Fluorescence quenching studies on binding fluoreno-9-spiro-oxazolidinedione to human serum albumin

Chem.-Biol. Interactions, 84 (1992) 221-228

221

Elsevier Scientific Publishers Ireland Ltd.

FLUORESCENCE QUENCHING STUDIES ON BINDING FLUORENO-9SPIRO-OXAZOLIDINEDIONE TO HUMAN SERUM ALBUMIN J. GONZALEZ-JIMENEZ a, H. JACQUOTTE b and I. CAYRE a

aDepartamento de Quimica-Fisica Farmaceutica, Facultad de Farm~cia, Universidad Complutense, Ciudad Univevsitaria, 28040 Madrid (Spain) and bFacult~ de Pharmacie, Universite de Lille H (France) (Received February 10th, 1992) (Revision received July 6th, 1992) (Accepted July 10th, 1992)

SUMMARY

Human serum albumin fluorescence quenching by fluorene-9-spiro-oxazolidinedione has been analyzed as a function of temperature. Such temperature dependence suggests that the mechanism of the quenching process is static in origin. This type of quenching implies that a non-fluorescent complex between oxazolidinedione and serum albumin has been formed and following the Stern-Volmer relationship we have calculated the binding constant. Thermodynamic parameters were also determined. The positive and large values of entropy and the negative value for enthalpy suggest that both hydrophobic and electrostatic interactions may play an important role in the stabilization of the complex. Finally, the irreversible changes in the spectral properties of HSA are interpreted in binding terms.

Key words: Quenching -

Fluorescence -

Binding --

Albumin -

Oxazolidinediona

INTRODUCTION

Fluorene-9-spiro-5'-oxazolidinedione (FSOD) was obtained by a new synthesis procedure keeping in mind the structure-activity relationship for these compounds, in an attempt to provide guidelines for the synthesis of spiro-5'oxazolidinedione substituents [1]. Substituents on the carbon in position 5 appear important for the selectivity of the oxazolidinediones both as antagonist of penCarrespondence to: J. Gonz~ez-Jim~nez, Departamento de Quimica-Fisica Farmaceutica, Facultad de Farmacia, Universidad Complutense, Ciudad Universitaria, 28040 Madrid, Spain. 0009-2797/92/$05.00 © 1992 Elsevier ScientificPublishers Ireland Ltd. Printed and Published in Ireland

222

tylenetetrazol in animals and as clinically useful agents in the therapy of absence seizures. This compound has been tested in mice for prevention of pentylenetetrazolinduced seizures, showing a low protective effect however, the motor activity was inhibited (Marcuerquiaga, unpublished). Although oxazolidinediones do not bind significantly to plasma proteins [2], it is important to provide information about binding studies because pharmacological activities of many natural and synthetic compounds are presumed to be due to their interaction with plasmatic proteins, mainly human serum albumin (HSA). In this paper, we have used the fluorescence-quenching method to study the binding of FSOD to HSA following the review of Eftink and Ghiron [3] and the works of Lehr [4] and Calhoun [5]. Thermodynamic parameters were determined from the temperature dependence of the apparent binding constant in order to analyse the binding process. Furthermore, the study of the temperature dependence on the fluorescence quenching can also be used to characterize the quenching as being predominantly dynamic or static [6]. EXPERIMENTAL

Fluorene-9-spiro-5'-oxazolidinedione was synthetized by Galvez and col. as reported in [1,7]. Its purity was checked by melting point and infrared, after recrystallization from ethanol. Furthermore, it was characterized by mass and NMR spectra. Human serum albumin, essentially fatty acid free, fraction V, was obtained from sigma Chemical Co. (St. Louis, MO.) and was used without further purification. Stock solution of HSA was prepared 0.01 mM in phosphate buffer (pH -7.4 ± 0.1), and its concentration determined spectrophotometrically with extinction coefficient at 277 nm of 36 600 L/mol cm [8]. Stock solution of FSOD was prepared in the same phosphate buffer containing 2.5 rag/100 ml and stored in the dark at 4°C. Dilutions of drug stock in the appropriate buffer were prepared immediately before use and their concentrations determined spectrophotometrically by using a molar extinction coefficient at 271 nm of 14 500 L/tool cm (Gonzalez, unpublished). Each experiment was performed at a constant protein concentration, - 0.055 mM, following the same method described earlier [9]. The concentration range for FSOD was 0.002-0.016 mM. Absorbanee measurements were taken in a UVICON-940 spectrophotometer. Fluorescence titration experiments were performed at 15 °, 25 ° and 37°C, using a Perkin-Elmer MPF-44-A spectrofluorometer equipped with a jacked cellholder, connected to a constant temperature bath thermostated to ± 0.5°C. The fluorescence intensity in the absence and in the presence of added FSOD was measured by exciting at 290 nm and recording the emission at 335 nm with a slitwidth of 4 nm.

