side (Derwenda et al., 1989) and pea lectin-methyl 3,6-di-O- .... the solution with o-toluidine reagent in trichloroacetic acid at 100 "C. (Sigma Diagnostic Kit 635).
THEJOURNAL OF BIOLOGICAL CHEMISTRY
Vol. 268, No. 11, Iseue of April 15, pp. 7668-7677, 1993 Printed in U.S.A.
Thermodynamics of Monosaccharide Binding to ConcanavalinA, Pea (Pisum sativum)Lectin, and Lentil (Lens culinaris)Lectin* (Received for publication, October 23, 1992)
Frederick P. Schwarz From the Center for Advanced Research in Biotechnology and the National Institute of Standards and Technology, 9600 Gudekky Drive, Rockville, Maryland 20850
Kamil Deep PuriS, Rama G. BhatS, and Avadhesha Surolia From the Molecular Biophysics Unit, Indian Instituteof Science, Bangalore 560 012, India
Titration calorimetry measurements of the binding of methyl a-D-mannopyranoside (MeaMan), D-mannOpyranoside (Man), methyl a-D-glucopyranoside (MeaGlu), and D-glucopyranoside (Glu) to concanavalin A (Con A), pea lectin,andlentillectinwere performed at 281 and 292 K in 0.01 M dimethylglutaric acid-NaOH buffer (pH 6.9)containing 0.15 M NaCl and Mn+' and Ca+=ions. The site binding enthalpies, AH, are the same at both temperatures and range from -28.4 f 0.9 (MeaMan) to -16.6 f 0.5 k J mol" (Glu) for Con A, from -26.2 f 1.1 (MeaMan) to -12.8 f 0.4 k J mol" (MeaGlu) for pea lectin, and from -16.6 f 0.7 (MeaMan) to -8.0 f 0.2 k J mol" (MeaGlu) for lentil lectin. The site binding constants range from 17 f 1 X los M" (MeaMan to Con A at 281.2 K) to 230 f 20 M-' (Glu to lentil lectinat 292.6 K) and exhibithigh they are in the specificity for Con A where MeaMan:Man:MeaGlu:Glu ratio of 21:4:5:1,while the 5:2:1.5:l for pea lectin and corresponding ratio is 4:2:2: 1 for lentil lectin. The higherspecificity for Con A indicates more interactions between the amino acid residues at the binding site and the carbohydrate ligand than for the pea and lentil lectin-carbohydrate complexes. Thecarbohydrate-lectinbindingresults exhibit enthalpy-entropycompensation in thatA H b ( k J mol") = -1.67 f 0.06 X lo4 + (1.30 f O.lZ)T(K)A& (J mol-'K-'). Differential scanning calorimetrymeasurements on the thermal denaturationof the lectins and their carbohydrate complexes show that the Con A tetramer dissociates into monomers, while the pea and lentil lectin dimers dissociate into two submonomer fragments. At the denaturation temperature,one carbohydrate binds toeach monomer of Con A and thepea and lentil lectins. Complexation with the carbohydrate increases the denaturation temperature of the lectin and the magnitude of the increasesyield binding constants in agreement with the determinations from titration calorimetry.
* This study was partially supported by a grant from the Department of Biotechnology, Government of India and by Grant G-021 from National Institute of Standards and Technology (to A. S.) and constitutes a study under the Indo-United States scientific collaboration program. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solelyto indicate this fact. $ Supported as a Senior Research Fellow of the Council of Scientific and Industrial Research and a post-doctoral fellow of the Department of Biotechnology, Government of India, respectively.
Lectins are proteins which bind carbohydrates with a high degree of specificity. This makes them useful as probes for carbohydrate structure in various biological systems such as cell membranes, and as ideal model systems for elucidating protein-carbohydrate interactions. Although the legume lectins, concanavalin A (Con A),' pea lectin, and lentillectin are very similar in their structuralproperties, they exhibit different binding affinities for saccharide derivatives of the mannopyranoside and glucopyranoside configuration (Goldstein et al., 1965; Debray et al., 1981). In solution, Con A below pH 5.8, pea lectin, and lentil lectin exist as dimers, composed of 25,000-30,000 molecular weight units (Mr),require Ca2+and Mn2+ions for carbohydrate binding, and have one carbohydrate binding site per monomer. X-ray analysis of the carbohydrate-lectin complexes, Con A-a-methyl-D-mannopyranoside (Derwenda et al., 1989) and pea lectin-methyl 3,6-di-O(a-D-mannopyranosy1)-a-D-mannopyranoside (Rini et al., 1989), show that the carbohydrate binding sites are near the Ca2+binding site which is adjacent to the Mn2+binding site. Above pH 6.9,Con A exists as a tetramer, and there are structural differences between the monomers of Con A and of pea and lentil lectin (Einspahr et al., 1986). The monomers of Con A consist of a continuous polypeptide chain of237 amino acid residues, while the monomers of pea and lentil lectin are each composed of two different polypeptide chains, one 6,000 M, (a-chain) andone 18,000M, (@-chain) (Einspahr et al., 1986).The thermodynamics of the carbohydrate binding reaction is importantin elucidating how these structural properties are related to the carbohydrate binding properties of the legume lectins. In this investigation, titration calorimetry was employedto determine the thermodynamics of the carbohydrate-lectin binding reaction in terms of the binding constant ( K b ) , and changes in the free energy AG, the binding enthalpy (A&,), and the binding entropy (AS*).The binding reaction between the lectin monomer ( L M ) and the carbohydrate (S)is LM
+S
c*
LM.S
(Eq. 1)
in 0.150 M NaCl + 0.01 M 3,3'-dimethyl glutaric acid-NaOH (DMG) buffer with 0.001 M MnC12 and 0.001 M CaC12 at pH 6.90 & 0.05. A few of the Con A and pea lectin reactions were performed at pH 7.4 with the 0.01 M DMG replaced by 0.02 M sodium phosphate buffer and without the Mn2+ and Ca2+ salts. The titrationcalorimetry measurements were performed 'The abbreviations used are: Con A, concanavalin A; Glu, Dglucopyranoside; DSC, differential scanning calorimetry; MeaMan, a-methyl-D-mannopyranoside; Man, D-mannopyranoside; MeaGlu, a-methyl-D-glucopyranoside;Tris, tris(hydroxymethy1)aminomethane.
