ + H, 30), 372 (M⁺, 43), 278 (100), 186 (51). HRMS (FAB): calcd. for C₁₉H₂₁N₂O₆ 373.1401, found 372.1395.

3.2. Spectrofluorimetric titrations

The linearity of the fluorescence emission vs. concentration was checked in the concentration range used (10⁻⁴-10⁻⁶ M). A correction for the absorbed light was performed when necessary. All spectrofluorimetric titrations were performed as follows: stock solutions of compounds 2a-g (ca. 10⁻³ M) were prepared in the corresponding solvents (absolute ethanol, acetonitrile, dioxane or cyclohexane, all of UVA-sol or HPLC grade) and used in the preparation of titration solutions by appropriate dilution. Titration of compounds 2a-g was carried out by addition of microliter amounts of standard solutions of the ions in acetonitrile or absolute ethanol. Protonation and deprotonation studies were performed by addition of HBF₄, methanesulphonic acid, triethylamine and tetrabutylammonium hydroxide. Cu(CF₃SO₃)₂, Ni(BF₄)₂, Zn(CF₃SO₃)₂, Hg(CF₃SO₃)₂, Ca(ClO₄)₂ and NaNO₃ were purchased from Alfa Aesar and used as received. Luminescence quantum yields were measured using as standard a solution of 1-naphthylamine in cyclohexane (0.1 M) (Φ_F = 0.46) [34], and were corrected for different refractive indexes of solvents. In all fluorimetric measurements the optical density did not exceed 0.1 [35].