1946
J. Phys. Chem. B 2008, 112, 1946-1952
Replacing the Cholesterol Hydroxyl Group with the Ketone Group Facilitates Sterol Flip-Flop and Promotes Membrane Fluidity Tomasz Ro´ g,†,‡ Lorna M. Stimson,§,¶ Marta Pasenkiewicz-Gierula,‡ Ilpo Vattulainen,§,#,£ and Mikko Karttunen*,¶ Faculty of Electrical and Communications Engineering, Helsinki UniVersity of Technology, Espoo, Finland, Department of Biophysics, Faculty of Biochemistry, Biophysics and Biotechnology, Jagiellonian UniVersity, Krako´ w, Poland, Laboratory of Physics and Helsinki Institute of Physics, Helsinki UniVersity of Technology, Espoo, Finland, Department of Applied Mathematics, The UniVersity of Western Ontario, London (ON), Canada, MEMPHYS-Center for Biomembrane Physics, UniVersity of Southern Denmark, Odense, Denmark, and Institute of Physics, Tampere UniVersity of Technology, Tampere, Finland ReceiVed: June 29, 2007; In Final Form: NoVember 20, 2007
The 3R-hydroxyl group is a characteristic structural element of all membrane sterol molecules, while the 3-ketone group is more typically found in steroid hormones. In this work, we investigate the effect of substituting the hydroxyl group in cholesterol with the ketone group to produce ketosterone. Extensive atomistic molecular dynamics simulations of saturated lipid membranes with either cholesterol or ketosterone show that, like cholesterol, ketosterone increases membrane order and induces condensation. However, the effect of ketosterone on membrane properties is considerably weaker than that of cholesterol. This is largely due to the unstable positioning of ketosterone at the membrane-water interface, which gives rise to a small but significant number of flip-flop transitions, where ketosterone is exchanged between membrane leaflets. This is remarkable, as flip-flop motions of sterol molecules have not been previously reported in analogous lipid bilayer simulations. In the same context, ketosterone is found to be more tilted with respect to the membrane normal than cholesterol. The atomic level mechanism responsible for the increase of the steroid tilt and the promotion of flip-flops is the decrease in polar interactions at the membrane-water interface. Interactions between lipids or water and the ketone group are found to be weaker than in the case of the hydroxyl group, which allows ketosterone to penetrate through the hydrocarbon region of a membrane.
Introduction Cholesterol (Chol) is a highly specialized molecule with a unique structure that is the result of refinement by extensive molecular evolution. The necessity of this specialized structure is demonstrated by the high concentration of cholesterol observed in biological membranes, as well as its numerous biological functions. Chol concentration in cell membranes is typically about 30 mol %, though in red blood cells, it may reach 50 mol %,1 and it may be as high as 70 mol % in specific membranes such as those found in the ocular lens.2 The biological role of Chol includes the maintenance of proper fluidity,3 the formation of glycosphingolipid-Chol-enriched raft domains,4 reduction of the passive permeability,5 and increase in the mechanical strength6 of the membrane (for recent reviews, see ref 7). The unique properties of Chol are also highlighted by the fact that, despite the large family of chemical species that sterols constitute, only Chol and the structurally similar ergosterol are found in substantial amounts in nature (Chol in animal and ergosterol in fungus membranes). * Address correspondence and reprint requests to Mikko Karttunen, E-mail:
[email protected], Web: www.softsimu.org. † Faculty of Electrical and Communications Engineering, Helsinki University of Technology. ‡ Jagiellonian University. § Laboratory of Physics and Helsinki Institute of Physics, Helsinki University of Technology. ¶ The University of Western Ontario. # University of Southern Denmark. £ Tampere University of Technology.
The Chol molecule consists of a planar tetracyclic ring system with the 3β-hydroxyl group and a short 8-carbon chain attached to C17 (see Figure 1). The ring system is asymmetric about the ring plane and has a flat side with no substituents (R-face) as well as a more rough side with two methyl groups (β-face). Being a smooth and rigid molecule, Chol is known to increase the order of saturated lipid acyl chains (the ordering effect)8 and the membrane surface density (the condensing effect).9 These properties have been recently confirmed in various atomistic simulation studies, including many-component systems related to lipid rafts.10,11 A variety of cholesterol analogues, on the other hand, have been reported to have a much weaker effect on membrane ordering and condensation. For instance, Chol precursors from the biosynthetic pathway like lanosterol,12-15 desmosterol,16 and 7-dehydrocholesterol17 influence lipid bilayers less than Chol. This is surprising considering that desmosterol and 7-dehydrocholesterol differ from Chol by just one double bond, located in the tail and ring system, respectively. Modifications to the polar head group of the Chol molecule also decrease the sterol effects on the structural properties of lipid membranes. This has been shown for cholesterol esters such as cholesterol sulfate,18 cholesterol hemisuccinate,19 doxycholesterols,20 and epicholesterol.21,22 Epicholesterol is the epimeric form of cholesterol, where the hydroxyl group is in the R conformation.21,22 The effects of structural modifications in sterols on the dynamic properties of membranes are poorly understood. Of
10.1021/jp075078h CCC: $40.75 © 2008 American Chemical Society Published on Web 01/25/2008
Cholesterol with Ketone as the Head Group
J. Phys. Chem. B, Vol. 112, No. 7, 2008 1947
Figure 2. The surface area per DPPC in the DPPC-Keto bilayer as a function of time.
