Journal of Pharmaceutical and Biomedical Analysis 91 (2014) 144–150
Contents lists available at ScienceDirect
Journal of Pharmaceutical and Biomedical Analysis journal homepage: www.elsevier.com/locate/jpba
Tryptophan 19 residue is the origin of bovine -lactoglobulin fluorescence Jihad René Albani ∗ , Julie Vogelaer, Loïc Bretesche, Daniel Kmiecik Laboratoire de Biophysique Moléculaire, Université des Sciences et Technologies de Lille, Bâtiment C6, 59655 Villeneuve d’Ascq Cédex, France
a r t i c l e
i n f o
Article history: Received 15 October 2013 Received in revised form 13 December 2013 Accepted 17 December 2013 Available online 3 January 2014 Keywords: Beta lactoglobulin Tryptophan Lifetime Calcofluor Substructures
a b s t r a c t -Lactoglobulin consists of a single polypeptide of 162 amino acid residues with 2 Trp residues, Trp 19 present in a hydrophobic pocket and Trp 61 present at the surface of the protein near the pocket. This study aimed to characterize the respective contribution of the two Trp residues to the overall fluorescence of the protein. We added for that calcofluor white, an extrinsic fluorophore, which, at high concentration compared to that of the protein, quenches completely emission of hydrophobic Trp residue(s). The study was performed at different pHs by recording fluorescence steady state spectra and measuring fluorescence lifetimes of the Trp-residues using Single Time Photon Counting method. Our results indicate that addition of calcofluor white does not induce a red shift of the tryptophan(s) emission peak (332 nm) but only a decrease in the fluorescence intensity. This means that Trp 61 residue does not contribute to the protein emission, tryptophan emission occurs from Trp 19 residue only. Also, excitation spectrum peak position (283 nm) of -lactoglobulin is not modified upon calcofluor white binding. These results mean that structural rearrangements within -lactoglobulin are not occurring upon calcofluor white binding. Energy transfer between Trp 19 residue and calcofluor white occurs with 100% efficiency, i.e. the two ˚ This energy transfer is not Forster type. fluorophores are very close one to each other (90%) was from Sigma–Aldrich (Saint Quentin Fallavier, France). Its concentration was determined at 278 nm with the following extinction coefficients: 17.6 × 103 M−1 cm−1 [24]. In the whole manuscript, lactoglobulin concentration is expressed in monomer. Absorbance data were obtained with a Varian DMS-100S (Les Ulis, France) spectrophotometer using 1-cm pathlength cuvettes. Fluorescence spectra were recorded with a Perkin-Elmer LS-5B spectrofluorometer (Perkin-Elmer, Courtaboeuf, France). The bandwidths used for the excitation and the emission were 5 nm. The quartz cuvettes had optical pathlengths equal to 1 and 0.4 cm for the emission and excitation wavelengths, respectively. Fluorescence
spectra were corrected for the background intensities of the buffer solution. Observed fluorescence intensities were first corrected for the dilution, and then corrections were made for the inner filter effect as described [25,26]. Fluorescence lifetime measurements were obtained with a Horiba Jobin Yvon FluoroMax-4-P (Horiba-Jobin Yvon, Longjumeau, France) using time correlated single photon counting method. A Ludox solution was used as scatter. Excitation was performed at 296 nm with a nanoLED. Each fluorescence decay was analyzed with one, two, three and four lifetimes and then values of 2 were compared in order to determine the best fit. A minimal value of 2 indicates the best fit. A 2 value that approaches 1 indicates a good fit. For example, let us consider the value of 2 equal to 1.054, 1.06 and 1.1 when analysis is done with 1, 2 and 3 lifetimes, respectively. One lifetime should be considered as the best description of the decay curve since there was no real improvement in 2 value when the experimental decay was fitted with one, two, three or four lifetimes [25,27–29]. The mean fluorescence lifetime calculated is the second order mean [25] 0 = ˙fi i
(1)
and fi =
ˇi i ˙ˇi i
(2)
where ˇi are the preexponential terms, i are the fluorescence lifetimes and fi the fractional intensities. All spectral experiments were performed at 20 ◦ C in 10 mM phosphate buffer and were all performed between 3 and 5 times. 3. Results and discussion 3.1. Origin of ˇ-lactoglobulin fluorescence Fluorescence emission spectra of -lactoglobulin recorded at pH 2 and 7 display a peak located at 331 ± 1 nm (Fig. 1). Thus, at these two pHs, -lactoglobulin is not denatured, otherwise emission peak will be located at 350–355 nm. This means that -lactoglobulin global conformation is identical or very close at both pHs, although 20
d a
Fluorescence intensity (a.u.)
emission peak of -lactoglobulin mutant type (W19A) is located around 305 nm and not at 340–345 nm, as it should be if emission of the mutant protein occurs from Trp 61 residue. In fact, since Trp 61 residue is at the protein surface, its emission peak should be redshifted compared to that of Trp 19 residue and not blue shifted. Also, emission peak observed at 305 nm for the W19A mutant (ex = 290 nm) corresponds to that of tyrosine residue. Tyrosine absorbs at 290 nm and thus in order to avoid its excitation, one should excite at least at 295 nm and beyond, since only tryptophan absorbs at this wavelength. Therefore, emission spectrum of W19A mutant, displayed in [13] corresponds to that occurring from tyrosine residue and not from tryptophan. Nevertheless, if Trp 61 residue is emitting, its emission intensity should be much smaller than that indicated by the authors (20% of the total wild type fluorescence) as the result of the emission from tyrosine residues of the -lactoglobulin mutant type (W19A). Tyrosine residues emission seems to be much more important than that of Trp 61 residue since emission peak of the mutant is located at 305 nm. Also, the weak or absence of Trp 61 residue fluorescence in wild type -lactoglobulin was explained as the result of the presence of a disulfide bridge (Cys66–Cys160) in proximity to this tryptophan ˚ inducing complete fluorescence quenching of Trp (≈3.65 ± 0.15 A) 61 residue [14] and/or to the self-quenching of Trp-61 by the nearby Trp-61 of the other monomer, in the -lactoglobulin dimeric form [15]. Other authors considered that structural modification of lactoglobulin modifies the distance between disulfide bridge and Trp 61 residue inducing a significant fluorescence from this Trp residue [16,17]. The aim of the present work is to find out whether both Trp residues contribute to -lactoglobulin fluorescence or not. Timeresolved studies and static quenching were performed in presence of high concentrations of calcofluor white, a fluorophore that is specific to both carbohydrate residues and hydrophobic sites in proteins. In fact, calcofluor white is a fluorescent probe capable of making hydrogen bonds with -(1 → 4) and -(1 → 3) polysaccharides [18]. It is commonly used to study the mechanism by which cellulose and other carbohydrate structures are formed in vivo and is also widely used in clinical studies [19,20]. Also, calcofluor white binds to hydrophobic sites of proteins such as ␣1 -acid glycoprotein and both human and bovine serum albumin [21,22]. In presence of high calcofluor concentrations compared to those of proteins (ratio 10–1), extrinsic fluorophore was able to quench completely emission of hydrophobic tryptophan residue [22,23], without modifying protein structures. Thus, we were able to demonstrate for example that both hydrophobic and hydrophilic tryptophan residues of bovine serum albumin contribute to the protein emission [22]. In the present work, static and time-resolved fluorescence studies were performed on -lactoglobulin in presence of high calcofluor white concentration. Experiments were performed at pH going from 2 to 12.
145
16
e c 12
8
b 4
f
0 300
320
340
360
380
400
420
Wavelength (nm) Fig. 1. Fluorescence emission spectra of 12 M -lactoglobulin at pH 2 in absence (a) and in presence (b) of 138 M calcofluor white and of 13 M -lactoglobulin at pH 7 in absence (d) and presence (f) of 149 M calcofluor white. Difference spectra (c and e) display emission peaks located at 332 nm. ex = 295 nm. Spectra recorded at pH 7 are normalized to those recorded at pH 2.
