Article pubs.acs.org/Langmuir

Molecular Interactions between Amantadine and Model Cell Membranes Fu-Gen Wu,*,†,‡ Pei Yang,‡ Chi Zhang,‡ Bolin Li,† Xiaofeng Han,†,‡ Minghu Song,§ and Zhan Chen*,‡ †

State Key Laboratory of Bioelectronics, School of Biological Science and Medical Engineering, Southeast University, Nanjing 210096, China ‡ Department of Chemistry, University of Michigan, 930 North University Avenue, Ann Arbor, Michigan 48109, United States § Compound Safety Prediction Group, Pfizer Inc., Groton, Connecticut 06340, United States S Supporting Information *

ABSTRACT: Sum frequency generation (SFG) vibrational spectroscopy was applied to study molecular interactions between amantadine and substrate supported lipid bilayers serving as model cell membranes. Both isotopically asymmetric and symmetric lipid bilayers were used in the research. SFG results elucidated how the watersoluble drug, amantadine, influenced the packing state of each leaflet of a lipid bilayer and how the drugs affected the lipid flip-flop process. It is difficult to achieve such detailed molecular-level information using other analytical techniques. Especially, from the flip-flop rate change of isotopically asymmetric lipid bilayer induced by amantadine, important information on the drug−membrane interaction mechanism can be derived. The results show that amantadine can be associated with zwitterionic PC bilayers but has a negligible influence on the flip-flop behavior of PC molecules unless at high concentrations. Different effects of amantadine on the lipid bilayer were observed for the negatively charged DPPG bilayer; low concentration amantadine (e.g., 0.20 mM) in the subphase could immediately disturb the outer lipid leaflet and then the lipid associated amantadine molecules gradually reorganize to cause the outer leaflet to return to the original orderly packed state. Higher concentration amantadine (e.g., 5.0 mM) immediately disordered the packing state of the outer lipid leaflet. For both the high and low concentration cases, amantadine molecules only bind to the outer PG leaflet and cannot translocate to the inner layer. The presence of amantadine within the negatively charged lipid layers has certain implications for using liposomes as drug delivery carriers for amantadine. Besides, by using PC or PG bilayers with both leaflets deuterated, we were able to examine how amantadine is distributed and/or oriented within the lipid bilayer. The present work demonstrates that SFG results can provide an in-depth understanding of the molecular mechanisms of interactions between water-soluble drugs and model cell membranes.

1. INTRODUCTION

the current and future pharmaceutical research and development. Amantadine (or 1-adamantylamine or 1-aminoadamantane) has been approved by US Food and Drug Administration for use as both an antiviral and an antiparkinsonian drug.4,5 It is a small amphiphilic (also lipophilic), water-soluble (in its protonated form) compound, consisting of an adamantane backbone that has an amino group substituted at one of the four methyne positions. Its molecular formula is shown in Figure 1. The interaction between amantadine and virus in the cell membrane has been investigated. For example, for influenza A virus, its M2 protein forms a transmembrane proton channel which is important for viral infection and replication. Amantadine can block this proton channel, inhibit the viral replication process, and thus can be used to treat influenza A infections.6,7 Amantadine was also reported to be able to induce conformational and dynamical changes of the influenza M2

When water-soluble drugs are injected into a human body or in vitro cultured cell media, a key step to ensure the drug efficacy is the passage of drugs through the plasma membrane, which is the vanguard of the cells to come into contact with the injected drugs. The ability of drugs to penetrate cell membranes to reach their targets will affect their pharmacokinetics and toxicology. Since many current water-soluble drugs are amphiphilic molecules, they can extensively interact with biological membranes composed of lipids with hydrophilic head groups and hydrophobic tails. The partitioning of drugs into membranes can also alter the structures and properties of cell membranes. In addition, many lipophilic drug molecules partitioning into membranes appear to have a preferred location and orientation that facilitates interactions with receptors through lateral diffusion.1,2 As the drug−membrane interaction is closely related to drug transport, distribution, accumulation, efficacy, and resistance,3 understanding various modes of the drug−membrane interaction is crucial to develop a better understanding of drug action, which is important for © 2014 American Chemical Society

Received: May 5, 2014 Revised: June 16, 2014 Published: June 30, 2014 8491

dx.doi.org/10.1021/la501718n | Langmuir 2014, 30, 8491−8499

Langmuir

Article

sensitive techniquesum frequency generation (SFG) vibrational spectroscopy. SFG was shown to be able to monitor the interaction dynamics between amantadine and lipid bilayers in situ and in real time. We constructed isotopically asymmetric lipid bilayers using hydrogenated/deuterated dipalmitoylphosphatidylcholine (DPPC) and hydrogenated/deuterated dipalmitoylphosphatidylglycerol (DPPG) leaflets. Such isotopically asymmetric lipid bilayers enabled us to investigate the interaction mechanism between drugs and lipid membranes through evaluation of the change of SFG spectra and lipid flipflop kinetics. The results revealed that amantadine did not significantly affect the flip-flop rate of neutral DPPC bilayer at the amantadine concentration below 10 mM. For the negatively charged DPPG bilayer, amantadine was found to only affect the packing state of the outer leaflet. Besides, by using lipid bilayers with two deuterated leaflets (dDPPC/dDPPC or dDPPG/ dDPPG), we were able to observe how amantadine was distributed and/or oriented inside the lipid bilayer.