time scales accessible to atomistic molecular dynamics simulations. Analyses of many structural properties largely clarify this behavior and point to the interplay of interaction mechanisms in the head group as well membrane core. The results provide insight into the exceptional properties of Chol as well as the conditions leading to promotion of flip-flops. Methods
Figure 1. Molecular structures of (a) DPPC, (b) cholesterol, and (c) ketosterone molecules with numbering of atoms and torsion angles. The chemical symbol for carbon atoms, C. In ketosterone, the 3-OH hydroxyl group is replaced with the 3-ketone group.
particular interest in this work is transition of sterols from one leaflet to another, a process known as flip-flop. Experiments have shown that the flip-flop of Chol is a rather slow process characterized by a lifetime on the order of minutes.23 For comparison, flip-flop time scales of phospholipids are considerably longer as their trans-bilayer movement takes place over hours instead of minutes.24 While the understanding of flipflops of sterols other than Chol is more limited, quasi-elastic neutron scattering measurements indirectly support the view that there are significant differences between the sterols: it has been observed that protrusions of Chol molecules are much more substantial compared with those of sterols such as lanosterol and ergosterol.25 The polar head groups of these sterols are identical, thus the differences in their protrusion behavior is likely to result from interactions and entropic contributions in the membrane core. Meanwhile, the effect of the polar head group on protrusions and flip-flops has not, to our knowledge, been studied yet. In this article, we explore what happens if the hydroxyl group of Chol is replaced with a ketone group, resulting in a molecule known as ketosterone (Keto). The steroid and tail structures of the molecules are not changed. The 3-ketone group is typical for steroid hormones, such as testosterone and aldosterone, but is not present in sterols that are constituents of cell membranes such as cholesterol or ergosterol. The 3-ketone group can also be present during metabolic processes for sterols in the membrane but is then reduced to the hydroxyl group.26,27 The implications of this modification are found to be significant. The transformation of Chol to Keto is observed to weaken the sterol effect on the bilayer propertiessboth the membrane order and condensation are affected less by Keto than Chol. This is essentially due to the fact that the tilt angle adopted by Keto is substantially larger than that of Chol,28 which results in considerable fluctuations in the Keto distribution in the membrane normal direction and a quite weak ordering of lipids around Keto. At the same time, we find intriguingly that Keto molecules flip-flop considerably more often than Chol. This sort of motion has not been previously observed for sterols in the
System Description and Parameters. We have performed atomistic molecular dynamics (MD) simulations of three membrane systems. The first bilayer was composed of 128 dipalmitoylphosphatidylcholine (DPPC) molecules, the second one of 128 DPPC and 32 Chol molecules (DPPC-Chol), and the third one of 128 DPPC molecules with 32 ketosterone molecules (DPPC-Keto). The only difference between Chol and Keto molecules lies in the head group since the hydroxyl group in Chol has been replaced with the ketone group in Keto (cf. Figure 1). All three bilayers were hydrated with 3500 water molecules. Further details concerning the construction, equilibration, and MD simulations have been given in ref 29. The initial structure of the DPPC-Keto bilayer was obtained by replacing the Chol 3β-hydroxyl group with a 3-ketone group in a DPPC-Chol bilayer, which has previously been equilibrated in a 50 ns MD simulation.29 Figure 1 shows the structure and numbering of atoms and torsion angles in DPPC and sterol molecules. The simulations were performed using the GROMACS software package.30 As for the force fields, we used the well-established and tested standard united atom force field parameters for DPPC molecules,31 where the partial charges were taken from the underlying model description.32 For water, we employed the simple point-charge (SPC) model.33 For the sterol force field, we used the description of Holtje et al.34 as described in ref 29. For the ketone group, we used the standard GROMACS parameters.35 Periodic boundary conditions with the usual minimum image convention were used in all three directions. The LINCS algorithm36 was used to preserve the bond length in the sterol hydroxyl group, and the SETTLE algorithm was used for water.37 The time step was set to 2 fs, and the simulations were carried out at constant pressure (1 atm) and temperature (323 K), which is above the main phase transition temperature of DPPC.38 The temperature and pressure were controlled using the Berendsen method39 with relaxation times set to 0.6 and 1.0 ps, respectively. The temperatures of the solute and solvent were controlled independently. The pressure was controlled semi-isotropically. The Lennard-Jones interactions were cut off at 1.0 nm. For the electrostatic interactions, we employed the particle-mesh Ewald method40 with a real space cutoff of 1.0 nm, β-spline interpolation (of order 5), and direct sum tolerance of 10-6. It has been shown that proper treatment of electrostatics is crucial in molecular simulations of biological systems.41,42
1948 J. Phys. Chem. B, Vol. 112, No. 7, 2008
Ro´g et al.