7668
Thermodynamics of Monosaccharide Binding to Lectirzs around 283 and 293 K. The carbohydrates were a-methyl-Dmannopyranoside (MeaMan), D-mannopyranoside (Man), amethyl-D-glucopyranoside (MeaGlu), and D-glucopyranoside (Glu). In addition, differences in the structural stabilities, particularly the affect of carbohydrate bindingon the thermal stabilities of the legume lectins, were investigated by differential scanning calorimetry (DSC). Previous DSC studies on the unfolding properties of Con A in the presence of carbohydrate ligands (Zahnley, 1981; Borrebaeck and Mattiasson, 1979) have shown that the presence of the ligand increases the thermalstability of the lectin. EXPERIMENTAL PROCEDURES
Materials-Con A, lentil lectin, 3,3'-dimethyl glutaric acid (DMG), Tris, and the monosaccharides were obtained from Sigma' and used without any furtherpurification. Pea lectin was prepared and purified to a purity> 95% as described previously (Bhattacharyya etal., 1985). Some samples of lentil lectin were also prepared and purified by a similar procedure (Bhattacharyyaet al., 1985). The specifications from Sigma stated that a 7% polyacrylamide electrophoresis gel for Con A with @-alaninebuffer (pH 4.5) exhibited one major and two faint minor bands and for lentil lectin with Tris-glycine buffer (pH 8.3) exhibited two diffuse major bands and one faint minor band. The two diffuse major bands for lentil lectin correspond tothe two isolectin dimers of lentil lectin, A and B, each with M, = 49,000. The binding properties of the isolectins are identical and isolectin B contains only 4 more lysine residues per monomer which accounts for their different electrophoretic properties (Howard et al., 1971). Since the binding properties are identical, the lentil lectin willbe considered as consisting of one species as was done in earlier studies (Bhattacharyya et al., 1985). A Pharmacia Phast native electrophoresis gel with 0.1 M Tris-HC1 buffer (pH 6.9) containing 0.2% bromphenol blue and 20% glycerolexhibited one major band for pea lectin and lentil lectin and one major band and a weak band for ConA. The glycerol, hydrochloric acid, bromphenol blue, Na2HP04, NaH'PO,, NaC1, MnC12, CaC12,and NaOH were of reagent quality. Preparation and Analysis of Solutions-The lectin solutions were prepared in the appropriate buffer by weight, dialyzed overnight in a large volume of the same buffer, and centrifuged to remove any insoluble lectin. The DMG buffer was prepared by adding dropwise aliquots of concentrated NaOH solution to 0.01 M 3,3'-Dimethylglutaric acid in 0.15 M NaCl to achieve the desired pH as determined by an Orion 811 pH meter equipped with a Corning EXL glass electrode. The phosphate buffer was prepared as described previously (Schwarz et al., 1991). The Tris buffer was prepared by adding HC1 dropwise to a 0.1 M solution of Tris to achieve pH 6.9 and then bromphenol and glycerol were added to make up, respectively, a 0.2 and a 20% solution for the native Phast gel. The concentrations of the protein solutions were determined at pH7.0 from UV spectroscopic measurements at 280 nm, where the absorbance of a 1%Con A solution is 12.4 (Goldstein et al., 1986), 1%pea lectin solution is 15.0 (Bhattacharyya et al., 1985), and a 1%lentil lectin solution is 12.6 (Goldstein et al., 1986). The monomer M, values were 25,000 for Con A (Edmundsen et al., 1971), 23,500 for pea lectin, and 24,500 for lentil lectin (Bhattacharyya et al., 1985). Solutions of the carbohydrates were prepared by weight in the dialysate to minimize differences between the protein buffer solution and ligand buffer solution in the titrationcalorimetry measurements. The Man and Glu solutions were analyzed by reacting an aliquot of the solution with o-toluidine reagent in trichloroacetic acid at 100 "C (Sigma Diagnostic Kit 635). The absorption at 650 nm of the bluegreen complex formed from reacting the carbohydrate with the otoluidine was compared to thatof standard glucose solutions to obtain the concentration of the sugar solution. The concentrations thus determined agreed to within 3% with the weight-calculated concentrations. Milligram quantities of the sugar were added directly to the 0.5 g of the lectin solution in the DSC cellfor the DSC measurements. Certain commercial equipment, instruments, and materials are identified in this paper in order to specify the experimental procedure as completely as possible. In no case does this identification imply a recommendation or endorsement by the National Institute of Standards and Technology, nor does it imply that thematerial, instrument, or equipment identified is necessarily the best available for the purpose.
7669
A native electrophoresis gel was run on the native and denatured samples of the legume lectins in 0.1 M Tris buffer containing 0.2% bromphenol blue and 20% glycerol at pH 6.8. The electrophoresis gelswere performed at 15 "C in a Pharmacia Phast System with native buffer strips. Calorimetric Titration Measurements-The calorimetric titrations of the carbohydrate ligand solutions into the lectin solution were performed with a Microcal Omega titration calorimeter as described previously (Schwarz et al., 1991). Briefly, 4-10-pl aliquots of the 530 mM ligand solution were added via a rotatingstirrer-syringe to the 0.1-1 mM protein solution contained in a 1.38-ml cell. The additions were 3 min apart to allow the exothermic heat peak accompanying each addition to return to the base line prior to the next addition. This was continued until the last titration peak was less than 25% of the initial peak. The total heat, Q,, was then fitted via a nonlinear least squares minimization method (Wiseman et al., 1989)to the total ligand concentration, X,, using the following equation (Yang, 1990), Qt
+ Xt/nMl + l/nKbMt - [ ( l + Xl/nMt + 1/nKbM,)*- 4Xt/nMt]'/')/2
= nMtAHaV{l
(Eq. 2)
where n is the number of sites, MI is the total protein monomer concentration, and V is the cell volume. The thermodynamic parameters are from the basic equation of thermodynamics, AG; =
- TASb
(Eq. 3)
where AGE = -nRTln[Kb]
(Eq. 4)
and n = number of moles, T is the absolute temperature, and R = 8.3151 J mol-' K-'. DSC Measurements-DSC measurements were performed with a Hart 7707 DSC heat conduction scanning microcalorimeter as described previously (Schwarz et al., 1991).The best least-squares fit of the two-state transition model, A nB,where A is the folded state and B is the unfolded state, to the data (Schwarz and Kirchhoff, 1988) yielded a van't Hoff enthalpy, A H " , from the ratio of the transition peak height (p) to the peak area (A,) as shown
-
A H u
= mRT;p/A,
(Eq. 5)
where for n = 1, m = 4.00 and for n = 4, m = 9.01, and Tp is the temperature of the peak maximum. The transition temperature, T,, is the temperature at half the peak area and thecalorimetric enthalpy is the peak area per number of moles of lectin monomer in the DSC cell. RESULTS
Titration Calorimetry Measurements-The results of a typical titration calorimetry measurement, which consisted of adding 5-pl aliquots of 7.0 & 0.2 mM MeaMan to 0.26 k 0.01 mM Con A in DMG buffer at pH 6.9 and 293.2 K, are shown in Fig. 1. The results exhibit a monotonic decrease in, the exothermic heat of binding from an initial heat of - 0.728 mJ with each addition. A plot of the total heat released as a function of total ligand concentration is shown in Fig. 2. Results of the best fits of Equation 2 to the slopes of similar plots for the carbohydrate ligands yield the Kb and values presented in Tables 1-111. Monomer concentrations of the lectin were used in the fitting procedures and the average stoichiometries from these fits were close to one. The results were independent of lectin concentration ( 5 1mM) and ligand concentration. The enthalpy of the carbohydrate dilution was negligible since no heat was observed upon the addition of 10 pl of 30 mM carbohydrate solution into the buffer solution. As can be seen in Tables 1-111, the binding reactions for Con A, pea lectin, and lentil lectin are all enthalpically driven. The enthalpies over the temperature range 281.2-298.2 K are the same indicating very little heat capacity change results from the binding reaction. With the exception of the ManCon A titrations at pH 7.4 and 289.5 K and Man-pea lectin titrations at pH 7.4 and 297.5 K, the binding enthalpies are
7670
Thermodynamics of Monosaccharide Bindingto Lectins
0
c
0
u W In L W
n
rn W
3
.-0 0
FIG. 1. Titration calorimeter results from adding 2 1 5 -aliquots ~1 of 7.0 f 0.2 mM MeM to 0.26 f 0.01 mM Con A in DMG buffer at pH 6.9 and 293.2 K.
