ViewVC Help
View File | Revision Log | Show Annotations | View Changeset | Root Listing
root/group/trunk/mattDisertation/lipid.tex
Revision: 1112
Committed: Thu Apr 15 02:45:10 2004 UTC (20 years, 3 months ago) by mmeineke
Content type: application/x-tex
File size: 32609 byte(s)
Log Message:
some more updates.

File Contents

# User Rev Content
1 mmeineke 971
2    
3 mmeineke 1105 \chapter{\label{chapt:lipid}PHOSPHOLIPID SIMULATIONS}
4 mmeineke 971
5     \section{\label{lipidSec:Intro}Introduction}
6    
7 mmeineke 1083 In the past 10 years, increasing computer speeds have allowed for the
8     atomistic simulation of phospholipid bilayers for increasingly
9 mmeineke 1087 relevant lengths of time. These simulations have ranged from
10 mmeineke 1089 simulation of the gel ($L_{\beta}$) phase of
11 mmeineke 1061 dipalmitoylphosphatidylcholine (DPPC),\cite{lindahl00} to the
12 mmeineke 1001 spontaneous aggregation of DPPC molecules into fluid phase
13 mmeineke 1061 ($L_{\alpha}$) bilayers.\cite{marrink01} With the exception of a few
14     ambitious
15 mmeineke 1089 simulations, \cite{marrink01:undulation,marrink:2002,lindahl00} most
16 mmeineke 1083 investigations are limited to a range of 64 to 256
17 mmeineke 1061 phospholipids.\cite{lindahl00,sum:2003,venable00,gomez:2003,smondyrev:1999,marrink01}
18 mmeineke 1083 The expense of the force calculations involved when performing these
19     simulations limits the system size. To properly hydrate a bilayer, one
20 mmeineke 1001 typically needs 25 water molecules for every lipid, bringing the total
21     number of atoms simulated to roughly 8,000 for a system of 64 DPPC
22 mmeineke 1061 molecules. Added to the difficulty is the electrostatic nature of the
23 mmeineke 1083 phospholipid head groups and water, requiring either the
24 mmeineke 1089 computationally expensive, direct Ewald sum or the slightly faster particle
25     mesh Ewald (PME) sum.\cite{nina:2002,norberg:2000,patra:2003} These factors
26     all limit the system size and time scales of bilayer simulations.
27 mmeineke 1001
28     Unfortunately, much of biological interest happens on time and length
29 mmeineke 1089 scales well beyond the range of current simulation technologies. One
30     such example is the observance of a ripple ($P_{\beta^{\prime}}$)
31     phase which appears between the $L_{\beta}$ and $L_{\alpha}$ phases of
32     certain phospholipid bilayers
33     (Fig.~\ref{lipidFig:phaseDiag}).\cite{katsaras00,sengupta00} These
34     ripples are known from x-ray diffraction data to have periodicities on
35     the order of 100-200~$\mbox{\AA}$.\cite{katsaras00} A simulation on
36     this length scale would have approximately 1,300 lipid molecules with
37     an additional 25 water molecules per lipid to fully solvate the
38     bilayer. A simulation of this size is impractical with current
39     atomistic models.
40 mmeineke 1001
41 mmeineke 1089 \begin{figure}
42     \centering
43     \includegraphics[width=\linewidth]{ripple.eps}
44     \caption[Diagram of the bilayer gel to fluid phase transition]{Diagram showing the $P_{\beta^{\prime}}$ phase as a transition between the $L_{\beta}$ and $L_{\alpha}$ phases}
45     \label{lipidFig:phaseDiag}
46     \end{figure}
47    
48 mmeineke 1083 The time and length scale limitations are most striking in transport
49 mmeineke 1089 phenomena. Due to the fluid-like properties of lipid membranes, not
50     all small molecule diffusion across the membranes happens at pores.
51     Some molecules of interest may incorporate themselves directly into
52     the membrane. Once there, they may exhibit appreciable waiting times
53     (on the order of 10's to 100's of nanoseconds) within the
54     bilayer. Such long simulation times are nearly impossible to obtain
55     when integrating the system with atomistic detail.
56 mmeineke 1001
57 mmeineke 1083 To address these issues, several schemes have been proposed. One
58 mmeineke 1089 approach by Goetz and Lipowsky\cite{goetz98} is to model the entire
59 mmeineke 1001 system as Lennard-Jones spheres. Phospholipids are represented by
60     chains of beads with the top most beads identified as the head
61     atoms. Polar and non-polar interactions are mimicked through
62 mmeineke 1089 attractive and soft-repulsive potentials respectively. A model
63     proposed by Marrinck \emph{et. al}.\cite{marrink04}~uses a similar
64     technique for modeling polar and non-polar interactions with
65 mmeineke 1001 Lennard-Jones spheres. However, they also include charges on the head
66     group spheres to mimic the electrostatic interactions of the
67 mmeineke 1089 bilayer. The solvent spheres are kept charge-neutral and
68 mmeineke 1001 interact with the bilayer solely through an attractive Lennard-Jones
69     potential.
