ViewVC Help
View File | Revision Log | Show Annotations | View Changeset | Root Listing
root/group/trunk/tengDissertation/Lipid.tex
Revision: 2905
Committed: Wed Jun 28 19:09:25 2006 UTC (18 years ago) by tim
Content type: application/x-tex
File size: 21116 byte(s)
Log Message:
more corrections.

File Contents

# Content
1 \chapter{\label{chapt:lipid}LIPID MODELING}
2
3 \section{\label{lipidSection:introduction}Introduction}
4
5 Under biologically relevant conditions, phospholipids are solvated
6 in aqueous solutions at a roughly 25:1 ratio. Solvation can have a
7 tremendous impact on transport phenomena in biological membranes
8 since it can affect the dynamics of ions and molecules that are
9 transferred across membranes. Studies suggest that because of the
10 directional hydrogen bonding ability of the lipid headgroups, a
11 small number of water molecules are strongly held around the
12 different parts of the headgroup and are oriented by them with
13 residence times for the first hydration shell being around 0.5 - 1
14 ns\cite{Ho1992}. In the second solvation shell, some water molecules
15 are weakly bound, but are still essential for determining the
16 properties of the system. Transport of various molecular species
17 into living cells is one of the major functions of membranes. A
18 thorough understanding of the underlying molecular mechanism for
19 solute diffusion is crucial to the further studies of other related
20 biological processes. All transport across cell membranes takes
21 place by one of two fundamental processes: Passive transport is
22 driven by bulk or inter-diffusion of the molecules being transported
23 or by membrane pores which facilitate crossing. Active transport
24 depends upon the expenditure of cellular energy in the form of ATP
25 hydrolysis. As the central processes of membrane assembly,
26 translocation of phospholipids across membrane bilayers requires the
27 hydrophilic head of the phospholipid to pass through the highly
28 hydrophobic interior of the membrane, and for the hydrophobic tails
29 to be exposed to the aqueous environment\cite{Sasaki2004}. A number
30 of studies indicate that the flipping of phospholipids occurs
31 rapidly in the eukaryotic endoplasmic reticulum and the bacterial
32 cytoplasmic membrane via a bi-directional, facilitated diffusion
33 process requiring no metabolic energy input. Another system of
34 interest is the distribution of sites occupied by inhaled
35 anesthetics in membrane. Although the physiological effects of
36 anesthetics have been extensively studied, the controversy over
37 their effects on lipid bilayers still continues. Recent deuterium
38 NMR measurements on halothane on POPC lipid bilayers suggest the
39 anesthetics are primarily located at the hydrocarbon chain
40 region\cite{Baber1995}. However, infrared spectroscopy experiments
41 suggest that halothane in DMPC lipid bilayers lives near the
42 membrane/water interface\cite{Lieb1982}.
43
44 Molecular dynamics simulations have proven to be a powerful tool for
45 studying the functions of biological systems, providing structural,
46 thermodynamic and dynamical information. Unfortunately, much of
47 biological interest happens on time and length scales well beyond
48 the range of current simulation technologies. Several schemes are
49 proposed in this chapter to overcome these difficulties.
50
51 \section{\label{lipidSection:model}Model and Methodology}
52
53 \subsection{\label{lipidSection:SSD}The Soft Sticky Dipole Water Model}
54
55 In a typical bilayer simulation, the dominant portion of the
56 computation time will be spent calculating water-water interactions.