L
u .d
E
m L
0
rn U
."
-c 13 U
W
I
0
a I
0780520260
I
I
I
I
I
1
1
FIG. 2. A plot of the total heat released as a function of total ligand concentration for the titration shown in Fig. 1. The solid line is the result of the best least squares fit of the data toEquation 2.
"
""
I-
I
0
2600
I
L
0
1040 1300 15602340 2080 1820 TIME (seconds)
.l
I
I
I
.2 .3 .4 TOTAL LIGAND CONCENTRATION [mM)
"Q-Q-4I
.5
.6
also independent of pH from 6.9 to 7.4 for Con A and pea amount of ligand required to inhibit agglutination of human lectin and from pH5.0 to 6.9 for MeaMan binding to ConA. red blood cells. In addition, the binding constants for each Con A> pea lectin > lentil lectin. The literature values (Ambrosin0 et al., 1987) of Kb = 2.5 f ligand decrease in the order The binding enthalpies for the legume lectins increase by 0.4 X lo3M" and A H b = 18 f 1 kJ mol" at pH4.5 and 298.15 almost a factor of three from MeaMan-Con A to Glu-lentil K for MeaGlu are the same as the correspondingaverage lectin binding. For Con A, the binding enthalpiesincrease in values at pH6.9 in Table I. The binding constants in going from MeaMan binding to the order of MeaMan c Man < MeaGlu c Glu, while for pea Glu binding at room temperature decrease by a factor of 21 and lentil lectin the order with regard to Glu > MeaGlu is for Con A, a factor of 5, for pea lectin, and a factor of 4 for reversed. For each ligand, the enthalpiesincrease in the order follow lentil lectin. In Table I, the binding constants close to room Con A < pea lectin< lentil lectin. The entropy changes temperature are in the order21 f 2:3.8 f 0.4:4.8 f 0.7:l.O for the same trends as the binding enthalpies. DSC Measurements-TypicalDSC scans of Con A, pea MeaMan:Man:MeaGlu:Glu which is close to the orderof the 0.01 M DMG buffer containing binding affinities of 16:4:4:1 from inhibition studies (Debray lectin, and lentil lectin in et al., 1981).In Table11, the corresponding order for pea lectin Mn2+, Ca2+, and0.150 M sodium chloride at pH 6.90 f 0.05 is 4.8 f 0.5:2.2 f 0.1:1.5 f 0.1:1 a t room temperature which are shown in Fig. 3. The concentrationsof the Mn2+ and Ca2+ is close to 4:2:1:1, the order of thebindingaffinity from ions were each 1 mM in thepea and lentil lectin solutions and inhibition studies (Debray et al., 1981). For lentil lectin, the 100 mM each in the Con A solutions. Lower concentrations order is 3.7 f 0.4:2.2 f 0.2:2.0 f 0.2:l.O which is the same as of the ions in the Con A solutions produced an additional 4:2:2:1,the orderof binding affinity from the inhibition stud-smaller, lower temperature transition peak (not shown) with ies (Debray et al. 1981).The relative affinities for these lectins a transition temperature, T,, close to that of the less theras obtainedby titration calorimetry are ingeneral agreement mally stable metallessform of Con A, apoconcanavalinA with the inhibition studies,which are based on the minimum (Schwarz, 1993). The metalless forms of pea and lentil lectin
7671
Thermodynamics of Monosaccharide Binding to Lectins
TABLEI Thermodynamicparameters of the binding of carbohydrates to concanavalin A in 0.15 M sodium chloridesolution and 0.01 M dimethylglutaricsodium hydroxidebuffer below p H 7 and 0.02 M sodiumphosphate buffer at p H 7.4
-
Carbohydrate
Methyl a-D-mannopyranoside
D-Mannopyranoside
Methyl a-D-glucopyranoside
D-Glucopyranoside
pH'
Temperature
Kb mM"
6.9 7.4 6.9 5.0, 7.4
K 281.3 f17 0.4 289.0 f 0.2 292.9f 0.5 298.0 f 0.4
-AG;
f1 11 f 1 11.9 f 0.7 7.0 f 0.3
22.8 22.4 22.9 21.9
- m b
kJ mol" f 0.1 27.5 4 0.2 23.1 f 0.1 29.2 f 0.1 26.5 28.4
f 2.2 f 0.3 f 0.3
-ASb
J mol" K" 17 f 8
Determinations
4 2 3 3
f 0.9'
2 f 1 22 f 1 15 f 19 20 f 3 10 f 4 -7 2f 4 18 f 12 14 f 4
3
0
3 2 3 1
f 5.7
6.9 7.4 6.9
283.2f 0.3 289.5f 0.8 292.6f 0.4
3.74 f 0.05 2.21 f 0.05 2.13 f 0.08
19.4 f 0 18.5 f16.4 0.1 18.6 f 0.1
22.3 f 1.1 f 1.2 23.9 f 3.6 23.1 f 0.8'
6.9 7.4 6.9 5.8
282.5 f 0.2 289.5 f 0.1 292.4 f 0.3 297.6 f 0.1
4.15 f 0.19 2.99 f 0.10 2.71 f 0.13 2.23 f 0.08
19.6 f 0.1 19.3 f 0.1 19.2 f 0.1 19.1 f 0.1
19.4 f 1.9 17.9 f 0 18.1 f 0.4 25.2 f 1.3 18.8 f 0.7'
6.9 6.9
2 8 2 . 6 f 0.5 292.3 f 0.1
0.82 f 0.08 0.56 f 0.04
15.8 f 0.2 15.4 f 0.2
16.1 f 2.2 17.1 f 1.4 16.6 f 0.5'
-5 f 1 -4 f 2 21 f 4 0 1f8 6 f 5 0
3
4 3
a At pH 6.9, the solution contained 1 mM each of MnC12 and CaC12.The concentrations of concanavalin A ranged from 0.18 to 0.43 mM, while the carbohydrate concentration ranged from 5.0 to 32.0 mM. Average values at pH 6.9.