Figure 4. Trajectories (along bilayer normal) of the ketone oxygen atoms of the six Keto molecules which successfully performed a full flip-flop between the two leaflets and of molecules that started to enter the membrane core but returned to the same leaflet. Other molecules did not express similar behavior.
Figure 3. Profiles of the atom density of sterol oxygen atoms in the core of DPPC-Chol (gray line) and DPPC-Keto (dashed line) bilayers (a); profiles of the atom density of Op (thin line), and sterol oxygen atoms (thick line) in DPPC-Chol (gray line) and DPPC-Keto (dashed line) membranes in upper (b) and lower (c) leaflets. The coordinate z ) 0 corresponds to the membrane center.
Analysis and Equilibration. In the discussion below, we consider the various quantities determined from the simulation data. Surface area/DPPC was calculated by dividing the total area of the membrane by the number of DPPC molecules in a single leaflet, that is, 64. The membrane thickness was determined from mass density profiles by considering the points where the mass densities of lipids and water are equal.43 The order parameter for the nth segment of an acyl chain, Smol, was calculated using
1 Smol ) 〈3 cos2 θn - 1〉 2 where θn is the instantaneous angle between the nth segmental vector, that is, (Cn-1, Cn+1) vector linking n - 1 and n + 1 carbon atoms in the acyl chain, and the bilayer normal; < > denotes the time average (and by ergodicity, the ensemble average). Smol provides essentially the same information as the commonly studied NMR order parameter SCD.44,45 For DPPC saturated chains, Smol ) 2|-SCD|. To characterize the orientation of sterols in a bilayer, we calculated the tilt of a sterol defined as the angle between the C3-C15 vector (cf. Figure 1 b) and the bilayer normal. To calculate the tilt angles for the acyl chains of DPPC, we averaged over segmental vectors g4 (where the nth segmental vector links carbon atoms n˜ 1 and n + 1 in the acyl chain) to obtain the average segmental vector. The tilt angle for a given acyl chain is then given by , where θ is the angle between the bilayer normal and the average segmental vector.46 In averaging conformational quantities in terms of gauche and trans states, only torsion angles 4-16 (see Figure 1) were taken into account.44 To analyze hydrogen bonding, water bridging, and charge pairing, we employed the same geometrical definitions as in our previous publications.47 A hydrogen bond between an OH group and an acceptor oxygen atom is defined
to exist when the O‚‚‚O distance (r) is e3.25 Å and the angle θ between the O‚‚‚O vector and the OH group (the O‚‚‚O-H angle) is e35°. The distance of 3.25 Å is the position of the first minimum in the radial distribution function (RDF) of the water oxygen atoms (Ow) relative to an oxygen atom of PC. A water bridge is defined to exist when a water molecule is simultaneously hydrogen bonded to two lipid oxygen atoms. Charge pairing, which essentially describes the electrostatic interaction between a positively charged molecular moiety (such as a methyl group in PC choline) and a negatively charged one (such as an oxygen atom in Keto or Chol OH group), complements our studies for atom-scale interaction mechanisms and is useful in describing interactions in the head group region. A charge pair is formed between a positively charged choline methyl group (N-CH3) and a negatively charged non-ester phosphate (Op) or a carbonyl oxygen (Oc) atom when they are located within 4.0 Å from each other. Standard errors given for all numerical values in the below text were estimated using block analysis described in ref 48. The simulation times were 100 ns for DPPC and DPPC-Chol systems and 200 ns for DPPC-Keto. To confirm that the systems equilibrated as expected, we monitored the time development of the potential energy and temperature of the system and the surface area/DPPC molecule. All quantities were found to stabilize to their equilibrium values within less than 10 ns. Figure 2 demonstrates this for the area per DPPC in the DPPC-Keto system. Consequently, the first 20 ns were considered as equilibration29 and only the remaining 80 and 180 ns of the trajectory were analyzed. In our previous paper,28 selected data (area per lipid, sterol tilt) were obtained from a 100 ns trajectory of a DPPC-Keto simulation. Current analysis was performed over extended trajectories and gave the same values for area per lipid and sterol tilt, proving the overall stability of these simulations. Results Flip-Flops of Sterol Molecules. Let us first concentrate on the most intriguing part of the present work, that is, the transbilayer motion of sterols. In later sections, we characterize the interactions that give rise to the flip-flop behavior observed here. First, to examine the vertical position of the sterols, we calculate the density profiles for oxygen atoms of the sterol and the PC phosphate group, Op, along the bilayer normal as shown in Figure 3. As in our previous papers, the profiles (Figure 3b,c) have been adjusted for membrane thickness.44 These profiles suggest that, relative to the phosphate group, the ketone group
Cholesterol with Ketone as the Head Group
J. Phys. Chem. B, Vol. 112, No. 7, 2008 1949
Figure 5. Snapshots of a Keto molecule during a flip-flop processes.