Figure 1. SFG experimental sample geometry to study drug−model cell membrane interactions.

transmembrane proton channel.7 Although the precise mode of how amantadine interacts with the channel is unknown, it is likely that this lipophilic amantadine reaches the channel after first partitioning into the phospholipid membrane.8 Direct interactions between amantadine and membrane are also important in other cases. For example, for treating neurodegenerative disorders, the permeation of amantadine across the membrane becomes an important issue since amantadine has to pass through the blood-brain barrier to reach the targets in the brain.8 Besides, since amantadine has a high affinity for the cyclic compound curcubituril [7] (CB7) (where amantadine can be included in the cavity of CB7), amantadine was also used to induce drug release in real cells in CB7-containing nanoparticle-based drug delivery systems. For example, intracellular triggering of the therapeutic effect of the AuNP-NH2-CB7 complex can be achieved through the administration of amantadine, removing CB7 from the nanoparticle surface (because amtandine has a higher affinity with CB7 than that with the AuNP’s surface amine group), causing the endosomal release and concomitant in situ cytotoxicity of AuNP-NH2.9 This amantadine-induced drug release in real cells shows that amantadine can at least partially pass through real cell membranes. Amantadine can also alter the membrane properties: It was shown that association between amantadine and cellular membrane can inhibit membrane fusion.10 Apart from direct medicinal uses, it has been shown that the adamantyl group of amtandine is capable of inserting into a lipid bilayer, which allows us to modify cell surfaces with various biomolecules that contain an adamantyl group.11 Thus, amantadine-containing (or adamantyl group containing) molecules can be very useful for cell surface modification, which is very important for the membrane fusion of cells and for the targeting of drugs to cells. These above applications and observations all involve the interactions between the amantadine groups and cell membranes. However, detailed molecular-level understanding of how amantadine influences the properties of cell membrane or model cell membrane still remains to be uncovered, which should be developed with urgent need. The studies on amantadine−cell membrane interactions will also provide important insights on the interaction mechanism of similar structured water-soluble drugs with cell membranes. Several biophysical techniques, such as neutron and X-ray diffraction,12 solution13 and solid14 NMR, EPR,15 and molecular dynamics simulations,8,14 have been applied to study the association between amantadine and cellular membranes. However, the interaction dynamics between amantadine and phospholipid bilayer is still unclear. In this work, the molecular interaction between amantadine and model lipid membranes was investigated using a submonolayer surface

2. EXPERIMENTAL SECTION 2.1. Experimental Details. 1,2-Dipalmitoyl-sn-glycero-3-phosphocholine (DPPC), 1,2-dipalmitoyl(d62)-sn-glycero-3-phosphocholine (DPPC-d62 or dDPPC), 1,2-dipalmitoyl-sn-glycero-3-phospho-(1′rac-glycerol) (sodium salt) (DPPG), and 1,2-dipalmitoyl(d62)-snglycero-3-[phospho-rac-(1-glycerol)] (sodium salt) (DPPG-d62, or dDPPG) were purchased from Avanti Polar Lipids. The molecular structures of these lipids are shown in Supporting Information (Figure S1). Amantadine hydrochloride was purchased from Sigma and its molecular structure is shown in Figure 1. Right-angle CaF2 prisms purchased from Chengdu YaSi Opteoelectronics Co, Ltd (Chengdu, China) were cleaned before the deposition of lipids.16−18 The Langmuir-Blodgett and Langmuir-Schaefer (LB/ LS) methods were used to deposit the proximal and distal leaflets of a single lipid bilayer onto the prisms, respectively. The lipid bilayer preparation details were published before16−18 and can be found in the Supporting Information. All experiments were carried out at room temperature (22 °C), at which the studied lipid molecules were all in gel phase. SFG is a second-order nonlinear optical spectroscopic technique that has submonolayer surface sensitivity,19,20 which makes SFG an ideal technique to monitor the structure and dynamics of surface molecules in situ. The details regarding SFG theories and measurements have been extensively published,21−41 and the details about our SFG experimental procedures can be found in the Supporting Information. 2.2. Calculation of Flip-Flop Dynamics. The intensity of the SFG signal coming from the lipid terminal CH3 or CD3 symmetric stretching vibration is related to the net population difference for the hydrogenated or deuterated lipid species, which can be expressed as42−44

ICH3orCD3 ∝ (Ninner − Nouter)2

(1)

where Ninner and Nouter respresent the numbers of the same type of lipid molecules in the inner leaflet and outer leaflet, respectively. From eq 1, we can see that before flip-flop, the bilayer composed of an isotopically asymmetric bilayer has the largest net population difference of the hydrogenated or deuterated lipid species in the two leaflets, and thus the SFG signal should have the largest value. As flipflop proceeds with time, the net population difference in the two leaflets decreases, leading to the decrease in ICH3(or ICD3). The lipid flip-flop kinetics can be analyzed by fitting the SFG timedependent signal using the following single exponential function:42,45−47

ICH3orCD3(t ) = Imax e−4kt + Imin 8492

(2)

dx.doi.org/10.1021/la501718n | Langmuir 2014, 30, 8491−8499

Langmuir

Article

Figure 2. Time-dependent SFG signals at 2070 and 2875 cm−1 detected from the dDPPC/DPPC bilayer. (A) Experimental measured (dots) and fitting result (line, using eq 2) for the 2070 cm−1 signal. (B) Experimental measured 2875 cm−1 signal (blue dots). For comparison purpose, the experimental measured 2070 cm−1 signal was also shown in (B).

result obtained by using the 2875 cm−1 signal gives the same k and t1/2 values as those obtained by using the 2070 cm−1 signal. Figure 2B shows that the time-dependent SFG signal decreases for the two leaflets have the same rate, indicating that the changes of the two leaflets have the same dynamics, which is a strong support for the occurrence of the flip-flop process. Therefore, it is clear that flip-flop occurs for the dDPPC/DPPC bilayer even without the interaction with amantadine. We then monitored the time-dependent SFG signal decrease detected from the dDPPC/DPPC bilayer in contact with amantadine solutions with different concentrations. Such timedependent data were fitted using the same method and the flipflop rate (k) and half-life (t1/2) of the bilayer were obtained. Figure 3 shows that for the amantadine solution concentration

where Imax and Imin are the maximum and minimum resonant SFG signal intensities from the symmetric stretching vibrations of the terminal CH3 or CD3 groups of the lipid tails, respectively, and k is the flip-flop rate. Besides, the half-life (t1/2) of flip-flop can be calculated from the above determined k:

t1/2 =

ln 2 2k

(3)