J.R. Albani et al. / Journal of Pharmaceutical and Biomedical Analysis 91 (2014) 144–150
local structural modifications could exist. For example, at pH 2, lactoglobulin is a monomer in the molten globule state [30]. The protein is stable although partially unfolded. -Lactoglobulin contains two Trp residues, Trp 19 residue located in the hydrophobic pocket of the protein and Trp 61 residue located at the hydrophilic aperture of the pocket. Thus, if protein fluorescence occurs from both residues, recorded emission spectrum will be the result of the contribution of each residue to the total fluorescence. Position of tryptophan residue emission peak and bandwidth of the spectrum calculated at half-intensity of the emission peak, would allow knowing whether one or two tryptophan residues are emitting. Position of the emission peak of Trp residue of -lactoglobulin is located at 331 nm and spectrum bandwidth is equal to 50 nm. These two values correspond to an emission spectrum occurring from a tryptophan residue located in a hydrophobic area of the protein [31]. At pH 2, emission occurs from a monomeric -lactoglobulin [30]. When a protein contains two classes of Trp residues, one at the surface and the second in hydrophobic environment and if both tryptophans contribute to the protein emission, fluorescence emission spectrum will be the result of the contribution of each class. For example, fluorescence emission spectrum of ␣1 -acid glycoprotein recorded with an excitation wavelength of 295 nm, shows a maximum located at 331 nm and a bandwidth of 53 nm. This spectrum is typical for protein containing Trp residues in both hydrophobic and hydrophilic areas of the protein [32]. Therefore, our data on -lactoglobulin clearly indicate that protein emission occurs from one tryptophan residue located in a hydrophobic area. In order to assess our result, we measured emission spectra of lactoglobulin at high concentration of calcofluor. Addition of high calcofluor white concentrations quenches fluorescence emission intensity without modifying position of the spectrum peak (Fig. 1). The difference between the two spectra (in absence and presence of calcofluor white) yields an emission spectrum with a peak located at 332 nm and not at 340 or 345 nm. This result clearly indicates that Trp 61 residue does not contribute to the protein emission. If Trp 61 residue contributes to the protein fluorescence, addition of high calcofluor concentration compared to that of lactoglobulin (ratio 10–1), will quench completely fluorescence emission of Trp 19 residue, inducing a red shift in the emission peak as we have already shown when experiments were performed on bovine serum albumin [22]. Result obtained at pH 2 was also obtained at pH 7, where lactoglobulin is in a dimeric state. Thus, absence of fluorescence from Trp 61 residue is independent of the ratio monomer/dimer present in solution and of any possible differences in the local structure of the protein around Trp residues. Also, since identical results were obtained whether the protein is in a 100% monomeric state (pH 2) or when it is in a dimeric state (pH 7), absence of Trp 61 residue fluorescence cannot be attributed to the selfquenching of Trp 61 by the nearby Trp 61 of the other monomer, in the -lactoglobulin dimeric form, as suggested in [15]. Identical results were obtained for other pHs , although degree of fluorescence quenching of tryptophan is not the same at all pHs (data not shown). Absence of emission of -lactoglobulin Trp 61 residue can be explained by the fact that this tryptophan is in proximity of the ˚ which is supposed to be Cys66–Cys160 disulfide moiety (≈3.7 A), a strong quencher of indole fluorescence [14]. Nevertheless, one should also expect that Trp 61 residue is not necessarily excited, as the result of its close interaction with other amino acids whether cysteine disulfide bridge or other amino acids [33]. Fig. 2 displays normalized excitation spectra of -lactoglobulin in solution in absence and presence of calcofluor white at pH 2. We observe that peak position (283 nm) is not modified upon calcofluor white binding. Also, global shape of the spectra is the same
1.0
Fluorescence intensity (a.u.)
146
0.8
0.6
0.4
0.2
0.0 250
260
270
280
290
300
310
Wavelength (nm) Fig. 2. Normalized fluorescence excitation spectra at pH 2 of 12 M -lactoglobulin in absence (continuous spectral line) and in presence of 138 M calcofluor white (dashed spectral lines). Peak position of both spectra is equal to 283 nm. em = 325 nm.
whether calcofluor white is present or not. These results mean that structural rearrangements within -lactoglobulin are not occurring upon calcofluor white binding. These results were also observed at pH 7. 3.2. Origin of fluorescence lifetimes of ˇ-lactoglobulin Trp 19 residue Fig. 3 displays fluorescence intensity decay of -lactoglobulin at pH 2 in absence of calcofluor white. Decay analysis shows the presence of three fluorescence lifetimes. In fact, fluorescence intensity decay I(, t), of -lactoglobulin Trp 19 residue can be adequately represented as I(, t) = 0.140e−t/0.48 + 0.697e−t/1.49 + 0.163e−t/4.29 where 0.140, 0.697 and 0.163 are the pre-exponential factors, 0.48, 1.49 and 4.29 ns are the decay times and is the emission wavelength (330 nm) (2 = 1.054). Addition of calcofluor white does not modify fluorescence lifetimes values. In fact, in presence of 116 M calcofluor, fluorescence intensity decay can be described as I(, t) = 0.175e−t/0.73 + 0.745e−t/2.032 + 0.098e−t/5.26 where 0.175, 0.745 and 0.098 are the pre-exponential factors, 0.73, 2.032 and 5.26 ns are the decay times and is the emission wavelength (330 nm) (2 = 0.99). These three fluorescence lifetimes were also observed at pH 7 and along the emission spectrum (310–410 nm). Fluorescence decays were analyzed with one, two and three lifetimes and 2 values were compared. The best values of 2 were obtained for three lifetimes. Analysis of the decays with four lifetimes did not improve 2 values, thus fluorescence intensity decays can be best described with three lifetimes [34]. Table 1 shows the fluorescence intensity decay data of lactoglobulin at different pH and observed at 330 nm in absence and presence of calcofluor white. Three fluorescence lifetimes were observed at all pH and along the emission spectrum (310–410 nm). Fluorescence decays were analyzed with one, two and three lifetimes and 2 values were compared. The best values of 2 were
J.R. Albani et al. / Journal of Pharmaceutical and Biomedical Analysis 91 (2014) 144–150
147
Fig. 3. Fluorescence intensity decay of 23 M -lactoglobulin in 10 mM phosphate buffer at pH 2 recorded at 330 nm. ex = 296 nm. Table 1 Fluorescence lifetimes and pre-exponential values of tryptophan 19 residue of lactoglobulin, measured in absence and presence of calcofluor white at different pHs . pH
2
6
8
10
11
1 (ns) 1 + calcofluor 2 (ns) 2 + calcofluor 3 (ns) 3 + calcofluor 0 (ns) 0 + calcofluor ˛1 ˛1 + calcofluor ˛2 ˛2 + calcofluor ˛3 ˛3 + calcofluor
0.48 0.73 1.89 2.032 4.29 5.36 2.648 2.762 0.1398 0.1753 0.6975 0.7449 0.1626 0.098
0.58 0.59 1.71 1.65 5.02 4.52 2.70 2.435 0.2536 0.2783 0.6283 0.6063 0.1181 0.1154
0.67 0.45 1.83 1.6 5.12 4.73 3.10 2.968 0.2707 0.2403 0.5606 0.5758 0.1687 0.1839
0.497 0.73 1.67 2.06 4.97 5.51 3.313 3.230 0.1933 0.342 0.5804 0.5106 0.2263 0.1475
0.3 0.57 1.62 1.83 4.65 4.76 3.053 2.96 0.2159 0.287 0.579 0.5376 0.2051 0.1755
obtained for three lifetimes. Analysis of the decays with four lifetimes did not improve 2 values (Table 2), thus fluorescence intensity decays can be best described with three lifetimes. This result clearly indicates that the sole Trp 19 residue emits with three fluorescence lifetimes independently of the percentage of monomers and dimers in solution. Also, presence of these Table 2 2 values obtained for -lactoglobulin tryptophan fluorescence decay analysis for one to four lifetimes. Conditions
2 (1)
2 (2)
2 (3)
2 (4)
pH 2 pH 2 + calcofluor pH 6 pH 6 + calcofluor pH 8 pH 8 + calcofluor pH 10 pH 10 + calcofluor pH 11 pH 11 + calcofluor
8.87 8.2 13.41 12.03 15.5 18.44 18.04 17.5 14.57 17.44
1.38 1.48 1.6 1.71 1.4 2.06 1.6 1.6 1.65 1.92
1.05 0.99 1.04 1.28 0.98 1.3 1.04 1.01 1.04 1.25
1.06 1.05 1.25 0.97 1.34 1.02 1.02 1.04 1.23
three lifetimes does not depend on the presence of calcofluor white and on any eventual local or global structural modification within -lactoglobulin. Recent studies [22,33,29,35] have shown that fluorescence lifetimes of tryptophan originate from sub-structures obtained at the excited state. Pre-exponential values characterize populations of these substructures. Tryptophan free in water emits with two lifetimes equal to 0.5 and 2.78 ns, these two lifetimes are observed for tryptophan residues in proteins, whether the macromolecules contain one or more Trp residues. Thus, these two lifetimes characterize two substates or substructures of the tryptophan, reached in the excited state, independently of the environment (solvent and/or Trp – amino acids interactions). These substructures differ from an environment to another, modifying values of the corresponding lifetimes amplitudes. The third lifetime observed in proteins is the result of the interaction between tryptophan residue(s) and amino acids environments inducing a third substructure with specific emission decay parameters. Each substructure is composed by tryptophan backbone with a specific electronic cloud. Electronic distribution around tryptophan backbone differs between a substructure and another [36,37]. 3.3. Binding of calcofluor white occurs close to Trp 19 residue Binding studies of calcofluor white to -lactoglobulin have been carried out by recording fluorescence emission and excitation spectra of both Trp 19 residue and calcofluor white. Fig. 4 displays emission spectra of -lactoglobulin at pH 2 recorded in presence of increasing concentrations of calcofluor white. A fluorescence intensity decrease is observed with calcofluor white concentrations increase. This intensity decrease is not accompanied by a shift of the emission peak (331 nm) showing clearly the absence of any local or global structural modification within the protein. Binding of calcofluor to -lactoglobulin can also be followed by recording fluorescence excitation spectra. In fact, binding of calcofluor white to the protein induces a decrease in the fluorescence intensity of the excitation spectrum (Fig. 5). This intensity decrease
148
J.R. Albani et al. / Journal of Pharmaceutical and Biomedical Analysis 91 (2014) 144–150
24
1.0
Fluorescence intensity (a.u.)