is located deeper than the hydroxyl group of Chol. A closer look at the density profiles for the sterol oxygen atoms (Figure 3a) shows that the density of the ketone group does not drop to zero in the core of the bilayer while the density of the hydroxyl group does. To better demonstrate that, we plotted the trajectories along the bilayer normal of oxygen atoms of six selected Keto molecules over a period of 200 ns (Figure 4). Two of the Keto molecules performed flip-flop motion, and four others entered the bilayer interior but returned back to the same interface region. All together, we observed two successful flipflop events and seven unsuccessful ones. For comparison, similar behavior was not observed for Chol. Furthermore, when we consider the many other sterols we have recently studied16,17,21,28,29,49 in membranes comprising DPPCs or DOPCs, we have never found flip-flops to occur (this represents a combined simulation time of 1500 ns performed under the same
conditions, and an additional 200 ns performed with different a force field).44 The time scales of the observed Keto flip-flops are rather short. The first event takes place over a period of about 40 ns, with the molecule first adopting a highly tilted orientation for an extended period, then turning to lie with its polar head group in the center of the membrane and finally pushing up into the other leaflet. The second flip-flop is faster (about 5 ns), but the steps of the process are largely identical with the first one. Figure 5 demonstrates a flip-flop event in detail by showing snapshots of a Keto molecule during a transition process. In qualitative terms, the dynamics of the systems simulated in this work show very convincingly the unstable positioning of Keto in the bilayer in comparison to Chol. The limited number of events observed implies, however, that quantitative analysis of the flip-flop motion of Keto molecules is challenging.
1950 J. Phys. Chem. B, Vol. 112, No. 7, 2008
Ro´g et al.
Figure 6. Partial density profiles along the bilayer normal for all bilayer atoms in DPPC (black line), DPPC-Chol (gray line), and DPPC-Keto (dashed line) bilayers.
Considering only the two complete flip-flop processes, one is tempted to propose a flip-flop rate on the order of nanoseconds per molecule. The several incomplete attempts are in favor of the idea that the time scale is indeed short compared to the usually assumed sterol flip-flop time of minutes or even hours. Nonetheless, for obvious reasons, the times proposed here should be considered as suggestive rather than quantitative. Sterol Effect on Bilayer Properties. Having found the rapid nature of flip-flops of Keto molecules, let us next characterize the two Keto and Chol containing membrane systems in more detail. Perhaps the most relevant quantity describing membrane systems is the area per molecule. The surface area per lipid is easy to calculate in a single-component bilayer by dividing the total area of the bilayer by the number of lipids in a single leaflet. For binary mixtures and many-component systems, the method for this calculation is no longer obvious as has been discussed in recent works.29,49 For this reason, in this work, we prefer to avoid this subtle issue by considering the total area divided by the number of DPPC molecules only; see Table 1. For our purposes, this is completely reasonable since our main objective is to compare the influence of Keto and Chol on the membrane system. The surface areas given in Table 1 show that the presence of either sterol leads to membrane condensation. Table 1 also shows that the decrease in the surface area is associated with an increase in the membrane thickness. The mass density profiles of lipids in each system, from which the membrane thickness can be found, are shown in Figure 6. From this figure and the data in Table 1, it is evident that the effect of Chol is stronger than that of the ketone analogue. Exchanging the Chol OH group with a ketone group leads to changes in the ordering of the DPPC acyl chains. This is illustrated by the molecular order parameter, Smol, whose profiles along the sn-1 and sn-2 chains of DPPC are shown in Figure 7. Mean values of Smol, averaged over segments 4-16, for the sn-1 and sn-2 chains are given in Table 1. Figure 7 and Table 1 clearly highlight the stronger ordering effect of Chol, the difference from Keto being about 10%. Nevertheless, both sterols increase Smol of the DPPC acyl chains at all depths in the membrane. Other parameters describing the order of the acyl chains, tilt angle of the chains, and average numbers of gauche states per sn-1 and sn-2 chains are given in Table 1. In agreement with our previous work,44 a lower value of Smol is associated with a larger tilt angle and an increasing number of gauche states per chain. Also the average lifetime of the trans conformation (given in Table 1) follows the trends observed in our previous studies; lower values of the Smol parameter are associated with shorter lifetime. Location and Orientation of Sterols in the Bilayer. As has been shown previously, modifications to the Chol structure can change the sterol orientation measured by the sterol ring tilt
Figure 7. Profiles of the molecular order parameter (Smol) calculated for (a) the DPPC sn-2 and (b) sn-1 chain in DPPC (black line), DPPCChol (gray line), and DPPC-Keto (dashed line) bilayers. Small segment numbers correspond to carbons close to the glycerol group.