70    
71     The model used in this investigation adds more information to the
72 mmeineke 1026 interactions than the previous two models, while still balancing the
73 mmeineke 1089 need for simplification of atomistic detail. The model uses
74     unified-atom Lennard-Jones spheres for the head and tail groups of the
75 mmeineke 1061 phospholipids, allowing for the ability to scale the parameters to
76 mmeineke 1026 reflect various sized chain configurations while keeping the number of
77     interactions small. What sets this model apart, however, is the use
78 mmeineke 1061 of dipoles to represent the electrostatic nature of the
79 mmeineke 1026 phospholipids. The dipole electrostatic interaction is shorter range
80 mmeineke 1089 than Coulombic ($\frac{1}{r^3}$ versus $\frac{1}{r}$), and therefore
81     eliminates the need for the costly Ewald sum.
82 mmeineke 1001
83 mmeineke 1089 Another key feature of this model is the use of a dipolar water model
84 mmeineke 1061 to represent the solvent. The soft sticky dipole ({\sc ssd}) water
85     \cite{liu96:new_model} relies on the dipole for long range electrostatic
86     effects, but also contains a short range correction for hydrogen
87 mmeineke 1089 bonding. In this way the simulated systems in this research mimic the
88     entropic contribution to the hydrophobic effect due to hydrogen-bond
89     network deformation around a non-polar entity, \emph{i.e.}~the
90     phospholipid molecules. This effect has been missing from previous
91     reduced models.
92 mmeineke 1026
93     The following is an outline of this chapter.
94 mmeineke 1083 Sec.~\ref{lipidSec:Methods} is an introduction to the lipid model used
95     in these simulations, as well as clarification about the water model
96     and integration techniques. The various simulations explored in this
97     research are outlined in
98     Sec.~\ref{lipidSec:ExpSetup}. Sec.~\ref{lipidSec:resultsDis} gives a
99     summary and interpretation of the results. Finally, the conclusions
100     of this chapter are presented in Sec.~\ref{lipidSec:Conclusion}.
101 mmeineke 1026
102 mmeineke 971 \section{\label{lipidSec:Methods}Methods}
103    
104 mmeineke 1026 \subsection{\label{lipidSec:lipidModel}The Lipid Model}
105    
106 mmeineke 971 \begin{figure}
107 mmeineke 1061 \centering
108     \includegraphics[width=\linewidth]{twoChainFig.eps}
109     \caption[The two chained lipid model]{Schematic diagram of the double chain phospholipid model. The head group (in red) has a point dipole, $\boldsymbol{\mu}$, located at its center of mass. The two chains are eight methylene groups in length.}
110 mmeineke 971 \label{lipidFig:lipidModel}
111     \end{figure}
112    
113     The phospholipid model used in these simulations is based on the
114     design illustrated in Fig.~\ref{lipidFig:lipidModel}. The head group
115     of the phospholipid is replaced by a single Lennard-Jones sphere of
116 mmeineke 1061 diameter $\sigma_{\text{head}}$, with $\epsilon_{\text{head}}$ scaling
117     the well depth of its van der Walls interaction. This sphere also
118     contains a single dipole of magnitude $|\boldsymbol{\mu}|$, where
119     $|\boldsymbol{\mu}|$ can be varied to mimic the charge separation of a
120 mmeineke 971 given phospholipid head group. The atoms of the tail region are
121 mmeineke 1083 modeled by beads representing multiple methyl groups. They are free
122     of partial charges or dipoles, and contain only Lennard-Jones
123     interaction sites at their centers of mass. As with the head groups,
124     their potentials can be scaled by $\sigma_{\text{tail}}$ and
125     $\epsilon_{\text{tail}}$.
126 mmeineke 971
127 mmeineke 1089 The possible long range interactions between atomic groups in the
128     lipids are given by the following:
129 mmeineke 971 \begin{equation}
130 mmeineke 1061 V_{\text{LJ}}(r_{ij}) =
131     4\epsilon_{ij} \biggl[
132     \biggl(\frac{\sigma_{ij}}{r_{ij}}\biggr)^{12}
133     - \biggl(\frac{\sigma_{ij}}{r_{ij}}\biggr)^{6}
134     \biggr]
135 mmeineke 971 \label{lipidEq:LJpot}
136     \end{equation}
137     and
138     \begin{equation}
139 mmeineke 1061 V_{\text{dipole}}(\mathbf{r}_{ij},\boldsymbol{\Omega}_{i},
140     \boldsymbol{\Omega}_{j}) = \frac{|\mu_i||\mu_j|}{4\pi\epsilon_{0}r_{ij}^{3}} \biggl[
141     \boldsymbol{\hat{u}}_{i} \cdot \boldsymbol{\hat{u}}_{j}
142     -
143 mmeineke 1083 3(\boldsymbol{\hat{u}}_i \cdot \mathbf{\hat{r}}_{ij}) %
144 mmeineke 1089 (\boldsymbol{\hat{u}}_j \cdot \mathbf{\hat{r}}_{ij} )\biggr]
145 mmeineke 971 \label{lipidEq:dipolePot}
146     \end{equation}
147     Where $V_{\text{LJ}}$ is the Lennard-Jones potential and
148     $V_{\text{dipole}}$ is the dipole-dipole potential. As previously
149     stated, $\sigma_{ij}$ and $\epsilon_{ij}$ are the Lennard-Jones
150     parameters which scale the length and depth of the interaction
151     respectively, and $r_{ij}$ is the distance between beads $i$ and $j$.