3. RESULTS AND DISCUSSION As shown in Figure 1, we constructed a dDPPC/DPPC bilayer on a CaF2 prism surface (the inner layer adhering to the CaF2 prism surface is composed of lipid dDPPC, while the outer layer facing water or aqueous solution is composed of DPPC) on the SFG experimental stage. The SFG ssp spectrum was collected from the dDPPC/DPPC bilayer in contact with water, which is the same as that detected from a dDSPC/DSPC bilayer.16 However, since the flip-flop of DPPC and dDPPC molecules within the bilayer is very fast, it is very difficult to obtain a pure “inner dDPPC + outer DPPC” bilayer without flip-flop. Thus, we do not show the detected SFG spectrum here. Instead of using the entire SFG spectrum, here we monitored the SFG signal intensity contributed from the CD3 symmetric stretching mode at 2070 cm−1 in the dDPPC leaflet and the CH3 symmetric stretching mode at 2875 cm−1 in DPPC leaflet to probe the inner and outer leaflets of the lipid bilayer, respectively. The initial isotopically asymmetric bilayer formed immediately after the contact of the two lipid monolayers or leaflets had the largest net population difference of the hydrogenated (or deuterated) lipid species in the two leaflets. As time went by, the transbilayer movement (flip-flop) of the lipid molecules within the bilayer led to the decrease in the net population difference of the hydrogenated (or deuterated) lipids in the two leaflets (as shown in eq 1), resulting in the decrease in SFG signal intensity ICH3 (or ICD3). Figure 2A and B shows the changes of SFG signal intensities ICD3 and ICH3 (along with ICD3 for comparison) versus time for the dDPPC/DPPC bilayer, respectively. The time-dependent SFG signal at 2070 cm−1 (the CD3 symmetric stretching vibrational mode) was detected from the inner leaflet of the dDPPC/DPPC bilayer and we refer it to ICD3. From the fitting result of the 2070 cm−1 signal using eq 2, we can deduce the flip-flop rate (k) and the half-life of the neat lipid bilayer (t1/2) to be (1.4 ± 0.2) × 10−4 s−1 and (2.5 ± 0.4) × 103 s. The fitting

Figure 3. Lipid flip-flop rate (k) observed from dDPPC/DPPC bilayer in contact with subphase amantadine solutions with different concentrations (Camantadine).

between 0.10 and 10.0 mM, the flip-flop rate (k) and half-life (t1/2) of the bilayer exhibit no substantial change (less than 30% change). Only when the amantadine concentration reached 10.0 mM did the measured k and t1/2 values have significant changes (k = (3.5 ± 1.0) × 10−4 s−1, t1/2 = (9.9 ± 0.3) × 102 s). Therefore, this study shows that amantadine does not substantially interact with the dDPPC/DPPC bilayer below the subphase amantadine concentration of 10.0 mM. At a higher amantadine concentration of 10.0 mM, amantadine may form aggregates within the interfacial region of the outer lipid leaflet, leading to a stronger disturbation effect on the lipid 8493

dx.doi.org/10.1021/la501718n | Langmuir 2014, 30, 8491−8499

Langmuir

Article

packing state, and thus the flip-flop rate of the PC bilayer decreases. To study whether the amantadine drugs were oriented in an orderly formation in the lipid bilayer, and to confirm if amantadine was adsorbed onto the lipid surface, SFG experiments were performed by using the dDPPC/dDPPC lipid bilayer before and after addition of amantadine to the bilayer subphase (Figures 4 and 5). The CD stretching

The CH or CD stretching signals, especially those from the terminal CH3 or CD3 groups in the lipid tails, have been extensively used to monitor interactions between lipid bilayers and other molecules.48−52 The OH stretching signals of the interfacial water molecules can also be used to study such interactions. In the lipid−water interfacial region, the water molecules can be aligned by the static electric field of the lipid bilayer or other molecules, resulting in well orientated water molecules which break the inversion symmetry of the water bulk. Through probing the OH stretching vibration modes of water, SFG spectroscopy can monitor the local water environment.53 The adsorption of charged molecules at a neutral lipid interface would generate an electric field, which will promote the ordered alignment of the interfacial water molecules.23,24,34,54−56 Here by investigating the interfacial water structure using SFG, we were able to probe the amantadine−membrane interactions. In the OH stretching frequency region, we can see that before the addition of amantadine to the subphase, the presence of a small peak (signal intensity ∼8) at 3180 cm−1 detected from the zwitterionic dDPPC/dDPPC bilayer shows that the interfacial water molecules were only slightly ordered at this bilayer interface. The introduction of amantadine to the subphase to reach the concentrations of 1.0 and 5.0 mM increased the SFG water signal to ∼15 and ∼26, respectively, indicating that the preferential alignment of water molecules at the lipid−subphase interface was enhanced by the electric field generated by the adsorbed charged drug molecules on the lipid surface. Therefore, for the PC bilayers, amantadine can be associated with the lipid bilayer at the subphase concentration of 1.0 mM and 5.0 mM, even though they did not alter the dDPPC/DPPC bilayer flip-flop rate as illustrated in Figure 3. The unchanged flip-flop rate of dDPPC/DPPC bilayer after the addition of 0− 5.0 mM amantadine can be explained as follows: the partition of drug molecules into the lipid bilayer did not disturb the original packing state of the lipid molecules because of the small size of amantadine, which makes it fitted well at the lipid interfacial region. However, when the drug concentration further increases to 10 mM, aggregates may form within the lipid interfacial region, leading to a loosened packing of the lipid molecules, and this will in turn cause a decrease in the lipid flip-flop rate of the lipid bilayer. To study how fast amantadine can be associated with the dDPPC/dDPPC bilayer, the time-dependent SFG signal changes after the addition of amantadine to reach 1.0 and 5.0

Figure 4. SFG spectra collected from the dDPPC/dDPPC bilayer in the frequency regions of 2750−3850 cm−1 before and after addition of amantadine to the subphase.

frequency region (1950−2300 cm−1) did not show evident peaks before (Camantadine = 0) and after addition of amantadine to reach the subphase concentration of 1.0 and 5.0 mM (Camantadine = 1.0 and 5.0 mM; data not shown). Very weak signals were detected in the CH stretching frequency region (2800−3100 cm−1) from the dDPPC/dDPPC bilayer in contact with amantadine solutions with different concentrations (Figure 4). Such signals are similar before and after the addition of amantadine to the subphase. These observations suggested that the addition of amantadine to the subphase did not cause the isotopical asymmetry between the two lipid leaflets. No CD signals indicated that either amantadine does not disrupt the bilayer or it disrupts the bilayer to the same degree. No detection of CH signal shows that either amantadine is not associated with the bilayer or it is associated with the bilayer but does not exhibit any order.