TABLE 1: Ordering and Condensing Effects of Sterolsa membrane Smol tilt (°) no. gauche lifetime (ps) area/DPPC (nm2) thickness (nm)
DPPC sn-2 sn-1 sn-2 sn-1 sterol sn-2 sn-1 sn-2 sn-1
DPPC-Chol
DPPC-Keto
0.28 ( 0.01 0.55 ( 0.01 0.51 ( 0.01 0.29 ( 0.01 0.58 ( 0.01 0.51 ( 0.01 23.8 ( 0.2 15.7 ( 0.2 17.3 ( 0.1 23.6 ( 0.2 16.0 ( 0.2 17.5 ( 0.1 19.8 ( 0.2 28.1 ( 0.2 3.0 ( 0.05 2.3 ( 0.05 2.4 ( 0.03 3.0 ( 0.05 2.3 ( 0.05 2.4 ( 0.03 84 ( 4 115 ( 4 107 ( 3 87 ( 4 119 ( 4 110 ( 3 0.660 ( 0.04 0.600 ( 0.04 0.630 ( 0.03 3.92 ( 0.06 4.69 ( 0.06 4.44 ( 0.04
a Average values of the molecular order parameter, S mol, chain tilt angle, number of gauche per acyl chain, and lifetimes of trans conformations. All results are given separately for the sn-1 and sn-2 chains of DPPC. Also given here are the average surface area per DPPC and the membrane thickness of DPPC, DPPC-Chol, and DPPC-Keto bilayers.
angle28 and its vertical position, where the latter is described by the position of the sterol oxygen relative to PC non-esterified phosphate oxygen atoms.21 Both these values are related to the magnitude of the sterol ordering effect. Keto was found to have an average tilt angle of 28.1°, which is considerably larger than the 19.8° tilt of Chol. The larger tilt of Keto is also intuitively in agreement with the expected correlation between the sterol tilt and the order of neighboring acyl chains.28 Membrane/Water Interface. To investigate further the effects of the ketone group, we next focused on the bilayer/ water interface and analyzed the atomic-level interactions of the ketone group with PC head groups and water molecules. In particular, we considered formation of hydrogen bonds (Hbonds), water bridges, and charge pairs47,50 and compared these results to those found for the Chol OH group. A closely related study focusing on the interface properties of phosphatidylcholine, phosphatidylglycerol, and phosphatidylethanolamine containing membranes can be found in ref 50. The hydroxyl group can act both as a hydrogen donor (Hbonds with water and PC) and as a hydrogen acceptor (H-bonds with water), while the ketone group can only be a hydrogen acceptor. Thus, for the ketone group, only H-bonds with water are possible. The average numbers of H-bonds formed by the two sterols are given in Table 2. The number of H-bonds between Keto and water is much smaller than between Chol
Cholesterol with Ketone as the Head Group
J. Phys. Chem. B, Vol. 112, No. 7, 2008 1951
TABLE 2: Interactions in the Membrane/Water Interfacea sterol-water H-bonds sterol-PC H-bonds sterol-PC water bridges sterol-PC charge pairs
DPPC-Chol
DPPC-Keto
0.38 0.82 0.32 1.15
0.09 0.06 0.84
a Sterol-water and sterol-DPPC H-bonds, water bridges and charge pairs in DPPC, DPPC-Chol, and DPPC-Keto bilayers. Errors are smaller than 0.02.