152     In $V_{\text{dipole}}$, $\mathbf{r}_{ij}$ is the vector starting at
153 mmeineke 1083 bead $i$ and pointing towards bead $j$. Vectors $\mathbf{\Omega}_i$
154 mmeineke 971 and $\mathbf{\Omega}_j$ are the orientational degrees of freedom for
155     beads $i$ and $j$. $|\mu_i|$ is the magnitude of the dipole moment of
156     $i$, and $\boldsymbol{\hat{u}}_i$ is the standard unit orientation
157 mmeineke 1087 vector rotated with Euler angles: $\boldsymbol{\Omega}_i$.
158 mmeineke 971
159 mmeineke 1089 The model also allows for the intra-molecular bend and torsion
160     interactions. The bond between two beads on a chain is of fixed
161     length, and is maintained using the {\sc rattle}
162     algorithm.\cite{andersen83} The bends are subject to a harmonic
163     potential:
164 mmeineke 971 \begin{equation}
165 mmeineke 1061 V_{\text{bend}}(\theta) = k_{\theta}( \theta - \theta_0 )^2
166 mmeineke 971 \label{lipidEq:bendPot}
167     \end{equation}
168 mmeineke 1061 where $k_{\theta}$ scales the strength of the harmonic well, and
169     $\theta$ is the angle between bond vectors
170     (Fig.~\ref{lipidFig:lipidModel}). In addition, we have placed a
171 mmeineke 1089 ``ghost'' bend on the phospholipid head. The ghost bend is a bend
172     potential which keeps the dipole roughly perpendicular to the
173     molecular body, where $\theta$ is now the angle the dipole makes with
174     respect to the {\sc head}-$\text{{\sc ch}}_2$ bond vector
175     (Fig.~\ref{lipidFig:ghostBend}). This bend mimics the hinge between
176     the phosphatidyl part of the PC head group and the remainder of the
177     molecule. The torsion potential is given by:
178 mmeineke 971 \begin{equation}
179 mmeineke 1061 V_{\text{torsion}}(\phi) =
180     k_3 \cos^3 \phi + k_2 \cos^2 \phi + k_1 \cos \phi + k_0
181 mmeineke 971 \label{lipidEq:torsionPot}
182     \end{equation}
183     Here, the parameters $k_0$, $k_1$, $k_2$, and $k_3$ fit the cosine
184     power series to the desired torsion potential surface, and $\phi$ is
185 mmeineke 1061 the angle the two end atoms have rotated about the middle bond
186     (Fig.:\ref{lipidFig:lipidModel}). Long range interactions such as the
187     Lennard-Jones potential are excluded for atom pairs involved in the
188 mmeineke 1089 same bond, bend, or torsion. However, long-range interactions for
189     pairs of atoms not directly involved in a bond, bend, or torsion are
190     calculated.
191 mmeineke 971
192 mmeineke 1083 \begin{figure}
193     \centering
194 mmeineke 1089 \includegraphics[width=0.5\linewidth]{ghostBendFig.eps}
195     \caption[Depiction of the ``ghost'' bend]{The ``ghost'' bend is a bend potential added to restrain the motion of the dipole on the {\sc head} group. The potential follows Eq.~\ref{lipidEq:bendPot} where $\theta$ is now the angle that the dipole makes with the {\sc head}-$\text{{\sc ch}}_2$ bond vector.}
196 mmeineke 1083 \label{lipidFig:ghostBend}
197     \end{figure}
198    
199 mmeineke 1089 All simulations presented here use a two-chain lipid as pictured in
200 mmeineke 1061 Fig.~\ref{lipidFig:lipidModel}. The chains are both eight beads long,
201 mmeineke 1026 and their mass and Lennard Jones parameters are summarized in
202     Table~\ref{lipidTable:tcLJParams}. The magnitude of the dipole moment
203 mmeineke 1089 for the head bead is 10.6~Debye (approximately half the magnitude of
204     the dipole on the PC head group\cite{Cevc87}), and the bend and
205     torsion parameters are summarized in
206     Table~\ref{lipidTable:tcBendParams} and
207 mmeineke 1061 \ref{lipidTable:tcTorsionParams}.
