Fluorescence of pterin, 6-formylpterin, 6-carboxypterin and folic acid in aqueous solution: pH e ﬀ ects

Steady-state and time-resolved studies have been performed on four compounds of the pterin family (pterin, 6-carboxypterin, 6-formylpterin and folic acid) in aqueous solution, using the single photon counting technique. The ﬂ uorescence characteristics (spectra, quantum yields, lifetimes) of these compounds and their dependence on the pH have been investigated. Most pterins can exist in two acid–base forms over the pH range between 3 and 13. Emission spectra and excitation spectra were obtained for both forms of each compound studied. Fluorescence quantum yields ( Φ F ) in acidic and basic media were measured. The Φ F of folic acid (<0.005 in both media) is very low compared to those of pterin (0.27 in basic media and 0.33 in acidic media), 6-carboxypterin (0.18 in basic media and 0.28 in acidic media) and 6-formylpterin (0.07 in basic media and 0.12 in acidic media). The variation in integrated ﬂ uorescence intensity and ﬂ uorescence lifetimes ( τ F ) was analysed as a function of pH. Dynamic quenching by OH  was observed and the corresponding bimolecular rate constants for quenching of ﬂ uorescence ( k q ) were calculated. The reported values for k q (M  1 s  1 ) are 3.6 × 10 9 , 1.9 × 10 9 and 1.1 × 10 10 M  1 s  1 for pterin, 6-carboxypterin and 6-formylpterin, respectively.


Introduction
Pterins are a family of heterocyclic compounds present in biological systems. These compounds are derived from 2-amino-4hydroxypteridine (pterin, Fig. 1). It is well-known that pterins participate in important biological functions, 1,2 e.g., folic acid (pteroylglutamic acid), a conjugated pterin, is a vitamin of the B group and acts as a coenzyme in reactions related to the synthesis of purinic and pyrimidinic bases. 3 It has been reported that some pterins are involved in different photochemical processes. The sensitivity of pterins to light has been known for several decades 4, 5 and they have been found in photosensitive organs, such as the eyes, of different animals. 6 Some reports suggest that pterins may act as blue antennas in superior plants 7 and other organisms such as the fungus Phycomyses blakesleeanus. 8 Moreover, it has been suggested that pterins may play some role in photosynthesis. 9 5,10- Methenyltetrahydrofolylpolyglutamate, a folic acid derivative, is the light-harvesting antenna of DNA photolyases, enzymes involved in DNA repair processes 10-12 that take place after UV irradiation. Recent studies have shown that pterin and some pterin derivatives are instrumental in the photosensitisation of DNA. Along this line, photoinduced cleavage of thymus calf DNA 13 and cleavage of plasmid DNA 14 have been reported.
In spite of the evident importance of pterins in photochemical processes that take place in biological systems, relatively few reports deal with the photochemistry 15-20 and photophysics 21-23 of this family of compounds. More basic studies on these topics are required in order to understand in detail the role of pterins in such photobiological processes.
Pterins behave as weak acids in aqueous solution, where several acid-base equilibria may be present. As reported by Albert 24 for several pterin derivatives, the dominant equilibrium in the pH range 4-12 involves the amide group (acid form) and the phenolate group (base form) (Fig. 1). The pK a of this equilibrium is around 8 for the various pterin derivatives studied. 19,24-26 Other functional groups of the pterin moiety (e.g. 2-amino group or ring nitrogen atoms) have pK a values lower than 2. 24 The photochemistry, as well as the photophysical behaviour, of the different acid-base forms of each compound presents significant differences, as reported in previous studies by us 18 and other groups. 15,16,21 In the context of our investigations on pterin derivatives, 14,18,19,25, 26 we have performed photophysical studies on a group of four compounds in aqueous solution: pterin (PT), 6carboxypterin (CPT), 6-formylpterin (FTP) and folic acid (FA) (Fig. 1). We report here results of measurements of steady-state and time-resolved fluorescence and discuss the dependence of the emission and excitation spectra, fluorescence quantum yields (Φ F ) and lifetimes (τ F ) on the pH.