Fluorescence intensity (a.u.)
1 20
16
12
8
4
0.8
0.6
0.4
0.2
12 0
0.0
300 310 320 330 340 350 360 370 380 390
0
4
Wavelength (nm)
12
16
20
24
28
[Calcofluor ] (µM)
Fig. 4. Fluorescence emission spectra of -lactoglobulin Trp 19 residue in presence of increased concentrations of calcofluor white. [-lactoglobulin] = 22.7 M. Calcofluor white concentration corresponding to spectrum 12 is equal to 27.91 M. Experimental conditions: 10 mM phosphate buffer, pH 2. ex = 295 nm.
is not accompanied by a shift of the spectrum peak (max = 283 nm), indicating that protein structure is not modified upon calcofluor white binding. Fig. 6 displays fluorescence intensity variation at both excitation and emission spectra peaks with calcofluor white concentrations. Decrease of the Trp 19 residue fluorescence intensities is accompanied by an increase of calcofluor white fluorescence intensity (Fig. 7). Fluorescence intensity variations of calcofluor white (intensity increase) and/or of Trp 19 residue (intensity decrease) allowed us to calculate the dissociation constant of -lactoglobulin–calcofluor complex. Fig. 8 displays fluorescence intensity increase of calcofluor white in presence of -lactoglobulin for different fluorophore concentrations (em = 440 nm and ex = 295 nm).
Fig. 6. Fluorescence intensity decrease of 22.7 M -lactoglobulin Trp 19 residue emission observed at 335 nm (squares) and excitation observed at 280 nm (circles) with calcofluor concentration at pH 2.
Analysis of the curve obtained can be performed with the following equation: IF =
IF(max) [calcofluor]
(3)
Kd + [calcofluor]
Kd value obtained is equal to 8.5 M. Fluorescence intensity decrease of Trp 19 residue is analyzed with the following equation: Io = 1 + Ka [calcofluor] I
(4)
where Io , I, Ka and [calcofluor] are respectively fluorescence intensity of the protein in absence of calcofluor white, protein fluorescence intensity in presence of specific calcofluor white
12
28 200
Fluorescence intensity (a.u.)
1 24
Fluorescence intensity (a.u.)
8
20 16 12 8 4
160
120
80
40
0
1
12 0 240
400 420 440 460 480 500 520 540 560 580
260
280
300
320
Wavelength (nm)
Wavelength (nm) Fig. 5. Fluorescence excitation spectra of -lactoglobulin Trp 19 residue in presence of increasing concentrations of calcofluor white at pH 2. [-lactoglobulin] = 22.7 M. Calcofluor white concentration corresponding to spectrum 12 is equal to 27.91 M. em = 335 nm and max = 283 nm.
Fig. 7. Fluorescence emission spectra of calcofluor white bound to -lactoglobulin in presence of increasing concentrations of calcofluor white at pH 2. [lactoglobulin] = 22.7 M. Calcofluor white concentration corresponding to spectrum 12 is equal to 27.91 M. Intensity of free calcofluor white has already been subtracted. ex = 295 nm.
J.R. Albani et al. / Journal of Pharmaceutical and Biomedical Analysis 91 (2014) 144–150
149
250
4
150
1/E
Fluorescence intensity (a.u.)
200
100
2
50
0 0
4
8
12
16
20
24
0 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
28
[Calcofluor White] (µM)
-1
1 / [Calcofluor] (µM )
Fig. 8. Titration of 22.7 M -lactoglobulin with calcofluor white. em = 440 nm and ex = 295 nm. 10 mM phosphate buffer, pH 2. Kd = 8.5 M.
concentration, association constant of -lactoglobulin–calcofluor white complex and concentration of added calcofluor white. Fig. 9 displays ratios of the emission intensities and mean fluorescence lifetimes of Trp 19 residue in absence and presence of different calcofluor white concentrations. The value of the association constant, obtained from the intensities values, was determined before the stoichiometry was reached. Ka was found equal to 0.119 M−1 , i.e. a Kd equal to 8.4 M. This value is identical to that found from the fluorescence intensity increase of calcofluor white (Figs. 7 and 8). Addition of calcofluor white to -lactoglobulin does not modify fluorescence lifetimes parameters (values and percentages) of Trp-19 residue and thus does not affect mean fluorescence lifetime (Fig. 9). This means that fluorescence intensity quenching of Trp 19
8
Io/ I or τo / τ
6
4 -1
Ka = 0.119 µM 2
Fig. 10. Energy transfer efficiency between Trp 19 residue of -lactoglobulin and calcofluor white at pH 2.
residue occurs via a static process as the result of calcofluor white binding to the protein. Energy transfer efficiency from Trp 19 residue and calcofluor white can be calculated with Eq. (5) E =1−
I Io
(5)
where I and Io are the fluorescence intensities in presence and absence of calcofluor white, respectively. Value of E calculated at infinite concentrations of calcofluor white was obtained by plotting 1/E as a function of 1/[calcofluor white]. Fig. 10 clearly shows that extrapolation at infinite concentration of calcofluor white yields a value of 1/E equal to 0.84, which means a value of E higher than 1 and thus an energy transfer higher than 100%. This means that distance between Trp 19 residue and calcofluor white is lower than ˚ In other terms, the two fluorophores are very close one to 5 A. each other and fluorescence intensity quenching of Trp 19 residue upon calcofluor white binding to -lactoglobulin does not occur via Förster energy transfer. This result can be explained by the fact that -lactoglobulin pocket binds easily hydrophobic ligands adopting their structure [38]. This confers to the protein pocket a high flexibility. Also, since -lactoglobulin contains two Trp residues, one within the hydrophobic pocket of the protein (Trp 19) and the second at the aperture of the pocket within a hydrophilic region (Trp 61), calcofluor white cannot be in contact simultaneously with both tryptophan residues. In absence of carbohydrate in the protein, calcofluor white binds to hydrophobic site [21] and therefore it is located within the -lactoglobulin pocket and thus is in close contact with Trp 19 residue as indicated by the calculated energy transfer efficiency. This leads us to the fact that fluorescence of Trp 61 residue is totally quenched or/and far from calcofluor white. 4. Conclusion In the present work, we demonstrated that
0 0
4
8
12
16
20
24
28
[Calcofluor ] (µM) Fig. 9. Variation of the emission intensity ratio (squares) and of mean fluorescence lifetime (circles) of Trp 19 residue with calcofluor white concentration, in 10 mM phosphate buffer, pH 2. [-lactoglobulin] = 22.7 M.
(1) only Trp 19 residue buried in the hydrophobic pocket of lactoglobulin emits, Trp 69 residue present at the protein surface near the aperture of the pocket does not emit. This result was observed at different pH going from 2 to 11. Therefore, these tryptophan fluorescence properties do not depend on the
150
J.R. Albani et al. / Journal of Pharmaceutical and Biomedical Analysis 91 (2014) 144–150
global or local structures of the protein or the percentage of monomers and dimers in solution. (2) Trp 19 residue emits with three lifetimes which values do not change with pH. (3) Each lifetime characterizes substructure of the tryptophan within the protein. The values of the pre-exponentials characterize the population of each substructure. These populations are modified with the pH as the result of a local structural modification. (4) Binding of calcofluor white to -lactoglobulin occurs within the hydrophobic pocket of the protein, in proximity to the Trp 19 residue.
