International Journal of Biological Macromolecules 23 (1998) 185 – 189
Electrostatics of the phospholipase-membrane interaction Marcelo D. Costabel a, Diego F. Vallejo b, J. Raul Grigera b,* b
a Departamento de Fı´sica, Uni6ersidad Nacional del Sur, Bahia Blanca, Argentina Instituto de Fı´sica de Lı´quidos y Sistemas Biolo´gicos (IFLYSIB), Uni6ersidad Nacional de La Plata, cc 565, 1900 La Plata, Argentina
Received 31 October 1997; received in revised form 2 April 1998
Abstract The electrostatic interaction of the Phospholipase A2 (PLA2)-membrane complex in the presence and absence of calcium is analysed by the computation of the electrostatic profiles of the components and the complex. The electrostatic potential was computed by using of the program MOLPOT that implement the boundary element method to solve the electrostatic problem. It considers a closed surface in three dimensions that contains the macromolecule that follows as close as possible the macromolecule shape. The results show that the presence of calcium ions contributes to the stability of the complex and at the same time creates a favourable electrostatic potential pattern that may be favourable for the lipolysis of the membrane components. © 1998 Elsevier Science B.V. All rights reserved. Keywords: Phospholipase; Membranes; Electrical potentials; Electrostatic; Enzyme binding
1. Introduction Electrostatic interactions play a relevant role in various intracellular processes. In consequence these interactions are important for a knowledge of the structure-function relationship of biological macromolecules. The electrostatic interactions often involve characteristic charged amino-acid residues but, in particular cases, are reinforced by the presence of ions, either positives or negatives, that favour the molecular reactivity. In this context we consider the analysis of the electrostatic interaction of the Phospholipase A2 (PLA2)-membrane complex. Phospholipase A2 hydrolyses specifically the 2ester bond of the 1,2-diacyl-3-sn-phosphoglycerides forming fatty acids and lysophospholipids whose products are part of the inflammatory response. This type of enzyme is characterised by * Corresponding author. Tel.: + 54 21 254904; fax: + 54 21 257317; e-mail:
[email protected] 0141-8130/98/$19.00 © 1998 Elsevier Science B.V. All rights reserved. PII S0141-8130(98)00045-2
working at alkaline pH, having its maximal enzymatic activity in the presence of an organised water–lipid interface and by its dependence on calcium [1]. Enzymatic action on lipids belonging to membranes involves first the binding to the membrane and then the catalytic action on the lipids. Careful calculations of the electrical potential profiles of the isolated enzyme have been done [2] but, as far as we know, there is no work considering the computation of electrical potential of both the enzyme and a membrane. Molecular dynamics studies of the complex have been reported [3,4], which ultimately may give a very good description of the process. The catalytic action of the enzyme can be separated from the enzyme-subtrate. Therefore, with the aim of clarifying the relevance of the calcium ions in the electrostatic profiles and its importance to the enzymatic action, we perform the analysis of the electrical field generated by the enzyme in solution. We analyse the profiles with and without
186
M.D. Costabel et al. / International Journal of Biological Macromolecules 23 (1998) 185–189
Fig. 1. Contour plots of the electrical potential generated by (a) The PLA2 in water having the two Ca2 + ions. The plane selected for the contour lines correspond to a plane passing through Ca2 + ion of the active site and a short distance from the location of the other. (b) The one produced by a piece of DMPC in a plane perpendicular to the membrane. The electrical potential was computed with MOLPOT (Ref. [4]).
calcium, i.e. the one produced by the membrane alone, and the situation with the complex. The result show that the presence of calcium ions contributes to the stability of the complex and at the same time creates a favourable electrostatic potential for the lipolysis of the membrane components.
2. Computational method The electrostatic potential was computed by using of the program MOLPOT due to Juffer et al. [5] that implement the boundary element method to solve the electrostatic problem. Briefly it considers a closed surface in three dimensions
that contains the macromolecule. This surface follows as close as possible the macromolecule shape. Inside the surface (region I) there are N charges qi located at points ri (i =1, 2,…, N). In the inner region I the electrical potential is 81 and the permittivity o1 while in the outer region (II) we have the electrical potential 82 and the permittivity o2. Inside the surface the Poisson’s equation is satisfied and we have N
9281 = − % qid(r−ri )/o1,
rRe I
(1)
i=1
where d (r−ri )is the Dirac’s delta in ri. The potential in the region II satisfies the linearized Poisson-Boltzmann equation 9282(r) =k 282(r)
rRe II
(2)
M.D. Costabel et al. / International Journal of Biological Macromolecules 23 (1998) 185–189
187
Fig. 2. Potential energy (arbitrary units) of the complex PLA2-membrane at different distances. The distance is taken form two arbitrary points, inside the protein and the membrane. (——) PLA2 with the two Ca2 + ions, (- - -) without any of the Ca2 + ions.
where k is the inverse Debye length. The Eqs. (1) and (2) with the appropriated boundary conditions (continuity of the electrical potential and continuity of the normal of the electrical displacement) and the additional condition of the regularity of the electrical potential at infinitum can be solved analytically only for simple geometry. The general cases can be solved numerically, as MOLPOT does. For the present case charges where assigned as the atomic partial charges, according to GROMOS force field [6] located in the centre of each atom. The dielectric permittivity of the macromolecule (Region I) was considered equal to 2.0 and the solvent (Region II) equals to 78.54 i.e. the water permittivity at 298.16 K. The Debye screening length was taken as 0.96223 nm, which corresponds to a physiological like solution. The protein co-ordinates of porcine pancreatic phospholipase A2 (E.C.3.1.1.4) [7] have been taken form the Brookhaven Protein Data Bank [8] and the membrane considered is a bilayer of Dimyristoyl phophatidylcholine (DMPC). As coordinates of the membrane we have chosen one configuration of the extensive molecular dynamics run performed by Chiu et al. [9] (kindly supplied by the authors).