Experimental
The pterins (Shircks Laboratories) were used without further purification. pH measurements were performed using a pHmeter Schott CG 843P with a pH-combination electrode Blue-Line 14pH (Schott). The pH of the aqueous solutions was adjusted by adding drops of HCl or NaOH from a micropipette. The concentrations of the acid and base used for this purpose ranged from 0.1 to 2 M. For experiments at pH lower than 11 the ionic strength was held constant at 10 3 M; for experiments at higher pH the ionic strength was of the same order as the HO concentration. UV-visible spectra were recorded on a Cary 5 (Varian) spectrophotometer.
Steady-state and time-resolved fluorescence measurements were performed using single photon counting equipment EAI-FS/FL900 (Edinburgh Analytical Instruments, UK). The quartz measurement cell (1 cm path length) was thermoregulated at 23.9 ± 0.2 C. Corrected fluorescence spectra were recorded between 350 and 650 nm at different excitation wavelengths using a high pressure Xe lamp (419 W). The excitation spectra of the compounds studied were recorded between 200 and 500 nm, monitoring the fluorescence intensity at 450 nm.
A N 2 excitation lamp (1.2 bar, operated at 6.3 kV and a frequency of 40 kHz) was employed for time-resolved studies. The single photon counting range of the equipment is 500 ps-500 μs, but the selected counting time window for the measurements reported in this study was 0-100 ns. The emission decays were monitored at 450 nm after excitation at 350 nm. Lifetimes were obtained from the monoexponential decays observed after deconvolution from the lamp background signal, using the software provided by Edinburgh Analytical Instruments. Our method of analysis of steady-state and time-resolved data has previously been described in detail. 27, 28 The fluorescence quantum yields were determined from the corrected fluorescence spectra using quinine bisulfate (Riedel-deHaen) in 0.5 M H 2 SO 4 as a reference 29 (Φ F = 0.546 30 ). In order to avoid inner filter effects, the absorptions of the solutions, at the excitation wavelength, were kept below 0.10.

Absorption and emission spectra
The absorption and fluorescence characteristics of pterin (PT) and three pterin derivatives (CPT, FPT, FA) have been investigated in the pH range 4-13. Under these conditions, the acidbase equilibrium to be considered involves the amide group in the acid form and the phenolate group in the base form 24 (Fig. 1). In the pH ranges 4.9-5.5 and 10.0-10.5, pterins are present at more than 99% in the acid and base forms, respectively. Although not of direct interest for biological systems, a large pH range was chosen for our investigations, because knowledge of the photophysical properties of the "pure" acid and base forms appeared to be of fundamental interest for understanding the photophysics of these compounds under less harsh pH conditions.
It should be noted that the absorption spectra of PT and CPT are similar, both in acidic and alkaline media, suggesting that the electronic distribution on the pterin moiety is only slightly affected by the presence of the -COOH group at position 6. In contrast, interactions between the substituent at position 6 and the pterin moiety affect considerably the absorption spectra of FPT (formyl) and FA (relatively long chain substituent). As can be observed in Fig. 2, a new absorption band centered at 310 nm appears in the spectrum of FPT in its acid form, and the band centered at approximately 350 nm has a higher relative intensity than in the case of PT and CPT, in both acidic and alkaline media. The differences that can be observed in the absorption spectra of FA are even more striking.
The fluorescence emission spectra of the four pterins show a dependence on the pH (Fig. 3). The emission spectra of the base forms obtained by excitation at 350 nm are red shifted in comparison with the spectra of the acid forms obtained by excitation at the same wavelength. The wavelengths of the fluorescence maxima (λ F ) are listed in Table 1.

Fig. 2
Absorption and excitation spectra of air-equilibrated aqueous solutions of pterin derivatives; solid lines: absorption spectra of acid forms (pH 5.5); dashed lines: excitation spectra of acid forms; dashed-dotted lines: absorption spectra of base forms (pH 10.5); dotted lines: excitation spectra of base forms (excitation spectra obtained by monitoring the fluorescence at 450 nm; for comparative purposes, each spectrum was normalized relative to the maximum of the lowest energy band). Table 1 Fluorescence quantum yields (Φ F ) in argon-saturated, air-equilibrated and oxygen-saturated aqueous solutions, wavelengths of fluorescence maxima (λ F ) and fluorescence lifetimes (τ F ) of the pterin derivatives a

Compound
Acid-base form a Measurements were carried out for the acid and base forms in the pH ranges 4.9-5.5 and 10.0-10.5, respectively (excitation wavelength: 350 nm; standard deviations are indicated in parenthesis).
Fluorescence spectra resulting from excitation at wavelengths shorter than 350 nm were also recorded for each compound in acidic and base solutions (results not shown). Wavelengths typically in the range between 230 and 280 nm were used for exciting the high energy band(s) of the pterins. In all cases, the fluorescence spectrum (normalized relative to the maximum emission value for comparative purposes) remained unchanged, irrespective of the excitation wavelength, suggesting that only one excited state contributes to the fluorescence. However, the fluorescence intensities decreased when exciting in the high energy absorption bands, i.e. at wavelengths shorter than 300 nm (for further discussion on this point, see Section 3.2).