57 As an efficient solvent model, the Soft Sticky Dipole (SSD) water
58 model\cite{Chandra1999,Fennell2004} is used as the explicit solvent
59 in this project. Unlike other water models which have partial
60 charges distributed throughout the whole molecule, the SSD water
61 model consists of a single site which is a Lennard-Jones interaction
62 site, as well as a point dipole. A tetrahedral potential is added to
63 correct for hydrogen bonding. The following equation describes the
64 interaction between two water molecules:
65 \begin{equation}
66 V_{SSD} = V_{LJ} (r_{ij} ) + V_{dp} (r_{ij} ,\Omega _i ,\Omega _j )
67 + V_{sticky} (r_{ij} ,\Omega _i ,\Omega _j )
68 \label{lipidSection:ssdEquation}
69 \end{equation}
70 where$r_{ij}$ is the vector between molecule $i$ and molecule $j$,
71 $\Omega _i$ and $\Omega _j$ are the orientational degrees of freedom
72 for molecule $i$ and molecule $j$ respectively. The potential terms
73 in Eq.~\ref{lipidSection:ssdEquation} are given by :
74 \begin{eqnarray}
75 V_{LJ} (r_{ij} ) &= &4\varepsilon _{ij} \left[ {\left(
76 {\frac{{\sigma _{ij} }}{{r_{ij} }}} \right)^{12} - \left(
77 {\frac{{\sigma _{ij}
78 }}{{r_{ij} }}} \right)^6 } \right], \\
79 V_{dp} (r_{ij} ,\Omega _i ,\Omega _j ) &= &
80 \frac{|\mu_i||\mu_j|}{4\pi\epsilon_{0}r_{ij}^{3}} \biggl[
81 \hat{u}_{i} \cdot \hat{u}_{j} - 3(\hat{u}_i \cdot \hat{\mathbf{r}}_{ij}) %
82 (\hat{u}_j \cdot \hat{\mathbf{r}}_{ij}) \biggr],\\
83 V_{sticky} (r_{ij} ,\Omega _i ,\Omega _j ) &=& v_0 [s(r_{ij}
84 )w(r_{ij} ,\Omega _i ,\Omega _j ) + s'(r_{ij} )w'(r_{ij} ,\Omega _i
85 ,\Omega _j )]
86 \end{eqnarray}
87 where $v_0$ is a strength parameter, $s$ and $s'$ are cubic
88 switching functions, while $w$ and $w'$ are responsible for the
89 tetrahedral potential and the short-range correction to the dipolar
90 interaction respectively:
91 \begin{eqnarray}
92 w(r_{ij} ,\Omega _i ,\Omega _j )& = &\sin \theta _{ij} \sin 2\theta _{ij} \cos 2\phi _{ij}, \\
93 w'(r_{ij} ,\Omega _i ,\Omega _j )& = &(\cos \theta _{ij} - 0.6)^2 (\cos \theta _{ij} + 0.8)^2 -
94 w_0.
95 \end{eqnarray}
96 Here $\theta _{ij}$ and $\phi _{ij}$ are the spherical polar angles
97 representing relative orientations between molecule $i$ and molecule
98 $j$. Although the dipole-dipole and sticky interactions are more
99 mathematically complicated than Coulomb interactions, the number of
100 pair interactions is reduced dramatically both because the model
101 only contains a single-point as well as "short range" nature of the
102 more expensive interaction.
103
104 \subsection{\label{lipidSection:coarseGrained}The Coarse-Grained Phospholipid Model}
105
106 Fig.~\ref{lipidFigure:coarseGrained} shows a schematic for our
107 coarse-grained phospholipid model. The lipid head group is modeled
108 by a linear rigid body which consists of three Lennard-Jones spheres
109 and a centrally located point-dipole. The backbone atoms in the
110 glycerol motif are modeled by Lennard-Jones spheres with dipoles.
111 Alkyl groups in hydrocarbon chains are replaced with unified
112 $\text{{\sc CH}}_2$ or $\text{{\sc CH}}_3$ atoms.
113 \begin{figure}
114 \centering
115 \includegraphics[width=3in]{coarse_grained.eps}
116 \caption[A representation of coarse-grained phospholipid model]{A
117 representation of coarse-grained phospholipid model. The lipid
118 headgroup consists of $\text{{\sc PO}}_4$ group (yellow),
119 $\text{{\sc NC}}_4$ group (blue) and a united C atom (gray) with a
120 dipole, while the glycerol backbone includes dipolar $\text{{\sc
121 CE}}$ (red) and $\text{{\sc CK}}$ (pink) groups. Alkyl groups in
122 hydrocarbon chains are simply represented by gray united atoms.}
123 \label{lipidFigure:coarseGrained}
124 \end{figure}
125
126 Accurate and efficient computation of electrostatics is one of the
127 most difficult tasks in molecular modeling. Traditionally, the
128 electrostatic interaction between two molecular species is
129 calculated as a sum of interactions between pairs of point charges,
130 using Coulomb's law:
131 \[
132 V = \sum\limits_{i = 1}^{N_A } {\sum\limits_{j = 1}^{N_B }
133 {\frac{{q_i q_j }}{{4\pi \varepsilon _0 r_{ij} }}} }
134 \]
135 where $N_A$ and $N_B$ are the number of point charges in the two
136 molecular species. Originally developed to study ionic crystals, the
137 Ewald sum method mathematically transforms this straightforward but
138 conditionally convergent summation into two more complicated but
139 rapidly convergent sums. One summation is carried out in reciprocal
140 space while the other is carried out in real space. An alternative
141 approach is the multipole expansion, which is based on electrostatic
142 moments, such as charge (monopole), dipole, quadrupole etc.
