Vitamin B2-sensitized degradation of the multifunctional drug Evernyl, in the presence of visible light – microbiological implications

Materials

Evernyl, riboflavin, deuterium oxide 99.9% (D₂O), superoxide dismutase (SOD) from bovine erythrocytes, catalase (CAT) from bovine liver, Rose Bengal (RB), furfuryl acetate (FAc) and furfuryl alcohol (FFA) were purchased from Sigma-Aldrich (MO, USA). Hydrogen peroxide (H₂O₂), 100 vol. was obtained from Parsol (Argentina). Ascorbic acid [(R)-3,4-dihydroxy-5-((S)-1,2-dihydroxyethyl)furan-2(5H)-one, Asa] was from Cicarelli (Argentina). Sabouraud dextrose agar medium was purchased from Laboratorios Britania S.A (Argentina). All these chemicals and the culture medium were used as received. Water was triply distilled. All the measurements were carried out at room temperature employing freshly prepared solutions. The working pH was 7±0.1, employing phosphate buffer.

Absorption and fluorescence measurements

Ground state absorption spectra were registered employing a Hewlett Packard 8453 diode array spectrophotometer.

Steady-state fluorescence was measured with a Spex Fluoromax spectrofluorometer at 25±1°C in air-equilibrated solutions and fluorescence lifetimes were determined by a time-correlated single photon counting technique (SPC) on an Edinburgh FL-900CD instrument, equipped with a blue LED (PicoQuant PLS 450). Excitation and emission wavelengths for Rf were 450 and 515 nm, respectively.

The rate constants for interactions of the excited singlet state of the Rf (¹Rf*) with Ev were determined employing a classical Stern–Volmer treatment of the data (eqs. 1 and 2).

Where I, I₀, ¹τ and ¹τ₀ are the respective steady-state intensities and lifetimes for Rf fluorescence in the presence and in the absence of Ev, being K_Q=¹k_q¹τ₀

Stationary photolysis

Stationary aerobic photolysis of aqueous solutions containing Ev and Rf was carried out in a homemade photolyzer, using blue (λ_max=467 nm) or green (λ_max=510 nm) LEDs as photoirradiation sources, for Rf- and RB-sensitization, respectively.

Oxygen uptake experiments

The steady state aerobic photolysis of aqueous solutions containing Ev and Rf was carried out in a homemade photolyser provided with a 150 W quartz-halogen lamp. A cut-off filter of 400 nm was used to select the irradiation wavelengths. Oxygen uptake in water was monitored with a polarographic 97-08 Orion electrode.

Kinetic measurements

The reactive rate constant for the chemical reaction of O₂(¹Δ_g) with Ev (k_r) was determined using a relative method (eq. 3), for which the knowledge of the reactive rate constant for the photo-oxidation of a reference compound R is required:

where slope and slope_R are the respective slopes of the first-order plots of Ev and reference consumption. The reference compound was FAc, with a reported k_r value of 5.5×10⁷ M^–1s^–1 [15].

The total quenching rate constant for O₂(¹Δ_g) deactivation by Ev (k_t), was determined employing the already described time-resolved phosphorescence detection (TRPD) equipment [5]. Usually, 15 shots were needed for averaging so as to achieve a good signal-to-noise ratio, from which the temporal decay curve was obtained. Air equilibrated solutions were employed in all cases.

O₂(¹Δ_g) lifetimes were evaluated in the presence (τ) and in the absence (τ₀) of Ev and the data was plotted as a function of quencher concentration, according to a simple Stern–Volmer treatment (eq. 4).

D₂O was employed as a solvent in the TRPD experiments, in order to enlarge the phosphorescence lifetime of O₂(¹Δ_g) within temporal range attainable by our photodetector [16].

The evaluation of Ev photostability was performed employing Asa as a reference compound. The experiments were made employing 0.04 mM Rf as a sensitizer, in pH 7 buffer, under blue-light irradiation. Absorbance changes for both substrates were monitored at 266 nm.