Figure 5. Time-dependent SFG signals collected at 2070 and 3200 cm−1 from the dDPPC/dDPPC bilayer starting immediately after the formation of the lipid bilayer (bilayer forms at t = 0 s). In (A), 1.0 mM amantadine solution in subphase (amantadine was added at t = 260 s), while in (B), 5.0 mM amantadine solution in the subphase (amantadine was added at t = 290 s). 8494

dx.doi.org/10.1021/la501718n | Langmuir 2014, 30, 8491−8499

Langmuir

Article

Figure 6. SFG spectra collected from the dDPPG/dDPPG bilayer after addition of 1.0 mM and 5.0 mM amantadine to the subphase. (A) 1950− 2300 cm−1, (B) 2750−3900 cm−1.

Figure 7. Time-dependent SFG signals collected at 2065 and 2875 cm−1 from the dDPPG/DPPG bilayer. In (A), 0.20 mM amantadine solution in the subphase (amantadine was added at t = 124 s), while in (B), 5.0 mM amantadine solution in the subphase (amantadine was added at t = 165 s).

amantadine was added to the subphase solution to reach 1.0 mM and 5.0 mM in contact with the dDPPG/dDPPG lipid bilayer. Upon the addition of drug molecules to the subphase, the SFG peak at 2065 cm−1 significantly increased in intensity as the concentration of the drug increased (Figures 6A). We believe that the increase in the intensity of the CD stretching signal suggests that some of the dDPPG molecules in the outer leaflet were disrupted or even displaced by the drug molecules, breaking the inversion symmetry of the two lipid layers. That is to say, upon the addition of amantadine molecules to the subphase, the inner leaflet of the bilayer was still composed of ordered dDPPG molecules while the outer leaflet might contain some less ordered dDPPG or might even contain a mixture of dDPPG and amantadine drug molecules. This conclusion will be further confirmed by the following results obtained from the dDPPG/DPPG bilayer interacting with amantadine in Figure 7. The two peaks at 2855 and 2930 cm−1 in Figure 6B are contributed by the CH stretching vibrations of the drug molecules associated with the lipid bilayer. For the isotopically symmetric dDPPG/dDPPG bilayer, no discernible CH stretching signals can be observed in this frequency region.48 Even for an isotopically asymmetric bilayer (such as dDPPG/ DPPG), the CH stretching signal peak centers reside at different positions (2875 and 2945 cm−1).16,48 Besides, the two peaks at 2855 and 2930 cm−1 become much more intense when the bilayer is in contact with 5.0 mM amantadine solution compared to that with the 1.0 mM amantadine solution, which further suggests that the signals are from the amantadine drug molecules. Thus, we believe that the observation of such signals is because amantadine molecules were asymmetrically dis-

mM subphase concentrations were monitored and shown in Figure 5. Addition of amantadine to the subphase (at t = 260 s) to reach 1.0 mM could gradually increase the SFG water signal at 3200 cm−1 from ∼9 to ∼13 within ∼900 s, while the signal of CD stretching vibration (2070 cm−1) still remained unchanged (Figure 5A). Differently, addition of amantadine to reach a higher concentration (5.0 mM) could immediately increase the SFG water signal from ∼8 to ∼30 within 10 s, and the 2070 cm−1 peak also remained unchanged (Figure 5B). The sudden jump of the SFG signal upon addition of 5 mM amantadine was due to the very fast drug adsorption process. The faster response of SFG water signal upon the addition of amantadine for the 5.0 mM case as compared to the 1.0 mM case suggests that the adsorption of drugs onto the bilayer is much faster at a higher drug concentration in the subphase. We also want to study molecular interactions between amantadine and negatively charged lipid bilayers, using DPPG as a model. SFG spectra collected from the dDPPG/dDPPG bilayer in the CH and CD stretching frequency regions have been published and show no discernible signal, indicating that the bilayer has a good order and possesses an inversion symmetry.48 For SFG spectra collected from the dDPPG/ DPPG bilayer, two strong peaks at 2065 and 2875 cm−1 could be observed (data not shown), and these two peak intensities did not change much with time within 5 h (Supporting Information Figure S2). This indicates that the dDPPG/DPPG bilayer is very stable and the flip-flop of the PG molecules within the lipid bilayers is almost negligible. To study whether the drug molecules are oriented in an orderly manner when interacting with the PG lipid bilayer, 8495

dx.doi.org/10.1021/la501718n | Langmuir 2014, 30, 8491−8499

Langmuir

Article

of the outer DPPG leaflet. The CH stretching time-dependent signal change result can be explained as follows: when amantadine molecules were added to the subphase, they could immediately insert into the lipid outer leaflet since these molecules are amphiphilic. Due to the electrostatic attractive interaction between the positively charged amine group of the drug and the negatively charged phosphate group of the lipid, amantadine tends to stay at the interfacial regions with its hydrophobic rings interacting with the carbonyl and the first several methylene groups of the lipid, which disturbs the tight, ordered packing of the lipid tails, resulting in the decrease of the outer leaflet CH SFG stretching signal. However, since the drug concentration in the subphase was low, only a limited number of drug molecules could be incorporated into the outer lipid leaflet. Amantadine is a small molecule which can interact with the PG molecules through hydrophobic and electrostatic interactions. These two interaction forces between amantadine and lipid molecules had not achieved balance, and as time went by, these drug molecules could reorganize to make the surrounding lipids more orderly-packed, reaching a final state that was not very different from the original packing state before the addition of the drug. Thus, the CH stretching signal gradually increased. If a higher concentration drug subphase solution (e.g., 5.0 mM) was in contact with the dDPPG/DPPG bilayer (Figure 7B), the association/partition of more drug molecules into the lipid outer membrane could immediately disturb the ordered packing of the lipid molecules. With so many drug molecules existing between the lipid molecules, they might form drug aggregates (Figure 8). Therefore, the lipid molecules were not

