Tautomeric Equilibria in Phenolic A-Ring Derivatives of Prodigiosin

The prodigiosin natural products contain a common 4-methoxy-pyrromethene chromophore that is attached to a pyrrole A-ring that has its lone-pair nitro...
0 downloads 0 Views 318KB Size
Download PDF
J. Phys. Chem. B 2007, 111, 11803-11811

11803

Tautomeric Equilibria in Phenolic A-Ring Derivatives of Prodigiosin Natural Products Jamie Q.-H. La, Alex A. Michaelides, and Richard A. Manderville* Departments of Chemistry and Toxicology, UniVersity of Guelph, Guelph, Ontario N1G 2W1, Canada ReceiVed: June 14, 2007; In Final Form: July 25, 2007

The prodigiosin natural products contain a common 4-methoxy-pyrromethene chromophore that is attached to a pyrrole A-ring that has its lone-pair nitrogen electrons in conjugation with the pyrromethene entity. This feature is known to play a key role in the biological activities (anticancer, antimicrobial, and immunosuppressive) of the prodigiosins. In an attempt to alter or improve upon the therapeutic potential of the prodigiosins, we have synthesized two new isomeric analogues that contain phenolic A-ring systems (a para (p)-phenol; an ortho (o)-phenol with respect to the pyrromethene) with lone-pair oxygen electrons in conjugation with the pyrromethene chromophore of the natural product. Herein, we report on the optical properties of the phenolic prodigiosin analogues that have been measured using absorbance and steady-state emission spectroscopy. For both analogues absorption measurements in aprotic solvents show that the neutral (L) ligands exist as the enol tautomers with λmax ∼ 460 nm, as noted for the parent prodigiosin natural product. However, in polar protic solvents the phenolic derivatives undergo ground-state prototropic tautomerization to generate keto tautomers with λmax ∼ 530 nm. This unique feature for a prodigiosin analogue involves proton transfer from the phenolic OH to the pyrromethene N1 proton acceptor atom. Tautomeric equilibrium constants (KT) of 1.4 in 1:4 MeCN/H2O (v/v) have been determined from examination of the absorption spectra. Titration of the o-phenolic derivative with Zn(II) in methanol yielded a 40-fold increase in fluorescence intensity (λmax 542 nm) and generated a new 1:1 complex with Zn(II) with a log K of 5.29, suggesting the potential utility of this analogue to act as a fluorescence probe in a biological matrix to monitor Zn(II) concentrations. Our results demonstrate that phenolic A-ring derivatives of prodigiosins possess some unique properties that may act to enhance the biological properties of the prodigiosin natural products.

1. Introduction Prodigiosin (1, Figure 1) belongs to a group of intensely colored pink/red alkaloids isolated from certain Serratia, Streptomyces, and Bacillus bacterial strains, that possess a common pyrrolylpyrromethene chromophore.1-3 Although they exhibit a wide range of biological activities,1-3 the natural derivatives isolated in the 1960s were not utilized as therapeutics due to their high systemic toxicity.4 However, with the development of a simple Pd-catalyzed cross-coupling procedure for prodigiosin synthesis,5 and new strategies for the preparation of the 2,2′-bipyrrole precursor,6-8 there has been renewed interest in the synthesis and development of prodigiosin analogues with enhanced biological activities and improved toxicological profiles. Some of the strategies employed include the generation of new derivatives with altered A-rings,9-11 C-rings,12 and exocyclic B-ring substituents,8,9 and these have improved the therapeutic potential of certain prodigiosin analogues. The immunosuppressive, antimicrobial, antimalarial, and anticancer properties of the prodigiosins have also stimulated research into their mechanism of action.1-3 Prodigiosins possess H+/Cl- symport activity13 that may trigger apoptosis in a variety of cancer cell lines14-16 by lowering intracellular pH. The ability of the protonated prodigiosin species to bind Cl- effectively17 is thought to play a key role in this activity that may also, in part, involve the transport of Cl- across cell membranes by an antiport mechanism.18 The redox properties of the prodigiosins could also play a role in their ability to induce apoptosis. * To whom correspondence should be addressed. Tel.: (519) 824-4120, ext. 53963. Fax: (519) 766-1499. E-mail: [email protected].

Figure 1. Prodigiosin (1) and A-ring phenolic derivatives 2 and 3.

Prodigiosins bind DNA effectively19,20 and generate a redoxactive complex with copper ions21 that facilitates oxidative DNA cleavage via π-radical cation formation at the electron-rich pyrroylpyrromethene moiety.22-24 The photooxidation of 1 and synthetic analogues also facilitates cell death25,26 and furnishes π-radical cation formation that triggers attachment to thiol nucleophiles.25 The ability of prodigiosins to act as tyrosine phosphatase inhibitors is also expected to derive from their redox properties in which prodigiosin-derived electrophiles target active-site cysteine residues.27 Structure-activity relationships for the prodigiosins have established that an electron-donating nitrogen-containing heterocyclic A-ring, with the lone-pair nitrogen electrons in conjugation with the tricyclic frame,9 and a C-4 alkoxy on the B-ring,4 are important for biological activity. The requirement for the C-4 alkoxy group prompted Jolicoeur and Lubell to prepare an amino prodigiosin analogue, in which the exocyclic alkoxy group of the natural prodigiosins is replaced by an electron-donating C-4 morpholino nitrogen (N) substituent, which may enhance the therapeutic potential of the prodigiosins.8

10.1021/jp074620z CCC: $37.00 © 2007 American Chemical Society Published on Web 09/15/2007