The thermally-driven experiments on the reaction H₂O₂+Ev were performed by adding the adequate amount of a solution of 100 vol H₂O₂ to 2 mL aqueous solution of 0.25 mM Ev contained in a spectrophotometric cell, under constant magnetic stirring. The absorbance changes at 266 nm were continuously monitored and plotted vs. time elapsed after mixing.

Laser flash photolysis experiments

Transient absorption spectra were determined in Argon-saturated 0.04 mM Rf aqueous solutions using a flash photolysis apparatus provided with a 150 W Xenon lamp as analyzing light. A 355 nm Nd:YAG laser output was employed as excitation source. The detection system comprised a PTI monochromator and a red-extended photomultiplier (Hamamatsu R666). The signal, acquired and averaged by a digital oscilloscope (Hewlett-Packard 54504A), was transferred to a PC via an HPIB parallel interface, where it was analyzed and stored. ³Rf* disappearance was monitored from the first-order decay of the absorbance at 670 nm, which is a zone where the interference of other species is negligible. To avoid self-quenching and triplet-triplet annihilation, ³Rf* decay was measured at low concentrations of the vitamin (typically 0.05 mM) and low enough laser energy.

For the determination of the rate constant for interaction of ³Rf*–Ev (³k_q), the Stern–Volmer expression shown in eq. 5 was employed.

Where ³τ and ³τ₀ are the experimentally determined lifetimes of ³Rf* in the presence and in the absence of Ev.

To perform the transient absorption spectra, the signals were fitted at each wavelength employing a bi-exponential decay function. The lifetime values were shared in the global fit using origin 8.0 (Originlab Corp., Northampton, MA, USA) software. The spectra were obtained by plotting the pre-exponential factors (a_i) for each lifetime using eq. 6 with i=2 [17, 18]:

Viability experiments on Candida albicans

Yeast strain and culture conditions: CA strain was obtained from a human skin clinical infection. The fungal strain was maintained in a Sabouraud dextrose agar Petri dish (Britania, Buenos Aires, Argentina) at 2°C for the time that the experiment lasted. In order to achieve a standard amount of approximately 10⁸ cells/mL the strain was cultured overnight in Sabouraud dextrose broth (Britania, Buenos Aires, Argentina) in orbital agitation at 37°C. Quantification of cells was performed both by Neubauer chamber counting and viable cell counting on the Petri dish surface.

Photosensitizer: Rf was used as sensitizer. The appropriate volume to achieve an absorbance of 0.3 at 445 nm in the test tube was added by filtration (0.2 μm filter, Merk Millipore, Darmstadt, Germany).

Ev stock solution: The Ev ethanolic solution was added to vials tubes in a concentration of 250 ppm, to maintain Ev concentration below the MIC (300 ppm), which was previously determined (data not shown).

Irradiation source: a LED light array system (FOG Electrónica, Argentina) was constructed using 32 LED units (3.3 V, 20 mA each one) that emitted blue light (λ=463–471 nm). An laser power meter (UNO, Gentec-Eo) was used to determine the LED irradiance at 465 nm. A value of 47.8 μW/cm² was measured for each blue LED. The light array was disinfected previously to each irradiation test using ethanol–water solution (70% v/v).

Effect of Ev and Rf on the viability of CA after blue LED light irradiation: to investigate the in vitro antimicrobial effects of blue light, Rf and Ev, alone and combined, on CA assays were performed in triplicate. A 100 μL aliquot of the overnight inoculum dilution (10³ CFU/mL) was placed inside sterile vials tubes containing 2.5 mL of sterile distilled water. One vial tube was used per each treatment. The fungal viability after treatment was assessed by viable cell counting using Petri plate surface method. Sabouraud dextrose agar was used as culture medium. CFU/mL were counted after 24 h of incubation at 37°C in dark.

In order to compare the antifungal effects of blue light, Ev and Rf, five treatments were performed as shown in Table 1. Each treatment was assayed in triplicate.