tributed in the lipid bilayer. Perhaps more oriented amantadine molecules were associated with the lipid outer leaflet. Moreover, the SFG OH stretching signal from H2O at around 3200 cm−1 became much weaker after the addition of amantadine to the subphase (data not presented). This is due to the charge neutralization of the positively charged amantadine and the negatively charged PO2− groups in the PG molecules, which makes the interfacial water more disordered. More detailed discussion on the SFG water OH stretching signal is presented below. From the dDPPG/dDPPG bilayer, two SFG peaks centered around 3200 and 3400 cm−1 could be observed before the introduction of amantadine to the subphase. These two peaks can be assigned to the OH stretching modes of hydrogen bonded interfacial water molecules.57,58 The difference of the SFG water spectra collected from the dDPPC/dDPPC and dDPPG/dDPPG bilayers reflects a different interfacial water environment: The ordered water molecules associated with the dDPPC/dDPPC bilayer reside in the region between cationic and anionic parts of the headgroup, while ordered water molecules in the dDPPG/dDPPG bilayer mainly orient at the charged lipid surface.27,56 After the addition of amantadine stock solution to the subphase of the dDPPG/dDPPG bilayer to reach amantadine concentration of 1.0 or 5.0 mM, the SFG water signal centered around 3200 and 3400 cm−1 significantly decreased. The electric charges of the negatively charged lipid surface were neutralized by the oppositely charged drug molecules and the orientation order of water at the lipid− amantadine solution interface therefore decreased substantially. Thus, from the changes in the SFG spectra collected in the OH stretching frequency region of the interfacial water molecules associated with the dDPPC/dDPPC and dDPPG/dDPPG bilayers, we demonstrated that amantadine molecules indeed can be associated with both kinds of lipid membranes. The time-dependent behaviors of the lipid signals upon the addition of drugs to the subphase were monitored by the CD stretching vibration of the inner dDPPG leaflet (2065 cm−1) and CH stretching vibration of the outer DPPG leaflet (2875 cm−1). In Figure 7A, we can see that before the contact of the two lipid leaflets, only weak CD stretching vibrational signal (∼18) contributed from the dDPPG monolayer in air was detected. When the dDPPG monolayer on the prism contacted the DPPG monolayer spread on water (at t = 50 s), a lipid bilayer formed, and the signal for both CD and CH significantly enhanced (CD signal changed from ∼18 to ∼115, CH signal changed from 0 to ∼110). At 124 s, the drug stock solution was injected into the subphase (while stirring the subphase using a magnetic stirrer). The final concentration of amantadine in the subphase reached 0.2 mM. The addition of amantadine at t = 124 s resulted in an immediate decrease of CH signal from ∼l10 to ∼25, while the CD signal remained at ∼106. The CD signal did not change with time, showing that the inner leaflet was not disrupted or displaced by the amantadine molecules. Therefore, we believe that amantadine molecules only interacted with the outer leaflet, as we discussed above about the dDPPG/dDPPG data. Different from the constant CD signal, the CH signal monitored at 2875 cm−1 gradually increased from ∼27 (at t = 135 s) to ∼65 (at t = 550 s), and it continued to increase to ∼96 at t = 7500 s. Even though the bilayer associated amantadine generates SFG CH stretching signals, but such signals are peaked at 2855 and 2930 cm−1. Here we observed the signal variation at 2875 cm−1, which was due to the change

Figure 8. Schematics showing the effect of amantadine on the packing state of the dDPPG/DPPG bilayer.

able to reorganize to achieve the original orderly-packed state, and thus the SFG CH stretching signal remained constant after the addition of the drug molecules. Different from the CH stretching signal, SFG CD stretching signal did not change, showing that amantadine did not interact with the inner leaflet even at this higher drug concentration. The present data clearly show that amantadine can have very distinct effects on the two leaflets of the PG bilayers. By investigating the influences (such as change of the flipflop rate, the packing state of the lipid molecules within the bilayer) exerted by drugs on lipid bilayers, and the orientational changes of the drugs themselves, we can obtain information on the interaction mechanisms between lipids and amantadine drugs at a molecular level. Previous results have demonstrated that amantadine readily partitions into lipid bilayers,11 and both protonated and deprotonated amantadine molecules have a preference for the interfacial region of the lipid bilayer, with the 8496

