Binding of methylumbelliferyl mannoside to concanavalin a under high pressure

Biochimica et Biophysica Acta, 790 (1984) 87-90 Elsevier

87

BBA Report BBA 30076

BINDING OF METHYLUMBELLIFERYLMANNOSIDE TO CONCANAVALIN A UNDER HIGH PRESSURE RICHARD B. THOMPSON * and JOSEPH R. LAKOWlCZ ** Department of Biological Chemistry, University of Maryland School of Medicine, 660 W. Redwood Street, Baltimore, AID 21201 (U.S.A.) (Received April 17th, 1984)

Key words: Carbohydrate binding; Concanavalin A; Mannose derivative," High pressure

The equilibrium binding of 4-methylumbelliferyl a-D-mannopyranoside to concanavalin A was measured by changes in fluorescence quenching observed at pressures ranging from 1 to 2000 bar (1974 atmospheres). From the pressure-induced changes in the apparent Ka, we calculated volume changes for the association reaction of - 2 . 5 and - 1.7 _ 1 m l / m o l for concanavalin A in its dimeric and tetrameric forms, respectively. This carbohydrate-binding reaction is less pressure-sensitive than other protein-ligand interactions that have been studied under pressure. The volume change observed is comparable to that expected for a reaction involving hydrogen bond formation, in a non-polar environment.

The effects of high pressure on biomolecules have been the focus of many recent experiments, due in part to the information that may be obtained about molecular interactions and their thermodynamics [1-3]. The structures of membranes, enzymes and nucleic acids, and the binding of some small molecules to the latter two groups have all received some study under pressure by various spectroscopic methods. The binding of carbohydrates to proteins has received little attention, however, despite the importance of such reactions in metabolism, cellular recognition and the immune response. While most methods for measuring equilibrium binding are difficult to adapt for use at high pressure, optical methods have proven widely useful for such studies [4-6]. We used fluorescence methods to measure the binding of the fluorescent-labeled monosaccharide, 4-methyl* Current address: Naval Research Laboratories, Washington, DC 20375, U.S.A. ** To whom correspondence should be sent. 0167-4838/$03.00 © 1984 Elsevier Science Publishers B.V.

umbelliferyl a-D-mannoside (MU-mannoside), to the lectin from jack bean, concanavalin A. The binding of MU-mannoside to concanavalin A is accompanied by essentially complete quenching of the fluorescence of the 4-methylumbelliferylmoiety [7], which itself contributes little to the binding [8]. It is therefore a convenient model for saccharide binding. This reaction has been well characterized at 1 atm by fluorescence, equilibrium dialysis, differential absorption and kinetic methods [8,9]. The change in fluorescence may be simply related to the association constant for the reaction [8]: log t/ F---Z-~-) r o - F ~ = logKa +log(e0

Lo( Fo -

To F ) )

(1)

F0, F and Foo are the fluorescence intensities of MU-mannoside in the presence of zero, the experimental, and infinite concentrations of concanavalin A, respectively; and P0 and L 0 are the total concentrations of concanavalin A and MU-mannoside, respectively. Thus by observing changes induced by pressure in the emission of an MU-

88

mannoside solution that is partly quenched by addition of concanavalin A, we can calculate the change in K~. Applying pressure may trivially affect the observed intensity by introducing certain minor artifacts. Among these are changes in the fluorophor's absorption [10], emission [11] and its quantum yield, changes in the pH of the solvent [15], distortion of the optics of the pressure vessel, or compression of the solvent. In fact, we found that the excitation and emission spectra of MUmannoside did shift approx. 1 nm at 2000 bar. Fortunately, the effects of these artifacts may be compensated for by comparing the fluorescence intensities of MU-mannoside samples under pressure in the presence and absence of concanavalin A. Since the MU-mannoside is excited at 318 nm, concanavalin A makes no significant contribution to the absorbance or emission of the sample. The fluorescence intensities observed for MU-

14

of° ~o/O/0~

Q: '< >.- 1 2

/ o i

w -

E I0

~.i~...~o

u.

