Femtosecond to Second Studies of a Water-Soluble Porphyrin

Feb 10, 2012 - The interactions of 5,10,15,20-tetrakis(4-sulfonatophenyl)-porphyrin (TSPP) with a quaternary ammonium modified β-cyclodextrin and hum...
5 downloads 4 Views 516KB Size
Download PDF
Article pubs.acs.org/Langmuir

Femtosecond to Second Studies of a Water-Soluble Porphyrin Derivative in Chemical and Biological Nanocavities Yilun Wang,† Boiko Cohen,† Laszlo Jicsinszky,‡ and Abderrazzak Douhal*,† †

Departamento de Química Física, Facultad de Ciencias Ambientales y Bioquímica, and INAMOL, Universidad de Castilla-La Mancha, Avenida Carlos III, S/N, 45071 Toledo, Spain ‡ Department of Synthesis, CYCLOLAB R&D Lab. Ltd., IX. Illatos ut 7, H-1097 Budapest, Hungary S Supporting Information *

ABSTRACT: The interactions of 5,10,15,20-tetrakis(4-sulfonatophenyl)-porphyrin (TSPP) with a quaternary ammonium modified β-cyclodextrin (QA-β-CD) and human serum albumin (HSA) protein in aqueous solutions at pH 7 were studied using steady-state, stopped-flow, and femtosecond to millisecond spectroscopy. TSPP forms 1:1 and 1:2 complexes with QAβ-CD (K1 = 1.9 × 105 M−1 and K2 = 7 × 103 M−1) at 293 K, whereas with the HSA protein only 1:1 complex (K1 = 1.7 × 106 M−1) has been found. The chemical and biological nanocavities have notable effects on the fluorescence lifetimes of the Q x state (from 9.3 to 11.1 ns in QA-β-CD and 11.6 ns in HSA). Furthermore, the rotational times (400 ps for the free TSPP, 1.6 and 19 ns for QA-β-CD and HSA protein complexes, respectively) clearly indicate the robustness of the formed entities. The confined environment does not affect much the fs dynamics (0.1−0.2 ps) of the encapsulated molecule. However, it clearly affect the ps one (1−2 ps (H2O) and 5−10 ps (QA-β-CD and HSA)). The effect of O2 on the relaxation of the triplet state of the free and encapsulated TSPP is also studied and the obtained results are discussed in light of the shielding effect provided by the chemical and biological cavities. The observed difference, longer triplet lifetime upon encapsulation, might be relevant to the efficiency of this porphyrin in photodynamic therapy. The presteady-state kinetics of the TSPP:HSA has been studied by the stopped-flow spectrometer, and a two-step model was proposed for the complexation processes. The results show the importance of the initial association step for the overall ligand recognition process. This first step occurs with rate constant of ∼4 × 105 M−1 s−1, which is about 5 orders of magnitude larger than the rate constant of the consecutive relaxation processes. We believe that our observations of molecular interaction between TSPP, QA-β-CD, and HSA protein from femtosecond to second at both ground and electronically first excited state give detailed information to improve our understanding of this kind of system and thus for a better design of drug delivery nanocarriers.

1. INTRODUCTION Porphyrins are heterocyclic macrocycles consisting of four interconnected modified pyrrole subunits. Synthetic porphyrins and their derivatives have drawn the interests of scientists from various fields and have been applied for example in photodynamic therapy (PDT),1 biosensors,2 and dye-sensitized solar cells.3 Among these applications, PDT is well-developed and has been approved for clinical use around 20 years ago. In PDT, porphyrins are usually introduced in the blood as relatively concentrated solutions (thus in aggregated forms), which may diminish their actions or even cause adverse effects. Interactions with macromolecules, such as proteins, may control the efficiency and the biodistribution of porphyrins, and therefore, it is of great importance to study this type of processes in order to formulate safe and effective drug dosages. 5,10,15,20Tetrakis(4-sulfonatophenyl)porphyrin (TSPP, Figure 1A) is a water-soluble porphyrin that has been already applied to treat cancer clinically.4 It has shown great efficiency in producing singlet oxygen, an active specie for PDT.5 Its interaction with chemical and biological nanocavities has been subject of several © 2012 American Chemical Society

studies, and it was shown that TSPP forms robust complexes with different stoichiometry depending on the experimental conditions.6,7 Cyclodextrins are good candidates to be used as drug delivery nanocarriers due to their ability to form stable inclusion complexes. During the last two decades, various kinds of cyclodextrin derivatives have been synthesized to enhance the physicochemical properties of cyclodextrin as high-performance drug carriers.8,9 Here, we used the cationic QA-β-CD ((2-hydroxy-3-N,N,N-trimethylamino) propyl-β-cyclodextrin chloride, Figure 1A) whose net positive charge helps in forming strong complexes with TSPP, through electrostatic interactions as a driving force for the complex formation.10 Human serum albumin (HSA), the most abundant protein in human blood plasma, is responsible for maintaining the osmotic pressure, buffering pH, and transporting hormones, fatty acid, Received: December 15, 2011 Revised: February 9, 2012 Published: February 10, 2012 4363

dx.doi.org/10.1021/la204949e | Langmuir 2012, 28, 4363−4372

Langmuir

Article

the fs−ps and ps−ns regime. The stopped-flow results elaborate the presteady-state dynamics of the TSPP:HSA complexation. We believe that these results could help to understand the photobehavior of porphyrins in nanocavities and thus help in improving the efficiency of photodynamic therapy by complexing with cationic cyclodextrin derivatives.