143
144 Here we consider a linear molecule which consists of two point
145 charges $\pm q$ located at $ ( \pm \frac{d}{2},0)$. The
146 electrostatic potential at point $P$ is given by:
147 \[
148 \frac{1}{{4\pi \varepsilon _0 }}\left( {\frac{{ - q}}{{r_ - }} +
149 \frac{{ + q}}{{r_ + }}} \right) = \frac{1}{{4\pi \varepsilon _0
150 }}\left( {\frac{{ - q}}{{\sqrt {r^2 + \frac{{d^2 }}{4} + rd\cos
151 \theta } }} + \frac{q}{{\sqrt {r^2 + \frac{{d^2 }}{4} - rd\cos
152 \theta } }}} \right)
153 \]
154 \begin{figure}
155 \centering
156 \includegraphics[width=3in]{charge_dipole.eps}
157 \caption[An illustration of split-dipole
158 approximation]{Electrostatic potential due to a linear molecule
159 comprising two point charges with opposite charges. }
160 \label{lipidFigure:chargeDipole}
161 \end{figure}
162 The basic assumption of the multipole expansion is $r \gg d$ , thus,
163 $\frac{{d^2 }}{4}$ inside the square root of the denominator is
164 neglected. This is a reasonable approximation in most cases.
165 Unfortunately, in our headgroup model, the distance of charge
166 separation $d$ (4.63 \AA in PC headgroup) may be comparable to $r$.
167 Nevertheless, we could still assume $ \cos \theta \approx 0$ in
168 the central region of the headgroup. Using Taylor expansion and
169 associating appropriate terms with electric moments will result in a
170 "split-dipole" approximation:
171 \[
172 V(r) = \frac{1}{{4\pi \varepsilon _0 }}\frac{{r\mu \cos \theta
173 }}{{R^3 }}
174 \]
175 where$R = \sqrt {r^2 + \frac{{d^2 }}{4}}$ Placing a dipole at point
176 $P$ and applying the same strategy, the interaction between two
177 split-dipoles is then given by:
178 \[
179 V_{sd} (r_{ij} ,\Omega _i ,\Omega _j ) = \frac{1}{{4\pi \varepsilon
180 _0 }}\left[ {\frac{{\mu _i \cdot \mu _j }}{{R_{ij}^3 }} -
181 \frac{{3\left( {\mu _i \cdot r_{ij} } \right)\left( {\mu _i \cdot
182 r_{ij} } \right)}}{{R_{ij}^5 }}} \right]
183 \]
184 where $\mu _i$ and $\mu _j$ are the dipole moment of molecule $i$
185 and molecule $j$ respectively, $r_{ij}$ is vector between molecule
186 $i$ and molecule $j$, and $R_{ij}$ is given by,
187 \[
188 R_{ij} = \sqrt {r_{ij}^2 + \frac{{d_i^2 }}{4} + \frac{{d_j^2
189 }}{4}}
190 \]
191 where $d_i$ and $d_j$ are the charge separation distance of dipole
192 and respectively. This approximation to the multipole expansion
193 maintains the fast fall-off of the multipole potentials but lacks
194 the normal divergences when two polar groups get close to one
195 another. The comparision between different electrostatic
196 approximation is shown in \ref{lipidFigure:splitDipole}. Due to the
197 divergence at the central region of the headgroup introduced by
198 dipole-dipole approximation, we discover that water molecules are
199 locked into the central region in the simulation. This artifact can
200 be corrected using split-dipole approximation or other accurate
201 methods.