Fluorescence quantum yields
For the four compounds investigated, the fluorescence quantum yields (Φ F ) were determined for both the acid and base forms in argon-saturated, air-equilibrated and oxygen-saturated solutions. The results for excitation at 350 nm are shown in Table 1. For PT and CPT, the values of Φ F are relatively high, Φ F (CPT) being slightly lower than Φ F (PT). It is noteworthy that Φ F (FPT) is lower by more than a factor of 2, both in acidic and alkaline media, whereas FA has very small Φ F values (<0.005). Therefore, the nature of the substituent at position 6 on the pterin moiety affects the deactivation pathways of the singlet excited states considerably (as is the case for the absorption spectra, Section 3.1). In particular, the long chain substituent at position 6 on the pterin moiety of FA might act as an "internal quencher", thus enhancing the radiationless deactivation of the singlet excited state. The values of Φ F for the acid forms of PT, CPT and FPT (0.33, 0.28 and 0.12, respectively) are higher than Φ F for the corresponding base forms (0.27, 0.18 and 0.07). The differences between the Φ F values determined in the presence or in the absence of O 2 were not significant, indicating that quenching of the singlet excited states by O 2 is negligible for the four compounds. To our knowledge, Φ F of pterins have not been reported previously, except in the case of PT in buffered aqueous solution at pH 10. 21 The very low value reported by the authors (0.057) probably results from fluorescence quenching by the buffer components. 32 The excitation spectra of the compounds studied in acidic and basic air-equilibrated solutions are shown in Fig. 2, together with corresponding absorption spectra. In both series of spectra, the wavelengths of the band maxima are similar. However, the intensities of the high energy bands relative to that of the lowest energy band are much lower in the excitation spectra. Fluorescence quantum yields obtained by exciting at different wavelengths (Φ F(λ) ) were calculated from the excitation spectra and from Φ F(350) ( Table 1) Fluorescence quantum yields calculated from the analysis of the excitation spectra, using eqn. (1), are listed in Table 2. The values of Φ F for the high energy bands are lower by a factor of at least 3 than those corresponding to the low energy bands (Tables 1 and 2). These results suggest that only a fraction of the energy of the upper excited state(s) (S 2 , S n ) is dissipated through internal conversion to the lowest singlet excited state (S 1 ). Therefore, intersystem crossing to the triplet manifold or photochemical reactivity should occur from an upper singlet excited state. Fig. 4 shows the variation of the integrated fluorescence intensities as a function of pH in the range 4-10.5. The behavior observed is due to the previously mentioned equilibria between the acid and the base forms (Fig. 1). The pK a values for these equilibria were determined from the changes in fluorescence    3 Values of K a and pK a for the acid-base equilibrium of the pterins shown in Fig. 1  intensities (I F ) integrated between 360 and 650 nm and corrected for absorbance. The experimental variation of I F as a function of pH at a given excitation wavelength could be fitted by eqn. (2), where, I a and I b are the integrated fluorescence intensities of the acid and base forms of the species involved in the acid-base equilibrium and K a is the dissociation constant. The corresponding fit for FA could not be carried out due to the very low fluorescence of this compound and the large experimental error in the titration. The pK a values determined in this work, as well as literature values obtained from spectrophotometric titration, 19,25,26 are listed in Table 3. No significant difference can be observed between either group of experimental data for PT and CPT. Two explanations may be suggested: 1) the pK a of the equilibria between the excited singlet states are the same as the corresponding pK a of the ground states; 2) deactivation of the excited singlet state of the acid (base) form is much faster than its deprotonation (protonation). In the case of FPT, however, the observed difference between the pK a values strongly suggests the existence of an acid-base equilibrium in the excited state, resulting in more acidic behavior than the ground state.

Influence of the pH on fluorescence emission
In all the cases studied, a strong decrease of the fluorescence intensity upon excitation at 350 nm was observed at a pH value higher than 11, although the wavelength of the emission maximum was not altered. The decrease of the fluorescence intensity as a function of the concentration of HO followed Stern-Volmer behavior (Fig. 5a). The corresponding Stern-Volmer constants (K SV ) ( Table 4) were compared to the K SV values obtained by analysis of the lifetime dependence on the concentration of HO (Section 3.4).