2. MATERIALS AND METHODS 2.1. Materials. TSPP (5,10,15,20-tetrakis(4-sulfonatophenyl)porphyrin) (Porphyrin System, 98%, Figure 1A), QA-β-CD ((2-hydroxy-3-N,N,N-trimethylamino) propyl-β-cyclodextrin chloride, Figure 1A) (CycloLab, 95%) and HSA (Human serum albumin) protein (Fluka Sigma-Aldrich, 99%) were used without further purification. A NaOH/KH2PO4 buffer solution (pH 7.0) was used in all cases. The buffer solution was prepared using doubly distilled water. All TSPP solutions contain 3% DMSO due to the poor solubility of TSPP in water. TSPP solutions were prepared at the concentration of ∼10−6 M except for the femtosecond fluorescence experiments, where 10−5 M solutions were used. 2.2. Experimental Setup. The steady-state absorption and fluorescence spectra have been measured using a JASCO V-670 spectrophotometer and FluoroMax-4 (Jobin-Yvone) spectrofluorometer, respectively. The picoseconds (ps) emission decays were measured using a time-correlated single-photon counting (TCSPC) system described elsewhere.14 The sample was excited by pulsed 40 ps diode laser centered at 433 nm (20 MHz repetition rate). The instrument response function (IRF) was typically 70 ps. In all cases, reduced χ2 was below 1.2 to achieve a good fit. The stopped-flow absorption measurements were done using an SX.18MV-R Stopped-Flow Reaction Analyzer (Applied Photophysics). The light source was a 150 W xenon arc lamp. The light first went through a monochromator fitted with a 250 nm holographic grating, then transmitted through a 20 μL cell cartridge with 2 mm path length and detected by a Hamamatsu R928 absorption photomultiplier. The cell had a dead time of around 1 ms. The transient traces were recorded at different time scales (500 ms and 20 s) and different wavelengths (every 2 nm between 390 to 440 nm). The kinetics was analyzed using multiexponential function and the time-resolved absorption spectra were constructed by the fitted kinetics traces. The time resolution of the system is 50 μs. Femtosecond (fs) emission transients were collected using the fluorescence up-conversion technique.15 The excitation wavelengths were 415 or 420 nm (∼0.1 nJ). To analyze the decays, a multiexponential function convoluted with the IRF was used to fit the experimental transients. The nano- to microsecond flash photolysis setup consists of LKS.60 laser flash photolysis spectrometer (Applied Photophysics) and Vibrant (HE) 355 II laser (Opotek) as a pump pulse source (5 ns duration).16 The excitation wavelengths were 415 or 420 nm. The transient decays were analyzed using exponential function and the transient absorption spectra were constructed for the experimental decay at single observation wavelength. All the experiments were done at 293 K. Further details about the experimental set-ups are given in the Supporting Information.

Figure 1. (A) Schematic representations of molecular structures of (left) TSPP and (right) QA-β-CD. (B) UV−visible absorption and emission spectra of TSPP in buffer (pH 7, black), with 10 mM of QAβ-CD (red) and 20 μM of HSA (blue) in water solutions. The excitation wavelength was 433 nm.

and drugs. The structure of the protein is well-defined by X-ray crystallography studies.11 HSA is a single polypeptide chain consisting of 585 amino acids and contains three homologous R-helical domains (I−III). The principal regions of ligand bindings to HSA are located in the hydrophobic cavities in subdomains IIA (binding site I) and IIIA (binding site II). The interactions between drugs and human serum albumin have been extensively studied by various methods.12 Recently we have shown how the interaction with HSA protein affects the excited state dynamics of 5,10,15,20-tetrakis(4-hydroxyphenyl)-21,23H-porphyrin (p-THPP). We found that the protein environment affects the B →Q y andQ y →Q x transition dynamics, as well as the lifetime of the relaxed Q x state, while the strongest effect is observed in the relaxation dynamics in the hot Qx state in HSA, which includes energy transfer to the protein in ∼1 ps and much slower solventassisted thermal equilibration component of about 20−30 ps.13 CDs are often used as drug delivery carriers, as well as, in many cases, as models for larger biological systems like the HSA protein.8,9 On the other hand, HSA protein is responsible for the transport of various ligands, including drugs, in the human blood circulatory system.11 Thus, in this work, we studied the femto- to millisecond photodynamics (fs−ms) of TSPP, in solution and complexed with QA-β-CD (Figure 1A) and the HSA protein. The results of combined steady-state absorption and emission, as well as fs-ms spectroscopies are discussed in terms of the excited state relaxation dynamics of TSPP in these media. Furthermore, the complexation between the cavities and TSPP is characterized by steady-state absorption and emission techniques and stoppedflow spectrometer. The effect of these cavities on the triplet (T1) of TSPP is studied under different experimental conditions. The emission dynamics of TSPP show a reasonable change in both