TABLEI1 Thermodynamicparameters of the binding of carbohydrates to pea lectin in 0.15 M sodium chloridesolution and 0.01 M dimethylglutaricsodium hydroxide buffer below p H 7 and 0.02 M sodium phosphate buffer at p H 7.4 Carbohydrate pH" Temperature Kb -AGE "b -ASb Determinations K mM" kJ mol" J K mol" Methyl a-D-mannopyranoside 6.9 282.3 f 0.8 3 . 2 8 f 0.01 19.0 f 0 25.1 f221.2 3f 4 7.4 287.6f 0.2 2.69 f 0.05 18.9 f 0 22.2 f 0.2 112f 7 1 . 9 f 0.16 18.3 f27.3 0.2 f 0.5 314f 2 6.9 291.9 f 0.9 1.65 f 0.14 18.4 f 0.2 24.7 f 0.3 21 f 10 2 7.4 298.3 f 0.1 26.2 f 1.1' 27 f 5 D-Mannopyranoside
6.9 7.4 292.3 6.9 7.4
Methyl a-D-glucopyranoside
D-Glucopyranoside
282.0 f 0.3 287.3 f 0.1 f 0.9 297.5 f O . 1
1.67 f 0.2 1.02 f 0.14 0.87 f 0.02 0.79 f 0.01
17.4 f 0.3 16.6 22.2 f 0.3 16.5 f 0.1 16.5 f 0
25.5 f 2.1 f 6.9 24.8 f 0.8 1 7 . 7 f 3.3 25.2 f 0.4'
29 3 f7 0 29 f 3 0 29 f 8
2 4 2
6.9 289.2 7.4 6.9
281.5 f 0.4 f 0.3 291.4 f 0.4
0.94k 0.05 0.77 f 0.04 0.61 f 0.03
16.0 f 0.1 1 6 . 0 f 0.1 1 5 . 5 f 0.1
12.4 f 0.7 1 0 f 1.7 13.1 f 0.1 12.8 f 0.4'
-13 4f 3 -20 f 6 -8 f 1 -11 f 3
2 3
6.9 7.4 6.9
281.3 f 0.4 287.4 f 0.2 291.4 f 0.1
0.74 f 0.04 0 . 3 6 f 0.09 0.4 f 0.03
15.4 f 0.1 14.0 f 0.6 14.5 f 0.2
15.7 f 0.7 13.6 f 1.5 1 3 . 3 f 1.4 14.5 f 1.2'
0 0 0 0
4 3 3
At pH 6.9, the solution contained 1 mM each of MnCI, and CaC1,. The concentrations of pea lectin ranged from 0.18 to 0.56 mM, while the carbohydrate concentration ranged from 12 to 27.0 mM. Average values at pH 6.9.
'
areunstable at this pH (Bhattacharyya et al., 1985) since dialysis of lentil lectin against EDTA in 1 M acetic acid to remove Mn2+also leads to dissociation of the lectin into its a and /3 chains (Fliegerova et al., 1974). Upon completion of the scan, a precipitatewas observed in the cells. Upon rescanning the solutions, the transitionsdid not reappear, indicatingthat the transitions are irreversible. All the peaks exhibit an asymmetry which corresponds to a proteinundergoing dissociation as it unfolds. As shown by the solid lines in Fig. 3, the peaks are best fitted to a two-state transitionof a protein unfolding and dissociating into four subfragments which may or may not be identical in size as follows. A
B'
c)
+ B" + B"' + B""
or 4B
(Eq. 6)
Con A exists as a tetramer at pH 6.9, while pea and lentil
lectins exist as dimers at this pH. Apparently, Con A denatures into its monomeric units, while pea and lentil lectin undergo dissociation into theira and /'3 chains. This is further substantiated by a nativeelectrophoresis gel run on the native and denatured solutions of the threelectins. A native electrophoresis gel ofthe denatured pea and lentil solutionsexhibited only two major bands well below the middle of the gel where the bands for the wild type lectins appear. These bands are presumably the a and /'3 chains. Samples of the denatured Con A solutions did not exhibit these bands. There is a slight transition temperature dependence on the scan rate as shown in Table IV. A maximum increase in the scan rate from 8 to 55 K h" increases the transition and maximum temperatures of the denaturation transitions by 3 & 1 K but had no effect on the shapes of the transitions. This
Thermodynamics of Monosaccharide Binding to Lectins
7672
TABLE111 Thermodynamic parametersof the binding of carbohydrates to lentil lectin in 0.15 M sodium chloride solution and 0.01 M dimethylglutaricsodium hydroxide buffer Carbohydrate
pH"
Temperature
Kb
mM"
Methyl a-D-mannopyranoside
6.9 6.9
K 281.1 f 0.5 292.7 f 0.7
1.19 f 0.03 0.84 f 0.04
16.6 f 0.1 1 6 . 4 f 0.1
D-Mannopyranoside
6.9 6.9
281.9 f 0.2 292.4 f 0.6
0.96 f 0.02 0 . 5 0 f 0.01
16.1 & 0 15.1 f 0.1
Methyl a-D-glucopyranoside
6.9 6.9
282.0 f 0.6 292.4 +- 0.4
0.52 f 0.05 0.45 f 0.03
14.9 k 0 14.8 f 0.1
D-Glucopyranoside
6.9 6.9
281.9f 0.2 292.6 f 0.8
0.33 f 0.01 0 . 2 3 f 0.02
13.6 f 0.1 13.2 f 0.2
-AGE
At pH 6.9, the solution contained 1 mM each of MnCI, and CaC1,. The concentrations the carbohydrate concentration ranged from 8.3 to 35.1 mM. *Average values.