223

RESULTS A N D DISCUSSION

Quenching data were treated by the usual S t e r n - Volmer relationship expressed as:

Fo]F =

1

+ KM

(1)

where F0 and F are the fluorescence intensities in the absence and presence of quencher, respectively, M is the concentration of quencher and K is the quenching constant. Both static and dynamic quenching mechanisms can be explained by the above equation. In the absence of lifetime measurements, studies of the temperature dependence on the quenching profiles can be used to characterize the quenching as being dynamic or static. Figure 1 shows the results of the fluorescence titrations at 15 °, 25 ° and 37°C plotted according to the Stern-Volmer equation, in the concentration range 0 - 0.016 mM quenching agent. In this representation the intercept of the linear plots for all three temperatures tend to 1.0 and the slopes become lower when the temperature increases, indicating that the fluorescence quenching mechanism observed in this study may be predominantly static. This is presumably due to the formation of a non-fluorescent complex between HSA and FSOD because 2.5

1.5

0.6

0

0

I

I

I

5

10 M e6

16

15C

I

25C

~

20

37C

Fig, 1. Stern-Volmer plot of the fluorescence quenching and its temperature dependence. Slopes a r e represented for 15 °, 25 ° and 37°C. The excitation wavelength 290 nm. The emission wavelength 335 nm.

224

its tryptophanyl fluorescence was quenched about 50-60% depending on temperature. Furthermore, the fluorescence maximum of HSA is located at 335 nm (Fig. 2) and shifts to lower wavelengths in the presence of FSOD at 0.016 mM concentration. These spectral differences were remarkably reproducible with each temperature. Hence the irreversible changes in both shape and the maximum fluorescence wavelength of HSA observed by adding FSOD could be associated with the interaction of the quencher with the protein [10]. Thus, the Stern-Volmer slopes in Fig. 1 can be interpreted as association constants or binding constants of the complex FSOD-HSA for the corresponding equilibrium: FSOD + HSA ~

FSOD-HSA

(2)

according to the model for static quenching described by Eftink and Ghiron [11] and more recently used by Swades [12]. ~ In order to better understand the quenching process, it is important to know that K = Kqro where Kq is the bimolecular quenching constant and r0 is the

F-

3(~0

350

400

nm

Fig. 2. Fluorescence emission spectra of HSA (A) in the absence, (B) in the presence of 0.016 mM FSOD. Fluorescence intensity is expressed in arbitrary units. Excitation wavelength 290 nm. Slit width 4 nm. Temperature 15°C. Fluorescence quenching 52%. The same pictures were obtained at 25 ° and 37°C.

225 lifetime of the HSA in the absence of quencher and assuming its published value of 6 ns (see p. 354 of Ref. 6), the rate of the Kq parameter summarized in Table I is faster than any diffusion rate constant in solution so, the quenching mechanism cannot be dynamic. This argument can also be used as a basis for comparison between static and dynamic quenching mechanisms. As can be seen in Table I, the binding constant decreases with increasing temperature. This is probably the result of a decreased stability of complex and thus lower values of the binding constant [13]. Based on the temperature dependence of the binding constant, we have carried out the thermodynamic study of the interaction between FSOD and HSA. The Gibbs free energy was determined from the binding constant according to the relationship: AG = -RTInK

(3)