11804 J. Phys. Chem. B, Vol. 111, No. 40, 2007 SCHEME 1

Likewise, we speculated that prodigiosin analogues 2 and 3 shown in Figure 1 containing phenolic A-ring systems would be of interest since they have lone-pair oxygen (O) electrons in conjugation with the tricyclic frame and the phenolic OH group of the ortho (o)-analogue 3 is in position to facilitate metal coordination by the pyrromethene moiety. The derivative 3 is also related to 2-(2′-hydroxyphenyl)benzimidazole (HBI) and the corresponding benzoxazole (HBO), and benzothiazole (HBT) series of compounds (HBX, X ) NH, O, S) that exist in a conformational equilibrium between the syn- (“closed”), anti-, and “open” enols in the ground state (Scheme 1).28-30 The planar “closed” syn-enol form that possesses an intramolecular hydrogen (H)-bond between the phenolic OH and the imine N undergoes an excited-state intramolecular proton transfer (ESIPT) process to generate the keto form, which generates a large Stokes-shifted fluorescence. This tautomerization process makes these molecules attractive for use as fluorescence sensors, laser dyes, molecular switches,30,31 models of DNA base pair tautomerization,32,33 and nonlinear optical materials.34 These reports suggested that phenolic A-ring derivatives of prodigiosin would exhibit some of these interesting properties and this could expand upon the medicinal applications of prodigiosin-based natural products. In this manuscript we focus on the optical properties of 2 and 3 and demonstrate their propensity to undergo ground-state tautomerization to generate the keto tautomer in protic solvents. The potential utility of 3 as a fluorescent probe for the detection of Zn(II) in a biological environment is also demonstrated. 2. Experimental Section 2.1. Materials and Methods. Unless otherwise noted, commercial compounds were used as received. DMF was obtained from Caledon Chemicals and distilled prior to use. Boronic acids were purchased from Frontier Scientific. Pd(PPh3)4 was purchased from Strem Chemical and stored under argon in a freezer at -20 °C. Pd3(OAc)6 was purchased from Sigma-Aldrich and also stored in the freezer. Tris-(m-sulfonatophenyl)phosphine trisodium (TPPTS) was purchased from Fluka. p-Methoxyanisaldehyde (PMA) was obtained from Aldrich and distilled under reduced pressure before use. 3-(NMorpholino)propanesulfonic acid (MOPS) and 3-{[tris(hydroxymethyl)methyl]amino}-propanesulfonic acid (TAPS) were purchased from Sigma. Quinine bisulfate·H2SO4 was obtained from Aldrich. Water used for buffers and spectroscopic solutions was obtained from a MilliQ filtration system (18.2 MΩ). Solvents not of spectroscopic grade were analyzed and corrected for spurious absorbance and emission intensities. Chromatography was performed on virgin silica obtained from Silicycle (230-400 mesh). Analytical thin-layer chromatography (TLC)

La et al. SCHEME 2

was carried out on glass-backed extra hard layer plates (60, F-254 indicator) also obtained from Silicycle. Compounds were visualized by UV light (λ ) 254 nm), oxidation with iodine, or using a PMA-based staining solution and heat. NMR spectra were recorded on a Bruker Avance-300 DPX spectrometer (1H, 300.1 MHz; 13C, 75.5 MHz) or a Bruker Avance-400 DPX spectrometer (1H, 400.1 MHz) in DMSO-d6, or CDCl3. 1H NMR spectra were referenced to the residual proton solvent signal, and 13C NMR spectra were referenced to the 13C NMR resonance of the deuterated solvent (DMSO-d6, δ ) 39.5 ppm and CDCl3, δ ) 77.0 ppm). All 13C NMR spectra were acquired using the J-modulated (JMOD) pulse sequence.35 Chemical shifts are given in ppm relative to TMS, and coupling constants (J) are reported in hertz (Hz). High-resolution mass spectra (HRMS) were conducted at the McMaster Regional Centre for Mass Spectrometry and were obtained on a Micromass/Waters Global Ultima quadrupole time-of-flight mass spectrometer using electrospray positive ionization at a resolving power of ∼8000. UV-vis spectra were recorded on a Cary 300-Bio UVvis spectrophotometer equipped with a Peltier block-heating unit and temperature controller. Fluorescence emission and excitation spectra were performed on a Cary Eclipse fluorescence spectrophotometer. Standard 10 mm light path quartz glass cells from Hellma GmbH & Co were used. All UV-vis spectra were recorded at 25 °C with baseline/background correction. Quantum yield values were determined in acetonitrile (MeCN) using the comparative method.36-38 The pH measurements were carried out using an Accumet portable AP63 pH meter equipped with an Accumet combination microelectrode with a calomel reference. Calibration was done using commercial buffers (Fisher Scientific, pH 4.00, 7.00, and 10.00, all (0.01). 2.2. Synthesis. The strategy used for the preparation of prodigiosin analogues 2 and 3 is outlined in Scheme 2. These reactions were conducted according to the literature,5,10 and the syntheses of 4 and 5 derived from 4-methoxy-3-pyrrolin-2-one (Aldrich) and 2-formyl-5-ethylpyrrole have been previously detailed.10 2.2.a. 2-(4-Hydroxy-phenyl)-4-methoxy-5-[(5-ethyl-2H-pyrrol2-ylidene)-methyl]-1H-pyrrole (2). The triflate 5 (70 mg, 0.20 mmol), Pd3(OAc)6 (4.0 mg, 0.006 mmol), TPPTS (7.5 mg, 0.013 mmol), Na2CO3 (21 mg, 0.39 mmol), and 4-hydroxyphenyl boronic acid (33 mg, 0.24 mmol) were dissolved in 4 mL of a 2:1 H2O/MeCN solvent mixture under an argon atmosphere. The reaction mixture was refluxed for 4 h and then diluted with 100 mL of H2O and extracted into CHCl3. The organic phase was separated, extracted with H2O and brine, and dried over Na2SO4. The solvent was then removed in vacuo, and the