Statistical treatment of data: statistical analysis of the experimental data was performed using one-way analysis of variance (ANOVA) and Origin 8.0 software (Originlab Corp., Northampton, MA, USA). All results showed homogeneity of variance. Mean comparisons were carried out with the Tukey test, which retains the overall significance level at 5% (p<0.05) [19].

Dark complexation Rf-Ev and interaction ¹Rf*-Ev

The absorption spectrum of Rf vs Rf+Ev shows the presence of a new band at 426 nm, evident in Fig. 1 (inset). Considering that no absorption from Ev is expected over 350 nm, the species absorbing in this region should be ascribed to the Rf-Ev dark complex [process (1)], being K_ass the complex association constant.

Fig. 1:

Stern–Volmer plots for the steady state (◆) and time-resolved (●) fluorescence quenching of ¹Rf* by Ev. Inset: (a) Absorption spectrum of 0.031 mM Rf+0.025 M Ev in water; (b) baseline.

Rf presents a fluorescence emission band centered at 515 nm, with a reported fluorescence quantum yield (Φ_F) of 0.25 [3]. In the presence of Ev, the fluorescence quenching ¹Rf* produces a decrease in the steady state emission intensity without change in shape of the fluorescence spectrum. The Stern–Volmer plots obtained from these measurements presents a neat positive curvature, typical for cases in which a fluorescent species is simultaneously affected by dark association in its ground state and by collisional quenching of its excited singlet state (Fig. 1) [5, 23]. The dynamic contribution was evaluated by SPC technique, obtaining a value of ¹k_q=4.5 × 10⁹ M^–1s^–1 [process (7), scheme 2]. The apparent association constant value for the system, was derived from the modified Stern–Volmer equation I₀/I=(1+K_D[Q])(1+K_ass[Q]) [5, 23], where K_D=¹k_q¹τ₀ accounts for the dynamic component of the fluorescence quenching, being ¹τ₀ the ¹Rf* lifetime. K_D was calculated from time-resolved measurements, with ¹τ₀=5.62 ns. A value of K_ass=9.3±0.5 was obtained by non-linear least square fitting employing the mentioned modified Stern–Volmer equation.

Photoirradiation of the mixture Rf plus Ev with visible light

The blue-light irradiation of the mixture 0.041 mM Rf+0.033 mM Ev in water produces the spectral changes shown in Fig. 2. In parallel oxygen uptake was observed, upon photoirradiation of a similar mixture [curve (a) Fig. 2, inset]. These results, not observable in the absence of light and oxygen, strongly suggest transformations in both Rf and Ev with participation of electronically excited states of the vitamin, already detected for ¹Rf*. Additionally should be taken into account the possible contribution of ROS such as O₂(¹Δ_g) (processes 5) and/or O₂^•− (processes 6) generated from ³Rf* which are produced with quantum yields of 0.49 and 0.009, respectively [24].

Fig. 2:

Absorption spectra of aqueous solutions 0.041 mM Rf+0.3 mM Ev vs. 0.041 mM Rf upon irradiation, under air-saturated conditions. Numbers on the spectra represent irradiation time in minutes. Inset: Oxygen uptake as a function of blue-light photoirradiation time of aqueous solutions containing: (a) 0.047 mM+0.25 mM Ev and (b)+20 nM SOD.

Laser flash photolysis experiments. Quenching of ³Rf* by Ev

In order to investigate the possible involvement ³Rf* in the Rf-photosensitized degradation of Ev, laser flash photolysis experiments were performed.

The transient absorption spectrum of the Rf solution was fitted to eq. 6 employing the lifetime values τ₁=14.6 μs and τ₂=184 μs. From the global fit, the decay-associated spectrum after Rf excitation for τ₁ is coincident with that attributed to ³Rf* [profile (a) in Fig. 3] [25–27], while the associated spectrum for τ₂ is similar to that reported to account for the species RfH^• [profile (b) in Fig. 3] [28, 29]. Subsequently, the transient absorption spectrum of a solution of Rf, but now in the presence of 0.3 mM Ev was fitted to eq. 6 with two lifetimes. An increase in the absorption of the species RfH^• was observed, with τ₂=148 μs [profile (c) in Fig. 3], whereas the absorption corresponding to ³Rf* (with τ₁=2.9 μs), totally overlaps with that obtained for ³Rf* in the absence of Ev [profile (d) in Fig. 3]. From these results, the transient Rf spectra determined in the absence in the presence of Ev, clearly represent the species ³Rf* and RfH^•. The latter is formed after an electron transfer process [step (8)] to yield the radical Rf^•–, rapidly protonated to the neutral radical RfH^•, a species for which a pK_a=8.3 has been reported [process (10)] [28, 29].