dx.doi.org/10.1021/la501718n | Langmuir 2014, 30, 8491−8499

Langmuir

Article

protonated species proving significantly more favorable.8 These results are in good agreement of our present SFG data. Although it was shown that amantadine could be partitioned well into lipid bilayers, the present work demonstrated that amantadine molecules did not exert a significant effect on the physical properties of the zwitterionic PC bilayer at concentrations below 10.0 mM. However, a small concentration of amantadine (such as 0.20 mM) in the subphase solution can have a large influence on the negatively charged PG bilayers. Compared to the phenothiazine-type antipsychotic drug chlorpromazine (CPZ) which we studied previously,16 although both amantadine and CPZ are positively charged, amantadine’s hydrophobic ring is much smaller than CPZ’s. This enables amantadine to be well fitted within the interlipid regions of PGs, binding tightly to the surrounding negatively charged phosphate groups of lipid molecules. By contrast, CPZ molecules can immediately bind to and disrupt the outer lipid leaflet and then gradually reduce the ordering of the inner lipid leaflet.16 Besides, distinct effects of amantadine and CPZ have also been observed for PCs: CPZ was found to significantly affect the flip-flop process of PC bilayers, while amantadine did not. These results point out the importance of the molecular size of the drug molecule in drug−membrane interactions. The present SFG data also revealed that amantadine acts as a glue in the outer lipid leaflet and cannot translocate to the inner lipid leaflet, while the large CPZ’s hydrophobic ring can immediately destroy the outer leaflet and then quickly disorder the inner layer.16 Amantadine is an amphiphilic, water-soluble small molecule, and it has been reported that amantadine can efficiently permeate through natural cell membrane. The natural cell membrane has a very complicated composition, including various lipids (such as phospholipids, sphingolipids, sugar-lipids, and sterols) and membrane proteins. The interaction between amantadine and the cell membrane also involves both lipids and proteins. Here in this present work we showed that although amantadine has a small effect on zwitterionic PCs, it can be trapped or “frozen” within the interfacial regions of negatively charged lipids such as PGs. This may partially explain amantadine’s side effects (its toxicity effects on cells) and may shed light on a series of similarly structured drug molecules used for various clinical applications. The present work also points out the possibility of using SFG spectroscopy to probe the effects of membrane composition on drug−membrane interactions, especially by combining the isotopic labeling technique. Besides, recently amantadine has been shown to have the potential to be used as a membrane anchor on which membrane active substances can be linked to the membrane surface for various applications. Such a strategy is very useful to modify the natural cell membrane. The specific interaction between amantadine and negatively charged phospholipids may account for the excellent anchoring ability of the amantadine moiety to the cell membrane.

amantadine interacts with PG and PC lipids quite differently. It does not have a strong interaction with the zwitterionic PC bilayers at a subphase solution less than 10.0 mM. With a higher subphase solution, it could interact with the dDPPC/ DPPC bilayer and increase the flip-flop rate. Distinct interactions were observed between amantadine and the negatively charged DPPG bilayer. Amantadine in a low concentration (e.g., 0.20 mM) solution in the subphase could immediately disturb the outer lipid leaflet and then gradually reorganize to cause the outer leaflet to return to an orderlypacked state. A higher concentration of amantadine (e.g., 5.0 mM) could immediately disorder the packing state of the outer lipid leaflet. For both cases, the inner PG leaflet was not affected by amantadine. Our result demonstrated that for the PG lipid bilayers we studied, amantadine could only disturb the packing state of the outer membrane, but leave the inner lipid layer intact. The presence of amantadine only in the outer lipid leaflet of the negatively charged lipid bilayer may have certain implications for using liposomes as drug delivery carriers for amantadine. The present work demonstrates another example in addition to those already published16,48,59,60 that SFG results can provide in-depth understanding on the molecular mechanisms of interactions between water-soluble drugs and model cell membranes.

■

ASSOCIATED CONTENT

S Supporting Information *

Detailed experimental procedures, molecular formulas of the phospholipids, and the time-dependent SFG signals collected from the dDPPG/DPPG bilayer. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by grants from the Natural Science Foundation of China (21303017), the Natural Science Foundation of Jiangsu Province (KB20130601), and a project funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (1107037001). We also thank the support from the University of Michigan and the financial aids from Southeast University.

■

REFERENCES

(1) Makriyannis, A.; Tian, X.; Guo, J. How lipophilic cannabinergic ligands reach their receptor sites. Prostaglandins & Other Lipid Mediators 2005, 77, 210−218. (2) Zhou, H. X.; Szabo, A. Enhancement of association rates by nonspecific binding to DNA and cell membranes. Phys. Rev. Lett. 2004, 93, 178101. (3) Seydel, J. K.; Coats, E. A.; Cordes, H. P.; Wiese, M. Drug membrane interaction and the importance for drug transport, distribution, accumulation, efficacy and resistance. Arch. Pharm. 1994, 327, 601−610. (4) Pinto, L. H.; Lamb, R. A. Controlling influenza virus replication by inhibiting its proton channel. Mol. Biosyst. 2007, 3, 18−23. (5) Blanchet, P. J.; Metman, L. V.; Chase, T. N. Renaissance of amantadine in the treatment of Parkinson’s disease. Adv. Neurol. 2003, 91, 251−257.

4. CONCLUSION SFG was used to investigate how amantadine interacts with zwitterionic PC and negatively charged PG bilayers, serving as model cell membranes. Lipid bilayers with a deuterated leaflet and a hydrogenated leaflet were used to study the behavior of each leaflet and the lipid flip-flop during the bilayer− amantadine interactions. Lipid bilayers with both leaflets deuterated were also used for SFG to monitor the C−H stretching signals from amantadine during the membrane interaction to understand its action. The results showed that 8497