Tautomeric Equilibria in Prodigiosin Derivatives resulting residue was purified over aluminum oxide (4:1 Hex/ EtOAc) to afford 2 as a dark red solid (10 mg, 17%). UV-vis (1:1 MeCN/H2O) λmax 455 nm, 455 14 143 cm-1 M-1. 1H NMR (CDCl3) δ (ppm): 10.69 (s, 1H), 10.25 (s, 1H), 7.82 (d, J ) 8.4, 1H), 6.96 (s, 1H), 6.85 (s, J ) 8.4, 1H), 6.73 (d, J ) 3.6, 1H), 6.12 (d, J ) 3.6, 1H), 6.05 (s, 1H), 3.93 (s, 3H), 2.78 (q, J ) 7.6, 2H), 1.31 (t; J ) 7.5, 3H). 13C NMR (CDCl3) δ: 164.8, 164.7, 157.8, 138.9, 138.4, 129.4, 128.7, 121.4, 116.5, 115.3, 115.3, 109.2, 94.4, 57.4, 20.8, 12.1. HRMS calcd for C18H18N2O2 [M + H]+ 295.1447, found 295.1440. 2.2.b. 2-(2-Hydroxy-phenyl)-4-methoxy-5-[(5-ethyl-2H-pyrrol2-ylidene)-methyl]-1H-pyrrole (3). A solution of the triflate 5 (176 mg, 0.435 mmol) in 1,4-dioxane (anhydrous) (7 mL) was treated in sequence with 2-hydroxyphenyl boronic acid (0.8 mmol) and potassium carbonate (1.6 mmol). 1,4-Dioxane was purged with argon for 10 min, treated with tetrakis(triphenylphosphine)palladium(0) (0.02 mmol), and heated at 90 °C with stirring for 24 h. The solvent was removed in vacuo, and the resulting residue was dissolved in CH2Cl2, washed with H2O and brine, and dried over Na2SO4. The solvent was then removed in vacuo, and the residue was dissolved in Et2O and then acidified with a solution of 1 M HCl in Et2O to effect precipitation of 3 in its protonated form. The salt was filtered, rinsed with Et2O, and then treated with 4 N NaOH and extracted with Et2O. The solvent was then removed in vacuo to produce 3 as a dark red solid (13.5 mg, 23%). UV-vis (1:1 H2O/MeCN) λmax 469 nm, 469 30 226 cm-1 M-1. 1H NMR (DMSO) δ (ppm): 14.60 (br s, 1H), 11.5 (br s, 1H), 7.45 (m, 1H), 7.26 (m, 1H), 7.06-6.91 (m, 2H), 6.63 (d, J ) 2.7, 1H), 6.10-6.06 (m, 2H), 3.92 (s, 3H), 2 0.77 (q, J ) 7.2, 2H), 1.34 (t, J ) 7.5, 3H). 13C NMR (CDCl3) δ: 167.3, 166.6, 158.9, 143.8, 136.8, 130.5, 127.3, 127.2, 120.5, 117.9, 116.8, 116.4, 116.1, 108.3, 93.2, 57.4, 20.5, 11.8. HMRS calcd for C18H18N2O2 [M + H]+ 295.1447, found 295.1442. 2.3. Determination of pKa Values. Apparent acidity constants (pKa) were determined spectrophotometrically by monitoring absorbance changes in the UV-vis spectra after additions of dilute HCl or NaOH solutions under constant temperature conditions (25 °C). Stock solutions of 2 and 3 (5.6 mM) were prepared in MeCN, and 5 µL of this solution was added to a quartz cuvette containing 2 mL total volume of 1:1 MeCN/ H2O (v/v), and the ionic strength (µ) was kept constant at a concentration of 0.1 M using NaCl. After temperature equilibration, solutions were blank corrected and pH measurements were taken directly in the cuvettes. Following each addition of acid or base, the pH of the solution was measured, and UV-vis spectra were recorded by the overlay method in the wavelength range of 375-650 nm. The pKa values were determined from a plot of the log(ionization ratio) versus pH, as described.10,39 2.4. Zinc Binding. Zinc complexation studies for 2 and 3 were carried out in methanol (MeOH) by absorption spectroscopy at 25 °C. Stock solutions of 2 and 3 were diluted in 2.5 mL of MeOH to a concentration of 14 µM in a quartz cuvette. An initial spectrum was taken ([Zn] ) 0) after which various Zn equivalents from a stock Zn(OAc)2 (6 mM) solution in MeOH were added to both the analytical and reference cuvettes. After each addition of Zn(OAc)2 the solution was stirred and allowed to equilibrate for a duration of 5 min after which UVvis spectra were recorded in the wavelength range of 375-600 nm. An apparent association (Kapp) constant for Zn complexation by 3 was determined from changes inherent to the UV-vis spectra and double-reciprocal analysis assuming a 1:1 binding isotherm.40,41

J. Phys. Chem. B, Vol. 111, No. 40, 2007 11805

Figure 2. UV-vis absorbance spectra for 2 as a function of pH in 1:1 MeCN/H2O (v/v) at 25 °C (0.1 M NaCl). (a) pH 4.3-8.5; arrow shows increased absorbance with increased acidity. (b) pH 8.5-11.3; arrow shows increased absorbance with increased basicity.

3. Results and Discussion 3.1. Protonation Equilibria. 3.1.a. Absorption Spectra. The prodigiosins are weakly basic (apparent pKa ∼ 7-8) due to the pyrromethene moiety; the protonated species appears red in color absorbing at ∼530 nm, whereas the free base appears yellow absorbing at ∼460 nm with reduced intensity.10,42,43 Previously, we measured the apparent pKa values of a series of prodigiosin phenyl A-ring analogues in 1:1 MeCN/H2O (v/v) using the spectrophotometric procedure; the MeCN cosolvent was needed due to the poor water solubility of the neutral prodigiosin.10 Under these conditions the apparent pKa of prodigiosin 1 was 7.98, whereas the corresponding pKa of a phenyl A-ring analogue was 7.05; attachment of a para (p)-methoxy substituent to the phenyl A-ring raised the pKa to 7.41.10 For the phenolic A-ring derivatives 2 and 3 UV-vis pH titrations showed two distinct regions involving reversible deprotonation equilibria, each containing a characteristic set of isosbestic points. Figure 2 shows pH titrations for the p-phenolic derivative 2. In the neutral to acidic pH region (Figure 2a) a pKa value of 7.42 was determined. As expected for N1 deprotonation,10 the protonated (LH+) form of 2 absorbed strongly at 506 nm, whereas the neutral ligand (L) absorbed at 455 nm with reduced intensity. The derived pKa value at 7.42 suggested that the p-phenolic A-ring in 2 affects that acidity of the pyrromethene moiety in a fashion analogous to a p-methoxyphenyl A-ring,10 as expected. Figure 2b shows a pH titration for 2 in the neutral to basic pH region, in which the second pKa value occurred at 9.96, which is in the range for deprotonation of a phenolic OH atom.44 The phenolate (L-) of 2 showed an absorption band at 473 nm with increased intensity compared to that of the neutral (L) ligand. Figure 3 shows the corresponding pH titrations for the o-phenolic derivative 3. In the neutral to acidic pH region (Figure 3a) the pKa ascribed to N1 deprotonation occurred at 6.05. Although this value is significantly lower than the

11806 J. Phys. Chem. B, Vol. 111, No. 40, 2007

Figure 3. UV-vis absorbance spectra for 3 as a function of pH in 1:1 MeCN/H2O (v/v) at 25 °C (0.1 M NaCl). (a) pH 4.2-7.5; arrow shows increased absorbance with increased acidity. (b) pH 7.5-13.6; arrow shows increased absorbance with increased basicity.

corresponding value obtained for 2, the UV-vis spectral changes (Figure 3a) were consistent with N1 deprotonation. The protonated LH+ species absorbed strongly at 504 nm, whereas the neutral ligand L absorbed at 469 nm with reduced intensity.10 Figure 3b shows the neutral to basic pH region, in which a second pKa at 12.55 was determined. This reversible process was attributed to deprotonation of the phenolic OH atom to generate the phenolate L-, which showed absorption at 531 nm with decreased intensity compared to that of the neutral ligand L. A compilation of the results including photophysical properties of the individual species is given in Table 1. The pKa value determined for deprotonation of N1 (6.05) in 3 was significantly lower than expected and is 1 pKa unit lower than the pKa recorded for a phenyl A-ring analogue (7.0510). For 2-(2′-hydroxy-phenyl)-benzimidazole the ortho OH group increases the pKa 0.2 pKa units compared to benzimidazole (pKa ) 5.68) that can be ascribed to stabilization of the protonated species by intramolecular H-bonding with the neighboring OH group.30 A similar trend is observed for 4,5-dimethyl-2-(2′hydroxyphenyl)imidazole, which has a pKa 7.43 that is 0.43 pKa units above the corresponding pKa of imidazole.45 Thus, the observed N1 pKa measured for the o-phenolic analogue 3 is not consistent with published results for other systems. However, as outlined by Rizzo et al.43 the pKa of prodigiosins is governed by the presence of two geometrical isomers (Rand β-forms) that have very different apparent pKa values [pKaR ∼ 8.23, pKaβ ∼ 5.4 in 1:1 MeCN/H2O (v/v)]. For prodigiosin 1 and the p-phenolic analogue 3 the recorded pKa values refer to the equilibrium mixture of the geometric isomers that have similar stabilities.43 In Scheme 3 is shown the protonation and rotomer interconversion equilibria for the o-phenolic derivative 3. Here it is readily apparent that the β-form benefits from intramolecular H-bonding in both the protonated (LH+) and neutral (L) species, and this stabilizing factor would be expected to shift the equilibrium in favor of the β-form over the R-form. Given that the pKa of the β-form of prodigiosin is ∼5.4, then the measured pKa of 6.05 for 3 is consistent with previously published results30,45 and the increase in pKa compared to that

