Biochemical Society Transactions ( 1 992) 20
Hydrophilic Surface Maps of Channel-Forming Peptides.
D1B
Dl1
323s
04
IAN D. KERR and MARK S.P. SANSOM Laboratory of Molecular Biophysics, The Rex Richards Buildin University of Oxford. South Parks Road, Oxford, OX1 3&. The primary step in constructing helix bundle models of ion channels is definition of the hydrophilic surface of an amphipathic helix [I]. To date, this has been achieved by inspection of helical wheel plots or via the use of hydrophobic moment calculations. Neither of these methods takes account of the conformations of hydrophilic amino acid sidechains and, therefore, may not identify the true hydrophilic face. Where detailed structural information is available there is the possibility of defining the amphipathic nature of a helix more
precise'& order to use available three dimensional structural information, we have developed a program that uses empirical energy calculations of the interactions of a water molecule with the surface of an amphipathic helix. The oxygen atom of a water molecule is placed at successive positions on a cylindrical by (i,r,$) where i is co-incident with the olar grid, . defined . elix axis, r is the distance from the centre of the helix and (I the angle around the (z,r) plane. At each grid point, with the peptide atoms and the oxygen atom fixed, an energy minimisation is erformed in order to determine the optimum orientation of t l e two hydrogen atoms. The peptide-water interaction energy, E(z,r,$), is then calculated according to equation I . E(ZJ,@W"+EE (1)
g.
-20
E, is the van der Waals energy and EE is the electrostatic energ For each (z,(I) the minimum value of E with respect to r is crosen. The resultant E,,,,,(z,$)file is shown as a contour plot which displays the h drophilic surface of the helix. The program has Keen used to analyse the hydrophilic surface of a number of channel-forming peptides for which there exists structural information, eg Staphylococcal &toxin:
fM-A-Q-D-I-I-S-T-I-G-D-L-V-K-W-I-1-D-T-V-N-F-T-K-K
+2.5
-E N
R
0
-2.5
- 180.0
&-toxin has been shown to form ion channels in planar lipid bilayers 121. Previous modelling studies have suggested that a hexameric bundle of &-toxin a-helices forms a transmembrane channel lined by the aspartate residues at positions 4,11 and 18 [I]. The calculation of the hydrophilic surface of &-toxin was performed using NMR determined coordinates of the peptide [3]. Figure 1 shows the hydrophilic surface of the peptide represented as a contour lot. The helix runs from z=-2.0 to +2.0 nm. From -2.0 to + I . f n m there is a favourable interaction of water with residues D4, D11 and D18. Favourable interactions in the C-terminal region are dispersed over a range of values. Averaging E,,,(z,$ ) over all z values generates a gra of vs. $ (figure 2). The minimum in this curve at +TOo defines the centre of the hydrophilic face of the helix. This information may subse uently be employed in construction of helix bundle models o't. &-toxin channels. The hydro hilic interactions at the C-terminus are consistent with a modefof &-toxin channel formation in which the N-termini of the helices cross the bilayer in response to a membrane potential whilst the C-termini remain anchored at the bilayerwater interface. This example demonstrates the use of the pro ram in defining more precisely the hydrophilic face of an amphpathic helix as a primary step in molecular modelling. The method has been applied to a range of channel-forming peptides, such as melittin and alamethicin, and to the putative transbilayer M2 helices of receptor channel proteins.
+ 180.0
IDK was supported by an SERClShell CASE studentship
I. Sansom, M.S.P., Kerr, I.D., Mellor, I.R. 1991. Eur.J.Biophys. 20 229-240 2. Mellor. I.R.. Thomas, D.H., Sansom, M.S.P. 1988. Biochim.
Biophys. Acta. 942: 280-294 3. Tappin, M.J.. Pastore. A., Norton, R.S.. Freer. J.H., Campbell, I.D. 1988. Biochemist? 27: 1643-1647
Hydrophilic Surface Maps of Channel-Forming Peptides.
D1B
Dl1
323s
04
IAN D. KERR and MARK S.P. SANSOM Laboratory of Molecular Biophysics, The Rex Richards Buildin University of Oxford. South Parks Road, Oxford, OX1 3&. The primary step in constructing helix bundle models of ion channels is definition of the hydrophilic surface of an amphipathic helix [I]. To date, this has been achieved by inspection of helical wheel plots or via the use of hydrophobic moment calculations. Neither of these methods takes account of the conformations of hydrophilic amino acid sidechains and, therefore, may not identify the true hydrophilic face. Where detailed structural information is available there is the possibility of defining the amphipathic nature of a helix more
precise'& order to use available three dimensional structural information, we have developed a program that uses empirical energy calculations of the interactions of a water molecule with the surface of an amphipathic helix. The oxygen atom of a water molecule is placed at successive positions on a cylindrical by (i,r,$) where i is co-incident with the olar grid, . defined . elix axis, r is the distance from the centre of the helix and (I the angle around the (z,r) plane. At each grid point, with the peptide atoms and the oxygen atom fixed, an energy minimisation is erformed in order to determine the optimum orientation of t l e two hydrogen atoms. The peptide-water interaction energy, E(z,r,$), is then calculated according to equation I . E(ZJ,@W"+EE (1)
g.
-20
E, is the van der Waals energy and EE is the electrostatic energ For each (z,(I) the minimum value of E with respect to r is crosen. The resultant E,,,,,(z,$)file is shown as a contour plot which displays the h drophilic surface of the helix. The program has Keen used to analyse the hydrophilic surface of a number of channel-forming peptides for which there exists structural information, eg Staphylococcal &toxin:
fM-A-Q-D-I-I-S-T-I-G-D-L-V-K-W-I-1-D-T-V-N-F-T-K-K
+2.5
-E N
R
0
-2.5
- 180.0
&-toxin has been shown to form ion channels in planar lipid bilayers 121. Previous modelling studies have suggested that a hexameric bundle of &-toxin a-helices forms a transmembrane channel lined by the aspartate residues at positions 4,11 and 18 [I]. The calculation of the hydrophilic surface of &-toxin was performed using NMR determined coordinates of the peptide [3]. Figure 1 shows the hydrophilic surface of the peptide represented as a contour lot. The helix runs from z=-2.0 to +2.0 nm. From -2.0 to + I . f n m there is a favourable interaction of water with residues D4, D11 and D18. Favourable interactions in the C-terminal region are dispersed over a range of values. Averaging E,,,(z,$ ) over all z values generates a gra of vs. $ (figure 2). The minimum in this curve at +TOo defines the centre of the hydrophilic face of the helix. This information may subse uently be employed in construction of helix bundle models o't. &-toxin channels. The hydro hilic interactions at the C-terminus are consistent with a modefof &-toxin channel formation in which the N-termini of the helices cross the bilayer in response to a membrane potential whilst the C-termini remain anchored at the bilayerwater interface. This example demonstrates the use of the pro ram in defining more precisely the hydrophilic face of an amphpathic helix as a primary step in molecular modelling. The method has been applied to a range of channel-forming peptides, such as melittin and alamethicin, and to the putative transbilayer M2 helices of receptor channel proteins.
+ 180.0
IDK was supported by an SERClShell CASE studentship
I. Sansom, M.S.P., Kerr, I.D., Mellor, I.R. 1991. Eur.J.Biophys. 20 229-240 2. Mellor. I.R.. Thomas, D.H., Sansom, M.S.P. 1988. Biochim.
Biophys. Acta. 942: 280-294 3. Tappin, M.J.. Pastore. A., Norton, R.S.. Freer. J.H., Campbell, I.D. 1988. Biochemist? 27: 1643-1647