dx.doi.org/10.1021/la501718n | Langmuir 2014, 30, 8491−8499

Langmuir

Article

(6) Wang, C.; Takeuchi, K.; Pinto, L. H.; Lamb, R. A.Ion channel activity of influenza A virus M2 protein: characterization of the amantadine block. J. Virol. 1993, 67, 5585−5594. (7) Cady, S. D.; Hong, M. Amantadine-induced conformational and dynamical changes of the influenza M2 transmembrane proton channel. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 1483−1488. (8) Chew, C. F.; Guy, A.; Biggin, P. C. Distribution and dynamics of adamantanes in a lipid bilayer. Biophys. J. 2008, 95, 5627−5636. (9) Kim, C.; Agasti, S. S.; Zhu, Z. J.; Isaacs, L.; Rotello, V. M. Recognition-mediated activation of therapeutic gold nanoparticles inside living cells. Nat. Chem. 2010, 2, 962−966. (10) Epand, R. M.; Epand, R. F.; McKenzie, R. C. Effects of viral chemotherapeutic agents on membrane properties. Studies of cyclosporin A, benzyloxycarbonyl-D-Phe-L-Phe-Gly and amantadine. J. Biol. Chem. 1987, 262, 1526−1529. (11) Versluis, F.; Voskuhl, J.; van Kolck, B.; Zope, H.; Bremmer, M.; Albregtse, T.; Kros, A. In situ modification of plain liposomes with lipidated coiled coil forming peptides induces membrane fusion. J. Am. Chem. Soc. 2013, 135, 8057−8062. (12) Duff, K. C.; Cudmore, A. J.; Bradshaw, J. P. The location of amantadine hydrochloride and free base within phospholipid multilayersa neutron and x-ray-diffraction study. Biochim. Biophys. Acta 1993, 1145, 149−156. (13) Wang, J. F.; Schnell, J. R.; Chou, J. J. Amantadine partition and localization in phospholipid membrane: a solution NMR study. Biochem. Biophys. Res. Commun. 2004, 324, 212−217. (14) Li, C.; Yi, M.; Hu, J.; Zhou, H. X.; Cross, T. A. Solid-state NMR and MD simulations of the antiviral drug amantadine solubilized in DMPC bilayers. Biophys. J. 2008, 94, 1295−1302. (15) Subczynski, W. K.; Wojas, J.; Pezeshk, V.; Pezeshk, A. Partitioning and localization of spin-labeled amantadine in lipid bilayers: An EPR study. J. Pharm. Sci. 1998, 87, 1249−1254. (16) Wu, F. G.; Yang, P.; Zhang, C.; Han, X. F.; Song, M. H.; Chen, Z. J. Phys. Chem. C; Submitted for publication. (17) Yang, P.; Wu, F. G.; Chen, Z. Dependence of alamethicin membrane orientation on the solution concentration. J. Phys. Chem. C 2013, 117, 3358−3365. (18) Yang, P.; Wu, F. G.; Chen, Z. Lipid fluid−gel phase transition induced alamethicin orientational change probed by sum frequency generation vibrational spectroscopy. J. Phys. Chem. C 2013, 117, 17039−17049. (19) Shen, Y. R. The Principles of Nonlinear Optics; Wiley: New York, 1984. (20) Perry, A.; Neipert, C.; Space, B.; Moore, P. B. Theoretical modeling of interface specific vibrational spectroscopy: methods and applications to aqueous interfaces. Chem. Rev. 2006, 106, 1234−1258. (21) Liljeblad, J. F. D.; Bulone, V.; Tyrode, E.; Rutland, M. W.; Johnson, C. M. Phospholipid monolayers probed by vibrational sum frequency spectroscopy: instability of unsaturated phospholipids. Biophys. J. 2010, 98, L50−L52. (22) Liljeblad, J. F. D.; Bulone, V.; Rutland, M. W.; Johnson, C. M. Supported phospholipid monolayers. the molecular structure investigated by vibrational sum frequency spectroscopy. J. Phys. Chem. C 2011, 115, 10617−10629. (23) Wurpel, G. W. H.; Sovago, M.; Bonn, M. Sensitive probing of DNA binding to a cationic lipid monolayer. J. Am. Chem. Soc. 2007, 129, 8420−8421. (24) Campen, R. K.; Ngo, T. T. M.; Sovago, M.; Ruysschaert, J. M.; Bonn, M. Molecular restructuring of water and lipids upon the interaction of DNA with lipid monolayers. J. Am. Chem. Soc. 2010, 132, 8037−8047. (25) Darwish, T. A.; Tong, Y.; James, M.; Hanley, T. L.; Peng, Q.; Ye, S. Characterizing the photoinduced switching process of a nitrospiropyran self-assembled monolayer using in situ sum frequency generation spectroscopy. Langmuir 2012, 28, 13852−13860. (26) Tong, Y.; Li, N.; Liu, H.; Ge, A.; Osawa, M.; Ye, S. Mechanistic studies by sum-frequency generation spectroscopy: hydrolysis of a supported phospholipid bilayer by phospholipase A2. Angew. Chem., Int. Ed. 2010, 49, 2319−2323.