202 \begin{figure}
203 \centering
204 \includegraphics[width=\linewidth]{split_dipole.eps}
205 \caption[Comparison between electrostatic
206 approximation]{Electrostatic potential map for two pairs of charges
207 with different alignments: (a) illustration of different alignments;
208 (b) charge-charge interaction; (c) dipole-dipole approximation; (d)
209 split-dipole approximation.} \label{lipidFigure:splitDipole}
210 \end{figure}
211
212 \section{\label{lipidSection:resultDiscussion}Results and Discussion}
213
214 \subsection{One Lipid in Sea of Water Molecules}
215
216 To tune our parameters without the inter-headgroup interactions,
217 atomistic models of one lipid (DMPC or DLPE) in sea of water
218 molecules (TIP3P) were built and studied using atomistic molecular
219 dynamics. The simulation was analyzed using a set of radial
220 distribution functions, which give the probability of finding a pair
221 of molecular species a distance apart, relative to the probability
222 expected for a completely random distribution function at the same
223 density
224 \begin{eqnarray}
225 g_{AB} (r) & = & \frac{1}{{\rho _B }}\frac{1}{{N_A }} <
226 \sum\limits_{i
227 \in A} {\sum\limits_{j \in B} {\delta (r - r_{ij} )} } >, \\
228 g_{AB} (r,\cos \theta ) & = & \frac{1}{{\rho _B }}\frac{1}{{N_A }} <
229 \sum\limits_{i \in A} {\sum\limits_{j \in B} {\delta (r - r_{ij} )}
230 } \delta (\cos \theta _{ij} - \cos \theta ) >.
231 \end{eqnarray}
232 From Fig.~\ref{lipidFigure:PCFAtom}, we can identify the first
233 solvation shell (3.5 \AA) and the second solvation shell (5.0 \AA)
234 from both plots. However, the corresponding orientations are
235 different. In DLPE, water molecules prefer to sit around $\text{{\sc
236 NH}}_3$ group due to the hydrogen bonding. In contrast, because of
237 the hydrophobic effect of the $ {\rm{N(CH}}_{\rm{3}}
238 {\rm{)}}_{\rm{3}} $ group, the preferred position of water molecules
239 in DMPC is around the $\text{{\sc PO}}_4$ Group. When the water
240 molecules are far from the headgroup, the distribution of the two
241 angles should be uniform. The correlation close to center of the
242 headgroup dipole in both plots tells us that in the closely-bound
243 region, the dipoles of the water molecules are preferentially
244 anti-aligned with the dipole of headgroup. When the water molecules
245 are far from the headgroup, the distribution of the two angles
246 should be uniform. The correlation close to center of the headgroup
247 dipole in both plots tell us that in the closely-bound region, the
248 dipoles of the water molecules are preferentially aligned with the
249 dipole of headgroup.
250 \begin{figure}
251 \centering
252 \includegraphics[width=\linewidth]{g_atom.eps}
253 \caption[The pair correlation functions for atomistic models]{The
254 pair correlation functions for atomistic models: (a)$g(r,\cos \theta
255 )$ for DMPC; (b) $g(r,\cos \theta )$ for DLPE; (c)$g(r,\cos \omega
256 )$ for DMPC; (d)$g(r,\cos \omega )$ for DLPE; (e)$g(\cos \theta,\cos
257 \omega)$ for DMPC; (f)$g(\cos \theta,\cos \omega)$ for DMLPE.}
258 \label{lipidFigure:PCFAtom}
259 \end{figure}
260
261 The initial configurations of coarse-grained systems are constructed
262 from the previous atomistic ones. The parameters for the
263 coarse-grained model in Table~\ref{lipidTable:parameter} are
264 estimated and tuned using isothermal-isobaric molecular dynamics.
265 Pair distribution functions calculated from coarse-grained models
266 preserve the basic characteristics of the atomistic simulations. The
267 water density, measured in a head-group-fixed reference frame,
268 surrounding two phospholipid headgroups is shown in
269 Fig.~\ref{lipidFigure:PCFCoarse}. It is clear that the phosphate end
270 in DMPC and the amine end in DMPE are the two most heavily solvated
271 atoms.