The binding enthalpy was estimated from the plot of the temperature dependence of the binding constant according to the van't Hoff relationship: (4)

d InK/d (l/T) = - A H / R

The entropy was calculated from the Gibbs free energy and the enthalpy as: AS =

-(AG

-

AH)/T

(5)

The results of the thermodynamic parameters are presented in Table I. Figure 3 shows the van't Hoff plot used to estimate the enthalpy of the FSOD binding reaction as a function of temperature. Such a positive linear dependence of lnK on 1/T is consistent with an exothermic molecular association process. The data were fitted to a straight line, thus neglecting possible heat-capacity changes that would result in curvature in the van't Hoff plots, especially given the probable small value of the heat capacity in analogy with other drug-HSA interactions [14]. Values for the Gibbs free energy presented in Table I show a very slight increase with temperature, as a result of the low lnK variation with temperature. TABLE I BINDING CONSTANTSAND THERMODYNAMICPARAMETERSFOR THE INTERACTION FSOD-HSA Each value represents the average of four experiments. In all cases the standard deviation was smaller than 5%. The value of enthalpy is - 2.05 kJ/tool. This value was obtained from the slope in Fig. 3. AG(kJ/m~)

AS (K)(J/kmol)

Kq1013(M-Is-1)

69.14 50.20

-26.67 -26.81

85.49 83.06

0.84

42.11

-27.44

81.90

0.70

T (K)

K

288 298

310

10-3(M -1)

1.15

226 12

InK

11.5

11

10.5

1 0 -

--

3.2

r

i

3.3

3.4

3.5

1/T exlo-3 Fig. 3. Dependence of InK on 1/T from a linear van't Hoff plot pertaining to the formation of FSOD - HSA complex over the temperature range studied. Values for K and T are given in Table I.

Following the same Table I, we find positive values of entropy in the range of 81.9- 85.5 J/K mol and they seem to decrease with temperature. The same order of magnitude found for binding chlorpheniramine to HSA [15] From the values of the thermodynamic parameters for the binding process between FSOD and HSA shown in Table I, it seems that the major contributing factor in the stabilization of the FSOD-HSA complex is entropic, rather than enthalpic in origin. It is also important to point out that human serum albumin contains only one tryptophan residue lying in a hydrophobic hole in the middle of the molecule [16] capable of carrying fats and other hydrophobic material. The values of the entropic parameters obtained in this study may be appropriate for the formation of hydrophobic bonds in the protein molecule [17]. On these bases, such hydrophobic interactions may be the dominant link in the formation of FSOD- HSA complex. However, a single type of interaction is not common in binding drugs to HSA [18,19]. FSOD carries the immune group which exhibits an acid character and, there is also an electronic dislocalization around its aromatic rings. Thus, we cannot discard some other types of interactions between ligand and protein. The foregoing considerations and the negative value for the enthalpy, -2.05 kJ/mol, lead us to think that electrostatic contributions and hydrogen bonds may take part in the binding process.