References [1] P.J. Kraulis, MOLSCRIPT: a program to produce both detailed and schematic plots of protein structures, J. Appl. Cryst. 24 (1991) 946–950. [2] T.V. Burova, Y. Choiset, C.K. Jankowski, T. Haertle, Conformational stability and binding properties of porcine pdorant binding protein, Biochemistry 38 (1999) 15043–15051. [3] G. Invernizzi, M.S. Amalikova, S. Brocca, M. Lotti, H. Molinari, R. Grandori, Comparison of bovine and porcine -lactoglobulin: a mass spectrometric analysis, J. Mass. Spectrom. 41 (2006) 717–727. [4] H. Molinari, L. Ragona, L. Varani, G. Musco, R. Consonni, L. Zetta, H.L. Monaco, Partially folded structure of monomeric bovine beta-lactoglobulin, FEBS Lett. 381 (1996) 237–243. [5] G. Panick, P. Malessa, R. Winter, Difference between the pressure and temperature denaturation and aggregation of -lactoglobulin, Biochemistry 38 (1999) 6512–6519. [6] S. Brownlow, J.H. Morais Cabral, R. Cooper, R.D. Flower, S.J. Yewdall, I. Polikarpov, A.C.T. North, L. Sawyer, Bovine -lactoglobulin at 1.8 A˚ resolution – still an enigmatic lipocalin, Structure 5 (1997) 481–495. [7] H.L. Monaco, G. Zanotti, P. Spadon, M. Bolognesi, L. Sawyer, E.E. Eliopoulos, Crystal structure of the trigonal form of bovine -lactoglobulin and of its complex with retinol at 2.5 A˚ resolution, J. Mol. Biol. 197 (1987) 695–706. [8] L. Sawyer, G. Kontopidis, The core lipocalin, bovine -lactoglobulin, Biochim. Biophys. Acta 1482 (2000) 136–148. [9] M.D. Perez, M.C. Diaz de Villegas, L. Sanchez, P. Aranda, J.M. Ena, M. Calvo, Interaction of fatty acids with -lactoglobulin and albumin from ruminant milk, Biochem. J. 106 (1989) 1094–1097. [10] A.A. Spector, J.E. Fletcher, Binding of long chain fatty acids to -lactoglobulin, Lipids 5 (1970) 403–411. [11] H.M. Farrell, M.J. Behe, J.A. Enyeart, Binding of p-nitrophenol phosphate and other aromatic-compounds to beta-lactoglobulin, J. Dairy Sci. 70 (1987) 252–258. [12] M. Ya, K. Sakurai, C. Kalidas, C. Batt, Y. Goto, Reversible unfolding of bovine -lactoglobulin mutants without a free thiol group, J. Biol. Chem. 278 (2003) 47009–47015. [13] Y. Cho, C.A. Batt, L. Sawyer, Probing the retinol-binding site of bovine lactoglobulin, J. Biol. Chem. 269 (1994) 11102–11107. [14] G. Palazolo, F. Rodriguez, B. Farruggia, G. Picó, N. Delorenzi, Heat treatment of -lactoglobulin: structural changes studied by partitioning and fluorescence, J. Agric. Food Chem. 48 (2000) 3817–3822. [15] D. Renard, J. Lefebvre, M.C.A. Griffin, W.G. Griffin, Effects of pH and salt environment on the association of -lactoglobulin revealed by intrinsic fluorescence studies, Int. J. Biol. Macromol. 22 (1998) 41–49.
[16] C.A.M. Portugal, J.C. Lima, J.G. Crespo, Effect of physicochemical conditions on the ultrafiltration of -lactoglobulin: fluorescence probing of induced structural change, J. Membr. Sci. 329 (2008) 69–80. [17] M.I. Viseu, T.I. Carvalho, S.M.B. Costa, Conformational transitions in lactoglobulin induced by cationic amphiphiles: equilibrium studies, Biophys. J. 86 (2004) 2392–2402. [18] I.D. Rattee, M.M. Breur, The Physical Chemistry of Dye Absorption, Academic Press, New York, 1974, pp. 180–182. [19] J.E. Gonzalez, G.M. York, G.C. Walker, Rhizobium meliloti exopolysaccharides: synthesis and symbiotic function, Gene 179 (1996) 141–146. [20] T.F. Hogan, R.S. Riley, J.G. Thomas, Rapid diagnosis of acute eosinophilic pneumonia (AEP) in a patient with respiratory failure using bronchoalveolar lavage (BAL) with calcofluor white (CW) staining, J. Clin. Lab. Anal. 11 (1997) 202–207. [21] J.R. Albani, Y. Plancke, Interaction between calcofluor white and carbohydrate of ␣1 -acid glycoprotein, Carbohydr. Res. 314 (1998) 169–175; J.R. Albani, Y. Plancke, Interaction between calcofluor white and carbohydrate of ␣1 -acid glycoprotein, Carbohydr. Res. 318 (1999) 194–200. [22] N. Tayeh, T. Rungassamy, J.R. Albani, Fluorescence spectral resolution of tryptophan residues in bovine and human serum albumins, J. Pharm. Biomed. Anal. 50 (2009) 109–116. [23] J.R. Albani, Effect of binding of calcofluor white on the carbohydrate residues of ␣1 -acid glycoprotein (orosomucoid) on the structure and dynamics of the protein moiety. A fluorescence study, Carbohydr. Res. 334 (2001) 141–151. [24] C.N. Pace, F. Vajdos, L. Fee, G. Grimsley, T. Gray, How to measure and predict the molar absorption coefficient of a protein, Prot. Sci. 4 (1995) 2411–2423. [25] J.R. Lakowicz, Principles of Fluorescence Spectroscopy, Kluwer Academic/Plenum, New York, 1999. [26] J.R. Albani, Principles and Applications of Fluorescence Spectroscopy, Blackwell, London, 2007. [27] M.G. Badea, L. Brand, Time-resolved fluorescence measurements, Methods Enzymol. 61 (1971) 378–425. [28] J. Yguerabide, Nanosecond fluorescence spectroscopy of macromolecules, Methods Enzymol. 26 (1972) 498–578. [29] J.R. Albani, Fluorescence lifetimes of tryptophan: structural origin and relation with So → 1 Lb and So → 1 La transitions, J. Fluoresc. 19 (2009) 1061–1071. [30] L. Ragona, F. Pusterla, L. Zetta, H.L. Monaco, H. Molinari, Identification of a conserved hydrophobic cluster in partially folded bovine -lactoglobulin at pH 2, Fold Design 2 (1997) 281–290. [31] E.A. Burstein, N.S. Vedenkina, M.N. Ivkova, Fluorescence and the location of tryptophan residues in protein molecules, Photochem. Photobiol. 18 (1973) 263–279. [32] J. Albani, Motions studies of the human ␣1 -acid glycoprotein (orosomucoid) followed by red-edge excitation spectra and polarization of 2-p-toluidinylnaphthalene-6-sulfonate (TNS) and of tryptophan residues, Biophys. Chem. 44 (1992) 129–137. [33] J.R. Albani, New insights in the interpretation of tryptophan fluorescence, J. Fluoresc. 17 (2007) 406–417. [34] J.R. Albani, Sub-structures formed in the excited state are responsible for tryptophan residues fluorescence in -lactoglobulin, J. Fluoresc. 21 (2011) 1683–1687. [35] M. Amiri, K. Jankeje, J.R. Albani, Origin of fluorescence lifetimes in human serum albumin. Studies on native and denatured protein, J. Fluoresc. 20 (2010) 651–656. [36] J.R. Albani, Origin of tryptophan fluorescence lifetimes. Part 1. Fluorescence lifetimes origin of tryptophan free in solution, J. Fluoresc. (2013) (in press). [37] J.R. Albani, Origin of tryptophan fluorescence lifetimes. Part 1. Fluorescence lifetimes origin of tryptophan in proteins, J. Fluoresc. (2013) (in press). [38] D. Kmiecik, J.R. Albani, Effect of 1-aminoanthracene (1-AMA) binding on the structure of three lipocalin proteins, the dimeric -lactoglobulin, the dimeric odorant binding protein and the monomeric ␣1 -acid glycoprotein. Fluorescence spectra and lifetimes studies, J. Fluoresc. 20 (2010) 973–983.