272 \begin{figure}
273 \centering
274 \includegraphics[width=\linewidth]{g_coarse.eps}
275 \caption[The pair correlation functions for coarse-grained
276 models]{The pair correlation functions for coarse-grained models:
277 (a)$g(r,\cos \theta )$ for DMPC; (b) $g(r,\cos \theta )$ for DLPE.}
278 \label{lipidFigure:PCFCoarse}
279 \end{figure}
280 \begin{figure}
281 \centering
282 \includegraphics[width=\linewidth]{EWD_coarse.eps}
283 \caption[Excess water density of coarse-grained
284 phospholipids]{Excess water density of coarse-grained
285 phospholipids.} \label{lipidFigure:EWDCoarse}
286 \end{figure}
287
288 \begin{table}
289 \caption{THE PARAMETERS FOR COARSE-GRAINED PHOSPHOLIPIDS}
290 \label{lipidTable:parameter}
291 \begin{center}
292 \begin{tabular}{lccccc}
293 \hline
294 % after \\: \hline or \cline{col1-col2} \cline{col3-col4} ...
295 Atom type & Mass & $\sigma$ & $\epsilon$ & charge & Dipole \\
296 \hline
297 $\text{{\sc CH}}_2$ & 14.03 & 3.95 & 0.0914 & 0 & 0 \\
298 $\text{{\sc CH}}_3$ & 15.04 & 3.75 & 0.195 & 0 & 0 \\
299 $\text{{\sc CE}}$ & 28.01 & 3.427& 0.294 & 0 & 1.693 \\
300 $\text{{\sc CK}}$ & 28.01 & 3.592& 0.311 & 0 & 2.478 \\
301 $\text{{\sc PO}}_4$ & 108.995& 3.9 & 1.88 & -1& 0 \\
302 $\text{{\sc HDP}}$ & 14.03 & 4.0 & 0.13 & 0 & 0 \\
303 $\text{{\sc NC}}_4$ & 73.137 & 4.9 & 0.88 & +1& 0 \\
304 $\text{{\sc NH}}_3$ & 17.03 & 3.25 & 0.17 & +1& 0\\
305 \hline
306 \end{tabular}
307 \end{center}
308 \end{table}
309
310 \subsection{Bilayer Simulations Using Coarse-grained Model}
311
312 A bilayer system consisting of 128 DMPC lipids and 3655 water
313 molecules has been constructed from an atomistic coordinate file.
314 The MD simulation is performed at constant temperature, T = 300K,
315 and constant pressure, p = 1 atm, and consisted of an equilibration
316 period of 2 ns. During the equilibration period, the system was
317 initially simulated at constant volume for 1 ns. Once the system was
318 equilibrated at constant volume, the cell dimensions of the system
319 was relaxed by performing under NPT conditions using Nos\'{e}-Hoover
320 extended system isothermal-isobaric dynamics. After equilibration,
321 different properties were evaluated over a production run of 5 ns.
322 \begin{figure}
323 \centering
324 \includegraphics[width=\linewidth]{bilayer.eps}
325 \caption[Image of a coarse-grained bilayer system]{A coarse-grained
326 bilayer system consisting of 128 DMPC lipids and 3655 SSD water
327 molecules.}
328 \label{lipidFigure:bilayer}
329 \end{figure}
330
331 \subsubsection{\textbf{Electron Density Profile (EDP)}}
332
333 Assuming a gaussian distribution of electrons on each atomic center
334 with a variance estimated from the size of the van der Waals radius,
335 the EDPs which are proportional to the density profiles measured
336 along the bilayer normal obtained by x-ray scattering experiment,
337 can be expressed by\cite{Tu1995}
338 \begin{equation}
339 \rho _{x - ray} (z)dz \propto \sum\limits_{i = 1}^N {\frac{{n_i
340 }}{V}\frac{1}{{\sqrt {2\pi \sigma ^2 } }}e^{ - (z - z_i )^2 /2\sigma
341 ^2 } dz},
342 \end{equation}
343 where $\sigma$ is the variance equal to the van der Waals radius,
344 $n_i$ is the atomic number of site $i$ and $V$ is the volume of the
345 slab between $z$ and $z+dz$ . The highest density of total EDP
346 appears at the position of lipid-water interface corresponding to
347 headgroup, glycerol, and carbonyl groups of the lipids and the
348 distribution of water locked near the head groups, while the lowest
349 electron density is in the hydrocarbon region. As a good
350 approximation to the thickness of the bilayer, the headgroup spacing
351 , is defined as the distance between two peaks in the electron
352 density profile, calculated from our simulations to be 34.1 \AA.