and water. However, if we consider only bonds where Chol acts as a hydrogen acceptor (25% of H-bonds), the numbers are similar. The small number of H-bonds formed by the ketone group with water suggests that the number of water bridges between Keto and PC (Table 2) will also be small. The negatively charged oxygen atom of the sterol can interact with the positively charged methyl group of the PC choline moiety (N-CH3) to form charge pairs. In DPPC-Chol and DPPC-Keto bilayers, the average number of O-N-CH3 charge pairs per sterol molecule is 1.15 and 0.84, respectively (Table 2). Summarizing, the interactions between the ketone groups and PC or water are significantly weaker than those between Chol and PC or water. This is demonstrated by the lack of direct H-bonds, small number of water bridges, and a reduction in the number of charge pairs by 27%. Discussion At the atomic level, the most striking difference between membrane cholesterol and ketosterone is the distinction in the pattern of sterol-PC interactions at the membrane-water interface. This relates to the interactions of the sterol head group with water and others lipids. Due to the lack of a hydrogen donor, the ketone group does not form H-bonds with lipids and is only a H-bond acceptor in interactions with water. Consequently, its hydration is 4 times lower than that of the hydroxyl group. The number of charge pairs between the ketone and choline groups is 27% smaller than in the case of the hydroxyl group. The present simulations indicate that these reduced interactions at the membrane-water interface result in the lack of the preferential location of Keto in the bilayer and give rise to a slow but continuous change in the sterol position and orientation in the bilayer. The sterol orientation, in particular, provides a good indicator of the sterol’s ability to order acyl chains in its vicinity: recent studies have shown that the tilt angle adopted on average by sterol molecules with respect to the membrane normal is indicative of the degree of its ordering and condensing effects in the membrane.28 On the basis of an extensive systematic comparison between five different sterols, the tilt angle of Chol has been found to be the smallest, and the ordering effect in turn the strongest.28 In this work, we have found that the tilt angle adopted by Keto is substantially larger than that adopted by Chol, thus the ordering and condensing effects are weaker for Keto. Previous experimental and theoretical studies have shown that modifications to the Chol polar head can affect the Chol location in the interfacial region. Most of these studies demonstrated that larger, more polar head groups tend to reside more deeply in the water phase.18,20 This location decreases the interactions between the sterol ring system and the hydrocarbon chains, thus decreasing the sterol’s ability to increase membrane order and promote condensation. Similar behavior was shown for epicholesterol, the epimeric form of cholesterol with the hydroxyl group in the R conformation.21 For the ketone group, we observe the opposite effect: the group participates in fewer interactions
at the water-membrane interface than the hydroxyl group, and as a result, it is located, on average, deeper in the bilayer. The deeper location of the hydroxyl group has been postulated to facilitate ordering and condensing effects in the case of sphingolipids and PC bilayers.51,52 However, in the case of ketosterone, not the location but the tilt seems to be the dominant factor. A surprising effect of the substitution of the hydroxyl with a ketone group is the promotion of the translocation of ketosterone molecules between the leaflets. These flip-flops were observed twice during the course of the simulation, and additionally, a considerable number of related attempts were obvious. The relevance of these findings is underlined by the fact that such flip-flop motions have never before been observed in MD simulations of bilayers containing Chol or other sterols. Due to the limited number of these events, only brief descriptions of the process can be given here, though. The rate of Keto flipflops is likely to be on a scale of microseconds, though this estimate is inevitably only suggestive. The time scale of complete flip-flop events is from nanoseconds to tens of nanoseconds. For incomplete processes, the time scale is obviously shorter; we have observed comparatively more events in which the molecule enters the membrane core and then returns back to the same interfacial region. The above findings are for sterol molecules in a saturated lipid bilayer; the time scales of similar processes in more disordered polyunsaturated membranes are likely to be considerably shorter. A recent