208 mmeineke 971
209 mmeineke 1061 \begin{table}
210 mmeineke 1092 \caption[Lennard-Jones parameters for the two chain phospholipids]{THE LENNARD JONES PARAMETERS FOR THE TWO CHAIN PHOSPHOLIPIDS}
211 mmeineke 1061 \label{lipidTable:tcLJParams}
212     \begin{center}
213 mmeineke 1089 \begin{tabular}{|l|c|c|c|c|}
214 mmeineke 1061 \hline
215 mmeineke 1089 & mass (amu) & $\sigma$($\mbox{\AA}$) & $\epsilon$ (kcal/mol) %
216     & $|\mathbf{\hat{\mu}}|$ (Debye) \\ \hline
217     {\sc head} & 72 & 4.0 & 0.185 & 10.6 \\ \hline
218     {\sc ch}\cite{Siepmann1998} & 13.02 & 4.0 & 0.0189 & 0.0 \\ \hline
219     $\text{{\sc ch}}_2$\cite{Siepmann1998} & 14.03 & 3.95 & 0.18 & 0.0 \\ \hline
220     $\text{{\sc ch}}_3$\cite{Siepmann1998} & 15.04 & 3.75 & 0.25 & 0.0 \\ \hline
221     {\sc ssd}\cite{liu96:new_model} & 18.03 & 3.051 & 0.152 & 0.0 \\ \hline
222 mmeineke 1061 \end{tabular}
223     \end{center}
224     \end{table}
225 mmeineke 1026
226 mmeineke 1061 \begin{table}
227 mmeineke 1089 \caption[Bend Parameters for the two chain phospholipids]{BEND PARAMETERS FOR THE TWO CHAIN PHOSPHOLIPIDS}
228 mmeineke 1061 \label{lipidTable:tcBendParams}
229     \begin{center}
230     \begin{tabular}{|l|c|c|}
231     \hline
232     & $k_{\theta}$ ( kcal/($\text{mol deg}^2$) ) & $\theta_0$ ( deg ) \\ \hline
233     {\sc ghost}-{\sc head}-$\text{{\sc ch}}_2$ & 0.00177 & 129.78 \\ \hline
234     $x$-{\sc ch}-$y$ & 58.84 & 112.0 \\ \hline
235     $x$-$\text{{\sc ch}}_2$-$y$ & 58.84 & 114.0 \\ \hline
236     \end{tabular}
237 mmeineke 1089 \begin{minipage}{\linewidth}
238     \begin{center}
239     \vspace{2mm}
240     All alkane parameters are based off of those from TraPPE.\cite{Siepmann1998}
241 mmeineke 1061 \end{center}
242 mmeineke 1089 \end{minipage}
243     \end{center}
244 mmeineke 1061 \end{table}
245    
246     \begin{table}
247 mmeineke 1089 \caption[Torsion Parameters for the two chain phospholipids]{TORSION PARAMETERS FOR THE TWO CHAIN PHOSPHOLIPIDS}
248 mmeineke 1061 \label{lipidTable:tcTorsionParams}
249     \begin{center}
250     \begin{tabular}{|l|c|c|c|c|}
251     \hline
252     All are in kcal/mol $\rightarrow$ & $k_3$ & $k_2$ & $k_1$ & $k_0$ \\ \hline
253     $x$-{\sc ch}-$y$-$z$ & 3.3254 & -0.4215 & -1.686 & 1.1661 \\ \hline
254     $x$-$\text{{\sc ch}}_2$-$\text{{\sc ch}}_2$-$y$ & 5.9602 & -0.568 & -3.802 & 2.1586 \\ \hline
255     \end{tabular}
256 mmeineke 1089 \begin{minipage}{\linewidth}
257     \begin{center}
258     \vspace{2mm}
259     All alkane parameters are based off of those from TraPPE.\cite{Siepmann1998}
260 mmeineke 1061 \end{center}
261 mmeineke 1089 \end{minipage}
262     \end{center}
263 mmeineke 1061 \end{table}
264    
265    
266     \section{\label{lipidSec:furtherMethod}Further Methodology}
267    
268 mmeineke 1026 As mentioned previously, the water model used throughout these
269 mmeineke 1089 simulations was the {\sc ssd/e} model of Fennell and
270     Gezelter,\cite{fennell04} earlier forms of this model can be found in
271     Ichiye \emph{et
272     al}.\cite{liu96:new_model,liu96:monte_carlo,chandra99:ssd_md} A
273 mmeineke 1061 discussion of the model can be found in Sec.~\ref{oopseSec:SSD}. As
274     for the integration of the equations of motion, all simulations were
275     performed in an orthorhombic periodic box with a thermostat on
276     velocities, and an independent barostat on each Cartesian axis $x$,
277 mmeineke 1089 $y$, and $z$. This is the $\text{NPT}_{xyz}$. integrator described in
278     Sec.~\ref{oopseSec:integrate}. With $\tau_B = 1.5$~ps and $\tau_T =
279     1.2$~ps, the box volume stabilizes after 50~ps, and fluctuates about
280     its equilibrium value by $\sim 0.6\%$, temperature fluctuations are
281     about $\sim 1.4\%$ of their set value, and pressure fluctuations are
282     the largest, varying as much as $\pm 250$~atm. However, such large
283     fluctuations in pressure are typical for liquid state simulations.
284 mmeineke 1026
285    
286     \subsection{\label{lipidSec:ExpSetup}Experimental Setup}
287    
288 mmeineke 1089 Two main classes of starting configurations were used in this research:
289     random and ordered bilayers. The ordered bilayer simulations were all
290     started from an equilibrated bilayer configuration at 300~K. The original
291     configuration for the first 300~K run was assembled by placing the
292     phospholipids centers of mass on a planar hexagonal lattice. The
293     lipids were oriented with their principal axis perpendicular to the plane.
294     The bottom leaf simply mirrored the top leaf, and the appropriate
295     number of water molecules were then added above and below the bilayer.
296 mmeineke 1026
297     The random configurations took more work to generate. To begin, a
298     test lipid was placed in a simulation box already containing water at
299 mmeineke 1089 the intended density. The water molecules were then tested against
300     the lipid using a 5.0~$\mbox{\AA}$ overlap test with any atom in the
301     lipid. This gave an estimate for the number of water molecules each
302     lipid would displace in a simulation box. A target number of water
303     molecules was then defined which included the number of water
304     molecules each lipid would displace, the number of water molecules
305     desired to solvate each lipid, and a factor to pad the initial box
306     with a few extra water molecules.