3. RESULTS AND DISCUSSION 3.1. Steady-State Measurements. Figure 1B shows the UV−visible absorption spectra of 1 μM TSPP in pH 7 buffer, 4364

dx.doi.org/10.1021/la204949e | Langmuir 2012, 28, 4363−4372

Langmuir

Article

complexed with QA-β-CD (10 mM) and with the HSA protein (20 μM) in buffer solutions. The reported value of pKa of TSPP is 4.8.17 It was shown that at acidic conditions (pH ∼2), TSPP forms J- or H-aggregates at low concentration.18,19 However, under neutral pH the monomeric form of TSPP is the predominant at low concentration. The absorption spectrum of TSPP in pH 7 buffer solution shows a typical behavior for a monomer that has a D2h symmetry with the B (Soret) band, corresponding to the S0−S2 transition, having a single peak with the maximum of absorption at 414 nm, while as a result of the symmetry breaking, the Q one, corresponding to the S0−S1 transition, splits into two bands: Qy (515 and 551 nm) andQ x (581 and 635 nm).20 No evidence was found for the existence of the protonated TSPP, which has a Soret band at 434 nm and nonsplitting Q-band. Interaction of TSPP with QA-β-CD results in a red shift of ∼5 nm in the absorption spectra, observed for both the B (maximum at 419 nm) and the Q bands (maxima at 517, 551, 586, and 641 nm). This shift is due to the confinement effect and the change of polarity of the local environment of TSPP as a result of the encapsulation by the hydrophobic pocket of the QA-β-CD. Similar red shift has been reported for other porphyrin-CD complexes.21,22 Figure 2A shows the change in the Soret band (414 nm) in the absorption spectrum of TSPP in buffer solution at different concentration of QA-β-CD. The figure clearly shows the existence of two isosbestic points at 416 and 418 nm, which is an evidence for the formation of 1:1 and 1:2 complexes of TSPP and QA-β-CD. This result is in agreement with previous absorption and NMR studies on the complexation behavior of TSPP with similar CD cavities, such as β-CD and heptakis(2,3,6-tri-O-methyl)β-CD (TMe-β-CD).7,23 To obtain the binding constants for these complexes, we used eq 1:24 A = (A 0 + K1A1[QA − β − CD] + K1K2A2 [QA − β − CD]2 )/(1 + K1[QA − β − CD] + K1K2[QA − β − CD]2 )

(1)

Figure 2. (A) UV−visible absorption spectra of TSPP in buffer (pH 7) upon addition of different amounts of QA-β-CD (0−10 mM). Note that the absorption spectra at higher concentrations of QA-β-CD (>1 mM) are overlapping. (Inset) Zoom of the absorbance variation of TSPP with QA-β-CD concentration observed at 414 and 419 nm. The solid lines are from the best fit using eq 1, assuming the formation of a 1:2 complex. The complete titration analysis is presented in Figure S1 in the Supporting Information. (B) UV−visible absorption spectra of TSPP in pH 7 buffer upon addition of different amounts of HSA protein (0−20 μM). (Inset up) Absorbance variation of TSPP with HSA protein concentration observed at 414 and 422 nm. The solid lines are from the best fit using eq 2, assuming the formation of a 1:1 complex. (Inset down) Job’s plot of absorption intensity of TSPP:HSA complexes at 422 nm upon increasing the molar fraction of HSA ( f HSA). The sum of the concentrations of HSA and TSPP was kept constant at 1 μM. (C) Proposed molecular structures of the complexes of TSPP, QA-β-CD and HSA protein.

where A is the absorbance of the system at different concentration of QA-β-CD, [QA-β-CD] is the concentration of QA-β-CD in the solution, A0 is the absorbance of TSPP without QA-β-CD, A1 is the absorbance of TSPP forming 1:1 complex with QA-β-CD (this parameter was left free during the fitting process), and A2 is the absorbance of TSPP forming 1:2 complex with QA-β-CD. From the best fit we obtained K1 = (1.9 ± 0.1) × 105 M−1 and K2 = (7 ± 1) × 103 M−1. Several studies on complexes of the free-base porphyrin and β-CD derivatives have shown formation of 1:1 complexes.21,22 However, in a recent NMR study it was shown that TSPP forms stable trans-type 1:2 complexes with TMe-β-CD with the binding constant in water being too large to be determined.7 Furthermore, 1:1 and 1:2 complexes of TSPP with the native β-CD with comparable binding constants as those found in the current study have been reported. It should be noted that the reported value for K1 for the native β-CD is an order of magnitude smaller (1.7 × 104 M−1) than the corresponding value for QA-β-CD (1.9 × 105 M−1).7 This shows that the presence of the positive charges in the latter favors the formation of more stable complexes with comparison to the unmodified β-CD. Upon complexation with HSA protein (Figure 2B), a red shift was also observed for both the B (maximum at 422 nm)