3. Results and discussion Fig. 1 shows contour plots of the electrical potential generated by the PLA2 in water having the two Ca2 + ions (Fig. 1a) and the one produced by a piece of DMPC membrane (Fig. 1b). The contour lines shown correspond to a plane passing through the Ca2 + ion of the active site and a plane perpendicular to the membrane respectively. Fig. 2 shows the electrical potential energy of the complex PLA2-membrane at different distances. Two arbitrary points, inside the protein and the membrane, have been taken as a reference of the distance. The figure shows the case of the enzyme with the to Ca2 + sites occupied and with both sites empty. It is clear the influence in the Ca2 + ions in the interaction energy. The function of both ions as determinants of the stability of the complex becomes apparent. Fig. 3 shows a cut of the complex PLA2-membrane with a plane perpendicular to the membrane passing through the Ca2 + ion located at the active site. The contour plot of the electrical potential is superimposed on the model. The active site is observed as showing a favourable electrical potential region for the further interaction with lipids belonging to the membrane at the point of minimum potential energy.
188
M.D. Costabel et al. / International Journal of Biological Macromolecules 23 (1998) 185–189
Fig. 3. Cut of the complex PLA2-membrane with a plane perpendicular to the membrane passing through the position of Ca2 + ion located at the active site showing the contour plot of the electrical potential. The protein-membrane distance corresponds to the minimum potential energy. (a) No Ca2 + ions present in the enzyme. (b) With the two Ca2 + sites occupied. The line corresponding to − 4.00 is marked with heavy line to show clearer the changes form one situation to the other. The large grey spheres indicate the calcium sites.
M.D. Costabel et al. / International Journal of Biological Macromolecules 23 (1998) 185–189
Through the analysis of the data it becomes apparent the relevance of the presence of calcium ions for both the stability of the complex, as shown by the potential energy curve and the production of an ‘electrical cliff’ that favours the insertion of the lipid into the enzyme active site. Similarly to the observations by Zhou and Shoulten [4] in human synovial PLA2 our results shown the effect on the surface potential due to the presence of Ca2 + ions and the preference of the bovine pancreatic PLA2 to bind to negative charged membranes. On the other hand it have been shown [10] that the binding of the phospholipase to the lipid interface is produced with dehydration of both surfaces. This reinforces our suggestion that the phospholipid can be transferred to the active site through an ‘electric channel’ generated by the presence of the positive charge of Ca2 + ions. It can also be concluded that a modification or mutation of residues taking place on the union of Ca2 + will affect in a determinant way the enzyme activity. Although this method gives only a static picture the present results are indicative of the importance of the calcium ions in the enzyme structure. The picture is, however, incomplete, because the charge movements may also modify the kinetics of the process [11].
.
189
Acknowledgements This work was partially supported by the Consejo Nacional de Investigaciones Cientficas y Te´cnicas of Argentina (CONICET). J.R.G. is Member of the Carrera del Investigador of CONICET.
References [1] Waite M, editor. The Phospholipases. Handbook of Lipid Research, vol. 5. New York: Plenum Press, 1987. [2] Scott CV, Mandel AM, Sigler PB, Honig B. Biophys J 1994;67:493. [3] Berendsen HJC, Engbert B, Marrink S-J, Ahlstro¨m P. In: Pullman A, et al., editors. Membrane Proteins: Structures, Interactions and Models. Kluwer, 1992:457–470. [4] Zhou F, Schulten K. Proteins 1996;25:12. [5] Juffer AH, Botta EFF, van Keuken BAM, van der Poleg A, Berendsen HJC. J Comput Phys 1990;94:1725. [6] van Gusteren WF, Berendsen HJC. Groningen Molecular Simulation. From BIOMOS n.v. Nijenborg 4, 9747 AG Groningen, 1987. [7] Dijkstra BW, Renetseder R, Kalk K, Hol WGJ, Drenth J. J Mol Biol 1983;168:163. PDB code 1p2p. [8] Berstein FC, Koetzle TF, Willams GJB, Meyer EF Jr., Brice MD, Rodgers JR, Kennard O, Shimanouchi T, Tasumi M. J Mol Biol 1977;112:535. [9] Chiu SW, Clark M, Subrammiam Scott HL, Jakobsson E. Biophys J 1995;69:1230. [10] Jain MK, Vaz WLC. Biochim Biophys Acta 1987;905:1. [11] Grigera JR, Vericat F, Ruderman G, de Xammar Oro JR. Chem Phy Lett 1989;156:615.