307 mmeineke 1026
308     Next, a cubic simulation box was created that contained at least the
309 mmeineke 1089 target number of water molecules in an FCC lattice (the lattice was for ease of
310 mmeineke 1026 placement). What followed was a RSA simulation similar to those of
311     Chapt.~\ref{chapt:RSA}. The lipids were sequentially given a random
312     position and orientation within the box. If a lipid's position caused
313 mmeineke 1083 atomic overlap with any previously placed lipid, its position and
314     orientation were rejected, and a new random placement site was
315 mmeineke 1026 attempted. The RSA simulation proceeded until all phospholipids had
316 mmeineke 1089 been placed. After placement of all lipid molecules, water
317 mmeineke 1083 molecules with locations that overlapped with the atomic coordinates
318     of the lipids were removed.
319 mmeineke 1026
320 mmeineke 1089 Finally, water molecules were removed at random until the desired water
321     to lipid ratio was achieved. The typical low final density for these
322     initial configurations was not a problem, as the box shrinks to an
323     appropriate size within the first 50~ps of a simulation under the
324     NPTxyz integrator.
325 mmeineke 1026
326 mmeineke 1083 \subsection{\label{lipidSec:Configs}Configurations}
327 mmeineke 1026
328 mmeineke 1089 The first class of simulations were started from ordered bilayers. All
329     configurations consisted of 60 lipid molecules with 30 lipids on each
330     leaf, and were hydrated with 1620 {\sc ssd/e} molecules. The original
331     configuration was assembled according to Sec.~\ref{lipidSec:ExpSetup}
332     and simulated for a length of 10~ns at 300~K. The other temperature
333     runs were started from a configuration 7~ns in to the 300~K
334     simulation. Their temperatures were modified with the thermostatting
335     algorithm in the NPTxyz integrator. All of the temperature variants
336     were also run for 10~ns, with only the last 5~ns being used for
337     accumulation of statistics.
338 mmeineke 1083
339     The second class of simulations were two configurations started from
340     randomly dispersed lipids in a ``gas'' of water. The first
341     ($\text{R}_{\text{I}}$) was a simulation containing 72 lipids with
342 mmeineke 1089 1800 {\sc ssd/e} molecules simulated at 300~K. The second
343     ($\text{R}_{\text{II}}$) was 90 lipids with 1350 {\sc ssd/e} molecules
344 mmeineke 1083 simulated at 350~K. Both simulations were integrated for more than
345 mmeineke 1089 20~ns to observe whether our model is capable of spontaneous
346     aggregation into known phospholipid macro-structures:
347     $\text{R}_{\text{I}}$ into a bilayer, and $\text{R}_{\text{II}}$ into
348     a inverted rod.
349 mmeineke 1083
350     \section{\label{lipidSec:resultsDis}Results and Discussion}
351    
352 mmeineke 1084 \subsection{\label{lipidSec:densProf}Density Profile}
353    
354     Fig.~\ref{lipidFig:densityProfile} illustrates the densities of the
355 mmeineke 1087 atoms in the bilayer systems normalized by the bulk density as a
356 mmeineke 1084 function of distance from the center of the box. The profile is taken
357 mmeineke 1089 along the bilayer normal (in this case the $z$ axis). The profile at
358 mmeineke 1087 270~K shows several structural features that are largely smoothed out
359 mmeineke 1089 at 300~K. The left peak for the {\sc head} atoms is split at 270~K,
360     implying that some freezing of the structure into a gel phase might
361     already be occurring at this temperature. However, movies of the
362     trajectories at this temperature show that the tails are very fluid,
363     and have not gelled. But this profile could indicate that a phase
364     transition may simply be beyond the time length of the current
365     simulation, and that given more time the system may tend towards a gel
366     phase. In all profiles, the water penetrates almost 5~$\mbox{\AA}$
367     into the bilayer, completely solvating the {\sc head} atoms. The
368     $\text{{\sc ch}}_3$ atoms, although mainly centered at the middle of
369     the bilayer, show appreciable penetration into the head group
370     region. This indicates that the chains have enough flexibility to bend
371     back upward to allow the ends to explore areas around the {\sc head}
372     atoms. It is unlikely that this is penetration from a lipid of the
373     opposite face, as the lipids are only 12~$\mbox{\AA}$ in length, and
374     the typical leaf spacing as measured from the {\sc head-head} spacing
375     in the profile is 17.5~$\mbox{\AA}$.