shift was determined by comparison with DSC measurements on the reversible unfolding transition of lysozyme in 0.2 M HC1-glycine-buffered solutions (pH 3.8) which also showed a slight transition temperatureincrease (1K) at thehigher scan rate when scannedunder the same conditions. Since this temperature increase is small and theshapes of the transitions are the same, the equilibrium two-state transitionmodel was applied to these transitions instead of the irreversible model of Sanchez-Ruiz (1989). This analysis of the irreversible DSC transitions in terms of a thermodynamic model is based on treating the transitions as a sequence of two processes; the reversible unfolding of the protein described by the unfolding temperature andthe calorimetric enthalpy followed by a slower irreversible process such as aggregation. This approach yields results for the overall process that are the same as for the reversible process (Manley et al., 1985). Values of the transition temperature(T"), the calorimetric enthalpy (AH,), and the van't Hoff enthalpy (AH")are presented in Table IV for each of the lectins at different pH values. The type of transition expressed in Equation 6is also supported by the increase with lectin concentration in both the transition temperature and the temperature at the peak maximum (T,) as shown below. The van't Hoff enthalpy is in terms of moles of tetramer for Con A and moles of dimer for pea and lentil lectin. The calorimetric enthalpy is in terms of moles of oligomer, tetramer for Con A and dimer for the pea and lentil lectins. Also presented in Table IV is the ratio of the calorimetric enthalpy to the van't Hoff enthalpy determined from the fit of the two-state transition model. The ratios are greaterthan one for Con A and close to one for pea and lentil lectin. The denaturationof the pea and lentil lectins follow the two-state model, while the higher ratios for Con A indicate the presence of an intermediate state, perhaps the metal-less form of Con A, apoconcanavalin A. Fukada et al.'s (1983) treatment for the denaturation peak temperature dependence on protein concentration of an oligomer dissociating into itsmonomer units canbe extended to a protein dissociating into four subunits upon denaturation, (AA'),
C)
2B
+ 2B'
(Eq. 7)
where A is the native form of one fragment, A' is the other fragment, and B is the unfolded states of these fragments. It can be shown that ln[(AA'),] = - A H , ( S ) / { R T , ( n - I)] + constant
(Eq. 8)
-ASS
"b
kJ mol-'
Determinations
J mol-' K"
15.9 f 0.6 17.3 f 1.6 16.6 f 0.7b
4 4
0 0 0
13.4 f 1.7 -10 4 f6 1 3 . 6 f 0.8 -5 4 &3 13.5 f O . l b -8 f 2 7.8 f 0.8 -24 4& 3 8 . 2 f 0.7 -23 4f 2 -24 f 1 8.0 f 0.2' 8 . 8 f 0.5 -174 f 2 1 0 . 6 f 1.0 -9 4f 4 -13 f 4 9.7 & 0.9b of lentil lectin ranged from 0.31 to 1.0 mM, while
where n = total number of moles of denatured species, 4. Plots of Equation 8 are shown in Fig. 4. In Table IV, the values of A H u determined from Equation 8, AHJS), are close to thecalorimetric enthalpies determinedwhen n = 4 for Con A and pea and lentil lectin. It should be emphasized that the temperatures at thepeak maxima are independent of the fits of the two-state transition model to the data and, thus, the overlap of AHH,with AH,(S) is independent confirmation that denaturation of the pea and lentil dimers results in dissociation of their monomers and denaturationof the Con A tetramer results in dissociation into its monomers. In Table IV, there is little dependence of the thermodynamic parameters on pH down to pH 5 for Con A and pea lectin, while lentil lectin results exhibit adecrease in T,,,from pH 6.9 to 5.0 along with corresponding decrease in the calorimetric and van't Hoff enthalpies. Lentil lectin is similar to globular proteins which exhibit decrease in the transitional enthalpies along with a decrease in the transition temperature with pH. This indicates that theprotein undergoes an increase in its heat capacity as it unfolds ( d ( A H ) / d T > 0). The transition temperature dependence on pH for lentil lectin reflects proton transfer between the lectin and solution as it unfolds. Lack of pH dependence of transition temperature for Con A and pea lectin indicates the absence of any proton transfer as they unfold. The thermal stabilityof the pea and lentil lectins are the same but less than that of Con A which unfolds at higher temperatures. In the presence of saturatingamounts of carbohydrate ligands, the transition temperatures and calorimetric enthalpies, and the van't Hoff enthalpies exhibit increases, while the calorimetric to van't Hoff enthalpyratios remain the same. The increase in both T,,, and T, with ligand concentration arisesfrom binding of the ligand to thelectin only in the folded form. The denaturation transition in the presence of bound ligand can be expressed as follows, (AA')&
C)
2B
+ 2B' + mL
(Eq. 9)
and at constantlectin concentration, ln[L] = -AH,(L)/[RT,m]
+ constant
(Eq. 10)
The best fits of ln[L] to 1/T, yield van't Hoff enthalpies close to thecalorimetric enthalpies for m = 4 for the Con A complex and m = 2 for the pea and lentil complexes. Thus, one carbohydrate binds to each of the four monomer Con A units and to each of the two pea and lentil lectin monomer units.