Contents lists available at ScienceDirect
Journal of Pharmaceutical and Biomedical Analysis journal homepage: www.elsevier.com/locate/jpba
Tryptophan 19 residue is the origin of bovine -lactoglobulin fluorescence Jihad René Albani ∗ , Julie Vogelaer, Loïc Bretesche, Daniel Kmiecik Laboratoire de Biophysique Moléculaire, Université des Sciences et Technologies de Lille, Bâtiment C6, 59655 Villeneuve d’Ascq Cédex, France
a r t i c l e
i n f o
Article history: Received 15 October 2013 Received in revised form 13 December 2013 Accepted 17 December 2013 Available online 3 January 2014 Keywords: Beta lactoglobulin Tryptophan Lifetime Calcofluor Substructures
a b s t r a c t -Lactoglobulin consists of a single polypeptide of 162 amino acid residues with 2 Trp residues, Trp 19 present in a hydrophobic pocket and Trp 61 present at the surface of the protein near the pocket. This study aimed to characterize the respective contribution of the two Trp residues to the overall fluorescence of the protein. We added for that calcofluor white, an extrinsic fluorophore, which, at high concentration compared to that of the protein, quenches completely emission of hydrophobic Trp residue(s). The study was performed at different pHs by recording fluorescence steady state spectra and measuring fluorescence lifetimes of the Trp-residues using Single Time Photon Counting method. Our results indicate that addition of calcofluor white does not induce a red shift of the tryptophan(s) emission peak (332 nm) but only a decrease in the fluorescence intensity. This means that Trp 61 residue does not contribute to the protein emission, tryptophan emission occurs from Trp 19 residue only. Also, excitation spectrum peak position (283 nm) of -lactoglobulin is not modified upon calcofluor white binding. These results mean that structural rearrangements within -lactoglobulin are not occurring upon calcofluor white binding. Energy transfer between Trp 19 residue and calcofluor white occurs with 100% efficiency, i.e. the two ˚ This energy transfer is not Forster type. fluorophores are very close one to each other (90%) was from Sigma–Aldrich (Saint Quentin Fallavier, France). Its concentration was determined at 278 nm with the following extinction coefficients: 17.6 × 103 M−1 cm−1 [24]. In the whole manuscript, lactoglobulin concentration is expressed in monomer. Absorbance data were obtained with a Varian DMS-100S (Les Ulis, France) spectrophotometer using 1-cm pathlength cuvettes. Fluorescence spectra were recorded with a Perkin-Elmer LS-5B spectrofluorometer (Perkin-Elmer, Courtaboeuf, France). The bandwidths used for the excitation and the emission were 5 nm. The quartz cuvettes had optical pathlengths equal to 1 and 0.4 cm for the emission and excitation wavelengths, respectively. Fluorescence
spectra were corrected for the background intensities of the buffer solution. Observed fluorescence intensities were first corrected for the dilution, and then corrections were made for the inner filter effect as described [25,26]. Fluorescence lifetime measurements were obtained with a Horiba Jobin Yvon FluoroMax-4-P (Horiba-Jobin Yvon, Longjumeau, France) using time correlated single photon counting method. A Ludox solution was used as scatter. Excitation was performed at 296 nm with a nanoLED. Each fluorescence decay was analyzed with one, two, three and four lifetimes and then values of 2 were compared in order to determine the best fit. A minimal value of 2 indicates the best fit. A 2 value that approaches 1 indicates a good fit. For example, let us consider the value of 2 equal to 1.054, 1.06 and 1.1 when analysis is done with 1, 2 and 3 lifetimes, respectively. One lifetime should be considered as the best description of the decay curve since there was no real improvement in 2 value when the experimental decay was fitted with one, two, three or four lifetimes [25,27–29]. The mean fluorescence lifetime calculated is the second order mean [25] 0 = ˙fi i
(1)
and fi =
ˇi i ˙ˇi i
(2)
where ˇi are the preexponential terms, i are the fluorescence lifetimes and fi the fractional intensities. All spectral experiments were performed at 20 ◦ C in 10 mM phosphate buffer and were all performed between 3 and 5 times. 3. Results and discussion 3.1. Origin of ˇ-lactoglobulin fluorescence Fluorescence emission spectra of -lactoglobulin recorded at pH 2 and 7 display a peak located at 331 ± 1 nm (Fig. 1). Thus, at these two pHs, -lactoglobulin is not denatured, otherwise emission peak will be located at 350–355 nm. This means that -lactoglobulin global conformation is identical or very close at both pHs, although 20
d a
Fluorescence intensity (a.u.)
emission peak of -lactoglobulin mutant type (W19A) is located around 305 nm and not at 340–345 nm, as it should be if emission of the mutant protein occurs from Trp 61 residue. In fact, since Trp 61 residue is at the protein surface, its emission peak should be redshifted compared to that of Trp 19 residue and not blue shifted. Also, emission peak observed at 305 nm for the W19A mutant (ex = 290 nm) corresponds to that of tyrosine residue. Tyrosine absorbs at 290 nm and thus in order to avoid its excitation, one should excite at least at 295 nm and beyond, since only tryptophan absorbs at this wavelength. Therefore, emission spectrum of W19A mutant, displayed in [13] corresponds to that occurring from tyrosine residue and not from tryptophan. Nevertheless, if Trp 61 residue is emitting, its emission intensity should be much smaller than that indicated by the authors (20% of the total wild type fluorescence) as the result of the emission from tyrosine residues of the -lactoglobulin mutant type (W19A). Tyrosine residues emission seems to be much more important than that of Trp 61 residue since emission peak of the mutant is located at 305 nm. Also, the weak or absence of Trp 61 residue fluorescence in wild type -lactoglobulin was explained as the result of the presence of a disulfide bridge (Cys66–Cys160) in proximity to this tryptophan ˚ inducing complete fluorescence quenching of Trp (≈3.65 ± 0.15 A) 61 residue [14] and/or to the self-quenching of Trp-61 by the nearby Trp-61 of the other monomer, in the -lactoglobulin dimeric form [15]. Other authors considered that structural modification of lactoglobulin modifies the distance between disulfide bridge and Trp 61 residue inducing a significant fluorescence from this Trp residue [16,17]. The aim of the present work is to find out whether both Trp residues contribute to -lactoglobulin fluorescence or not. Timeresolved studies and static quenching were performed in presence of high concentrations of calcofluor white, a fluorophore that is specific to both carbohydrate residues and hydrophobic sites in proteins. In fact, calcofluor white is a fluorescent probe capable of making hydrogen bonds with -(1 → 4) and -(1 → 3) polysaccharides [18]. It is commonly used to study the mechanism by which cellulose and other carbohydrate structures are formed in vivo and is also widely used in clinical studies [19,20]. Also, calcofluor white binds to hydrophobic sites of proteins such as ␣1 -acid glycoprotein and both human and bovine serum albumin [21,22]. In presence of high calcofluor concentrations compared to those of proteins (ratio 10–1), extrinsic fluorophore was able to quench completely emission of hydrophobic tryptophan residue [22,23], without modifying protein structures. Thus, we were able to demonstrate for example that both hydrophobic and hydrophilic tryptophan residues of bovine serum albumin contribute to the protein emission [22]. In the present work, static and time-resolved fluorescence studies were performed on -lactoglobulin in presence of high calcofluor white concentration. Experiments were performed at pH going from 2 to 12.
145
16
e c 12
8
b 4
f
0 300
320
340
360
380
400
420
Wavelength (nm) Fig. 1. Fluorescence emission spectra of 12 M -lactoglobulin at pH 2 in absence (a) and in presence (b) of 138 M calcofluor white and of 13 M -lactoglobulin at pH 7 in absence (d) and presence (f) of 149 M calcofluor white. Difference spectra (c and e) display emission peaks located at 332 nm. ex = 295 nm. Spectra recorded at pH 7 are normalized to those recorded at pH 2.