227

Finally, it is interesting to consider that the static quenching is a measure of the ability of a quencher molecule to exist next to the tryptophan residue in the HSA. The fact that changes either in the shape or the wavelength of HSA fluorescence spectrum were observed by adding FSOD, suggest the possibility that the tryptophan is present in the protein binding site [20]. Therefore, we think of the blue shift in the emission spectra can be due to the perturbation in the tryptophan microenvironment by the FSOD and subsequent unfolding of the globular structure [3]. The same effects have been observed by Chen in the HSA by mercuric ions [21]. He has suggested that the emission spectrum of the partially quenched HSA and significantly shifted compared with the native protein, may be related to tryptophan fluorescence which is largely quenched by mercuric ions while the tyrosine fluorescence appears to be relatively intact, because each one of these groups lies in different microenviroments. Taking into account the above considerations, the K values found in this work and the very small concentration range of quencher used, the FSOD is apparently able to penetrate into the tryptophan hydrophobic microenvironment generating a conformational change in the albumin since globular proteins, and particularly serum albumin due to its flexibility, may undergo changes in protein conformation [3,10,21]. REFERENCES 1 E. Galvez, G.G. Trigo, M. Martinez and N. Cabezas, Synthesis and structural study of ciclopentane, indene and fluorene spiro-derivatives, J. Heterocyclic. Chem., 20 (1983) 13-16. 2 A. Goodman and T. Gilmans, The Pharmacological Basis of Therapeutics, Macmillan Press, USA, 1990, p. 453. 3 M.R. Eftink and C.A. Ghiron, Fluorescence quenching studies with proteins, Anal. Biochem., 114 (1981) 119-227. 4 S. Lehrer, Solute perturbation of protein fluorescence. The quenching of the tryptophyl fluorescence of model compounds and of lysozyme by iodide ion, Biochemistry, 10 (1971) 3254-3263. 5 D.B. Calhoun, J.M. Vanderkooi and S.W. Englader, Penetration of small molecules into proteins studied by quenching of phosphorescence and fluorescence, Biochemistry, 22 (1983) 1533-1539. 6 J.R. Lacowicz, Principles of Fluorescence Spectroscopy, Plenum Press, NY, 1983. 7 G.G. Trigo, M. Martinez, E. Galvez and N. Cabezas, Synthesis and structural study of quinuclidine spiro-derivatives, J. Heterocyclic Chem., 18 (1981) 1507-1511. 8 T. Peters, Jr. The Plasma Proteins, Vol. 1, 2nd edn., F.W. Putman (Ed.), Academic Press, NY, 1975, p. 147. 9 J. Gonzalez-Jimenez, G. Frutos, I. Cayre and M. Cortijo, Chlorpheniramine binding to human serum albumin by fluorescence quenching measurements, Biochimie, 73 (1981) 551- 556. 10 D.C. Clark, J. Smith and D.R. Wilson, A spectroscopic study of the conformational properties of foamed bovine serum albumin, J. Colloid Interface Sci., 121 (1988) 136-147. 11 M.R. Eftink and C.A. Ghiron, Exposure of tryptophanyl residues and protein dynamics, Biochemistry, 25 (1977) 5546-5551. 12 J.K. Swades, P.W. Mui and A. Scheraga, Thermodynamics of the quenching of tyrosyl fluorescence by dithiothreitol, Biochemistry, 26 (1987) 5761- 5769. 13 M.R. Eftink and C.A. Ghiron, Fluorescence quenching of indole and model micelle system, J. Phys. Chem., 5 (1976) 486-493.

228

14 J.B. Chaires, Thermodynamics of the Daunomycin-DNA interactions: ionic strength dependence of the enthalpy and entropy, Biopolymers, 24 (1985) 403-419. 15 J. Gonz~tlez-Jim~nez, G. Frutos and I. Cayre, Influencia de los acidos grasos en la unibn dei maleato de clorfeniramina a la seroalbfLmina humana, Anal. Chem, 86 (1990) 363-365. 16 A.G. Walton and F.C. Maenpa, Application of fluorescence spectroscopy to the study of proteins at interfaces, J. Colloid Interface Sci., 2 (1979) 265-278. 17 J.B. Chaires, N. Dattagupta and D.M. Crothers, Deoxyribonucleic acid. Equilibrium binding studies on interaction of Daunomycin with deoxyribonucleic acid, Biochemistry, 21 (1982) 3940 - 3946. 18 J. Nishijo, N. Morita, S. Asada, H. Nakae and E. Iwamote, Interaction of Theophylline with bovine serum albumin and competitive displacement by benzoic acid, Chem. Pharm. Bull., 33 (1985) 2648-2653. 19 A. Shrake and F.D. Ross, Biphasic denaturation of human albumin due to ligand. Redistribution during unfolding, J. Biol. Chem., 263 (1988) 15392-15399. 20 P. Midoux, P. Walh, J.C. Auchet and M. Monsigny, Fluorescence quenching of tryptophan by Trifluoroacetamida, Biochim. Biophys. Acta, 281 (1984) 16-25. 21 F.R. Cben, Fluorescence quenching due to mercuric ion interaction with aromatic aminoacids and proteins, Arch. Biochem. Biophys., 142 (1971) 552-564.