Thermodynamics of Monosaccharide Lectins Binding to
L 346
b
-
L
_
7673
" "
356 36 1 TEMPERATURE IK)
35 1 I
366
37 1
350
355
I
FIG. 3. a, DSC scan of a 0.5-g sample of 0.53 mM Con A in DMG buffer containing 100 mM each of Ca2+and Mn2+ ions at a scan rate of 13 K h-'. The solid lines are thebest least squarestits of the DSC data to theA CI 4B two state transition model. b, DSC scan of a 0.5-g sample of 0.28 mM pea lectin in DMG buffer containing 1mM each of Ca2+and MnZ+ions a t a scan rate of 13 K h-'. The solid lines are the best least squares tits of the DSC data to theA 0 4B two state transition model. c, DSC scan of a 0.5-g sample of 0.28 mM lentil lectin in DMG buffer containing 1mM each of Ca2+and Mn2+ions a t a scanrate of 13 K h-'. The solid lines are the best least squares fits of the DSC data to theA CI 4B two state transition model. 330
345 335 -
C
332
337
340 TEMPERATURE (K) I
T I
342 347 TEMPERATURE (K)
1
352
J
35'7
Thermodynamics of MonosaccharideBinding to Lectins
7674
pH
Con A 6.90 f 0.05 6.90 f 0.05b 5.30 f 0.05 5.00 f 0.05
TABLEIV Thermodynamic dataon the denaturation transitions of the legume lectins Concentration T, MH, AH" A H C I A H "
mM
K
0.071-0.712 0.844 0.680 0.680
360.3-363.9 364.6 f 0.2 365.5 f 0.1 364.2 f 0.1
2376 f 212 2312 f 40 2228 k 48 2120 f 100
1480 f 77 1664 f 8 1509 f 19 1581 f 28
1.69 f 0.17 1.39 f 0.04 1.48 f 0.05 1.36 f 0.08
Pea lectin 7.40 f 0.05 0.085-0.268 343.9-345.1 1380 f 60 1064 f 160 0.84 f 0.10 6.90 & 0.05 0.176-0.209 347.0-347.2 1381 f 36 1124 f 20 0.89 f 0.03 6.90 f 0.05' 0.281 350.2 1660 f 60 1200 f 100 0.78 f 0.17 6.00 f 0.05 0.046-0.180 344.2-345.5 965 f 12 840 f 92 0.92 f 0.13 5.00 f 0.05 0.190 345.7 f 0.6 802 f 140 1037 f 3 0.81 f 0.14 Lentil lectin 6.90 f 0.05 0.120-0.635 345.9-347.8 1070 f 164 1504 f 14 0.73 f 0.151 6.90 f 0.348 0.05b 348.9 1360 f 160 1320 f 20 0.53 f 0.04 5.0 f 0.05 0.147-0.415 340.3-341.5 686 f 78 1289 f 21 0.63 f 0.140 The A H @ ) values were determined from Equation 8 with n = 4. DSC measurements were performedat a scan rate of 55 K h-', while the usual scan rate was 13 K h-'.
It has been shown (Schwarz, 1988) that the increase in the
TP
K
kJ mol"
AH(S)a kJ mol"
361.9-368.6 368.5 f 0.2 366.4 f 0.1 364.9 f 0.1
2556 f 153
344.6-345.7 347.9-348.6 351.2 345.3-346.7 346.5 f 0.1
1362 f 362
347.1-348.7 350.2 340.3-342.1
2016 f 399
complete picture of the thermodynamics of the binding process. The loss in free energy of the ligand-lectin complex upon its formation involves not only enthalpy changes but also entropy changes which can be quite significant as shown by the titration calorimetry results. For example, the affinity of Glu binding to lentil lectin, A@ = -13.2 f 0.2 kJ mol", consists of a decrease in the enthalpy, A& = -9.7 f 0.9 kJ &(Tm) = explAH[l/Tm(0) - l/Tml/Rl/{l - [Lll (Eq. 11) mol", and a substantialincrease in the entropy of the system, where AH is the denaturation enthalpy, T,,,(O)is the denatur- TASb = 3.9 f 1.2 kJ mol", upon binding. Changes in the ation temperature in the absence of ligand, and [L] is the enthalpy can be interpreted in terms of differences in the total ligand concentration. Using the calorimetric enthalpies interactions between the solvent and lectin and between the in Table V, and T,,,(O)= 362.4 f 0.1 K for Con A, 348.6 f 0.1 ligand and lectin. These interactionsexplicitly involve hydroK at pH 6.9 and 345.9 K at pH 7.4 for pea lectin, and 346.5 gen bonding, electrostatic, and van der Waals interactions which can be identified from knowledge of the positioning of & 0.1 K for lentil lectin, values for the binding constant at the transition temperature,Kb(T,,,), were determined. In Table the ligand at thelectin binding site and from energy minimiV, these values are compared to thevalues for Kb determined zation calculations (Imberty et al., 1991). It is, however, from extrapolating the titrationcalorimetry results in Tables difficult to explicitly identify the sources of the entropy con1-111 using Kb(Tm)= &(T)eXp[AHb(l/T - l/Tm)/R]. The tributions to the binding reaction. In fact, earlier interpretaTJO) values used in the determination ofKb(T,,,) yield the tions of the binding affinities (Bhattacharyya and Brewer, most consistent values for the Kb for each carbohydrate and 1988) ignored the entropic contributions andrelied on describwere within experimental error of the transition temperatures ing changes in AGb in terms of pair-wise interactions between of the ligand free lectin. The results are shown in Table V in the carbohydrate OH groups and the amino acid residues at terms of the ratio which is close to one, indicating the binding site. As can be seen with the Glu-lentil lectin that the titration calorimetry results agree with the DSC interactions, neglecting entropy effects can substantially alter results for the determination of Kb. The DSC results are also the estimation of the binding affinity, AGb. The thermodynamics of the binding reaction presented in independent of pH from 6.9 to 7.4 and of scan rate as shown in Table V for the pea lectin. This lack of pH dependence for Tables 1-111 show that the changes in enthalpy and entropy the binding constant was also observed with titration calorim- upon binding appear to be compensatory. This "compensaetry at room temperature. tory" effect is shown in Fig. 5 where A f f b values for all the There are also some additional changes in the thermody- lectinsareplottedasa function of T A s b at 292 K. The namic unfolding parameters in the presence of carbohydrate compensatory correlation, ligands. The calorimetric enthalpies arehigher but do overlap AHb = -1.67 f 0.06 X lo' + (1.30 f 0.12)TASb (Eq. 12) with the calorimetric enthalpies in the absence of ligand for Con A and lentil lectin, while they are, however, higher by is quite good in that almost all the data lie within three about 45 f 11% for the pea lectin. The higher calorimetric standard deviations of ASb predicted from Equation 12 and enthalpies, especially for the pea lectin, can reflect a positive that d(Affb)/d(TASb)is close to 1. This correlation not Only dependence of the transitionenthalpy on temperatureas applies to thebinding of each carbohydrate to a specific lectin observed with the globular proteins. but extends to thebinding of the carbohydrate ligands to all three lectins. Similar enthalpy-entropy compensation with DISCUSSION the slope of TASb close to 1 has been observed in many Earlier interpretations (Goldstein et al., 1965; Poretz and protein-ligand interactions where the experimental conditions Goldstein, 1970; Bhattacharyyaand Brewer, 1988) of the are fixed and the ligand structure is varied (congener series) binding mechanism of carbohydrates to the legume lectins (Lumry and Rajender, 1970; Belleau and Lavoie, 1968). This have relied on comparisons of the binding constants. To has been interpreted (Leffler and Grunwald, 1963) to mean elucidate the binding mechanism, however, requires a more that as theligand becomes more tightly bound to theprotein transition temperature with ligand concentration at [L] >>> [protein] depends on the association constant and that the association constant at the denaturation temperature can be determined from the increase in the denaturation and ligand concentration as shown,
Thermodynamics of Monosaccharide Binding to Lectins
-2.90
7675
f
t
I I " " 1
0.00274
l/To
0.00277
i n 1/K
L
-1
k
\
FIG. 4. u, a plot of ln[lectin concentration] us. l / T pfor Con A. b, a plot of ln(1ectin concentration] us. l / T pfor pea lectin. c, a plot of ln[lectin concentration] us. 1/T, for lentil lectin.