La et al. of the β-form of 1 is explained by stabilization of the protonated species by the OH group. The phenolic OH pKa (12.55) of 3 is consistent with the presence of an intramolecular H-bond between the phenolic OH and N1 of 3 (β-form, Scheme 3). For phenolic groups containing ortho substituents, stabilization of the neutral parent species compared to the anion caused by an intramolecular H-bond is known to greatly influence the pKa.46 Theoretical calculations indicate that o-methoxyphenol will possess a pKa value +2.4 pKa units above phenol due to the intramolecular H-bond between the phenolic OH and the methoxy O atom.46 For 3 the ∆pKa compared to 2 is 2.59. In DMSO-d6 the p-phenolic derivative 2 shows the presence of two exchangeable proton resonances at δ 10.69 and 10.25 ppm, whereas for the o-phenolic derivative 3 the exchangeable protons were observed as broad resonances at δ 14.60 and 11.5 ppm. The downfield proton in 3 at δ 14.60 ppm was ascribed to the phenolic OH proton, and its large downfield shift is indicative of a strong intramolecular H-bond and is expected to increase the stability of the neutral species by ∼4.3 kcal/mol.47 Thus, stabilization of the neutral ligand L of 3 due to the presence of the intramolecular H-bond would raise the phenolic pKa, as observed. An additional factor for phenolic deprotonation of 3 would be electronic repulsion between the developing phenolate negative charge and the N1 lone-pair electrons, which would require a conformational change upon removal of the phenol OH proton (Scheme 3). 3.1.b. Fluorescence Spectra. The emission spectra of the three ligands (LH+, L, L-) for 2 and 3 were recorded in 1:1 MeCN/ H2O (v/v) (0.1 M NaCl) (Figure 4). For 2, distinct emission spectra were observed for LH+ (534 nm), L (514 nm), and L(592 nm) (Figure 4a). The Stokes shifts for the individual species were 1036, 2523, and 4250 cm-1, respectively. The Stokes shift noted for the phenolate of 2 is in the range observed for other conjugated phenolates48 and suggests delocalization of the phenolate negative charge into the pyrromethene chromophore in the excited state. For 3 the emission spectra for LH+ (555 nm) and L (555 nm) were practically identical with the exception that LH+ gave a broader emission spectrum (Figure 4b). For the neutral ligand L the Stokes shift was 3304 cm-1 that was 781 cm-1 greater that the Stokes shift noted for L of the para isomer 2. This observation combined with the similarity of the fluorescence of neutral 3 (L) to the fluorescence of its protonated (LH+) form suggested that L underwent an ESIPT process and that migration of the phenolic OH atom to N1 occurred in the excited state. The phenolate L- of 3 showed a distinct emission spectrum with maxima (λem) at 582 nm. It is important to point out that 2 and 3 were weakly fluorescent with quantum yields in MeCN ranging from 0.02 to 0.06 (Table 1) indicating that these free ligands are not suitable for use as fluorescent probes. 3.2. Tautomeric Equilibria. 3.2.a. Absorption Spectra. The absorption spectra of L recorded for 2 and 3 in 1:1 MeCN/H2O (v/v) with λmax at ∼460 nm (Table 1) showed an additional absorption with lower intensity at ∼530 nm. Under analogous conditions prodigiosin 1 shows a single absorption band at 468 nm, as do 2 and 3 in MeCN, DMSO, CHCl3, and other aprotic solvents (data not shown). Figure 5 shows UV-vis absorbance spectra of 2 (Figure 5a) and 3 (Figure 5b) in different volume ratios of MeCN/H2O at 25 °C that contained 10 mM aqueous buffer (TAPS pH 8.5 for 2 and MOPS pH 7.0 for 3) to ensure formation of the neutral species L in the solvent mixtures. For both 2 and 3 increasing the water content leads to an increase in the absorption at ∼530 nm (λmax ) 538 nm in 80% aqueous buffer for both 2 and 3) with isosbestic points at 475 nm for 2 and 491 nm for 3. These results were classical evidence for

Tautomeric Equilibria in Prodigiosin Derivatives

J. Phys. Chem. B, Vol. 111, No. 40, 2007 11807

TABLE 1: Protonation Constants and Photophysical Data of Prodigiosin (1), 2, and 3a species N1 pKa PhOH pKa LH+

L

L-

d

data +

+

[LH ]/[L][H ] [L]/[L-][H+] absorption λmax (nm), log  emission λmax (nm) quantum yieldd Stokes shift (cm-1) absorption λmax (nm), log  emission λmax (nm) quantum yield Stokes shift (cm-1) absorption λmax (nm), log  emission λmax (nm) Stokes shift (cm-1)

prodigiosin (1)b

p-PhOH-, 2c

o-PhOH-, 3c

7.98 ( 0.02 n.a. 533, 5.05 n.d. n.d. n.d. 468, 4.60

7.42 ( 0.03 9.96 ( 0.05 506, 4.60 534 0.02 1036 455, 4.15 514 0.03 2523 473, 4.32 592 4250

6.05 ( 0.03 12.55 ( 0.09 504, 4.89 555 0.06 1823 469, 4.48 555 0.04 3304 531, 4.40 582 1650

n.a.

a 1:1 H2O/MeCN, µ 0.1 M NaCl, 25 °C. b Taken from ref 10. c Errors in pKa values were derived from UV-vis titrations done in triplicate. Quantum yields were determined in MeCN.