376 mmeineke 1084
377     \begin{figure}
378     \centering
379     \includegraphics[width=\linewidth]{densityProfile.eps}
380     \caption[The density profile of the lipid bilayers]{The density profile of the lipid bilayers along the bilayer normal. The black lines are the {\sc head} atoms, red lines are the {\sc ch} atoms, green lines are the $\text{{\sc ch}}_2$ atoms, blue lines are the $\text{{\sc ch}}_3$ atoms, and the magenta lines are the {\sc ssd} atoms.}
381     \label{lipidFig:densityProfile}
382     \end{figure}
383    
384    
385     \subsection{\label{lipidSec:scd}$\text{S}_{\text{{\sc cd}}}$ Order Parameters}
386    
387 mmeineke 1083 The $\text{S}_{\text{{\sc cd}}}$ order parameter is often reported in
388 mmeineke 1087 the experimental characterizations of phospholipids. It is obtained
389 mmeineke 1083 through deuterium NMR, and measures the ordering of the carbon
390     deuterium bond in relation to the bilayer normal at various points
391     along the chains. In our model, there are no explicit hydrogens, but
392     the order parameter can be written in terms of the carbon ordering at
393     each point in the chain:\cite{egberts88}
394     \begin{equation}
395     S_{\text{{\sc cd}}} = \frac{2}{3}S_{xx} + \frac{1}{3}S_{yy}
396     \label{lipidEq:scd1}
397     \end{equation}
398     Where $S_{ij}$ is given by:
399     \begin{equation}
400 mmeineke 1084 S_{ij} = \frac{1}{2}\Bigl\langle(3\cos\Theta_i\cos\Theta_j
401     - \delta_{ij})\Bigr\rangle
402 mmeineke 1083 \label{lipidEq:scd2}
403     \end{equation}
404 mmeineke 1084 Here, $\Theta_i$ is the angle the $i$th axis in the reference frame of
405     the carbon atom makes with the bilayer normal. The brackets denote an
406     average over time and molecules. The carbon atom axes are defined:
407 mmeineke 1089 \begin{itemize}
408     \item $\mathbf{\hat{z}}\rightarrow$ vector from $C_{n-1}$ to $C_{n+1}$
409     \item $\mathbf{\hat{y}}\rightarrow$ vector that is perpendicular to $z$ and
410     in the plane through $C_{n-1}$, $C_{n}$, and $C_{n+1}$
411     \item $\mathbf{\hat{x}}\rightarrow$ vector perpendicular to
412 mmeineke 1083 $\mathbf{\hat{y}}$ and $\mathbf{\hat{z}}$.
413 mmeineke 1089 \end{itemize}
414     This assumes that the hydrogen atoms are always in a plane
415     perpendicular to the $C_{n-1}-C_{n}-C_{n+1}$ plane.
416 mmeineke 1083
417     The order parameter has a range of $[1,-\frac{1}{2}]$. A value of 1
418     implies full order aligned to the bilayer axis, 0 implies full
419 mmeineke 1087 disorder, and $-\frac{1}{2}$ implies full order perpendicular to the
420 mmeineke 1083 bilayer axis. The {\sc cd} bond vector for carbons near the head group
421 mmeineke 1087 are usually ordered perpendicular to the bilayer normal, with tails
422     farther away tending toward disorder. This makes the order parameter
423 mmeineke 1083 negative for most carbons, and as such $|S_{\text{{\sc cd}}}|$ is more
424     commonly reported than $S_{\text{{\sc cd}}}$.
425    
426 mmeineke 1084 Fig.~\ref{lipidFig:scdFig} shows the $S_{\text{{\sc cd}}}$ order
427     parameters for the bilayer system at 300~K. There is no appreciable
428     difference in the plots for the various temperatures, however, there
429 mmeineke 1089 is a larger difference between our model's ordering, and the
430     experimentally observed ordering of DMPC. As our values are closer to
431     $-\frac{1}{2}$, this implies more ordering perpendicular to the normal
432     than in a real system. This is due to the model having only one carbon
433     group separating the chains from the top of the lipid. In DMPC, with
434     the flexibility inherent in a multiple atom head group, as well as a
435     glycerol linkage between the head group and the acyl chains, there is
436     more loss of ordering by the point when the chains start.
437 mmeineke 1083
438     \begin{figure}
439     \centering
440     \includegraphics[width=\linewidth]{scdFig.eps}
441     \caption[$\text{S}_{\text{{\sc cd}}}$ order parameter for our model]{A comparison of $|\text{S}_{\text{{\sc cd}}}|$ between our model (blue) and DMPC\cite{petrache00} (black) near 300~K.}
442     \label{lipidFig:scdFig}
443     \end{figure}
444    
445 mmeineke 1087 \subsection{\label{lipidSec:p2Order}$P_2$ Order Parameter}
446 mmeineke 1083
447 mmeineke 1087 The $P_2$ order parameter allows us to measure the amount of
448 mmeineke 1089 directional ordering that exists in the bodies of the molecules making
449     up the bilayer. Each lipid molecule can be thought of as a cylindrical
450     rod with the head group at the top. If all of the rods are perfectly
451     aligned, the $P_2$ order parameter will be $1.0$. If the rods are
452     completely disordered, the $P_2$ order parameter will be 0. For a
453     collection of unit vectors pointing along the principal axes of the
454     rods, the $P_2$ order parameter can be solved via the following
455 mmeineke 1087 method.\cite{zannoni94}