353 This value is close to the x-ray diffraction experimental value 34.4
354 \AA\cite{Petrache1998}.
355
356 \begin{figure}
357 \centering
358 \includegraphics[width=\linewidth]{electron_density.eps}
359 \caption[The density profile of the lipid bilayers]{Electron density
360 profile along the bilayer normal. The water density is shown in red,
361 the density due to the headgroups in green, the glycerol backbone in
362 brown, $\text{{\sc CH}}_2$ in yellow, $\text{{\sc CH}}_3$ in cyan,
363 and total density due to DMPC in blue.}
364 \label{lipidFigure:electronDensity}
365 \end{figure}
366
367 \subsubsection{\textbf{$\text{S}_{\text{{\sc cd}}}$ Order Parameter}}
368
369 Measuring deuterium order parameters by NMR is a useful technique to
370 study the orientation of hydrocarbon chains in phospholipids. The
371 order parameter tensor $S$ is defined by:
372 \begin{equation}
373 S_{ij} = \frac{1}{2} < 3\cos \theta _i \cos \theta _j - \delta
374 _{ij} >
375 \end{equation}
376 where $\theta$ is the angle between the $i$th molecular axis and
377 the bilayer normal ($z$ axis). The brackets denote an average over
378 time and molecules. The molecular axes are defined:
379 \begin{itemize}
380 \item $\mathbf{\hat{z}}$ is the vector from $C_{n-1}$ to $C_{n+1}$.
381 \item $\mathbf{\hat{y}}$ is the vector that is perpendicular to $z$ and
382 in the plane through $C_{n-1}$, $C_{n}$, and $C_{n+1}$.
383 \item $\mathbf{\hat{x}}$ is the vector perpendicular to
384 $\mathbf{\hat{y}}$ and $\mathbf{\hat{z}}$.
385 \end{itemize}
386 In coarse-grained model, although there are no explicit hydrogens,
387 the order parameter can still be written in terms of carbon ordering
388 at each point of the chain\cite{Egberts1988}
389 \begin{equation}
390 S_{ij} = \frac{1}{2} < 3\cos \theta _i \cos \theta _j - \delta
391 _{ij} >.
392 \end{equation}
393
394 Fig.~\ref{lipidFigure:Scd} shows the order parameter profile
395 calculated for our coarse-grained DMPC bilayer system at 300K. Also
396 shown are the experimental data of Tu\cite{Tu1995}. The fact that
397 $\text{S}_{\text{{\sc cd}}}$ order parameters calculated from
398 simulation are higher than the experimental ones is ascribed to the
399 assumption of the locations of implicit hydrogen atoms which are
400 fixed in coarse-grained models at positions relative to the CC
401 vector.
402
403 \begin{figure}
404 \centering
405 \includegraphics[width=\linewidth]{scd.eps}
406 \caption[$\text{S}_{\text{{\sc cd}}}$ order parameter]{A comparison
407 of $|\text{S}_{\text{{\sc cd}}}|$ between coarse-grained model
408 (blue) and DMPC\cite{Petrache2000} (black) near 300~K.}
409 \label{lipidFigure:Scd}
410 \end{figure}
411
412 %\subsection{Bilayer Simulations Using Gay-Berne Ellipsoid Model}
413
414 \section{\label{lipidSection:Conclusion}Conclusion}
415
416 Atomistic simulations have been used in this study to determine the
417 preferred orientation and location of water molecules relative to
418 the location and orientation of the PC and PE lipid headgroups.
419 Based on the results from our all-atom simulations, we developed a
420 simple coarse-grained model which captures the essential features of
421 the headgroup solvation which is crucial to transport process in
422 membrane system. In addition, the model has been explored in a
423 bilayer system was shown to have reasonable electron density
424 profiles, $\text{S}_{\text{{\sc cd}}}$ order parameter and other
425 structural properties. The accuracy of this model is achieved by
426 matching atomistic result. It is also easy to represent different
427 phospholipids by changing a few parameters of the model. Another
428 important characteristic of this model distinguishing itself from
429 other models\cite{Goetz1998,Marrink2004} is the computational speed
430 gained by introducing a short range electrostatic approximation.
431 Further studies of this system using z-constraint method could
432 extend the time length of the simulations to study transport
433 phenomena in large-scale membrane systems.