SCHEME 3

Figure 4. Normalized emission spectra for (a) 2 and (b) 3 in 1:1 MeCN/H2O (v/v) at 25 °C (0.1 M NaCl). The spectra of the protonated (LH+, solid lines), neutral (L, dotted lines), and deprotonated (L-, dashed lines) ligand were obtained during pH titrations.

prototropic tautomerization45,49 of the enol form of 2 and 3 with λmax at ∼460 nm to generate the keto form with λmax ∼ 530 nm, which, until now, has never been observed for the free base of a prodigiosin analogue. For related enolimine systems it is well-established that the enolimine (enol form) is typically favored over the keto form (enaminone tautomer) due to the aromatic character of the phenyl ring.30-34,45 In this regard, we recently examined the optical properties of the nucleoside adduct, 8-(2′′-hydroxyphenyl)-2′-deoxyguanosine,50 and noted that the molecule fluctuates between a twisted and planar enol conformation in the ground state and does not form the keto structure, the pKa of the N7 nitrogen being ∼3, which is 3-4 pKa units below the N1 pKa observed for the prodigiosin analogues 2 and 3. In contrast, Bra¨uer et al.45 found that 4,5-dimethyl-2-(2-hydroxyphenyl)imidazole, which possess a pKa ∼ 7.43 for the protonated imidazole, generates the keto form (λmax ) 342 nm) in neutral water, whereas the enol form (λmax ) 312 nm) is favored in ethanol (EtOH) and all aprotic solvents. This was considered a remarkable result since systems bearing a phenolic ring are known to favor the enol form in the ground state.30-34,50 In

contrast, systems bearing naphthols have long been known to undergo ground-state tautomerization with the keto tautomer being favored in polar protic solvents.51-54 The HOMA (harmonic oscillator model of aromaticity) index of naphthalene (0.827) is lower than that of benzene (0.974), and this difference in aromatic character is thought to explain why nonaromatic keto forms of 2-naphthol are more important in tautomeric mixtures than the keto forms in tautomeric phenol.55 For 4,5dimethyl-2-(2-hydroxyphenyl) the basicity of the imidazole N atom was thought to play a role in lowering the energy of the keto form in neutral aqueous solutions.45 Prodigiosin analogues 2 and 3 have N1 basicities that are similar to the corresponding N basicity of 4,5-dimethyl-2-(2hydroxyphenyl)imidazole, but unlike 4,5-dimethyl-2-(2-hydroxyphenyl)imidazole, the proton acceptor pyrromethene is a conjugated chromophore that may stabilize the keto form, similar to conjugated naphthol donor groups. This prompted examination of the optical properties of 2 and 3 in a wider range of protic solvents to draw comparison to the behavior of 4,5dimethyl-2-(2-hydroxyphenyl)imidazole that does not form the keto form in EtOH. Figure 6 shows UV-vis absorbance spectra of the o-phenolic derivative 3 in i-PrOH, EtOH, and MeOH in which the absorbance at 538 nm for the keto tautomer is present

11808 J. Phys. Chem. B, Vol. 111, No. 40, 2007

Figure 5. UV-vis absorbance spectra of (a) 2 and (b) 3 in different volume ratios of MeCN/buffered H2O (10 mM TAPS pH 8.5 for 2 and 10 mM MOPS pH 7.0 for 3) at 25 °C, where the solid line corresponds to the spectrum in 100% MeCN and the dotted lines are spectra from solutions containing 20% increment additions of buffered H2O (indicated by the arrow) up to 80%.

Figure 6. UV-vis absorbance spectra of 3 in i-PrOH (dotted trace), EtOH (bold trace), and MeOH.

in all protic solvents but is prevalent in MeOH. For the p-phenolic derivative 2 the absorbance at ∼530 nm for the keto tautomer could be observed in MeOH, but was not detected in EtOH and i-PrOH (not shown), indicating that the o-phenolic analogue 3 undergoes tautomerization more readily, which is ascribed to the importance of the intramolecular H-bond in 3.54 Because the absorbance ascribed to the keto tautomer at ∼530 nm was distinct from the enol absorption at ∼460 nm it was possible to provide an estimate of the tautomeric equilibrium constant (KT) from the UV-vis data recorded in the various solvent systems using the equation KT ) % keto (K)/(100 %K) where %K ) (Kabs)/(Kabs + enol (E)abs) × 100.51,52 The KT values determined for 2 and 3 along with solvatochromic parameters for the various solvents are given in Table 2. The Kamlet-Taft π* parameter is a dipolarity/polarizability term that measures the endoergic effects of solute-solvent dipoledipole and dipole-induced interactions, whereas the R and β values are H-bonding terms that measure the exoergic effects of the solvent acting as a H-bond donor (R) or H-bond acceptor (β) with the solute.56,57 The Dimroth and Reichardt normalized “solvent polarity” ET parameter58 is a measure of solvent polarity. Values for these parameters have also been determined in MeCN/H2O mixtures at 25 °C59,60 and were used to provide

La et al. an estimate of the properties of the H2O/MeCN mixtures used in this study by converting the mixtures into mole fraction of H2O in order to use the reported values. Mills and Beak have examined solvent effects on keto-enol equilibria and found that KT values correlate with Kamlet-Taft parameters; for systems with intramolecular H-bonds the π* term dominates, whereas for systems with no internal H-bond the equilibrium is controlled by the solvent β term.61 For KT values estimated for 2 and 3 in H2O/MeCN a better correlation with π* (R ) 0.9) than with ET (R ) 0.86) (R values are greater than 0.98 when only the 20-60% MeCN range is considered) and no correlation with the H-bonding terms was found. In alcohol solvents KT values for 3 also correlate with both ET and π* (R ) 0.94), but now a better correlation is found with R (R ) 0.98) but not with β (data not shown). Overall, our results are consistent with previous findings that demonstrate the importance of solvent polarity,45,61 intermolecular H-bonding from the solvent,45,54 and intramolecular H-bonding in the substrate,45,54 as factors that stabilize the keto tautomer relative to the enol tautomer. 3.2.b. Fluorescence Spectra. The emission spectra of the keto forms of 2 and 3 were recorded in the 20% MeCN/80% H2O (0.1 M NaCl) mixture containing 10 mM buffer at pH 8.5 and 7.0, respectively. For 2, excitation at 538 nm generated a distinct emission spectra with λmax ) 615 nm (Figure 7a) that was redshifted by 23 nm compared to the emission spectra noted for the phenolate L- (592 nm) (Figure 4a). In contrast, excitation at 538 nm for the keto form of 3 yielded a spectrum with λmax ) 555 nm that resembled closely the emission spectra recorded for LH+ (555 nm) and L (555 nm) (Figure 4b), with the exception that a more prominent shoulder peaking at 591 nm was also present (Figure 7b). Now that the full range of emission spectra have been recorded for 2 and 3 (LH+, L, L-, and keto) the optical properties observed for the ortho isomer 3 can be rationalized through examination of the structures outlined in Scheme 4. The neutral ligand (L) is expected to possess a strong intramolecular H-bond between the phenolic OH and N1 in the enol structure that is favored in aprotic solvents. Prototropic tautomerization of the enol into the keto structure is observed in polar protic solvents in the ground state. The emission spectra of the protonated (LH+), neutral (L), and keto structures have identical λmax values (555 nm), which was not the case for the para isomer 2, in which distinct fluorescence spectra were observed for the different forms of 2. For the o-phenolic derivative 3 we propose that the zwitterion shown in Scheme 4 is the dominant structure formed in the excited state. This structure would not play a role in the emissive properties of the p-phenolic analogue 2, because the intramolecular H-bonding interaction between the phenolate and the protonated N1, as shown for 3 in Scheme 4, is not possible for 2. Thus, excitation of the enol of 3 generates the zwitterion by an ESIPT process, but it is not a prototropic tautomerization in the excited state because the proton transfer is not accompanied by migration of π-electrons.49 Excitation of the protonated LH+ species of 3 would also be expected to generate the zwitterion through loss of the acidic phenolic proton in the excited state (it is commonly accepted that aromatic hydroxyl groups become more acidic in the excited state in comparison to the ground state45) to form the stable intramolecular H-bond present in the zwitterion species. Migration of π-electrons upon excitation of the keto structure of 3 would also yield the zwitterion. This provides a reasonable explanation for the emissive properties of 3.