experimental study of Chol in membranes with varying polyunsaturation is in favor of this view.53 Let us summarize our observations and finding in the following way: unlike Chol, Keto lacks the ability to function as a hydrogen donor. Thus, Keto is less hydrophilic and resides deeper inside the membrane and, in comparison to Chol, becomes more detached from the water-bilayer interface. On the basis of our simulations, that seems to be the dominant mechanism in keto flip-flop. The situation with Chol may be more complicated, but as we have not observed Chol flip-flops, we cannot verify the predominant mechanism(s). First, detachment from the interface is one possibility. Due to hydrogen bonding, however, it occurs in much longer time scales as fluctuations are not easily able to move Chol deeper inside the membrane core and initiate the flip-flop as the free energy cost is very high. Another possibility is a flip-flop mediated by defects:54 the membrane surface may become locally perturbed by water molecules toward the membrane core. When these defects become large enough, they may facilitate cholesterol flip-flop similarly to the translocation of zwitterionic phospholipids. Free energy calculations are necessary to explore the mechanisms of Chol flip-flop. That is a topic of a separate study. All the above data show that the ketone group does not effectively anchor the steroid molecules at the membrane-water interface and show that this group is not appropriate for sterol constituents in the cell membrane. At the same time, the ketone group can facilitate transmembrane mobility of steroid hormones, preventing them from a prolonged residence time in one of the membrane leaflets. Acknowledgment. This work was supported by the Academy of Finland, the Emil Aaltonen Foundation, and the Natural Sciences and Engineering Research Council of Canada (NSERC). Computational resources provided by the Finnish IT Center for Science (CSC), the HorseShoe supercluster computing facility at the University of Southern Denmark, and the SharcNet grid computing facility (www.sharcnet.ca) are gratefully acknowl-
1952 J. Phys. Chem. B, Vol. 112, No. 7, 2008 edged. T.R. holds a Marie Curie Intra-European Fellowship (024612-Glychol). References and Notes (1) Sackmann, E. Biological Membranes Architecture and Function. In Structure and Dynamics of Membranes; Lipowsky, R., Sackmann, E., Eds.; Elsevier: Amsterdam, 1995; pp 1-64. (2) Li, L. K.; So, L.; Spector, A. J. Lipid Res. 1985, 26, 600. (3) Mouritsen, O. G.; Jørgensen, K. Chem. Phys. Lipids 1994, 73, 3. (4) Simons, K.; Ikonen, E. Nature 1997, 387, 569. (5) Subczynski, W. K.; Wisniewska, A.; Yin, J.-J.; Hyde, J. S.; Kusumi, A. Biochemistry 1994, 33, 7670. (6) Bloom, M.; Mouritsen, O. G. The Evolution of Membrane. In Structure and Dynamics of Membranes; Lipowsky, R., Sackmann, E., Eds.; Elsevier: Amsterdam, 1995; pp 65-95. (7) Ohvo-Rekila¨, H.; Ramstedt, B.; Leppima¨ki, P.; Slotte, J. P. Prog. Lipid Res. 2002, 41, 66. (8) Oldfield, E.; Meadows, M.; Rice, D.; Jacobs, R. Biochemistry 1978, 17, 2727. (9) Smaby, J. M.; Momsen, M. M.; Brockman, H. L.; Brown, R. E. Biophys. J. 1997, 73, 1492. (10) Niemela¨ P. S.; Ollila, S.; Hyvo¨nen, M. T.; Karttunen, M.; Vattulainen, I. PLoS Comput. Biol. 2007, 3, e34. (11) Aittoniemi, J.; Niemela¨, P. S.; Hyvo¨nen, M. T.; Karttunen, M.; Vattulainen, I. Biophys. J. 2007, 92, 1125. (12) Nielsen, M.; Thewalt, J.; Miao, L.; Ipsen, J. H.; Bloom, M.; Zuckermann, M. J.; Mouritsen, O. G. Europhys. Lett. 2000, 52, 368. (13) Miao, L.; Nielsen, M.; Thewalt, J.; Ipsen, J. H.; Bloom, M.; Zuckermann, M. J.; Mouritsen, O. G. Biophys. J. 2002, 82, 1429. (14) Urbina, J. A.; Moreno, B.; Arnold, W.; Taron, C. H.; Orlean, P.; Oldfield, E. Biophys. J. 1998, 75, 1372. (15) Yeagle, P. L.; Martin, R. B.; Lala, A. K.; Lin, H. K.; Bloch, K. Proc. Natl. Acad. Sci. U.S.A. 1977, 74, 4924. (16) Vainio, S.; Jansen, M.; Koivusalo, M.; Ro´g, T.; Karttunen, M.; Vattulainen, I.; Ikonen, E. J. Biol. Chem. 2006, 281, 1121. (17) Ollila, S.; Ro´g, T.; Karttunen, M.; Vattulainen, I. J. Struct. Biol. 2007, 159, 311. (18) Faure, C.; Tranchant, J. F.; Dufourc, E. J. Biophys. J. 1996, 70, 1380. (19) Massey, J. Biochim. Biophys. Acta 1998, 1415, 193. (20) Karolis, C.; Coster, H. G.; Chilcott, T. C.; Barrow, K. D. Biochim. Biophys. Acta 1998, 1368, 247. (21) Ro´g, T.; Pasenkiewicz-Gierula, M. Biophys. J. 2003, 84, 1818. (22) Dufourc, E. J.; Parish, E. J.; Chitrakorn, S;, Smith, I. C. P. Biochemistry 1984, 23, 6062. (23) Hamilton, J. A. Curr. Opin. Lipidol. 2003, 14, 263. (24) Pomorski, T.; Holthuis, J. C. M.; Herrmann, A.; van Meer. G. J. Cell Sci. 2004, 117, 805. (25) Endress, E.; Heller, H.; Casalta, H.; Brown, M. F.; Bayerl, T. M. Biochemistry 2002, 41, 13078.