Fig. 3:

Decay-associated spectra after excitation of 40 μM Rf in Ar-saturated aqueous solution in the absence and in the presence of 0.3 mM Ev. The spectra were calculated by globally fitting all the individual-wavelength data using eq. 6 (see experimental section) with two lifetime values. Rf solution: (•) τ₁=14.6 μs; (■) τ₂=184 μs. Rf+Ev solution: (Δ) τ₁=12.9 μs; (□) τ₂=148 μs. Inset: Stern–Volmer plot for the quenching of ³Rf* by Ev.

A systematic reduction of ³Rf* lifetime was observed in the presence of increasing concentrations of Ev, supporting the occurrence of reaction (8) between the cosmetic-additive and ³Rf*, for which a rate constant value ³k_q=6.2±0.3×10⁹ M^–1s^–1 was graphically determined as shown in Fig. 3, inset.

Involvement of O₂(¹Δ_g) in Ev photodegradation. Determination of k_t and k_r

The possible involvement of O₂(¹Δ_g) in Ev photodegradation was investigated employing the exclusive O₂(¹Δ_g)-generator RB. This xanthenic dye generates the oxidative species with a quantum yield of 0.75 in water [30]. Rf was not used in this kinetic determination in order to avoid possible interferences by other oxidative species produced by the vitamin.

Photoirradiation of the mixture of RB [A(549)=0.4]+0.033 mM Ev in water, generates spectral changes, shown in Fig. 4, in principle attributable to a O₂(¹Δ_g)-mediated chemical transformation of Ev as described by reactions (13)–(17).

Fig. 4:

Absorption spectra of aqueous RB (A(549)=0.4)+0.3 mM Ev vs. RB [A(549)=0.4] upon photoirradiation at 510 nm under air-saturated conditions. Numbers on the spectra represent irradiation time in minutes. Inset A: Stern–Volmer plot for the quenchingof O₂(¹Δ_g) phosphorescence by Ev. Inset B: First order plot for oxygen uptake by aqueous solutions of RB (Abs₅₄₉=0.4) plus 0.3 mM Ev (a) and 0.3 mM FFA (b), upon visiblelight irradiation (cut-off 400 nm). Inset C: Changes in the molar concentration of individual solutions of 0.35 mM Asa (c), and 0.35 mM Ev (b) as a function of blue-light-irradiation time in the presence of 0.04 mM riboflavin as a sensitizer, in pH 7 buffered aqueous solution. Trace (a) corresponds to changes in molar concentration of the sensitizer solution in the absence of additives. Monitoring wavelength: 265 nm.

Being k_q+k_r=k_t

The presence of Ev in the sub-mM concentration range quenches the IR phosphorescence emission of O₂(¹Δ_g), as detected by TRPD. These experiments clearly demonstrates the existence of the interaction O₂(¹Δ_g)-Ev, which may be physical in nature [process (17)] and/or reactive [process (16)]. The overall quenching constant value of k_t=(3.6±0.3) 10⁶ M^–1s^–1 was obtained for Ev (Fig. 4, Inset A). In parallel a value of (1.1±0.2)×10⁶ M^–1s^–1 was determined for the reactive rate constant k_r, employing the actinometric method described in the experimental section (Fig. 4, Inset B).

The possible role of the ROS O₂^•– and H₂O₂ in the Rf-sensitized oxidation of Ev

The rate of oxygen consumption by a photoirradiated solution of 0.040 mM Rf+0.25 mM Ev was increased when performed in the presence of 20 nM SOD, as shown in Fig. 2 (inset). The enzyme specifically produces the dismutation of the species O₂^•– through process (18) [31].