Tautomeric Equilibria in Prodigiosin Derivatives

J. Phys. Chem. B, Vol. 111, No. 40, 2007 11809

TABLE 2: Solvatochromic Parameter Values and Tautomeric Equilibrium Constants for 2 and 3a solvent 80% MeCN/20% H2O 60% MeCN/40% H2O 40% MeCN/60% H2O 20% MeCN/80% H2O MeOH EtOH i-PrOH

ET c

0.75 0.80 0.83 0.89 0.762e 0.654 0.546

π* d

0.82 0.87 0.97 1.08 0.60f 0.54 0.48

R

β

KT (2/3)

∆GT (2/3)b

0.87 0.90 0.90 0.97 0.93 0.83 0.76

0.59 0.60 0.61 0.61 0.62 0.77 0.95

0.06/0.13 0.15/0.21 0.45/0.46 1.40/1.37 0.14/0.52 n.d./0.28 n.d./0.19

1.67/1.21 1.12/0.92 0.47/0.46 -0.20/-0.19 1.16/0.39 n.d./0.75 n.d./0.98

a H2O contained 10 mM TAPS buffer, pH 8.5 for 2 and 10 mM MOPS buffer, pH 7.0 for 3, each with µ 0.1 M NaCl. b In kcal mol-1. c ET values in MeCN/H2O taken from ref 60. d Kamlet-Taft parameters in MeCN/H2O taken from average values in ref 59. e ET values of alcohols taken from ref 58. f Kamlet-Taft parameters for alcohols taken from ref 56.

Figure 7. Emission spectra of 2 (a) and 3 (b) recorded in 80% H2O/ 20% MeCN with 10 mM buffer at pH 8.5 and 7.0, respectively. Emission spectra recorded with excitation at 538 nm; λmax for the keto tautomer.

SCHEME 4

Figure 8. (a) UV-vis titration of 3 with Zn(II) acetate in MeOH at 25 °C; the arrow points to increased additions of Zn(II). (b) Emission spectrum of 3 (solid line) in MeOH and the 1:1 complex of 3 with Zn(II) (dotted line) in MeOH at ambient temperature; the arrow points to the increase in emission of 3 upon Zn(II) coordination.

3.3. Complexation Studies. We have demonstrated that prodigiosin 1 coordinates Cu(II) ions to form a 1:1 complex in which all three N atoms of 1 attach to Cu(II) to form a square planar complex, as evidenced by X-ray crystallography.21 In contrast, prodigiosin 1 coordinates Zn(II) ions to form a 2:1 Zn(1)2 complex in which the A-ring does not participate in Zn(II) coordination; the X-ray crystal structure of the Zn(1)2 complex showed that the pyrromethene moiety of two prodigiosin ligands bound to Zn(II) in a tetrahedral arrangement.21 This background work coupled with the report by Henary and Fahrni30 that a 2-(2′-hydroxyphenyl)-benzoxazole derivative coordinates Zn(II) (log K 3.93) to form a 1:1 complex exhibiting a 50-fold increase in fluorescence intensity (λmax 411 nm), and that such a ligand may be suitable as a fluorescence probe in a biological environment to monitor Zn(II) coordination, prompted examination of the Zn(II) binding properties of the phenolic prodigiosin analogues 2 and 3. A UV-vis titration of 2 with Zn(II) in MeOH formed a complex with λmax 504 nm and a shoulder at 480 nm in a 1.5:1 ratio (data not shown). The UV-vis spectrum of the Zn(II) complex of 2 resembled the spectrum of the protonated species LH+, and the titration could not be fit to a 1:1 binding isotherm.

The fluorescence spectrum of the Zn(II) complex of 2 exhibited a 3-fold increase in fluorescence intensity compared to the neutral ligand in MeOH (data not shown), and overall these results were completely consistent with 2 binding Zn(II) in a 2:1 fashion, as observed for prodigiosin 121 and other prodigiosin analogues12 where the A-ring cannot participate in Zn(II) binding. In contrast, the UV-vis titration of 3 with Zn(II) ions in MeOH shown in Figure 8a exhibited λmax at 521 nm (a red shift of 17 nm compared to LH+ of 2) that was accompanied by a distinct shoulder peak at 489 nm in a ratio of 2:1 with respect to λmax at 521 nm. Examination of the titration using the 1:1 binding isotherm showed excellent correlation (R ) 0.999), and an apparent binding constant K of 1.95 × 105 M-1 (log K 5.29) was determined for Zn(II) coordination by 3. These results were consistent with participation of Zn(II) binding by the phenolic A-ring of 3 to form a new 1:1 complex. Figure 8b shows the change is emission upon Zn(II) coordination by 3. The Zn(II) complex of 3 shows an emission spectrum with λmax 542 nm that exhibits a 40-fold increase in emission intensity compared to that of the free ligand. For prodigiosin 1 we previously demonstrated that it exhibits weak fluorescence in water and that the emission yield is increased as the polarity of solvent decreased, or upon noncovalent binding to DNA, suggesting that DNA binding likely involved a reduction in the exposure of 1 to water.19 Fluores-