and Q bands (maxima at 518 nm, 553 nm, 591 and 647 nm). The observed red shift is due to the interaction of TSPP with the amino acid residues of the protein pocket. It has been reported that the nonprotonated nitrogen atoms of the porphyrin can interact with the hydrogen atoms of Ser489 and Lys414 residues of the HSA protein.25 Similar red shift has been observed by others for systems consisting of porphyrins and protein complexes.13,26 Figure 2B shows the change in the absorption spectra of TSPP in buffer solution at different concentrations of 4365

dx.doi.org/10.1021/la204949e | Langmuir 2012, 28, 4363−4372

Langmuir

Article

the HSA protein. The figure shows one isosbestic point at 419 nm, which indicates 1:1 stoichiometry. To get the stoichiometry of the complex, we used the Job’s plot technique, plotting the change in the absorbance of the complex versus the molar fraction of HSA (f HSA) in the solutions (inset of Figure 2B). The sum of the concentrations of TSPP and HSA is kept constant at 1 μM, which results in a change only in the molar fraction. The plot clearly shows that the maximum of the absorbance intensity appears at f HSA = 0.5, indicating that the complex has 1:1 stoichiometry. We used eq 224 to determine the binding constant between TSPP and the HSA protein: A=

(A 0 + K1A1[HSA]) (1 + K1[HSA])

(2)

where A is the absorbance of the system at different concentration of HSA, [HSA] is the concentration of HSA in the solution, A0 is the absorbance of TSPP without HSA, and A1 is the absorbance of TSPP forming 1:1 complex with HSA. The best fit gives K1 = (1.7 ± 0.1) × 106 M−1 and ΔG0 = −34.9 ± 0.1 kJ/mol at 293 K (calculated as ΔG0 = −RT ln K1). Binding constants from 104 to 107 have been reported for porphyrin and albumin complexes.25−27 Similar binding constants have been reported for the interaction of TSPP with HSA under physiological conditions.28 The high binding constant indicates the formation of a robust complex between TSPP and HSA. Figure 1B also shows also the comparison of the emission spectra of TSPP in different media after 433 nm excitation. The main features of the fluorescence spectrum of TSPP in pH 7 buffer are two peaks at 643 and 704 nm, assigned to the emission from theQ x band. TheQ x fluorescence emission of TSPP in QA-β-CD solution also exhibits two emission peaks at 645 and 709 nm, and they show a red shift (2 and 5 nm; 48 and 100 cm−1, respectively) compared to those of TSPP in the pH 7 buffer. TheQ x fluorescence emission of TSPP in presence of HSA displays two emission peaks at 649 and 716 nm and shows a red shift (6 and 12 nm; 144 and 238 cm−1, respectively) compared to those of TSPP in buffer. Upon encapsulation by QAβ-CD and by HSA protein, the ratio of intensity of the Qx (0,1) band to that of the Qx (0,0) strongly decreases. This is most probably due to a change in the electronic transition moment of TSPP upon encapsulation, and it reflects a conformational change of the sulfonatophenyl substituent. In a 1:2 complex of TSPP with CD of trans-type, the rotational motion of the phenyl rings sandwiched by the CD moieties is completely restricted, whereas the encapsulated counterparts still have probability of rotation.7 Studies have suggested that the interaction of the sulfonate groups of the encapsulated dyes can interact with the amino acid residues of the protein pocket thus affecting the behavior of the guest by restricting the rotational/ conformational freedom and/or altering its electronic distribution.27,29 Such suggestion is based on the direct observation of sulfonate azo-dyes with several amino acids.30 3.2. Stopped-Flow Absorption Measurements. In order to acquire information on the presteady-state kinetics of the binding of TSPP with QA-β-CD and HSA protein, we performed stopped-flow absorption measurements. No kinetics has been observed for the mixing of 10 μM TSPP buffer solution and 10 mM QA-β-CD solution, most probably because it is so fast that it happens within the dead time of the system (1 ms). The time-resolved spectra of the mixing of 10 μM TSPP and 200 μM HSA protein solutions are shown in Figure 3A. The difference between the spectra of TSPP in buffer alone and the

Figure 3. (A) Time-resolved stopped-flow UV−visible absorption spectra of TSPP:HSA complexes in comparison with their steady-state spectra of the complex. (B) Variation of the sum (□, A) and the product (O, B) of the inverse shorter (τ1) and longer (τ2) relaxation times with the sum of the concentrations of HSA protein and TSPP.