The increase in the rate of oxygen uptake induced by the presence of the enzime supports the mechanistic involvement of the species O₂^•–, as stated in Scheme 2. Nevertheless, a neat decrease in that rate should be expected for cases in which O₂^•– were the prevailing oxidative species, according to reaction (19).

It is well known that in O₂^•–-mediated mechanism, SOD can either preclude or stimulate the oxidation of a given substrate, depending on the particular mechanistic role. That is, the ultimate effect will depend on whether the ROS acts either as a final oxidative species or as an intermediate in the subsequent generation of other ROS [32].

It has been reported that the presence of SOD favors the H₂O₂-mediated oxidation of a series of biologically-relevant hydroxyl-aromatic derivatives including gallic acid and hydroquinones [32]. A similar behavior, through reaction (20), can be assumed in the case of the phenolic compound Ev,

In order to assess this possibility, the photolysis of the mixture 0.043 Mm Rf+0.25 mM Ev in the absence and in the presence of 0.04 and 0.4 nM CAT was performed. H₂O₂ is decomposed by CAT according to reaction (21).

Spectral changes of the photolyzed solution, monitored at 266 nm and similar to those shown in Fig. 2, experimented a decrease in the Ev decomposition rate as CAT concentration increased. This result, shown in Fig. 5, strongly suggests the involvement of H₂O₂ in the Rf-sensitized photodegradation of Ev.

Fig. 5:

The effect of CAT on the photosensitized oxidation of Ev at 266 nm for the visible-light irradiation of aqueous solutions containing: (■) 36 μM Rf; 36 μM Rf+0.25 mM Ev and CAT: (•) 0 mM, (▲) 0.04 nM and (▼) 0.4 nM. Inset: Changes in Abs at 266 nm for the reaction of 0.25 mM Ev and H₂O₂: (■) 3.6 mM and (•) 18 mM.

The viability of the reaction of H₂O₂+Ev was tested through a simple thermally-driven assay, by plotting the 266 nm-spectral decrease of the Ev absorption band as a function of time elapsed after the mixture H₂O₂-Ev. The decomposition rate of Ev rises with the increase of H₂O₂ concentration, as show in Fig. 5, inset.

It is important to note that the individual presence of SOD and CAT, dissolved in a D₂O solution in similar conditions as employed in the described auxiliary O₂^•– and H₂O₂ tests, did not change the lifetime of O₂(¹Δ_g). It was evaluated through TRPD using RB [A(549)=0.4] as a photogenerator of the ROS.

Evaluation of Ev photostability against ascorbic acid as a reference compound

So far, experimental evidence clearly indicates that Ev is photooxidized in aqueous solution through a Rf-photosensitized process, with participation of ROS. Nevertheless, this result is some vague in photostability terms. Since several reactive species are simultaneously involved in Ev photodegradation, it is interesting to establish a comparative standard for the whole oxidative event. Hence, in order to perform a quantitative evaluation in relative terms, the overall photodecomposition rates of Ev and Asa were evaluated. Asa, transparent to visible light and currently known as vitamin C, is probably the most extensively studied biologically-relevant substrate in relation to Rf-photosensitized degradations. This pathway has been suggested as the main cause for the in-vivo photodegradation of vitamin C [21, 33–36].

Aqueous solutions of Rf as a sensitizer in the presence of identical concentrations of Ev and Asa were individually photolized with blue-light. The respective molar degradation rates (Rate_Ev and Rate_Asa), were evaluated from the initial slopes of the degradation profiles represented in inset C of Fig. 4. Results indicate a ratio Rate_Asa/Rate_Ev~2.

Effects of Ev, blue light and Rf on CA viability

A resume of results on control runs and treatments for the in vitro assays of CA viability in the presence of different combinations of blue-light, Rf and Ev are shown in Table 1.