11810 J. Phys. Chem. B, Vol. 111, No. 40, 2007 cence enhancements observed for molecules upon coordination to substrates have been ascribed to decreases in degrees of molecular motion and shielding of the excited state from water molecules, among others.62 For prodigiosin 1 we ascribed its weak emission in water to the fact that it possesses NH and N atoms that can form exciplexes with water in the excited state that can quench the emission of the fluorophore.19,63 Thus, removal of 1 from an aqueous environment increased fluorescence intensity. For 2 and 3 we observed weak emission yield regardless of the solvent, and this is ascribed to the presence of the phenolic A-ring system that can act like water and quench the emission intensity. Thus, the large fluorescence increase observed for 3 upon Zn(II) coordination is completely consistent with coordination by the phenolic OH atom of 3, which removes its quenching attributes in the excited state. Increased rigidity by restriction of motion in the 1:1 Zn(II) complex of 3 would also be expected to increase the emission intensity. That 3 shows a 40-fold increase in fluorescence intensity upon Zn(II) complexation suggests that it may serve as a probe to gauge Zn(II) concentrations in a biological matrix.30 In fact, with a log K of 5.29 for Zn(II) complexation by 3 suggests that it would serve as a more sensitive probe than the previous ligand (log K 3.93) reported by Henary and Fahrni.30 4. Conclusions The prodigiosin analogues 2 and 3 that contain isomeric phenolic A-ring systems show unique characteristics compared to the family of natural and synthetic prodigiosin analogues studied to date. Unlike all other prodigiosins, 2 and 3 undergo prototropic tautomerization in polar protic solvents to generate keto tautomers with λmax ∼ 530 nm; the enol tautomers with λmax ∼ 460 nm are the only tautomers formed in aprotic solvents. The o-phenolic analogue 3 also possesses new properties in terms of metal coordination since it binds Zn(II) to form a 1:1 complex with a log K of 5.29. The natural prodigiosin 1 binds Zn(II) to form a tetrahedral Zn(1)2 complex in which the A-pyrrole ring of 1 does not participate in Zn(II) coordination, suggesting that the phenolic A-ring of 3 coordinates to Zn(II). The new 1:1 Zn(II) structure of 3 is very emissive and may serve as a fluorescence probe to monitor Zn(II) concentrations in a biological matrix. The complex may also serve as a model to gain an understanding of substrate (DNA or protein) binding by the 1:1 Cu(II) complex of the natural product 1 in which the A-pyrrole ring participates in Cu(II) binding. The neutral ligand (L) of 3 also has unique properties that stem from the intramolecular H-bond between the phenolic OH and N1 of the pyrromethene chromophore. The pKa of the protonated ligand (LH+) of 3 is ∼6.0, which is ∼1.4 pKa units below the pKa determined for the p-phenolic analogue 2. This change in pKa has been ascribed to stabilization of the β-form of prodigiosin due to intramolecular H-bonding that is available to 3 in both L and LH+. Since natural prodigiosins fluctuate between R- and β-forms that are predicted to have very different substrate (DNA or protein) binding affinities, the derivative 3 may serve as a useful probe to determine substrate binding by the β-form. Work on determining the biological properties of 2 and 3 is currently underway to discover whether the unique properties reported herein enhance or inhibit the therapeutic potential of the prodigiosin natural products. Acknowledgment. Financial support for this work was provided by the Natural Sciences and Engineering Research Council of Canada (NSERC), University of Guelph, the Canada Foundation for Innovation (CFI), and the Ontario Innovation

La et al. Trust Fund (OIT). R.A.M. is grateful to Dr. Wojciech Gabryelski (UOG) for helpful suggestions. References and Notes (1) Manderville, R. A. Curr. Med. Chem.: AntisCancer Agents 2001, 1, 195-218. (2) Montaner, B.; Pe´rez-Toma´s, R. Curr. Cancer Drug Targets 2003, 3, 57-65. (3) Fu¨rstner, A. Angew. Chem., Int. Ed. 2003, 42, 3582-3603. (4) Boger, D. L.; Patel, M. J. Org. Chem. 1988, 53, 1405-1415. (5) D’Alessio, R.; Rossi, A. Synlett 1996, 6, 513-514. (6) Wasserman, H. H.; Xia, M. D.; Wang, J. J.; Petersen, A. K.; Jorgensen, M.; Power, P.; Parr, J. Tetrahedron 2004, 60, 7419-7425. (7) Dairi, K.; Tripathy, S.; Attardo, G.; Lavalle´e, J. F. Tetrahedron Lett. 2006, 47, 2605-2606. (8) Jolicoeur, B.; Lubell, W. D. Org. Lett. 2006, 8, 6107-6110. (9) D’Alessio, R.; Bargiotti, A.; Carlini, O.; Colotta, F.; Ferrari, M.; Gnocchi, P.; Isetta, A.; Mongelli, N.; Motta, P.; Rossi, A.; Rossi, M.; Tibolla, M.; Vanotti, E. J. Med. Chem. 2000, 43, 2557-2565. (10) Melvin, M. S.; Tomlinson, J. T.; Park, G.; Day, C. S.; Saluta, G. R.; Kucera, G. L.; Manderville, R. A. Chem. Res. Toxicol. 2002, 15, 734741. (11) Baldino, C. M.; Parr, J.; Wilson, C. J.; Ng, S.-C.; Yohannes, D.; Wasserman, H. H. Bioorg. Med. Chem. Lett. 2006, 16, 701-704. (12) Regourd, J.; Al-Sheikh Ali, A.; Thompson, A. J. Med. Chem. 2007, 50, 1528-1536. (13) Sato, T.; Konno, H.; Tanaka, Y.; Kataoka, T.; Nagai, K.; Wasserman, H. H.; Ohkuma, S. J. Biol. Chem. 1998, 273, 21455-21462. (14) Yamamoto, C.; Takemoto, H.; Kuno, K.; Yamamoto, D.; Tsubura, A.; Kamata, K.; Hirata, H.; Yamamoto, A.; Kano, H.; Seki, T.; Inoue, K. Hepatology 1999, 30, 894-902. (15) Yamamoto, D.; Kiyozuka, Y.; Uemura, Y.; Yamamoto, C.; Takemoto, H.; Hirata, H.; Tanaka, K.; Hioki, K.; Tsubura, A. J. Cancer Res. Clin. Oncol. 2000, 126, 191-197. (16) Castillo-Avila, W.; Abal, M.; Robine, S.; Pe´rez-Toma´s, R. Life Sci. 2005, 78, 121-127. (17) Sessler, J. L.; Eller, L. R.; Cho, W.-S.; Nicolaou, S.; Aguilar, A.; Lee, J. T.; Lynch, V. M.; Magda, D. J. Angew. Chem., Int. Ed. 2005, 44, 5989-5992. (18) Seganish, J. L.; Davis, J. T. Chem. Commun. 2005, 5781-5783. (19) Melvin, M. S.; Ferguson, D. C.; Lindquist, N.; Manderville, R. A. J. Org. Chem. 1999, 64, 6861-6869. (20) Montaner, B.; Castillo-Avila, W.; Martinell, M.; O ¨ llinger, R.; Aymami, J.; Giralt, E.; Pe´rez-Toma´s, R. Toxicol. Sci. 2005, 85, 870-879. (21) Park, G.; Tomlinson, J. T.; Melvin, M. S.; Wright, M. W.; Day, C. S.; Manderville, R. A. Org. Lett. 2003, 5, 113-116. (22) Melvin, M. S.; Tomlinson, J. T.; Saluta, G. R.; Kucera, G. L.; Lindquist, N.; Manderville, R. A. J. Am. Chem. Soc. 2000, 122, 63336334. (23) Melvin, M. S.; Wooton, K. E.; Rich, C. C.; Saluta, G. R.; Kucera, G. L.; Lindquist, N.; Manderville, R. A. J. Inorg. Biochem. 2001, 87, 129135. (24) Melvin, M. S.; Calcutt, M. W.; Noftle, R. E.; Manderville, R. A. Chem. Res. Toxicol. 2002, 15, 742-748. (25) Tomlinson, J. T.; Park, G.; Misenheimer, J. A.; Kucera, G. L.; Hesp, K.; Manderville, R. A. Org. Lett. 2006, 8, 4951-4954. (26) Park, G.; Tomlinson, J. T.; Misenheimer, J. A.; Kucera, G. L.; Manderville, R. A. Bull. Korean Chem. Soc. 2007, 28, 49-52. (27) Fu¨rstner, A.; Reinecke, K.; Prinz, H.; Waldmann, H. ChemBioChem 2004, 5, 1575-1579. (28) Mosquera, M.; Penedo, J. C.; Rios Rodrı´guez, M. C.; Rodrı´guezPrieto, F. J. Phys. Chem. 1996, 100, 5398-5407. (29) Abou-Zied, O.; Jimenez, R.; Thompson, E. H. Z.; Millar, D. P.; Romesberg, F. E. J. Phys. Chem. A 2002, 106, 3665-3672. (30) Henary, M. M.; Fahrni, C. J. J. Phys. Chem. A 2002, 106, 52105220. (31) de Vivie-riedle, R.; De Waele, V.; Kurtz, L.; Riedle, E. J. Phys. Chem. A 2003, 107, 10591-10599. (32) Abou-Zied, O. K.; Jimenez, R.; Romesberg, F. E. J. Am. Chem. Soc. 2001, 123, 4613-4614. (33) Dupradeau, F.-Y.; Case, D. A.; Yu, C.; Jimenez, R.; Romesberg, F. E. J. Am. Chem. Soc. 2005, 127, 15612-15617. (34) Seo, J.; Kim, S.; Park, S. Y. J. Am. Chem. Soc. 2004, 126, 1115411155. (35) Sanders, J. K. M.; Hunter, B. K. Modern NMR Spectroscopy; Oxford University Press: Oxford, 1987; pp 93-207. (36) Demas, J. N.; Crosby, G. A. J. Phys. Chem. 1971, 75, 991-1024. (37) Williams, A. T. R.; Winfield, S. A.; Miller, J. N. Analyst 1983, 108, 1067-1071. (38) Fery-Forgues, S.; Lavabre, D. J. Chem. Educ. 1999, 76, 12601264. (39) Patterson, G. S. J. Chem. Educ. 1999, 76, 395-398.