Ro´g et al. (26) Breitling R.; Krazeisen, A.; Moller, G.; Adamski, J. Mol. Cell Endocrinol. 2001, 171, 199. (27) Gachotte, D.; Sen, S. E.; Eckstein, J.; Barbuch, R.; Krieger, M.; Ray, B. D.; Bard, M. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 12655. (28) Aittoniemi. J.; Ro´g, T.; Niemela¨, P.; Pasenkiewicz-Gierula, M.; Karttunen, M.; Vattulainen, I. J. Phys. Chem. B 2006, 110, 25562. (29) Ro´g, T.; Vattulainen, I.; Pasenkiewicz-Gierula, M.; Karttunen, M. Biophys. J. 2007, 92, 3346. (30) Lindahl, E.; Hess, B.; van der Spoel, D. J. Mol. Model. 2001, 7, 306. (31) Berger, O.; Edholm, O.; Jahnig, F. Biophys. J. 1997, 72, 2002. (32) Tieleman, D. P.; Berendsen, H. J. C. J. Chem. Phys. 1996, 105, 4871. (33) Berendsen, H. J. C.; Postma, J. P. M.; Van Gunsteren, W. F.; Hermans, J. Interaction Models for Water in Relation to Protein Hydration. In Intermolecular Forces; Pullman, B., Ed.; Reidel: Dordrecht, The Netherlands, 1981. (34) Holtje, M.; Forster, T.; Brandt, B.; Engels, T.; von Rybinski, W.; Holtje, H.-D. Biochim. Biophys. Acta 2001, 1511, 156. (35) Van Gunsteren, W. F.; Berendsen, H. J. C. Gromos-87 manual; Biomos BV Nijenborgh 4, 9747 AG Groningen, The Netherlands, 1987. (36) Hess, B.; Bekker, H.; Berendsen, H. J. C.; Fraaije, J. G. E. M. J. Comput. Chem. 1997, 18, 1463. (37) Miyamoto, S.; Kollman, P. A. J. Comput. Chem. 1992, 13, 952. (38) Vist, M. R.; Davis, J. H. Biochemistry 1990, 29, 451. (39) Berendsen, H. J. C.; Postma, J. P. M.; Van Gunsteren, W. F.; DiNola, A.; Haak, J. R. J. Chem. Phys. 1984, 81, 3684. (40) Essman, U.; Perera, L.; Berkowitz, M. L.; Darden, H. L. T.; Pedersen, L. G. J. Chem. Phys. 1995, 103, 8577. (41) Patra, M.; Karttunen, M.; Hyvo¨nen, M. T.; Falck, E.; Vattulainen, I. J. Phys. Chem. B 2004, 108, 4485. (42) Smith, P. E.; Pettitt, B. M. J. Chem. Phys. 1991, 95, 8430. (43) Patra, M.; Salonen, E.; Tera¨ma¨, E.; Vattulainen, I.; Faller, R.; Lee, B. W.; Holopainen, J.; Karttunen, M. Biophys. J. 2006, 90, 1121. (44) Ro´g, T.; Pasenkiewicz-Gierula, M. Biophys. J. 2001, 81, 2190. (45) Davis, J. H. Biochim. Biophys. Acta 1983, 737, 117. (46) Ro´g, T.; Murzyn, K.; Gurbiel, R.; Takaoka, Y.; Kusumi, A.; Pasenkiewicz-Gierula, M. J. Lipid Res. 2004, 45, 326. (47) Pasenkiewicz-Gierula, M.; Ro´g, T.; Kitamura, K.; Kusumi, A. Biophys. J. 2000, 78, 1376. (48) Hess, B. J. Chem. Phys. 2002, 116, 209. (49) Falck, E.; Patra, M.; Karttunen, M.; Hyvo¨nen, M. T.; Vattulainen, I. Biophys. J. 2004, 87, 1076. (50) Murzyn, K.; Zhao, W.; Karttunen, M.; Kurdziel, M.; Ro´g, T. BioInterphases 2006, 1, 98. (51) Ro´g, T.; Pasenkiewicz-Gierula, M. Biophys. J. 2006, 91, 3756. (52) Ro´g, T.; Pasenkiewicz-Gierula, M. Biochimie 2006, 88, 449. (53) Harroun, T. A.; Katsaras, J.; Wassall, S. R. Biochemistry 2006, 45, 1227. (54) Tieleman, D. P.; Marrink, S. J. J. Am. Chem. Soc. 2006, 128, 12462.