i

I

o .----~'-~°~ h

400

i

i

800 1200 PRESSURE, BAR

I

1600

I

2000

Fig. 1. Relative fluorescence intensities of MU-mannoside ([2) and MU-mannoside plus concanavalin A ( O ) as a function of pressure at 2 5 ° C and pH 5.5. Triangles (v) indicate values observed as pressure was released. 4-methylumbelliferyl a-Dmannopyranoside and concanavalin A, type V, were both from Sigma; the latter was freed of nicked chains and stored as previously described [12]. Experiments were performed at pH 5.5 and 7.5 in the acetate and Tris buffers used by Senear and Teller [12]. Pressure experiments were performed in an apparatus essentially identical to that of Paladini and Weber [24]. Fluorescence intensities were measured on an SLM 8000 fluoremeter, in comparison with an external standard of 2,5-diphenyloxazole. Excitation was at 318 nm (4 nm bandpass) through a Coming 7-54 filter; emission was observed through a O-51 cutoff filter to remove Raman scatter. The fluorescence lifetime was measured as 0.4 ns as described previously [25]; phase sensitive detection [26] showed that a homogeneous population of emitters was present.

mannoside in the presence and absence of concanavalin A at pH 5.5 and 25°C are depicted in Fig. 1. We chose concentrations of MU-mannoside and concanavalin A such that the former would be approx. 30-50% quenched (and therefore 30-50% bound), to ensure the maximum signal change for any change in K a [2]. From the relatively small difference pressure causes in the intensity of both the quenched and unquenched samples, it is apparent that pressure has little effect on the binding. If the pressure had strongly enhanced the binding, the intensity of the MU-mannoside in the presence of concanavalin A would have decreased sharply in comparison to the unquenched sample. The data shown in Fig. 1 were measured at pH 5.5; similar results were obtained at pH 7.5. At pH 5.5 and 7.5, concanavalin A is present as a dimer and a tetramer, respectively [12]; the latter has a slightly higher affinity for MU-mannoside [8]. Pressure can dissociate oligomeric proteins [13,14], and thus conceivably cause a pressure-dependent change in apparent Ka. However, we note that at these concentrations concanavalin A's degree of association depends solely on pH [12]. From the small changes in pK~ with pressure known for acetate (pH 5.5) and Tris (pH 7.5) buffers [15], we expect pressure to have a negligible effect on concanavalin A's degree of association. Association c o n s t a n t s calculated from pressure-dependent fluorescence intensities using Eqn. 1 are depicted in Fig. 2. The K~ values measured at 1 atm are in good agreement with the values of Loontiens et al. [8]. It is clear from these data that pressure has only a slight effect on the equilibrium constant of the binding reaction. The pressure sensitivity of the equilibrium constant of a reaction is related to the volume difference (AV) between reactant(s) and product(s) by: d(ln Ka) dP

-AV RT

(2)

By substituting the values in Fig. 2 into Eqn. 2 and expressing pressure in units of cal/ml (1 cal/ml = 41.8 bar), we calculate volume changes of - 2 . 6 and - 1 . 7 _ 1 m l / m o l for the association at pH 5.5 and 7.5, respectively. The negative sign indicates that pressure strengthens the association. Also indicated in Fig. 2 is the pressure dependence

89

5.0C

,

/

/

4.00

/

/

/

// /

"o •i K- 3 . 0 0

•

•

~

o

-o

o

•

2.00 X.....~.

& v = + Io m l / m O l l ""~

""-~

.~.~.

___.~ ._..._~

1.00

PRESSURE, BAR

Fig. 2. Effect of pressure on the association constant for MU-mannoside binding to concanavalln A at pH 5.5 ((3, v) and pH 7.5 (O, v), 25 o C. The similarity of values obtained while compressing the sample (O, 0) to those obtained as pressure was released (v v) suggests that no irreversible denaturation of the protein has taken place. The two dashed lines indicate the pressure dependence expected for K a if AV were + 10 ml/mol.

expected for K a if the volume difference for the reaction were + 10 ml/mol. Our results are distinctly different from those of Fitos et al. [16], who found AV = + 20 + 7 m l / m o l for the same reaction using ultraviolet differential absorbance spectra. Although both our method and theirs yielded similar results at 1 atm [8], it is uncertain how pressure affects the application of each. Our method should eliminate the effect of the artifacts mentioned above, but we have no pressure-independent intensity standard to test this. Moreover, we have assumed that F~o remains zero (concanavalin A completely quenches MUmannoside) as the pressure is increased. We note that the small changes in fluorescence excitation and emission spectra we observed are unlikely to affect our results because of the comparative method employed. By comparison, ultraviolet difference spectra are sensitive to such shifts. However, since we have no independent measure of the absorbance of either the free or the bound MUmannoside, we cannot identify with certainty the source of the difference between our results and those of Fitos et al. [16].

