Mesoscopic models of lipid bilayers and bilayers with embedded ...

Mesoscopic models of lipid bilayers and bilayers with embedded ... Mesoscopic models of lipid bilayers and bilayers with embedded ...

science.uva.nl
from science.uva.nl More from this publisher
30.08.2013 Views

40 Structural characterization of lipid bilayers the same as in equation 4.1, but the angle φ is now defined as the angle between the orientation of the vector along two beads in the chain and the normal to the bilayer plane: cos φ = rij · ^n rij = zij rij (4.2) where ^n is a unit vector normal to the bilayer, and rij = ri − rj is the vector between beads i and j (rij = |rij|). The order parameter has value 1 if this vector is on average parallel to the bilayer normal, 0 if the orientation is random, and −0.5 if the bond is on average parallel to the bilayer plane. With this definition of the angle φ, we can compute the order parameter for a vector between any two beads in the lipid. In particular we are interested in characterizing the overall order of the chains and the local order. For the first quantity we define the indexes of the vector rij in equation 4.2 as: i = tn and j = t1, where tn is the last bead in the lipid tail and t1 is the first one. We call S tail the corresponding order parameter. For the local order we define i = tm+1 and j = tm with the index m increasing going toward the tail end, and call the corresponding order parameter Sm. If m is taken progressively from the headgroup to the tail-end of the molecule, a plot of the corresponding order parameters, Sm, gives an indication of the persistence of order from the interfacial region to the bilayer core. Area per lipid and bilayer thickness The area per lipid can be experimentally estimated from X-ray or neutron scattering [84, 85], or from the just described lipid order parameter profiles [86]. We compute the area per lipid, AL, by dividing the total bilayer projected area by half the number of lipids in the bilayer, since we find that, on average, there is an equal number of lipids in each monolayer. Experimentally the bilayer (total) thickness is computed from the peak-to-peak headgroup distance measured by X-ray diffraction. In simulations the same approach can also be used, and the thickness can be computed as the distance between the peaks in the density profile. To compute the thickness of the bilayer hydrophobic core, Dc, we consider the average distance along the bilayer normal (which we assume to be the z-axis) between the tail bead (or beads in case of double-tail lipids) connected to the headgroup of the lipids in one monolayer and the ones in the opposite monolayer: where zi t1 (i = 1, 2). Dc = z 1 t1 − z 2 t1 (4.3) is the average z position of the first tail beads of the lipids in monolayer i It is also useful to consider the lipid end-to-end distance, Lee, defined as the distance between the positions of the first bead(s), rt1 , and the last bead(s) rtn , of the

4.3 Computational details 41 lipid tail(s): Lee = 〈|rtn − rt1 |〉 (4.4) and its projection, L n ee, onto the normal to the bilayer plane: L n ee = 〈|ztn − zt1 |〉 (4.5) where the bilayer is taken parallel to the xy-plane. If the bilayer consists of two opposing monolayers which are in contact by the terminal carbons in the tails, the area per lipid and bilayer thickness are related by VL = ALDc/2, where VL is the volume of one lipid; and the thickness and the (projected) end-to-end distance are related by Dc = 2L n ee. However, if the two monolayers are interdigitated, the above relations do not hold [86]. For example, for partially interdigitated bilayers it will be Dc < 2L n ee and for fully interdigitated bilayers Dc = L n ee. To investigate the presence of an interdigitated phase we define a measure for the extent of interpenetration of the hydrophobic cores (tails) of the lipids on opposite sides of the bilayer by defining the chain overlap D overlap, as Doverlap = 2Lnee − Dc Ln . (4.6) ee where Dc and L n ee are defined in equations 4.3 and 4.4, respectively. 4.3 Computational details We first study the structural properties of single-tail lipids, in which we vary the length of the hydrophobic tail, the chain stiffness, and the headgroup interaction parameter. All the studied bilayers consist of 400 lipids, and approximately 5000 water beads, with a total bead density of ρ = 3. The non-bonded interactions between the beads are represented by the soft repulsion of equation 2.2, with the parameter set derived by Groot in [46] and reported in table 4.3. The reduced temperature was T ∗ = 1, and at this temperature all the considered bilayers are in the fluid phase. All the bilayers aij w h t w 25 15 80 h 15 35 (15) 80 t 80 80 25 Table 4.1: Repulsion parameters aij (see equation 2.2) used in our simulations. Water beads are indicated as w, hydrophilic head beads as h and hydrophobic tail beads as t. The parameters are in units of kBT. The value in parenthesis corresponds to a repulsion parameter between the headgroups which results in a non interdigitated bilayer (see text).