-2.90
\
1
1 0.00292
0.00289 -0.20
1/Tp I
I
I
I
i n l/K I
I 1 I" I I
0.00286
( i e . larger decrease in A H b ) , its rotational and vibrational degrees of freedom will become morerestricted and, thus,the entropy will decrease. Another explanation(Eftink et al., 1983) for this compensation can arise from the assumption that theprotein is in equilibrium between two different states, Po PI and that the ligand binds to either state with different
-
1/Tp
I
I
I
I
I
I
I
I
in / l K
0.00289
affinities to produce mixed POLand PIL complexes. This coupling between the binding process and the transition between the two states results in enthalpy-entropy compensation. Since the observed heat capacity change upon carbohyPI is drate binding is 0, the enthalpy change for the Po close to zero. Whether this enthalpy-entropy compensation
-
Thermodynamics of Monosaccharide Binding to Lectins
7676
TABLEV Thermodynamic data onthe thermal transition of the legume lectins in thepresence of carbohydrate ligands Ligand
Ligand conc.1 lectin conc.
T,
mM/mM
K
22-65/0.721 11-44/0.712 1&55/0.712 21-66/0.712
367.4-369.0 364.8-366.7 366.5-367.8 365.1-366.2
2882 f 68 2784 f 200 2744 f 120 2392 f 176
1574 f 38 1530 f 99 1628 f 51 1673 f 33
19-74/0.165 350.9-352.8 1696 pH 21-44/0.266 7.4 349.1-350.1 1522 pH 7.4' 23/0.016 350.1 f 0.2 Man 18-76/0.154 350.2-351.8 1742 MeaGlu 21-69/0.165 350.7-352.2 Glu 18-76/0.150 349.8-351.4
f 40 f 20 1538 f 28 f 139 1494 f 130 1670 f 34
1494 f 31 1574 f1.00 60 1704 f 18 1602 f 11 1554 f 81 1506 f 74
AHH,/AH"
AH" kJ mol"
TP
mu(L)"
K
kJ mol"
Kt./& (DSC)'
Con A MeaMan
Man MeaGlu Glu
Pea lectin MeaMan
1.8 f 0.1 1.8 f 0.2 1.6 f 0.1 1.4 f 0.1
368.4-370.0 365.9-367.4 367.1-368.6 365.9-367.0
2970 f 150 3326 f 1146 3312 f 400 3202 f 400
1.00 f 0.05 351.8-354.1 1209 f 0.05 350.0-351.1 0.95 f 0.15 353.1 f 0.1 1.10 f 0.251 350.5-352.9 1069 1.05 f 0.10 351.8-353.0 1946 1.1f 0.1 350.0-352.2 1342
1.01 f 0 1.01 f 0 1.02 f 0 1.19 f 0
f 200 0.85
f0 0.81 f 0 0.70 f 0.1 f0 1.28 f 0 1.34 f 0
f 200 1.04 f 232 f 160
Lentil lectin 350.0-351.2 1318 f 128 1347 f 73 MeaMan 1.0 f 0.05 350.8-352.3 1495 f 25 0.81 f0 24-73/0.635 Man 1606 f 41 349.2-351.1 1200 f 60 0.7 f 0.1 349.5-351.4 1267 f 250 1.12 f0 24-78/0.635 MeaGlu 1452 f 57 0.9 f 0.1 349.2-351.4 1226 f 54 349.7-351.8 1285 f 88 1.17 f 0 20-75/0.635 Glu 1463 f 60 0.9 f 0.1 349.4-351.0 348.6-350.5 1240 f 80 1663 f 233 0.96 0 21-80/0.635 a The M ( L ) values were determined from Equation 10 with rn = 4 for the con A complexes and n = 2 for the pea and lentilcomplexes. The values for Kb(DSC)were determined from Equation 11, while those for Kg were determined from the Kb and AHb values at room temperature (see text). e The scan rate was 45 K h-l. while the usual scan ratewas 13 K h-I.
*
FIG. 5. A plot of A H b us. TASb at room temperature for all the lectin complexes. The straight line is the best least squares fit of the data to a straight line.
-31 I -9.1
I
I
-5.90,
-2.70
T4S
1
,499 (kJ per mole)
I
Ll
3.7
6.9
extends to other carbohydrateligands and lectins remains to stronger interactions between the ligand and lectin. A comparison of the Man and MeLvGlu results yields almost the be seen. This relationship shows that formonosaccharide same binding constant for each lectin which arises from the binding to the legume lectins it is possible to estimate the entropy changes in the binding reactionfrom estimations of compensatory changes in the binding enthalpy and entropy, the enthalpy changes. The enthalpy changes can then be resulting in similar values for AGb. The only difference beestimated from the pair wise interactions and energy minim- tween Glu and Man is the positioning of the OH group on ization methods suchas the docking of MeaMan and MeaGlu C2: in Glu it is equatorial, while in Man it is axial to the weakens to Con A as described by Imberty et al. (1991). Furthermore, pyranose ring. This difference is significant in that it enthalpy-entropy compensationshows that, again, changes in substantially the binding interaction (enthalpy) for the D AGO are notas specific a criterion of the binding interactions glucopyranoside derivatives to all three lectins. From the xray crystallographic binding study of MeaMan to Con A, as is since AG shows less variation for thebinding Denvenda et al. (1989) observed that thisOH group is involved reactions. Comparison of the binding constants among the three lec- in water bridges to Ser-168 and Thr-226 or Asp-71 of an tins show that in going from MeaMan toGlu, the specificity adjacent tetramer,so that in solutionwhere the Con A tetraincreases in the order Con A > pea lectin > lentil lectin. In mers are far apart it is not clear how this group affects the comparing the D-mannopyranoside derivative results, there is binding parameters. Minimum energycalculations (Sekhalittle change in the entropyfor Con A and pea lectin so that rudu and Rao, 1984; Rao et al., 1989) show that this group in the increase in the binding affinity upon the addition of a MeLvMan and also, by inference, Man can interact withAsnthis group in i.e. 14 of Con A in allconfigurations,whereas methyl group results from an increase in the enthalpy,
Thermodynamics of Monosaccharide Binding to Latins
7677