No inhibitory effect on the strain of CA yeast cell viability was observed when 0.03 mM Rf was added in the dark to the culture medium, in accord with previous reports [37–39]. In parallel, a MIC of 300 ppm Ev was determined for the CA strain. This result is in agreement with previous studies on compounds possessing a phenolic OH attached to the aromatic ring and able to inhibit the growth of the CA strain. [40] Hence, a concentration of 250 ppm was employed in the present series of experimental assays.

Control runs on CA strain viavility (Control 2, Table 1) were performed employing only blue-light irradiation. Results show a decrease in the strain viability with irradiation time (Fig. 6). Statistically similar results –according to Tukey test (p<0.05)– were obtained for 0.03 mM Rf without Ev and 0.03 mM Rf with Ev added (treatments 1 and 2 in Fig. 6 and Table 1).

Fig. 6:

Percentage of Candida albicans reduction in CFU/mL vs. Blue- lirradiation time in different treatments. (a): cellular suspension+light irradiation (L+ Rf- Ev-). (b): Celular suspension+light irradiation+250 ppm Ev (L+Rf–Ev+). (c): Celular suspension+light irradiation+0.026 mM Rf (L+Rf+Ev–). (d): Celular suspension+light irradiation+250 ppm Ev+0.026 mM Rf (L+Rf+Ev+). Inset: Means of Candida albicans CFU/mL. A: without irradiation and B: after 30 min of blue-light irradiation. 1: L+Rf–Ev–, 2: L+Rf+Ev–, 3: L+Rf–Ev+ and 4: L+Rf+Ev+. A different symbol, either * or ** indicates significant differences according to the Tukey test (p<0.05). Error bars represent standard deviations.

The higher reduction in CA viavility (ca. 95% after 30 min light irradiation), statitically different from run control 2 and treatments 1 and 2 occurred with treatment 3 (Table 1), it is in the simultaneous presence of light, Rf and Ev. In other words, even when photoirradiation in the absence and in the presence of Rf produces a moderate inhibition in cell viability of CA yeast, the addition of Ev, in a sub-MIC condition, constitutes the key factor for an significant inhibitory effect on the strain (Table 1 and Fig. 6).

Comparative evaluation of Ev photostability

In a recent paper we show that Asa, in the presence of Rf and visible light, participates in the generation and subsequent deactivation the ROS O₂(¹Δ_g), O₂^•–, H₂O₂ and OH^• [36]. As in the present case for Ev, the main reactive pathway is mediated by O₂^•– production after the quenching of ³Rf* by Ev, in detriment to the O₂(¹Δ_g)-mediated mechanism.

According to our results the overall rate for Rf-sensitized oxygenation of Asa is only two-fold higher than exhibited by Ev. On this basis, the latter should be also considered as a labile compound toward photoinduced oxidation.

Microbiological implications of Rf+Ev photoirradiation in the presence of CA strain

It is well known the general antimicrobial activity of Ev and its specific action against certain fungal strains [7, 8, 11, 12]. Nevertheless, it is worth to mention that after Rf-photosensitized irradiation, the antimycotic effect of this perfuming agent does not decrease; on the contrary, it shows a significant increase. In fact, more than 90% of inhibition on CA strain viability rate was reached after the simultaneous application of Ev, Rf and blue light photoirradiation.

It has been recently reported a notorious reduction in CA viability employing only blue-light irradiation [46]. This photo-inhibitory effect, checked by ourselves in control run 2 (Table 1), is attributed to the presence of CA intracellular sensitizers such as flavins and flavoproteins, which are responsible for photopromoted intracellular generation of ROS after blue light irradiation, being CA highly susceptible to H₂O₂ [46, 47].

The observed effect on the combined action of blue light, Rf and Ev can be evaluated as an increase in ca. 30% inhibitory effectiveness by comparing control run 2 and treatment 3 in Table 1. This may be due to the Ev-stimulated generation of ROS in the Rf-photosensitization, whith the consequent accumulation of H₂O₂ in the tests solutions, a process already discussed on kinetic basis in previous paragraphs.