Time-resolved study
Fluorescence decays were analyzed for both the acid and base forms of the compounds studied. A first-order rate law was observed for all the decays. A typical trace recorded for FPT is shown in Fig. 6. Fluorescence lifetimes (τ F ) were obtained by averaging at least three values taken in the pH range 3.0-6.2 (for the acid forms) and in the pH range 9.0-11.0 (for the base forms). The results are shown in Table 1.  The fluorescence lifetimes of the base forms also decreased at pH values higher than 11. The dependence of this parameter on the HO concentration showed Stern-Volmer behavior. The corresponding K SV (Table 4) are comparable to those obtained from fluorescence intensity measurements. These results reveal that quenching of pterin fluorescence by HO is a purely dynamic quenching. Knowledge of the K SV and τ F values permitted calculation of the bimolecular rate constants for the quenching of fluorescence by HO (k q ). An average of the values obtained from plots of I Fo /I F vs. HO and of τ Fo /τ F vs. HO was used for K SV . These results are also shown in Table 4. The values of k q decrease in the order k q (FPT) > k q (PT) > k q (CPT).
Assuming that no acid-base equilibrium occurs in the excited state (Section 3.3), the dependence of the fluorescence lifetimes on the pH in the range 4-13 results from the reaction scheme shown in Fig. 7.
If the steady-state hypothesis is applied to the excited species, the following expression relating τ F to [H ] can be deduced: where k a and k b are the rate constants of fluorescence decay for the acid and base forms, respectively; the other parameters have the same meaning as previously defined.
Since all the constants in eqn. (3) have been determined for PT, FPT and CPT (Tables 1, 3 and 4), the evolution of 1/τ F as a function of pH may be predicted using this equation. The experimental and predicted values for 1/τ F are in good agreement, as shown in Fig. 8. Therefore, this result supports the mechanism proposed in Fig. 7.

Conclusions
The fluorescence properties of pterin (PT), 6-carboxypterin (CPT), 6-formylpterin (FTP) and folic acid (FA) have been studied in aqueous solution over the pH range 4-13. Under these conditions, the pterins participate in an acid-base equilibrium that involves the amide group (acid form) and the phenolate group (base form).
The fluorescence emission spectra of the four pterins showed a dependence on the pH, the emission spectra of the base forms being red shifted in comparison with the spectra of the acid forms. The normalized emission spectra remained unchanged, irrespective of the excitation wavelength, suggesting that only the lowest excited singlet state contributes to the fluorescence. The fluorescence quantum yields (Φ F ), however, depend on the excitation wavelength, decreasing at shorter wavelengths. Values of Φ F (determined by exciting in the absorption band of lowest energy) and lifetimes (τ F ) for the acid forms were higher than the corresponding values for the base forms, and were considerably affected by the nature of the substituent at position 6 on the pterin moiety. In particular, FA showed very small Φ F values suggesting that its relatively long chain substituent might act as an "internal quencher", enhancing the radiationless deactivation of the singlet excited state.
No significant differences were observed between the pK a values of PT and CPT as determined from the changes in integrated fluorescence intensities and those obtained from spectrophotometric titration, suggesting that the pK a values of the equilibria between the excited singlet states are the same as the corresponding pK a values of the ground states, or that deactivation of the excited singlet state of the acid (base) form is much faster than its deprotonation (protonation).
Above pH 11, the fluorescence of the pterins was efficiently quenched by hydroxide ions (HO ). The Stern-Volmer quenching constants (K SV ) obtained from the analyses of fluorescence lifetimes and intensities were comparable showing that quenching of pterin fluorescence by HO is a dynamic process. The differences in the values of the bimolecular quenching rate constants k q (Table 4) may be related to the different size and charge of the molecules. In fact, the rate constant for a diffusion-controlled process for charged species A Z A and B Z B , with diffusion coefficients D A and D B and apparent molecular radii r A and r B , can be expressed as 33 : with D AB = D A D B , R = r A r B and r 0 =7.1 × 10 10 m.
QSAR calculations allow the estimation of the size of the molecules investigated using the semi-empirical AM1 method: thus, we calculated value of 452.7, 519.2 and 509.2 Å 3 for PT, FPT and CPT, respectively. The apparent size of the HO anion was assumed to remain constant. Therefore, for a diffusioncontrolled process, a lower size should lead to a smaller secondorder quenching rate constant if the ionic species have identical charges. As PT has a smaller volume than FPT and the same charge of 1, it is expected to show a lower k q value than FPT, as observed. In the case of colliding species having similar sizes, the charge difference becomes the dominant factor. Therefore, the k q value for CPT, which has a formal charge of 2, is expected to be lower than that observed for the quenching of monovalent anions by HO .