J.R. Albani et al. / Journal of Pharmaceutical and Biomedical Analysis 91 (2014) 144–150
local structural modifications could exist. For example, at pH 2, lactoglobulin is a monomer in the molten globule state [30]. The protein is stable although partially unfolded. -Lactoglobulin contains two Trp residues, Trp 19 residue located in the hydrophobic pocket of the protein and Trp 61 residue located at the hydrophilic aperture of the pocket. Thus, if protein fluorescence occurs from both residues, recorded emission spectrum will be the result of the contribution of each residue to the total fluorescence. Position of tryptophan residue emission peak and bandwidth of the spectrum calculated at half-intensity of the emission peak, would allow knowing whether one or two tryptophan residues are emitting. Position of the emission peak of Trp residue of -lactoglobulin is located at 331 nm and spectrum bandwidth is equal to 50 nm. These two values correspond to an emission spectrum occurring from a tryptophan residue located in a hydrophobic area of the protein [31]. At pH 2, emission occurs from a monomeric -lactoglobulin [30]. When a protein contains two classes of Trp residues, one at the surface and the second in hydrophobic environment and if both tryptophans contribute to the protein emission, fluorescence emission spectrum will be the result of the contribution of each class. For example, fluorescence emission spectrum of ␣1 -acid glycoprotein recorded with an excitation wavelength of 295 nm, shows a maximum located at 331 nm and a bandwidth of 53 nm. This spectrum is typical for protein containing Trp residues in both hydrophobic and hydrophilic areas of the protein [32]. Therefore, our data on -lactoglobulin clearly indicate that protein emission occurs from one tryptophan residue located in a hydrophobic area. In order to assess our result, we measured emission spectra of lactoglobulin at high concentration of calcofluor. Addition of high calcofluor white concentrations quenches fluorescence emission intensity without modifying position of the spectrum peak (Fig. 1). The difference between the two spectra (in absence and presence of calcofluor white) yields an emission spectrum with a peak located at 332 nm and not at 340 or 345 nm. This result clearly indicates that Trp 61 residue does not contribute to the protein emission. If Trp 61 residue contributes to the protein fluorescence, addition of high calcofluor concentration compared to that of lactoglobulin (ratio 10–1), will quench completely fluorescence emission of Trp 19 residue, inducing a red shift in the emission peak as we have already shown when experiments were performed on bovine serum albumin [22]. Result obtained at pH 2 was also obtained at pH 7, where lactoglobulin is in a dimeric state. Thus, absence of fluorescence from Trp 61 residue is independent of the ratio monomer/dimer present in solution and of any possible differences in the local structure of the protein around Trp residues. Also, since identical results were obtained whether the protein is in a 100% monomeric state (pH 2) or when it is in a dimeric state (pH 7), absence of Trp 61 residue fluorescence cannot be attributed to the selfquenching of Trp 61 by the nearby Trp 61 of the other monomer, in the -lactoglobulin dimeric form, as suggested in [15]. Identical results were obtained for other pHs , although degree of fluorescence quenching of tryptophan is not the same at all pHs (data not shown). Absence of emission of -lactoglobulin Trp 61 residue can be explained by the fact that this tryptophan is in proximity of the ˚ which is supposed to be Cys66–Cys160 disulfide moiety (≈3.7 A), a strong quencher of indole fluorescence [14]. Nevertheless, one should also expect that Trp 61 residue is not necessarily excited, as the result of its close interaction with other amino acids whether cysteine disulfide bridge or other amino acids [33]. Fig. 2 displays normalized excitation spectra of -lactoglobulin in solution in absence and presence of calcofluor white at pH 2. We observe that peak position (283 nm) is not modified upon calcofluor white binding. Also, global shape of the spectra is the same
1.0
Fluorescence intensity (a.u.)
146
0.8
0.6
0.4
0.2
0.0 250
260
270
280
290
300
310
Wavelength (nm) Fig. 2. Normalized fluorescence excitation spectra at pH 2 of 12 M -lactoglobulin in absence (continuous spectral line) and in presence of 138 M calcofluor white (dashed spectral lines). Peak position of both spectra is equal to 283 nm. em = 325 nm.
whether calcofluor white is present or not. These results mean that structural rearrangements within -lactoglobulin are not occurring upon calcofluor white binding. These results were also observed at pH 7. 3.2. Origin of fluorescence lifetimes of ˇ-lactoglobulin Trp 19 residue Fig. 3 displays fluorescence intensity decay of -lactoglobulin at pH 2 in absence of calcofluor white. Decay analysis shows the presence of three fluorescence lifetimes. In fact, fluorescence intensity decay I(, t), of -lactoglobulin Trp 19 residue can be adequately represented as I(, t) = 0.140e−t/0.48 + 0.697e−t/1.49 + 0.163e−t/4.29 where 0.140, 0.697 and 0.163 are the pre-exponential factors, 0.48, 1.49 and 4.29 ns are the decay times and is the emission wavelength (330 nm) (2 = 1.054). Addition of calcofluor white does not modify fluorescence lifetimes values. In fact, in presence of 116 M calcofluor, fluorescence intensity decay can be described as I(, t) = 0.175e−t/0.73 + 0.745e−t/2.032 + 0.098e−t/5.26 where 0.175, 0.745 and 0.098 are the pre-exponential factors, 0.73, 2.032 and 5.26 ns are the decay times and is the emission wavelength (330 nm) (2 = 0.99). These three fluorescence lifetimes were also observed at pH 7 and along the emission spectrum (310–410 nm). Fluorescence decays were analyzed with one, two and three lifetimes and 2 values were compared. The best values of 2 were obtained for three lifetimes. Analysis of the decays with four lifetimes did not improve 2 values, thus fluorescence intensity decays can be best described with three lifetimes [34]. Table 1 shows the fluorescence intensity decay data of lactoglobulin at different pH and observed at 330 nm in absence and presence of calcofluor white. Three fluorescence lifetimes were observed at all pH and along the emission spectrum (310–410 nm). Fluorescence decays were analyzed with one, two and three lifetimes and 2 values were compared. The best values of 2 were
J.R. Albani et al. / Journal of Pharmaceutical and Biomedical Analysis 91 (2014) 144–150
147
Fig. 3. Fluorescence intensity decay of 23 M -lactoglobulin in 10 mM phosphate buffer at pH 2 recorded at 330 nm. ex = 296 nm. Table 1 Fluorescence lifetimes and pre-exponential values of tryptophan 19 residue of lactoglobulin, measured in absence and presence of calcofluor white at different pHs . pH
2
6
8
10
11
1 (ns) 1 + calcofluor 2 (ns) 2 + calcofluor 3 (ns) 3 + calcofluor 0 (ns) 0 + calcofluor ˛1 ˛1 + calcofluor ˛2 ˛2 + calcofluor ˛3 ˛3 + calcofluor
0.48 0.73 1.89 2.032 4.29 5.36 2.648 2.762 0.1398 0.1753 0.6975 0.7449 0.1626 0.098
0.58 0.59 1.71 1.65 5.02 4.52 2.70 2.435 0.2536 0.2783 0.6283 0.6063 0.1181 0.1154
0.67 0.45 1.83 1.6 5.12 4.73 3.10 2.968 0.2707 0.2403 0.5606 0.5758 0.1687 0.1839
0.497 0.73 1.67 2.06 4.97 5.51 3.313 3.230 0.1933 0.342 0.5804 0.5106 0.2263 0.1475
0.3 0.57 1.62 1.83 4.65 4.76 3.053 2.96 0.2159 0.287 0.579 0.5376 0.2051 0.1755
obtained for three lifetimes. Analysis of the decays with four lifetimes did not improve 2 values (Table 2), thus fluorescence intensity decays can be best described with three lifetimes. This result clearly indicates that the sole Trp 19 residue emits with three fluorescence lifetimes independently of the percentage of monomers and dimers in solution. Also, presence of these Table 2 2 values obtained for -lactoglobulin tryptophan fluorescence decay analysis for one to four lifetimes. Conditions
2 (1)
2 (2)
2 (3)
2 (4)
pH 2 pH 2 + calcofluor pH 6 pH 6 + calcofluor pH 8 pH 8 + calcofluor pH 10 pH 10 + calcofluor pH 11 pH 11 + calcofluor
8.87 8.2 13.41 12.03 15.5 18.44 18.04 17.5 14.57 17.44
1.38 1.48 1.6 1.71 1.4 2.06 1.6 1.6 1.65 1.92
1.05 0.99 1.04 1.28 0.98 1.3 1.04 1.01 1.04 1.25
1.06 1.05 1.25 0.97 1.34 1.02 1.02 1.04 1.23
three lifetimes does not depend on the presence of calcofluor white and on any eventual local or global structural modification within -lactoglobulin. Recent studies [22,33,29,35] have shown that fluorescence lifetimes of tryptophan originate from sub-structures obtained at the excited state. Pre-exponential values characterize populations of these substructures. Tryptophan free in water emits with two lifetimes equal to 0.5 and 2.78 ns, these two lifetimes are observed for tryptophan residues in proteins, whether the macromolecules contain one or more Trp residues. Thus, these two lifetimes characterize two substates or substructures of the tryptophan, reached in the excited state, independently of the environment (solvent and/or Trp – amino acids interactions). These substructures differ from an environment to another, modifying values of the corresponding lifetimes amplitudes. The third lifetime observed in proteins is the result of the interaction between tryptophan residue(s) and amino acids environments inducing a third substructure with specific emission decay parameters. Each substructure is composed by tryptophan backbone with a specific electronic cloud. Electronic distribution around tryptophan backbone differs between a substructure and another [36,37]. 3.3. Binding of calcofluor white occurs close to Trp 19 residue Binding studies of calcofluor white to -lactoglobulin have been carried out by recording fluorescence emission and excitation spectra of both Trp 19 residue and calcofluor white. Fig. 4 displays emission spectra of -lactoglobulin at pH 2 recorded in presence of increasing concentrations of calcofluor white. A fluorescence intensity decrease is observed with calcofluor white concentrations increase. This intensity decrease is not accompanied by a shift of the emission peak (331 nm) showing clearly the absence of any local or global structural modification within the protein. Binding of calcofluor to -lactoglobulin can also be followed by recording fluorescence excitation spectra. In fact, binding of calcofluor white to the protein induces a decrease in the fluorescence intensity of the excitation spectrum (Fig. 5). This intensity decrease
148
J.R. Albani et al. / Journal of Pharmaceutical and Biomedical Analysis 91 (2014) 144–150
24
1.0
Fluorescence intensity (a.u.)