corresponding one to a 0.1 ms after mixing TSPP and HSA solutions indicates the occurrence of fast processes within this time range. Note that changes shorter than 0.1 ms cannot be detected by our system due to the limitation of the resolution of the apparatus (50 μs). The observed initial step can be assigned to diffusion processes that have been reported to have a rate constant around 109 M−1 s−1.31 The spectrum at 20 s after the mixing of TSPP and HSA solutions is comparable to the steady-state spectrum of TSPP in HSA solution, which means that the interaction and the consecutive structural change is over by that time. Furthermore, no change was observed between 0.1 ms and 20 s when the absorption was monitored in the range from 250 to 340 nm, and which corresponds to the HSA protein spectral absorption (data not shown). This indicates that the structural change of HSA protein upon complexation with TSPP does not occur, or may not be observed due to the dead-time of the instrumental setup (1 ms). Note, that another possible explanation for the lack of change in the absorption spectra of the HSA protein could be due to the fact that the protein is in excess and the large background contribution might be from uncomplexed HSA. The kinetics of the binding of TSPP to HSA, monitored at 414 nm for two protein concentrations, is fitted by biexponential function (Figure S2). The first component changes from 120 to 30 ms when the concentration of the HSA protein increases from 10 to 100 μM. The longer component changes from 5.1 to 3.6 s. Based on the results from the fit and the 4366

dx.doi.org/10.1021/la204949e | Langmuir 2012, 28, 4363−4372

Langmuir

Article

(assuming that the average diffusion time is slower than the rotational one).34 The molecular recognition step results in a 4nm red shift (∼230 cm−1) when compared to the free TSPP in buffer, and in a molecular extinction coefficient of the caged TSPP that is also much lower. The small value of the bimolecular dissociation constant, k−1, indicates that complex I is quite stable and robust. The relaxation steps masked by diffusion and recognition of TSPP inside the protein cavity are defined by k+2 and k−2. In the time-resolved spectra, these steps result in additional stabilization of the complex that induces additional red shift of 4 nm (∼220 cm−1) from 418 to 422 nm of the visible spectra with no change in the molecular extinction coefficient. Through relaxations, the molecular structure of TSPP is influenced by the local environment and thus it changes with the time when specific and nonspecific interactions should play a key role in making the complex II more robust. The overall equilibrium constant was calculated following eq 5

concentration dependence of the obtained components we suggest the following 2-step model: k +1

TSPP + HSA ↽ ⎯⎯⎯⇀ ⎯⎯⎯ (TSPP − HSA)I k−1 k +2

⎯⎯⎯⇀ ↽ ⎯⎯⎯ (TSPP − HSA)II k−2

where k+1 is the bimolecular association constant, k−1 is the dissociation constant. k+2 and k−2 are the rate constants for the formed complex relaxation steps. Complexes I and II are conformational isomers of the complexes between TSPP and the HSA protein. The association of TSPP and HSA protein leads to the formation of complex I (k+1 and k−1), which evolves into complex II through conformational relaxation and structural stabilization (k+2 and k−2) Similar model has been suggested by others for the binding interactions of different ligands with proteins; and the rate constants can be derived from the following eq 3 and 4, assuming that the first step equilibrates much faster than the others:32,33

k k K ov(kinetics) = +1 +2 k−1 k−2

(5)

and it has a value of (1.6 ± 0.6) × 106 M−1, which is in a good agreement with the equilibrium constant, (1.7 ± 0.1) × 106 M−1, obtained from the steady-state titration experiments. 3.3. Picosecond Time-Resolved Emission Measurements. To get information on the excited-state dynamics of TSPP in these media, we measured the fluorescence decays by exciting at 433 nm and observing at 640 nm. The results are shown in Table 1 and the comparison of the decays is shown in Figure S3. For the buffer solution, the biexponential fit gives a longer component around 9.3 ns (58%) and a shorter one of 1.4 ns (42%). The longer time constant can be assigned to the fluorescence lifetime of the TSPP monomer. This is in agreement with previously published results, where the reported fluorescence lifetime of TSPP in water is 10.3 ns.35 However, it was found that in pH 7 phosphate buffered saline solution this component decreased to 9.5 ns, in agreement with our results.36 This phenomenon was attributed to the effect of the ionic strength in the phosphate buffer. The contribution of the shorter component increased from 42% at low concentration (1 μM) to 69% upon increasing the concentration (10 μM). Hence we assign this component to formation of aggregates. Similar behavior has been reported for different porphyrins in solution.13,18,37 TSPP easily forms J- and H-aggregates at acidic pH, whereas in neutral water it is present mostly in its monomeric

1 1 + = k+1(cHSA + cTSPP) + k−1 + k+2 + k−2 τ1 τ2 (3)

1 1 × = k+1(cHSA + cTSPP)(k+2 + k−2) + k−1k−2 τ1 τ2 (4)