Tautomeric Equilibria in Prodigiosin Derivatives (40) Borah, S.; Melvin, M. S.; Lindquist, N.; Manderville, R. A. J. Am. Chem. Soc. 1998, 120, 4557-4562. (41) Conners, K. A. Binding Constants. The Measurement of Molecular Complex Stability; John Wiley & Sons: New York, 1987; p 152. (42) Hearn, W. R.; Medina-Castro, J.; Elson, M. K. Nature 1968, 220, 170-171. (43) Rizzo, V.; Morelli, A.; Pinciroli, V.; Sciangula, D.; D’Alessio, R. J. Pharm. Sci. 1999, 88, 73-78. (44) Perrin, D. D.; Dempsey, B.; Serjeant, E. P. pKa Prediction for Organic Acids and Bases; Chapman and Hall Ltd.: New York, 1981; p 137. (45) Bra¨uer, M.; Mosquera, M.; Pe´rez-Lustres, J. L.; Rodriguez-Prieto, F. J. Phys. Chem. A 1998, 102, 10736-10745. (46) Himo, F.; Noodleman, L.; Blomberg, M. R. A.; Siegbahn, P. E. M. J. Phys. Chem. A 2002, 106, 8757-8761. (47) Gilli, P.; Bertolasi, V.; Ferretti, V.; Gilli, G. J. Am. Chem. Soc. 2000, 122, 10405-10417. (48) Il’ichev, Y. V.; Perry, J. L.; Manderville, R. A.; Chignell, C. F.; Simon, J. D. J. Phys. Chem. B 2001, 105, 11369-11376. (49) Raczyn´ska, E. D.; Kosin´ska, W.; Os´mialowski, B.; Gawinecki, R. Chem. ReV. 2005, 105, 3561-3612. (50) McLaughlin, C. K.; Lantero, D. R.; Manderville, R. A. J. Phys. Chem. A 2006, 110, 6224-6230.

J. Phys. Chem. B, Vol. 111, No. 40, 2007 11811 (51) Stoyanov, S.; Antonov, L. Dyes Pigm. 1988, 10, 33-45. (52) Antonov, L.; Stoyanov, S.; Stoyanova, T. Dyes Pigm. 1995, 27, 133-142. (53) Antonov, L.; Fabian, W. M. F.; Nedeltcheva, D.; Kamounah, F. S. J. Chem. Soc., Perkin Trans. 2 2000, 1173-1179. (54) Fabian, W. M. F.; Antonov, L.; Nedeltcheva, D.; Kamounah, F. S.; Taylor, P. J. J. Phys. Chem. A 2004, 108, 7603-7612. (55) Os´mialowski, B.; Raczyn´ska, E. D.; Krygowski, T. M. J. Org. Chem. 2006, 71, 3727-3736. (56) Kamlet, J.; Doherty, R. M.; Abraham, M. H.; Marcus, Y.; Taft, R. W. J. Org. Chem. 1983, 48, 2877-2887. (57) Kamlet, M. J.; Doherty, R. M.; Abraham, M. H.; Marcus, Y.; Taft, R. W. J. Phys. Chem. 1988, 92, 5244-5255. (58) Reichardt, C. Chem. ReV. 1994, 94, 2319-2358. (59) Marcus, Y.; Migron, Y. J. Phys. Chem. 1991, 95, 400-406. (60) Krygowski, T. M.; Wrona, P. K.; Zielkowska, U.; Reichardt, Tetrahedron 1985, 41, 4519-4527. (61) Mills, S. G.; Beak, P. J. Org. Chem. 1985, 50, 1216-1224. (62) Frankewich, R. P.; Thimmaiah, K. N.; Hinze, W. L. Anal. Chem. 1991, 63, 2924-2933. (63) Hoshino, M.; Imamura, M.; Ikehara, K.; Hama, Y. J. Phys. Chem. 1981, 85, 1820-1823.