Fluorescence intensity (a.u.)
1 20
16
12
8
4
0.8
0.6
0.4
0.2
12 0
0.0
300 310 320 330 340 350 360 370 380 390
0
4
Wavelength (nm)
12
16
20
24
28
[Calcofluor ] (µM)
Fig. 4. Fluorescence emission spectra of -lactoglobulin Trp 19 residue in presence of increased concentrations of calcofluor white. [-lactoglobulin] = 22.7 M. Calcofluor white concentration corresponding to spectrum 12 is equal to 27.91 M. Experimental conditions: 10 mM phosphate buffer, pH 2. ex = 295 nm.
is not accompanied by a shift of the spectrum peak (max = 283 nm), indicating that protein structure is not modified upon calcofluor white binding. Fig. 6 displays fluorescence intensity variation at both excitation and emission spectra peaks with calcofluor white concentrations. Decrease of the Trp 19 residue fluorescence intensities is accompanied by an increase of calcofluor white fluorescence intensity (Fig. 7). Fluorescence intensity variations of calcofluor white (intensity increase) and/or of Trp 19 residue (intensity decrease) allowed us to calculate the dissociation constant of -lactoglobulin–calcofluor complex. Fig. 8 displays fluorescence intensity increase of calcofluor white in presence of -lactoglobulin for different fluorophore concentrations (em = 440 nm and ex = 295 nm).
Fig. 6. Fluorescence intensity decrease of 22.7 M -lactoglobulin Trp 19 residue emission observed at 335 nm (squares) and excitation observed at 280 nm (circles) with calcofluor concentration at pH 2.
Analysis of the curve obtained can be performed with the following equation: IF =
IF(max) [calcofluor]
(3)
Kd + [calcofluor]
Kd value obtained is equal to 8.5 M. Fluorescence intensity decrease of Trp 19 residue is analyzed with the following equation: Io = 1 + Ka [calcofluor] I
(4)
where Io , I, Ka and [calcofluor] are respectively fluorescence intensity of the protein in absence of calcofluor white, protein fluorescence intensity in presence of specific calcofluor white
12
28 200
Fluorescence intensity (a.u.)
1 24
Fluorescence intensity (a.u.)
8
20 16 12 8 4
160
120
80
40
0
1
12 0 240
400 420 440 460 480 500 520 540 560 580
260
280
300
320
Wavelength (nm)
Wavelength (nm) Fig. 5. Fluorescence excitation spectra of -lactoglobulin Trp 19 residue in presence of increasing concentrations of calcofluor white at pH 2. [-lactoglobulin] = 22.7 M. Calcofluor white concentration corresponding to spectrum 12 is equal to 27.91 M. em = 335 nm and max = 283 nm.
Fig. 7. Fluorescence emission spectra of calcofluor white bound to -lactoglobulin in presence of increasing concentrations of calcofluor white at pH 2. [lactoglobulin] = 22.7 M. Calcofluor white concentration corresponding to spectrum 12 is equal to 27.91 M. Intensity of free calcofluor white has already been subtracted. ex = 295 nm.
J.R. Albani et al. / Journal of Pharmaceutical and Biomedical Analysis 91 (2014) 144–150
149
250
4
150
1/E
Fluorescence intensity (a.u.)
200
100
2
50
0 0
4
8
12
16
20
24
0 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
28
[Calcofluor White] (µM)
-1
1 / [Calcofluor] (µM )
Fig. 8. Titration of 22.7 M -lactoglobulin with calcofluor white. em = 440 nm and ex = 295 nm. 10 mM phosphate buffer, pH 2. Kd = 8.5 M.
concentration, association constant of -lactoglobulin–calcofluor white complex and concentration of added calcofluor white. Fig. 9 displays ratios of the emission intensities and mean fluorescence lifetimes of Trp 19 residue in absence and presence of different calcofluor white concentrations. The value of the association constant, obtained from the intensities values, was determined before the stoichiometry was reached. Ka was found equal to 0.119 M−1 , i.e. a Kd equal to 8.4 M. This value is identical to that found from the fluorescence intensity increase of calcofluor white (Figs. 7 and 8). Addition of calcofluor white to -lactoglobulin does not modify fluorescence lifetimes parameters (values and percentages) of Trp-19 residue and thus does not affect mean fluorescence lifetime (Fig. 9). This means that fluorescence intensity quenching of Trp 19
8
Io/ I or τo / τ
6
4 -1
Ka = 0.119 µM 2
Fig. 10. Energy transfer efficiency between Trp 19 residue of -lactoglobulin and calcofluor white at pH 2.
residue occurs via a static process as the result of calcofluor white binding to the protein. Energy transfer efficiency from Trp 19 residue and calcofluor white can be calculated with Eq. (5) E =1−
I Io
(5)
where I and Io are the fluorescence intensities in presence and absence of calcofluor white, respectively. Value of E calculated at infinite concentrations of calcofluor white was obtained by plotting 1/E as a function of 1/[calcofluor white]. Fig. 10 clearly shows that extrapolation at infinite concentration of calcofluor white yields a value of 1/E equal to 0.84, which means a value of E higher than 1 and thus an energy transfer higher than 100%. This means that distance between Trp 19 residue and calcofluor white is lower than ˚ In other terms, the two fluorophores are very close one to 5 A. each other and fluorescence intensity quenching of Trp 19 residue upon calcofluor white binding to -lactoglobulin does not occur via Förster energy transfer. This result can be explained by the fact that -lactoglobulin pocket binds easily hydrophobic ligands adopting their structure [38]. This confers to the protein pocket a high flexibility. Also, since -lactoglobulin contains two Trp residues, one within the hydrophobic pocket of the protein (Trp 19) and the second at the aperture of the pocket within a hydrophilic region (Trp 61), calcofluor white cannot be in contact simultaneously with both tryptophan residues. In absence of carbohydrate in the protein, calcofluor white binds to hydrophobic site [21] and therefore it is located within the -lactoglobulin pocket and thus is in close contact with Trp 19 residue as indicated by the calculated energy transfer efficiency. This leads us to the fact that fluorescence of Trp 61 residue is totally quenched or/and far from calcofluor white. 4. Conclusion In the present work, we demonstrated that
0 0
4
8
12
16
20
24
28
[Calcofluor ] (µM) Fig. 9. Variation of the emission intensity ratio (squares) and of mean fluorescence lifetime (circles) of Trp 19 residue with calcofluor white concentration, in 10 mM phosphate buffer, pH 2. [-lactoglobulin] = 22.7 M.
(1) only Trp 19 residue buried in the hydrophobic pocket of lactoglobulin emits, Trp 69 residue present at the protein surface near the aperture of the pocket does not emit. This result was observed at different pH going from 2 to 11. Therefore, these tryptophan fluorescence properties do not depend on the
150
J.R. Albani et al. / Journal of Pharmaceutical and Biomedical Analysis 91 (2014) 144–150
global or local structures of the protein or the percentage of monomers and dimers in solution. (2) Trp 19 residue emits with three lifetimes which values do not change with pH. (3) Each lifetime characterizes substructure of the tryptophan within the protein. The values of the pre-exponentials characterize the population of each substructure. These populations are modified with the pH as the result of a local structural modification. (4) Binding of calcofluor white to -lactoglobulin occurs within the hydrophobic pocket of the protein, in proximity to the Trp 19 residue.