The value of AV we obtained, although smaller than most AV values for protein-ligand binding [3], is reasonable if the intermolecular interactions likely to be involved are considered. In particular, concanavalin A is known to bind particular saccharide steroisomers with great specificity, and others not at all [17]. Also, the thermodynamic driving force for the binding is mainly enthalpy [8]. These facts suggest that hydrogen bonds form between concanavalin A and saccharide ligands during binding. X-ray crystallographers have observed carbohydrate-protein hydrogen bonds in lysozyme [18] and hexokinase [19], two enzymes which also bind saccharides with great stereospecificity. Since AG for the binding reaction is - 6 . 2 kcal/mol [8], we may reasonably assume that MU-mannoside binding is accompanied by net formation of perhaps two hydrogen bonds. Hydrogen bond formation has been found by several experimenters to be favored by pressure [20-23]; the volume changes measured upon formation ranged from - 1 . 3 to - 4 . 6 ml/mol. Thus one might expect, as a first approximation, that a binding reaction involving formation of two hydrogen bonds would be weakly pressure-dependent, exhibiting a volume change of - 2 to - 1 0 ml/mol. The authors are grateful to Drs. Ronald J. Doyle and Harry L.T. Mobley for helpful discussions, and the latter for gel electrophoresis of the concanavalin A preparations. This work was supported by grant GM-293118 from the National Institutes of Health, with instrumentation purchased in part by the National Science Foundation. References 1 Jaenicke, R. (1981) Annu. Rev. Biophys. Bioeng. 10, 1-67 2 Weber, G. and Drickamer, H.G. (1983) Q. Rev. Biophys. 16, 89-112 3 Heremans, K. (1982) Annu. Rev. Biophys. Bioeng. 11, 1-21 4 Torgerson, P.M., Drickamer, H.G. and Weber, G. (1979) Biochemistry 18, 3079-3083 5 Li, T.M., Hook, J.W., Drickamer, H.G. and Weber, G. (1976) Biochemistry 15, 5571-5508 6 Torgerson, P.M., Drickamer, H.G. and Weber, G. (1980) Biochemistry 19, 3957-3960 7 Dean, B.R. and Homer, R.B. (1973) Biochim. Biophys. Acta 322, 141-144

90 8 Loontiens, F.G., Clegg, R.M. and Jovin, T.M. (1977) Biochemistry 16, 159-166 9 Clegg, R.M., Loontiens, F.G. and Jovin, T.M. (1977) Biochemistry 16, 167-175 10 Brey, L.A., Schuster, G.B. and Drickamer, H.G. (1979) J. Chem. Phys. 71, 2765-2772 11 Politis, T.G. and Drickamer, H.G. (1981) J. Chem. Phys. 75, 3203-3210 12 Senear, D.F. and Teller, D.C. (1981) Biochemistry 20, 3076-3083 13 Paladini, A.A. and Weber, G. (1981) Biochemistry 20, 2587-2593 14 Thompson, R.B. and Lakowicz, J.R. (1984) Biochemistry 23, 3411-3417 15 Neuman, R.C., Kauzmann, W.C. and Zipp, A. (1973) J. Phys. Chem. 77, 2687-2691 16 Fitos, I., Heremans, K. and Loontiens, F.G. (1979) React.' K.inet. Catal. Lett. (Budapest) 12, 393-397 17 Goldstein, I.J., Hollerman, C.E. and Smith, E.E. (1965) Biochemistry 4, 876-883

18 Blake, C.C.F., Koenig, D.F., Mair, G.A., North, A.C.T., Philipps, D.C. and Sarma, V.A. (1965) Nature 206, 757-763 19 Anderson, C.M., Stenkamp, R.E., McDonald, R.C. and Steitz, T.A. (1978) J. Mol. Biol. 123, 207-219 20 Suzuki, K. and Tsuehiya, M. (1975) Bull. Chem. Soc. Jpn. 48, 1701-1704 21 Josefiak, C. and Schneider, G.M. (1979) J. Phys. Chem. 83, 2126-2128 22 Josefiak, C. and Schneider, G.M. (1980) J. Phys. Chem. 84, 3004-3007 23 Fishman, E. and Drickamer, H.G. (1956) J. Chem. Phys. 24, 548-553 24 Paladini, A.A. and Weber, G. (1981) Rev. Sci. Instrum. 52, 419-427 25 Lakowicz, J.R., Cherek, H. and Baiter, A. (1981) J. Biochem. Biophys. Methods 5, 131-146 26 Lakowicz, J.R. and Cherek, H. (1981) J. Biochem. Biophys. Methods 5, 19-35