40 Structural characterization <strong>of</strong> <strong>lipid</strong> <strong>bilayers</strong><br />

the same as in equation 4.1, but the angle φ is now defined as the angle between the<br />

orientation <strong>of</strong> the vector along two beads in the chain <strong>and</strong> the normal to the bilayer<br />

plane:<br />

cos φ = rij · ^n<br />

rij<br />

= zij<br />

rij<br />

(4.2)<br />

where ^n is a unit vector normal to the bilayer, <strong>and</strong> rij = ri − rj is the vector between<br />

beads i <strong>and</strong> j (rij = |rij|). The order parameter has value 1 if this vector is on average<br />

parallel to the bilayer normal, 0 if the orientation is r<strong>and</strong>om, <strong>and</strong> −0.5 if the bond is<br />

on average parallel to the bilayer plane. With this definition <strong>of</strong> the angle φ, we can<br />

compute the order parameter for a vector between any two beads in the <strong>lipid</strong>. In particular<br />

we are interested in characterizing the overall order <strong>of</strong> the chains <strong>and</strong> the local<br />

order. For the first quantity we define the indexes <strong>of</strong> the vector rij in equation 4.2 as:<br />

i = tn <strong>and</strong> j = t1, where tn is the last bead in the <strong>lipid</strong> tail <strong>and</strong> t1 is the first one. We<br />

call S tail the corresponding order parameter. For the local order we define i = tm+1<br />

<strong>and</strong> j = tm <strong>with</strong> the index m increasing going toward the tail end, <strong>and</strong> call the corresponding<br />

order parameter Sm. If m is taken progressively from the headgroup to the<br />

tail-end <strong>of</strong> the molecule, a plot <strong>of</strong> the corresponding order parameters, Sm, gives an<br />

indication <strong>of</strong> the persistence <strong>of</strong> order from the interfacial region to the bilayer core.<br />

Area per <strong>lipid</strong> <strong>and</strong> bilayer thickness<br />

The area per <strong>lipid</strong> can be experimentally estimated from X-ray or neutron scattering<br />

[84, 85], or from the just described <strong>lipid</strong> order parameter pr<strong>of</strong>iles [86]. We compute<br />

the area per <strong>lipid</strong>, AL, by dividing the total bilayer projected area by half the number<br />

<strong>of</strong> <strong>lipid</strong>s in the bilayer, since we find that, on average, there is an equal number <strong>of</strong><br />

<strong>lipid</strong>s in each monolayer.<br />

Experimentally the bilayer (total) thickness is computed from the peak-to-peak<br />

headgroup distance measured by X-ray diffraction. In simulations the same approach<br />

can also be used, <strong>and</strong> the thickness can be computed as the distance between the<br />

peaks in the density pr<strong>of</strong>ile.<br />

To compute the thickness <strong>of</strong> the bilayer hydrophobic core, Dc, we consider the average<br />

distance along the bilayer normal (which we assume to be the z-axis) between<br />

the tail bead (or beads in case <strong>of</strong> double-tail <strong>lipid</strong>s) connected to the headgroup <strong>of</strong><br />

the <strong>lipid</strong>s in one monolayer <strong>and</strong> the ones in the opposite monolayer:<br />

where zi t1<br />

(i = 1, 2).<br />

Dc = z 1 t1<br />

− z 2 t1<br />

(4.3)<br />

is the average z position <strong>of</strong> the first tail beads <strong>of</strong> the <strong>lipid</strong>s in monolayer i<br />

It is also useful to consider the <strong>lipid</strong> end-to-end distance, Lee, defined as the distance<br />

between the positions <strong>of</strong> the first bead(s), rt1 , <strong>and</strong> the last bead(s) rtn , <strong>of</strong> the

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!