References [1] P.J. Kraulis, MOLSCRIPT: a program to produce both detailed and schematic plots of protein structures, J. Appl. Cryst. 24 (1991) 946–950. [2] T.V. Burova, Y. Choiset, C.K. Jankowski, T. Haertle, Conformational stability and binding properties of porcine pdorant binding protein, Biochemistry 38 (1999) 15043–15051. [3] G. Invernizzi, M.S. Amalikova, S. Brocca, M. Lotti, H. Molinari, R. Grandori, Comparison of bovine and porcine -lactoglobulin: a mass spectrometric analysis, J. Mass. Spectrom. 41 (2006) 717–727. [4] H. Molinari, L. Ragona, L. Varani, G. Musco, R. Consonni, L. Zetta, H.L. Monaco, Partially folded structure of monomeric bovine beta-lactoglobulin, FEBS Lett. 381 (1996) 237–243. [5] G. Panick, P. Malessa, R. Winter, Difference between the pressure and temperature denaturation and aggregation of -lactoglobulin, Biochemistry 38 (1999) 6512–6519. [6] S. Brownlow, J.H. Morais Cabral, R. Cooper, R.D. Flower, S.J. Yewdall, I. Polikarpov, A.C.T. North, L. Sawyer, Bovine -lactoglobulin at 1.8 A˚ resolution – still an enigmatic lipocalin, Structure 5 (1997) 481–495. [7] H.L. Monaco, G. Zanotti, P. Spadon, M. Bolognesi, L. Sawyer, E.E. Eliopoulos, Crystal structure of the trigonal form of bovine -lactoglobulin and of its complex with retinol at 2.5 A˚ resolution, J. Mol. Biol. 197 (1987) 695–706. [8] L. Sawyer, G. Kontopidis, The core lipocalin, bovine -lactoglobulin, Biochim. Biophys. Acta 1482 (2000) 136–148. [9] M.D. Perez, M.C. Diaz de Villegas, L. Sanchez, P. Aranda, J.M. Ena, M. Calvo, Interaction of fatty acids with -lactoglobulin and albumin from ruminant milk, Biochem. J. 106 (1989) 1094–1097. [10] A.A. Spector, J.E. Fletcher, Binding of long chain fatty acids to -lactoglobulin, Lipids 5 (1970) 403–411. [11] H.M. Farrell, M.J. Behe, J.A. Enyeart, Binding of p-nitrophenol phosphate and other aromatic-compounds to beta-lactoglobulin, J. Dairy Sci. 70 (1987) 252–258. [12] M. Ya, K. Sakurai, C. Kalidas, C. Batt, Y. Goto, Reversible unfolding of bovine -lactoglobulin mutants without a free thiol group, J. Biol. Chem. 278 (2003) 47009–47015. [13] Y. Cho, C.A. Batt, L. Sawyer, Probing the retinol-binding site of bovine lactoglobulin, J. Biol. Chem. 269 (1994) 11102–11107. [14] G. Palazolo, F. Rodriguez, B. Farruggia, G. Picó, N. Delorenzi, Heat treatment of -lactoglobulin: structural changes studied by partitioning and fluorescence, J. Agric. Food Chem. 48 (2000) 3817–3822. [15] D. Renard, J. Lefebvre, M.C.A. Griffin, W.G. Griffin, Effects of pH and salt environment on the association of -lactoglobulin revealed by intrinsic fluorescence studies, Int. J. Biol. Macromol. 22 (1998) 41–49.
[16] C.A.M. Portugal, J.C. Lima, J.G. Crespo, Effect of physicochemical conditions on the ultrafiltration of -lactoglobulin: fluorescence probing of induced structural change, J. Membr. Sci. 329 (2008) 69–80. [17] M.I. Viseu, T.I. Carvalho, S.M.B. Costa, Conformational transitions in lactoglobulin induced by cationic amphiphiles: equilibrium studies, Biophys. J. 86 (2004) 2392–2402. [18] I.D. Rattee, M.M. Breur, The Physical Chemistry of Dye Absorption, Academic Press, New York, 1974, pp. 180–182. [19] J.E. Gonzalez, G.M. York, G.C. Walker, Rhizobium meliloti exopolysaccharides: synthesis and symbiotic function, Gene 179 (1996) 141–146. [20] T.F. Hogan, R.S. Riley, J.G. Thomas, Rapid diagnosis of acute eosinophilic pneumonia (AEP) in a patient with respiratory failure using bronchoalveolar lavage (BAL) with calcofluor white (CW) staining, J. Clin. Lab. Anal. 11 (1997) 202–207. [21] J.R. Albani, Y. Plancke, Interaction between calcofluor white and carbohydrate of ␣1 -acid glycoprotein, Carbohydr. Res. 314 (1998) 169–175; J.R. Albani, Y. Plancke, Interaction between calcofluor white and carbohydrate of ␣1 -acid glycoprotein, Carbohydr. Res. 318 (1999) 194–200. [22] N. Tayeh, T. Rungassamy, J.R. Albani, Fluorescence spectral resolution of tryptophan residues in bovine and human serum albumins, J. Pharm. Biomed. Anal. 50 (2009) 109–116. [23] J.R. Albani, Effect of binding of calcofluor white on the carbohydrate residues of ␣1 -acid glycoprotein (orosomucoid) on the structure and dynamics of the protein moiety. A fluorescence study, Carbohydr. Res. 334 (2001) 141–151. [24] C.N. Pace, F. Vajdos, L. Fee, G. Grimsley, T. Gray, How to measure and predict the molar absorption coefficient of a protein, Prot. Sci. 4 (1995) 2411–2423. [25] J.R. Lakowicz, Principles of Fluorescence Spectroscopy, Kluwer Academic/Plenum, New York, 1999. [26] J.R. Albani, Principles and Applications of Fluorescence Spectroscopy, Blackwell, London, 2007. [27] M.G. Badea, L. Brand, Time-resolved fluorescence measurements, Methods Enzymol. 61 (1971) 378–425. [28] J. Yguerabide, Nanosecond fluorescence spectroscopy of macromolecules, Methods Enzymol. 26 (1972) 498–578. [29] J.R. Albani, Fluorescence lifetimes of tryptophan: structural origin and relation with So → 1 Lb and So → 1 La transitions, J. Fluoresc. 19 (2009) 1061–1071. [30] L. Ragona, F. Pusterla, L. Zetta, H.L. Monaco, H. Molinari, Identification of a conserved hydrophobic cluster in partially folded bovine -lactoglobulin at pH 2, Fold Design 2 (1997) 281–290. [31] E.A. Burstein, N.S. Vedenkina, M.N. Ivkova, Fluorescence and the location of tryptophan residues in protein molecules, Photochem. Photobiol. 18 (1973) 263–279. [32] J. Albani, Motions studies of the human ␣1 -acid glycoprotein (orosomucoid) followed by red-edge excitation spectra and polarization of 2-p-toluidinylnaphthalene-6-sulfonate (TNS) and of tryptophan residues, Biophys. Chem. 44 (1992) 129–137. [33] J.R. Albani, New insights in the interpretation of tryptophan fluorescence, J. Fluoresc. 17 (2007) 406–417. [34] J.R. Albani, Sub-structures formed in the excited state are responsible for tryptophan residues fluorescence in -lactoglobulin, J. Fluoresc. 21 (2011) 1683–1687. [35] M. Amiri, K. Jankeje, J.R. Albani, Origin of fluorescence lifetimes in human serum albumin. Studies on native and denatured protein, J. Fluoresc. 20 (2010) 651–656. [36] J.R. Albani, Origin of tryptophan fluorescence lifetimes. Part 1. Fluorescence lifetimes origin of tryptophan free in solution, J. Fluoresc. (2013) (in press). [37] J.R. Albani, Origin of tryptophan fluorescence lifetimes. Part 1. Fluorescence lifetimes origin of tryptophan in proteins, J. Fluoresc. (2013) (in press). [38] D. Kmiecik, J.R. Albani, Effect of 1-aminoanthracene (1-AMA) binding on the structure of three lipocalin proteins, the dimeric -lactoglobulin, the dimeric odorant binding protein and the monomeric ␣1 -acid glycoprotein. Fluorescence spectra and lifetimes studies, J. Fluoresc. 20 (2010) 973–983.