The results from the fit are shown in Figure 3B, and the obtained rate constants are k+1= (2.8 ± 0.2) × 105 M−1 s−1, k−1 = (4.9 ± 0.4) s−1, k+2 = (3.1 ± 0.3) × 10−1 s−1 and k−2 = (1.1 ± 0.2) × 10−2 s−1. cHSA and cTSPP are the concentrations of the HSA protein and TSPP, respectively. k+1 has a value that is 5 orders of magnitude bigger than the remaining rate constants, which can be attributed to the molecular recognition processes between both entities. Time scale of this kind of molecular recognition has been suggested to be ∼100 μs at concentrations of 10−4 M and hence, the time-resolved absorption spectrum at 0.1 ms could be assigned to the structure of TSPP:HSA complexes just after recognition.34 Notice that time might be limited by the time resolution of the used stopped-flow apparatus (50 μs). The two rate constants critical for the final recognition are those describing the reorientation of the ligand in its solvent shell and the one assigned to a single encounter

Table 1. (Top) Values of the Fluorescence Lifetimes (τi), Fluorescence Rotational Times (φi), and Their Pre-Exponential Factors (Ai) Obtained from Single- or Multi-Exponential Fits of the Fluorescence Emission and Anisotropy Decays of TSPP in Indicated Media and (Bottom) Values of Time Constants (in μs) of the Exponential Functions Used in Fitting the Laser Flash Photolysis Transients of TSPP in Neutral Water, Aqueous QA-β-CD, and HSA Solutions Having Different Oxygen Content fluorescence lifetimes

fluorescence rotational times

medium

τ1/ns

A1/%

τ2/ns

buffer pH 7 QA-β-CD HSA

9.3 11.1 11.6

58 86 90

1.4 1.3 1.9

deoxygenated buffer pH 7 QA-β-CD HSA

A2/% 42 14 10 triplet lifetimes

φ1/ns

A1/%

φ2/ns

A2/%

0.41 0.41 0.41

100 83 61

1.6 19.1

17 39

as prepared

τ1/μs

A1/%

τ1/μs

A1/%

88 700 1900

100 100 100

1.5 8.3 4.8

100 100 45

τ2/μs

26 4367

O2 saturated A2/%

τ1/μs

A1/%

τ2/μs

A2/%

55

0.4 1.9 2.1

100 100 32

8.9

68

dx.doi.org/10.1021/la204949e | Langmuir 2012, 28, 4363−4372

Langmuir

Article

form.18,19 However, the formation of aggregates at physiological pH in buffer solutions is still possible and in similarity with the decrease in the fluorescence lifetime of the monomer it was explained with the increase in the ionic strength of the solution due to the presence of phosphate ions.36 The fluorescence decays of the TSPP (1 μM) complexes with QA-β-CD (10 mM) and the HSA protein (20 μM) are biexponential in similarity with the decays in pH 7 buffer solutions, and the fit gives two components of ∼1.5 and ∼11 ns. However, the relative contribution is different from what is observed for the free TSPP. The relative contribution of the long time component increases from 58% in pH 7 buffer solution to 86 and 92% in presence of 10 mM QA-β-CD and 20 μM HSA protein, respectively. This increase is concomitant with notable increase in the lifetime from 9.3 ns for the free TSPP to 11.1 and 11.6 ns when interacting with QA-β-CD and HSA protein, respectively. The decrease in the relative contribution of the short time component, assigned to the presence of aggregates is in agreement with previous reports, where the interaction with CD and protein cavities has been found to reduce the formation of dimers and aggregates.10,13,18,37 Furthermore the increase in the monomer lifetime can be explained with the screening effect of the nanocavities. Thus, the encapsulation of the dye by the hosts does not have prominent effect on the lifetime of the relaxed Q x state. 3.4. Picosecond Time-Resolved Emission Anisotropy Measurements. In order to get information on the rotational times of TSPP, TSPP/QA-β-CD and TSPP/HSA complexes, we performed ps time-resolved emission anisotropy measurements exciting at 433 nm and observing at 640 nm. The results are shown in Table 1 and the comparison of the decays is shown in Figure S4. For buffer solution, a single-exponential decay with a rotational time of φ = (410 ± 40) ps was recorded. Values between 377 to 500 ps were obtained with the same dye under different excitation, pH and solvent conditions.10,36,38 With the molecular modeling of TSPP, we calculated the hydrodynamic volume as 760 Å3. Comparable value has been reported by others.19 Using hydrodynamic theory, we got φ = 1.1 ns and 430 ps under stick- and slip-boundary limits, respectively.39 The rotational time we got from the experiments is closer to the value calculated under slip boundary conditions. This suggests that the rotational motion of the molecule is only influenced by the displacement of the solvent molecules. Upon complexation with 10 mM QA-β-CD, the anisotropy decay displays a biexponential behavior with φ1 = (410 ± 40) ps (83%) and φ2 = (1.6 ± 0.3) ns (17%). At 10 mM of QA-βCD, the equilibrium between TSPP and QA-β-CD strongly shifts to the formation of 1:2 complexs. Using the longer rotational time, we calculated the hydrodynamic volume to be 3200 Å3. We used molecular modeling to calculate the volume of 1:1 and 1:2 complex and we got 2400 and 3100 Å3, respectively. These calculations indicate that the 1:2 complex of TSPP and QA-β-CD contributes to the rotational motion because the hydrodynamic volume calculated by the experimental rotational time agrees with the volume of 1:2 complex calculated by molecular modeling. Therefore, the first component is due to “monomer-like” TSPP in solution, while the second one is due to the global rotational motion of the 1:2 complexes in agreement with the steady-state results. Using 10 mM QA-β-CD and 1 μM of TSPP, we estimated from the obtained equilibrium constant values, 98% of TSPP forms 1:2 complex, 1.9% forms 1:1 complex and 0.1% exists as free drug