456 mmeineke 1083
457 mmeineke 1089 Define an ordering tensor $\overleftrightarrow{\mathsf{Q}}$, such that,
458 mmeineke 1087 \begin{equation}
459 mmeineke 1089 \overleftrightarrow{\mathsf{Q}} = \frac{1}{N}\sum_i^N %
460 mmeineke 1087 \begin{pmatrix} %
461     u_{ix}u_{ix}-\frac{1}{3} & u_{ix}u_{iy} & u_{ix}u_{iz} \\
462     u_{iy}u_{ix} & u_{iy}u_{iy}-\frac{1}{3} & u_{iy}u_{iz} \\
463     u_{iz}u_{ix} & u_{iz}u_{iy} & u_{iz}u_{iz}-\frac{1}{3} %
464     \end{pmatrix}
465     \label{lipidEq:po1}
466     \end{equation}
467     Where the $u_{i\alpha}$ is the $\alpha$ element of the unit vector
468     $\mathbf{\hat{u}}_i$, and the sum over $i$ averages over the whole
469 mmeineke 1089 collection of unit vectors. This allows the tensor to be written:
470 mmeineke 1087 \begin{equation}
471 mmeineke 1089 \overleftrightarrow{\mathsf{Q}} = \frac{1}{N}\sum_i^N \biggl[
472     \mathbf{\hat{u}}_i \otimes \mathbf{\hat{u}}_i
473     - \frac{1}{3} \cdot \mathsf{1} \biggr]
474 mmeineke 1087 \label{lipidEq:po2}
475     \end{equation}
476 mmeineke 1083
477 mmeineke 1089 After constructing the tensor, diagonalizing
478     $\overleftrightarrow{\mathsf{Q}}$ yields three eigenvalues and
479     eigenvectors. The eigenvector associated with the largest eigenvalue,
480     $\lambda_{\text{max}}$, is the director axis for the system of unit
481     vectors. The director axis is the average direction all of the unit vectors
482     are pointing. The $P_2$ order parameter is then simply
483 mmeineke 1087 \begin{equation}
484     \langle P_2 \rangle = \frac{3}{2}\lambda_{\text{max}}
485     \label{lipidEq:po3}
486     \end{equation}
487 mmeineke 1083
488 mmeineke 1087 Table~\ref{lipidTab:blSummary} summarizes the $P_2$ values for the
489     bilayers, as well as the dipole orientations. The unit vector for the
490     lipid molecules was defined by finding the moment of inertia for each
491     lipid, then setting $\mathbf{\hat{u}}$ to point along the axis of
492     minimum inertia. For the {\sc head} atoms, the unit vector simply
493     pointed in the same direction as the dipole moment. For the lipid
494     molecules, the ordering was consistent across all temperatures, with
495     the director pointed along the $z$ axis of the box. More
496     interestingly, is the high degree of ordering the dipoles impose on
497 mmeineke 1089 the {\sc head} atoms. The directors for the dipoles themselves
498     consistently pointed along the plane of the bilayer, with the
499     directors anti-aligned on the top and bottom leaf.
500 mmeineke 1087
501 mmeineke 1083 \begin{table}
502 mmeineke 1089 \caption[Structural properties of the bilayers]{BILAYER STRUCTURAL PROPERTIES AS A FUNCTION OF TEMPERATURE}
503 mmeineke 1087 \label{lipidTab:blSummary}
504 mmeineke 1083 \begin{center}
505     \begin{tabular}{|c|c|c|c|c|}
506     \hline
507 mmeineke 1084 Temperature (K) & $\langle L_{\perp}\rangle$ ($\mbox{\AA}$) & %
508     $\langle A_{\parallel}\rangle$ ($\mbox{\AA}^2$) & %
509     $\langle P_2\rangle_{\text{Lipid}}$ & %
510     $\langle P_2\rangle_{\text{{\sc head}}}$ \\ \hline
511 mmeineke 1083 270 & 18.1 & 58.1 & 0.253 & 0.494 \\ \hline
512     275 & 17.2 & 56.7 & 0.295 & 0.514 \\ \hline
513     277 & 16.9 & 58.0 & 0.301 & 0.541 \\ \hline
514     280 & 17.4 & 58.0 & 0.274 & 0.488 \\ \hline
515     285 & 16.9 & 57.6 & 0.270 & 0.616 \\ \hline
516     290 & 17.0 & 57.6 & 0.263 & 0.534 \\ \hline
517     293 & 17.5 & 58.0 & 0.227 & 0.643 \\ \hline
518     300 & 16.9 & 57.6 & 0.315 & 0.536 \\ \hline
519     \end{tabular}
520     \end{center}
521     \end{table}
522 mmeineke 1087
523 mmeineke 1089 \subsection{\label{lipidSec:miscData}Further Structural Data}
524 mmeineke 1087
525     Also summarized in Table~\ref{lipidTab:blSummary}, are the bilayer
526 mmeineke 1089 thickness ($\langle L_{\perp}\rangle$) and area per lipid ($\langle
527     A_{\parallel}\rangle$). The bilayer thickness was measured from the
528     peak to peak {\sc head} atom distance in the density profiles. The
529 mmeineke 1087 area per lipid data compares favorably with values typically seen for
530 mmeineke 1089 DMPC (60.0~$\mbox{\AA}^2$ at 303~K)\cite{petrache00}. Although our
531 mmeineke 1087 values are lower this is most likely due to the shorter chain length
532     of our model (8 versus 14 for DMPC).
533    
534 mmeineke 1089 \subsection{\label{lipidSec:diffusion}Lateral Diffusion Constants}
535    
536     The lateral diffusion constant, $D_L$, is the constant characterizing
537     the diffusive motion of the lipid molecules within the plane of the bilayer. It
538     is given by the following Einstein relation:\cite{allen87:csl}
539     \begin{equation}
540     D_L = \lim_{t\rightarrow\infty}\frac{1}{4t}\langle |\mathbf{r}(t)
541     - \mathbf{r}(0)|^2\rangle
542     \end{equation}
543     Where $\mathbf{r}(t)$ is the $xy$ position of the lipid at time $t$
544     (assuming the $z$-axis is parallel to the bilayer normal).