MeaGlu does not interact with any of the amino acid residues very little temperature shiftof the major transition peak upon in most of the binding configurations. In addition, Imberty et going from 1to 100 mM metal ion concentration for the Con al. (1991) show that the C2-OH group interacts with the A DSC scans. This lower stability may be attributed, again, binding site (Leu-99 residue) only in the Man configuration. to theless integrated structureof the pea and lentil monomers The decrease in from the D-mannopyranoside to D- which readily dissociate in acid pH with EDTA or as these lectins absorbs heat. Binding of the carbohydrates increases glucopyranoside derivatives can, thus, be attributed to the lack of interaction between C2-OH of the D-glucopyranoside the thermal stability of the lectin and K b at thedenaturation temperature agrees with the titration calorimetry results at derivatives and Con A. The addition of a methyl group on the C1-OH group to room temperature. Con A, pea, and lentil lectin exhibit single form the a-methyl derivatives enhances the binding enthal- peak transitions which are indicative of globular-type propies for the mannopyranosides, while decreasing the binding teins. enthalpies for the pea and lentillectin-glucopyranosides. This REFERENCES is likely the result of an inductive effect on the distribution Belleau, B., and Lavoie, J. L. (1968) Can. J. Biochem. 4 6 , 1397-1409 of the charge density on the pyranoside ring since the methyl Bhattacharyya, L., Brewer, C.F., Brown, R. D., and Koenig, S. H. (1985) Biochemistry 24,4974-4980 group can act as a slightly charge withdrawing group. For the Bhattacharyya, L., and Brewer, C. F. (1988) Eur. J. Biochem. 1 7 6 , 207-212 Borrehaeck, C., and Mattiasson, B. (1979) Eur. J. Biochem. 107,67-71 pea and lentil lectins, its effect on the mannopyranosides, Debray, H., Decont, D., Strecker, G., Spik, G., and Montreuil, J. (1981) Eur. J. where the C2-OH is axial, is opposite to its effect on the Biochem. 117,41-55 Z., Yariv, J., Helliwell, J. R., Kalb (Gilboa) A. J., Dodson, E. J., glucopyranosides, where the C2-OH is equatorial to the pyr- Derewenda, Papiz, M. Z., Wan, T., andCampbell, J. (1989) EMBd J. 8,2189-2193 anoside rings. Edmundson, A. B., Ely, K. R., Sly, D. A., Westholm, F. A,, Powers, D. A,, and Liener, I. E. (1971) Biochemistry 10,3554-3559 The decrease in carbohydrate binding specificity in going Eftink, M.R., Anusiem, A. C., and Biltonen, R. L. (1983) Biochemistry 2 2 , 9QQAL9QQfi from Con A to lentil lectin can be attributed to less carbohy" " """ Einspahr, H., Parks, E. H., Suguna, K., Subramanian, E., and Suddath, F. L. drate-amino acid interactions in the complexes of the pea and (1986) J. Biol. Chem. 261,16518-16527 lentil lectins. Rini et al. (1989) identified hydrogen bond Fliegerova, O., Salvetova, A., Ticha, M., and Kocourek, J. (1974) Bioehirn. Baophys. Acta 3 6 1 , 416-426 formation between pea lectin and only the C3-OH, C4-OH, Fukada, H., Sturtevant, J. M., and Quiocho, F. A. (1983) J. Biol. Chem. 2 6 8 , 13193-13198 and C5-OH groups of the terminal mannopyranoside of Goldstein, I. J., Hollerman, C. E., and Smith,E. E. (1965) Biochemistry 4,876methyl 3,6-di-O-(a-D-mannopyranosyl)-a-D-mannopyrano- 88.7 side, while Derwenda et al. (1989) identified additional hydro- Howard, I. K., Sage, H. J., Stein, M.D., Young, N. M., Leon, M.A., and Dyckes, D. F. (1971) J. Biol. Chem. 246,1590-1595 gen bonds between Con A and the C6-OH group of MeaMan. Imberty, A., Hardman, K. D., Carver, J. P., and Perez, S. (1991) Glycobiology 1.631 2 -,- - - 4 -4- The DSC results show that at the denaturation temperaLeffler, J. E., and Grunwald, E. (1963)Rates and Equilibria of Organic Reactions, ture, the pea and lentil lectin dimers dissociate ultimately pp. 315-402, John Wiley and Sons, Inc., New York Lumry, R., and Rajender, S. (1970) Biopolymers 9,1125-1227 into sub-monomer fragments upon unfolding. Earlier DSC Manly, S. P., Matthews, K. S., and Sturtevant, J. M. (1985) Biochemistry 2 4 , results (Zahnley, 1981) on Con A indicate that denaturation 3842-3846 Poretz, R. D., and Goldstein, I. J. (1970) Biochemistry 9 , 2890-2896 leads to demetallization since the denaturation temperature Rao, V. S. R., Biswas, M., Mukhopadhyay, and Balaji, P. V. (1989) J. Mol. increases with metallization of the metalless form of Con A. Struct. 194,203-214 Rini, J. M., Hardman, K. D., Einspahr, H. M., Suddath, F. L., and Carver, J. The dissociation of the pea lectin upon denaturation, thus, P. (1989) Trans. Am. CrystallographficAssoc. 2 6 , 51-63 agrees with the observation that demetallization of pea lectin Sanchez-Ruiz, J. M., Lopez-Lacomba, J. L., Cortijo, M., and Mateo, P. L. (1988) Biochemistry 2 7 , 1648-1652 at low pH results in cleavage of the monomer units into their Schwarz, F. P. (1993) J. Inorg. Biochem., in press a and fi subunits. The structuralintegrity of the pea and lentil Schwarz, F. P. (1988) Biochemistry 27,8429-8436 Schwarz, F. P., and Kirchhoff, W. H. (1988) Therrnochirnica Acta 1 2 8 , 267lectins is apparently enhanced by metal binding. Since the 295 residues of pea lectin which bind to the metal ions are all Schwarz, F. P., Puri, K., and Surolia, A. (1991) J. Biol. Chem. 2 6 6 , 2434424350 present only in the /3 chain, this induced stabilization by the Sekharudu, Y. C., and Rao, V. S. R. (1984) J. Biomol. Struct. and Dyn. 2 , 4153 metal binding must involve some indirect effect. The thermal Wiseman, T.,Williston, S., Brandts, J. F., and Lin,L. N. (1989)Anal. Biochem. stability of pea and lentil lectins are about the same, while 1 7 9 , 131-137 Yang, C. P. (1990) Omega Data in Origin, p. 66, Microcal Inc., Northhampton, Con A is higher. This enhanced stability of Con A does not MA result from the increase metal concentration since there is Zahnly, J. (1981) J. Inorg. Eiochem. 1 6 , 67-78