in the water solution. The excitation wavelength is 433 nm, where mainly 1:2 complexes absorb and the free TSPP very weakly absorbs. Yet the anisotropy measurements reveal 83% contribution of the component assigned to the population of free TSPP molecules. Several contributions to this component can account for this behavior. One can not assume a homogeneous distribution of the formed 1:2 complexes. In other words, some of the phenyl groups of TSPP can be weakly docked inside the QA-β-CD cavities, and in that case TSPP can rotate within their cavities, giving rise to a free-like component in the overall anisotropy. Another possibility is the interaction between the positively charged QA-β-CD and the negatively charged TSPP, thus forming complexes through electrostatic forces, where the TSPP might not be included in the cavities of QA-β-CD. Finally, it should be taken into account that the formation of the 1:2 complexes might induce fast dephasing, which will affect the early time behavior of the anisotropy decay (vide infra). In the presence of 20 μM HSA, the anisotropy decay significantly changes. It displays a double-exponential behavior, with rotational time φ1 = (410 ± 40) ps (61%) and φ2 = (19 ± 1) ns (39%). The first component is due to the “freelike” TSPP in solution, while the second one is assigned to the global rotational motion of the TSPP/HSA complex. The assignment is in good agreement with previous study of anisotropy of HSA where a rotational time of 23 ns has been reported.40 The result indicates that the TSPP/HSA forms a robust complex in agreement with the large binding constant obtained from the steady-state measurements. The equilibrium constant acquired from the static experiments suggests that 97% of TSPP (1 μM) form complexes with HSA in the presence of 20 μM HSA, and the excitation wavelength was set at 433 nm where mainly the complexes absorb. However, the anisotropy results show that 61% of the contribution is arising from “free-like” TSPP. In similarity with the observations for the anisotropy of the TSPP/QA-β-CD complexes, this type of behavior can be explained with the high heterogeneity of the protein system. We can not exclude that some of the TSPP molecules are forming complexes on the surface of the protein moiety, instead of being inside its pocket, and they may rotate like the TSPP molecules in water. Moreover, TSPP can rotate inside the protein cage in a “loose” type complex. The initial anisotropies, r0, for TSPP in buffer, QA-β-CD, and HSA are 0.15, 0.26, and 0.13, respectively. In the case of TSPP monomer at neutral pH, the transition moment of B (excited S2 state) and Q bands (emitting S1 state) is split into two components, x and y. Even though the emitting Q state is the x-polarized, due to significant overlap with the B states and the high probability that these are simultaneously excited, the relative polarization of the absorption dipole is not easily quantified.36 Furthermore, fast dephasing can also occur in the relaxation from S2 to S1 state that will influence the value of the initial anisotropy.41 Recently we have shown that the interaction of a chromophore (trisodium 8-hydroxypyrene-1,3,6trisulfonate (pyranine, HPTS)) through its SO3− groups with the amino acid residues of the HSA protein cavity can influence the mixing between the 1Lb and 1La states and in that manner affect the value of the initial anisotropy.27 The initial value of the anisotropy in the 1:2 complexes of TSPP with QA-β-CD is higher (0.26) in comparison with the values of the free monomer (0.15) and of the encapsulated by the HSA protein one (0.13). This most probably is due to a geometry change in the porphyrin ring due to the complete restriction of rotational 4368

dx.doi.org/10.1021/la204949e | Langmuir 2012, 28, 4363−4372

Langmuir

Article

motion of the phenyl rings sandwitched between the two CD entities, while the encapsulated phenyl rings still have some degree of rotational freedom.7 3.5. Femtosecond Time-Resolved Emission Measurements. To get information on the ultrafast dynamics, we performed fs-ps emission experiment. We excited TSPP (10 μM) in buffer at 415 nm and TSPP in QA-β-CD (10 mM) and HSA protein (20 μM) at 420 nm. The femtosecond excitation puts the excited TSPP at low vibrational levels of the S2 electronic state. We gated at different wavelengths from 510 to 720 nm. Representative decays are shown in Figure 4A. The emission