545    
546     Fig.~\ref{lipidFig:diffusionFig} shows the lateral diffusion constants
547     as a function of temperature. There is a definite increase in the
548     lateral diffusion with higher temperatures, which is exactly what one
549     would expect with greater fluidity of the chains. However, the
550     diffusion constants are two orders of magnitude smaller than those
551     typical of DPPC.\cite{Cevc87} This is counter-intuitive as the DPPC
552     molecule is sterically larger and heavier than our model. This could
553     be an indication that our model's chains are too interwoven and hinder
554     the motion of the lipid or that the dipolar head groups are too
555     tightly bound to each other. In contrast, the diffusion constant of
556     the {\sc ssd} water, $9.84\times 10^{-6}\,\text{cm}^2/\text{s}$, is
557     reasonably close to the bulk water diffusion constant ($2.2999\times
558     10^{-5}\,\text{cm}^2/\text{s}$).\cite{Holz00}
559    
560     \begin{figure}
561     \centering
562 mmeineke 1112 \includegraphics[width=\linewidth]{msdFig.eps}
563     \caption[Lateral mean square displacement for the phospholipid at 300~K]{This is a representative lateral mean square displacement for the center of mass motion of the phospholipid model. This particular example is from the 300~K run. The box is drawn about the region used in the calculation of the diffusion constant.}
564     \label{lipidFig:msdFig}
565     \end{figure}
566    
567     \begin{figure}
568     \centering
569 mmeineke 1089 \includegraphics[width=\linewidth]{diffusionFig.eps}
570     \caption[The lateral diffusion constants versus temperature]{The lateral diffusion constants for the bilayers as a function of temperature.}
571     \label{lipidFig:diffusionFig}
572     \end{figure}
573    
574 mmeineke 1087 \subsection{\label{lipidSec:randBilayer}Bilayer Aggregation}
575    
576     A very important accomplishment for our model is its ability to
577     spontaneously form bilayers from a randomly dispersed starting
578     configuration. Fig.~\ref{lipidFig:blImage} shows an image sequence for
579     the bilayer aggregation. After 3.0~ns, the basic form of the bilayer
580     can already be seen. By 7.0~ns, the bilayer has a lipid bridge
581     stretched across the simulation box to itself that will turn out to be
582     very long lived ($\sim$20~ns), as well as a water pore, that will
583     persist for the length of the current simulation. At 24~ns, the lipid
584 mmeineke 1089 bridge has broken, and the bilayer is still integrating the lipid
585     molecules from the bridge into itself. However, the water pore is
586     still present at 24~ns.
587 mmeineke 1087
588     \begin{figure}
589     \centering
590     \includegraphics[width=\linewidth]{bLayerImage.eps}
591     \caption[Image sequence of the bilayer aggregation]{Image sequence of the bilayer aggregation. The blue beads are the {\sc head} atoms the grey beads are the chains, and the red and white bead are the water molecules. A box has been drawn around the periodic image.}
592     \label{lipidFig:blImage}
593     \end{figure}
594    
595     \subsection{\label{lipidSec:randIrod}Inverted Rod Aggregation}
596    
597     Fig.~\ref{lipidFig:iRimage} shows a second aggregation sequence
598     simulated in this research. Here the fraction of water had been
599     significantly decreased to observe how the model would respond. After
600     1.5~ns, The main body of water in the system has already collected
601     into a central water channel. By 10.0~ns, the channel has widened
602 mmeineke 1089 slightly, but there are still many water molecules permeating the
603     lipid macro-structure. At 23.0~ns, the central water channel has
604     stabilized and several smaller water channels have been absorbed by
605     the main one. However, there is still an appreciable water
606     concentration throughout the lipid structure.
607 mmeineke 1087
608     \begin{figure}
609     \centering
610     \includegraphics[width=\linewidth]{iRodImage.eps}
611     \caption[Image sequence of the inverted rod aggregation]{Image sequence of the inverted rod aggregation. color scheme is the same as in Fig.~\ref{lipidFig:blImage}.}
612     \label{lipidFig:iRimage}
613     \end{figure}
614    
615     \section{\label{lipidSec:Conclusion}Conclusion}
616    
617 mmeineke 1089 We have presented a simple unified-atom phospholipid model capable of
618     spontaneous aggregation into a bilayer and an inverted rod
619     structure. The time scales of the macro-molecular aggregations are
620     approximately 24~ns. In addition the model's properties have been
621     explored over a range of temperatures through prefabricated
622     bilayers. No freezing transition is seen in the temperature range of
623     our current simulations. However, structural information from 270~K
624     may imply that a freezing event is on a much longer time scale than
625     that explored in this current research. Further studies of this system
626     could extend the time length of the simulations at the low
627     temperatures to observe whether lipid crystallization can occur within
628     the framework of this model.
629    
630     Potential problems that may be obstacles in further research, is the
631     lack of detail in the head region. As the chains are almost directly
632     attached to the {\sc head} atom, there is no buffer between the
633     actions of the head group and the tails. Another disadvantage of the
634     model is the dipole approximation will alter results when details
635     concerning a charged solute's interactions with the bilayer. However,
636     it is important to keep in mind that the dipole approximation can be
637     kept an advantage by examining solutes that do not require point
638     charges, or at the least, require only dipole approximations
639     themselves. Other advantages of the model include the ability to alter
640     the size of the unified-atoms so that the size of the lipid can be
641     increased without adding to the number of interactions in the
642     system. However, what sets our model apart from other current
643     simplified models,\cite{goetz98,marrink04} is the information gained
644     by observing the ordering of the head groups dipole's in relation to
645     each other and the solvent without the need for point charges and the
646     Ewald sum.