Table 2. Values of Fluorescence Lifetimes (τi) and PreExponential Factors (Ai) Obtained from a Single- or MultiExponential Fit of the Emission Decays of TSPP in Neutral Water, Aqueous QA-β-CD, and HSA Solutions λobs/nm

decay

buffer pH 7

QA-β-CD

HSA

510

τ1/fs (A1) τ2/ps (A2) τ3/ns (A3) τ1/fs (A1) τ2/ps (A2) τ3/ns (A3) τ1/fs (A1) τ2/ps (A2) τ3/ns (A3) τ1/fs (A1) τ2/ps (A2) τ3/ns (A3) τ1/fs (A1) τ2/ps (A2) τ3/ns (A3) τ1/fs (A1) τ2/ps (A2) τ3/ns (A3) τ1/fs (A1) τ2/ps (A2) τ3/ns (A3) τ1/fs (A1) τ2/ps (A2) τ3/ns (A3)

90 (0.98) 1.5 (0.01) 9.3 (0.01) 100 (0.93) 1.6 (0.06) 9.3 (0.01) 140 (0.85) 1.7 (0.14) 9.3 (0.01) 100 (0.85) 1.3 (0.08) 9.3 (0.07) 190 (0.59) 1.7 (0.03) 9.3 (0.38)

50 (0.98) 5.0 (0.01) 11.1 (0.01) 100 (0.93) 5.0 (0.06) 11.1 (0.01) 110 (0.87) 5.0 (0.11) 11.1 (0.02) 130 (0.85) 5.4 (0.06) 11.1 (0.09) 150 (0.68) 5.5 (0.11) 11.1 (0.21)

50 (0.98) 5.0 (0.01) 11.6 (0.01) 70 (0.91) 5.0 (0.08) 11.6 (0.01) 110 (0.89) 5.0 (0.10) 11.6 (0.01) 110 (0.89) 5.9 (0.10) 11.6 (0.01) 100 (0.62) 6.7 (0.32) 11.6 (0.06)

1.8 (0.04) 9.3 (0.96)

5.0 (0.11) 11.1 (0.89)

5.1 (0.16) 11.6 (0.84)

1.8 (0.04) 9.3 (0.96)

10.0 (0.07) 11.1 (0.93)

10.0 (0.16) 11.6 (0.84)

1.8 (0.09) 9.3 (0.91)

10.0 (0.09) 11.1 (0.91)

10.0 (0.09) 11.6 (0.91)

550

590

620

640

650

700

720

have been reported for free base porphyrins, and they were assigned to the ultrafast internal conversion from B state toQ y state.13,20,43 In the 570−650 nm region, which is due to the emission from the Q y state, the fast decaying component increases to 140−190 fs. This time constant can be assigned to the internal conversion processes of Q y → Q x. For the free base porphyrin, 5,10,15,20-tetrakis-(4-hydroxyphenyl)-21,23Hporphyrin (p-THPP) similar time constant for the internal conversion has been reported.13 Fast dephasing between two nearly degenerate states with perpendicular transition dipole moments, as is the case of TSPP, can take place on ultrafast time scale, which will result in a significant decrease in the initial anisotropy. Consecutive relaxation processes from the B state to Qy and then to Qx states within the initial 200 fs have been proposed for TSPP in water, as well as for other structurally similar free base porphyrins.13,43 The second component in the emission transients has a value of 1.3−1.9 ps with a small amplitude around 10% in the whole spectral region. In similarity with other ultrafast studies, we assign this component to a vibrational relaxation/cooling in the Q x state.43 A mechanism involving three decaying processes for the ultrafast relaxation of free base porphyrin systems has been proposed that involves intramolecular vibrational energy redistribution (IVR) with time constants of 100−200 fs, a single ps component, assigned to elastic collisions with solvent molecules, and finally 10−20 ps one attributed to thermal equilibration or cooling by energy transfer to the solvent.13 For the observed time window, the longer component of the transients, fixed to 9.3 ns in the 570−720 nm range and obtained from the TCSPC measurements, is the value of the fluorescence lifetime of the relaxed Q x state and is assigned to intersystem crossing. Scheme 1 shows the energy diagram of the above photodynamics.

Figure 4. (A) Femtosecond emission transients of TSPP in (□) buffer (pH 7), (O) in presence of 10 mM QA-β-CD, and (Δ) in 20 μM HSA, observed at (1) 510, (2) 620, (3) 640, and (4) 700 nm. The excitation wavelength was 415 nm for free TSPP and 420 nm for the other ones. (B) Normalized transient absorption decays of TSPP in (□) buffer (pH 7), (O) in presence of 10 mM QA-β-CD, and (Δ) in 20 μM HSA under different concentrations of molecular oxygen: (1) as prepared (without purging N2 or O2), (2) deoxygenated (purging with N2), and (3) saturated with molecular oxygen (purging with O2). The excitation and observation wavelengths were 415 and 470 nm, respectively. The solid lines represent the best fits using single- or twoexponential functions.

transients were fitted using multiexponential function convoluted to the IRF obtained. The results of the fit are given in Table 2. 3.5.1. TSPP in Aqueous Buffer Solution. The fastest fluorescence decay time (100 ± 50 fs) is obtained at 510 nm, which corresponds to the emission from the Soret (B) band. Similar time constant (200 fs) for the decay of the B band has been reported for the same molecule.42 The difference in the two values can be explained with the different experimental conditions in the two studies. Comparable times (even