(27) Chen, X.; Hua, W.; Huang, Z.; Allen, H. C. Interfacial water structure associated with phospholipid membranes studied by phasesensitive vibrational sum frequency generation spectroscopy. J. Am. Chem. Soc. 2010, 132, 11336−11342. (28) Ma, G.; Allen, H. C. DPPC langmuir monolayer at the air− water interface: probing the tail and head groups by vibrational sum frequency generation spectroscopy. Langmuir 2006, 22, 5341−5349. (29) Ma, G.; Allen, H. C. Condensing effect of palmitic acid on DPPC in mixed langmuir monolayers. Langmuir 2007, 23, 589−597. (30) Fu, L.; Ma, G.; Yan, E. C. Y. In situ misfolding of human islet amyloid polypeptide at interfaces probed by vibrational sum frequency generation. J. Am. Chem. Soc. 2010, 132, 5405−5412. (31) Liu, W.; Wang, Z. G.; Fu, L.; Leblanc, R. M.; Yan, E. C. Y. Lipid compositions modulate fluidity and stability of bilayers: characterization by surface pressure and sum frequency generation spectroscopy. Langmuir 2013, 29, 15022−15031. (32) Fu, L.; Liu, J.; Yan, E. C. Y. Chiral sum frequency generation spectroscopy for characterizing protein secondary structures at interfaces. J. Am. Chem. Soc. 2011, 133, 8094−8097. (33) Xiao, D. Q.; Fu, L.; Liu, J.; Batista, V. S.; Yan, E. C. Y. Amphiphilic adsorption of human islet amyloid polypeptide aggregates to lipid/aqueous interfaces. J. Mol. Biol. 2012, 421, 537−547. (34) Watry, M. R.; Tarbuck, T. L.; Richmond, G. L. Vibrational sumfrequency studies of a series of phospholipid monolayers and the associated water structure at the vapor/water interface. J. Phys. Chem. B 2003, 107, 512−518. (35) Mondal, J. A.; Nihonyanagi, S.; Yamaguchi, S.; Tahara, T. Three distinct water structures at a zwitterionic lipid/water interface revealed by heterodyne-detected vibrational sum frequency generation. J. Am. Chem. Soc. 2012, 134, 7842−7850. (36) Mondal, J. A.; Nihonyanagi, S.; Yamaguchi, S.; Tahara, T. Structure and orientation of water at charged lipid monolayer/water interfaces probed by heterodyne-detected vibrational sum frequency generation spectroscopy. J. Am. Chem. Soc. 2010, 132, 10656−10657. (37) Ye, S. J.; Li, H. C.; Yang, W. L.; Luo, Y. Accurate determination of interfacial protein secondary structure by combining interfacialsensitive amide I and amide III spectral signals. J. Am. Chem. Soc. 2014, 136, 1206−1209. (38) Ma, S. L.; Li, H. C.; Tian, K. Z.; Ye, S. J.; Luo, Y. In situ and real time SFG measurements reveal organization and transport of cholesterol analog 6-ketocholestanol in cell membrane. J. Phys. Chem. Lett. 2014, 5, 419−424. (39) Wei, F.; Li, H. C.; Ye, S. J. Specific ion interaction dominates over hydrophobic matching effects in peptide-lipid bilayer interactions: the case of short peptide. J. Phys. Chem. C 2013, 117, 26190−26196. (40) Ye, S. J.; Li, H. C.; Wei, F.; Jasensky, J.; Boughton, A. P.; Yang, P.; Chen, Z. Observing a model ion channel gating action in model cell membranes in real time in situ: membrane potential change induced alamethicin orientation change. J. Am. Chem. Soc. 2012, 134, 6237− 6243. (41) Han, X. F.; Liu, Y. W.; Wu, F. G.; Jansensky, J.; Kim, T.; Wang, Z. L.; Brooks, C. L., III; Wu, J. F.; Xi, C. W.; Mello, C. M.; Chen, Z. J. Phys. Chem. B 2014, 118, 2904−2912. (42) Anglin, T. C.; Conboy, J. C. Lateral pressure dependence of the phospholipid transmembrane diffusion rate in planar-supported lipid bilayers. Biophys. J. 2008, 95, 186−193. (43) Anglin, T. C.; Conboy, J. C. Kinetics and thermodynamics of flip-flop in binary phospholipid membranes measured by sumfrequency vibrational spectroscopy. Biochemistry 2009, 48, 10220− 10234. (44) Liu, J.; Brown, K. L.; Conboy, J. C. The effect of cholesterol on the intrinsic rate of lipid flip-flop as measured by sum-frequency vibrational spectroscopy. Faraday Discuss. 2013, 161, 45−61. (45) Liu, J.; Conboy, J. C. Direct measurement of the transbilayer movement of phospholipids by sum-frequency vibrational spectroscopy. J. Am. Chem. Soc. 2004, 126, 8376−8377. (46) Liu, J.; Conboy, J. C. 1,2-Diacyl-phosphatidylcholine flip-flop measured directly by sum-frequency vibrational spectroscopy. Biophys. J. 2005, 89, 2522−2532. 8498

dx.doi.org/10.1021/la501718n | Langmuir 2014, 30, 8491−8499

Langmuir

Article

(47) Anglin, T. C.; Cooper, M. P.; Li, H.; Chandler, K.; Conboy, J. C. Free energy and entropy of activation for phospholipid flip-flop in planar supported lipid bilayers. J. Phys. Chem. B 2010, 114, 1903− 1914. (48) Chen, X. Y.; Wang, J.; Kristalyn, C. B.; Chen, Z. Real-time structural investigation of a lipid bilayer during its interaction with melittin using sum frequency generation vibrational spectroscopy. Biophys. J. 2007, 93, 866−875. (49) Yang, P.; Ramamoorthy, A.; Chen, Z. Membrane orientation of MSI-78 measured by sum frequency generation vibrational spectroscopy. Langmuir 2011, 27, 7760−7767. (50) Ding, B.; Chen, Z. Molecular interactions between cell penetrating peptide Pep-1 and model cell membranes. J. Phys. Chem. B 2012, 116, 2545−2552. (51) Ye, S. J.; Nguyen, K. T.; Chen, Z. Interactions of alamethicin with model cell membranes investigated using sum frequency generation vibrational spectroscopy in real time in situ. J. Phys. Chem. B 2010, 114, 3334−3340. (52) Ye, S. J.; Nguyen, K. T.; Le Clair, S. V.; Chen, Z. In situ molecular level studies on membrane related peptides and proteins in real time using sum frequency generation vibrational spectroscopy. J. Struct. Biol. 2009, 168, 61−77. (53) Verreault, D.; Hua, W.; Allen, H. C. From conventional to phase-sensitive vibrational sum frequency generation spectroscopy: probing water organization at aqueous interfaces. J. Phys. Chem. Lett. 2012, 3, 3012−3028. (54) Ong, S. W.; Zhao, X. L.; Eisenthal, K. B. Polarization of water molecules at a charged interface: second harmonic studies of the silica/ water interface. Chem. Phys. Lett. 1992, 191, 327−335. (55) Zhao, X. L.; Ong, S. W.; Eisenthal, K. B. Polarization of water molecules at a charged interface. Second harmonic studies of charged monolayers at the air/water interface. Chem. Phys. Lett. 1993, 202, 513−520. (56) Sung, W.; Seok, S.; Kim, D.; Tian, C. S.; Shen, Y. R. Sumfrequency spectroscopic study of langmuir monolayers of lipids having oppositely charged headgroups. Langmuir 2010, 26, 18266−18272. (57) Wei, X.; Miranda, P. B.; Zhang, C.; Shen, Y. R. Sum-frequency spectroscopic studies of ice interfaces. Phys. Rev. B 2002, 66, 085401/ 1−085401/13. (58) Du, Q.; Superfine, R.; Freysz, E.; Shen, Y. R. Vibrational spectroscopy of water at the vapor/water interface. Phys. Rev. Lett. 1993, 70, 2313−2316. (59) Chen, X. Y.; Tang, H. Z.; Even, M. A.; Wang, J.; Tew, G. N.; Chen, Z. Observing a molecular knife at work. J. Am. Chem. Soc. 2006, 128, 2711−2714. (60) Avery, C. W.; Som, A.; Xu, Y. J.; Tew, G. N.; Chen, Z. Dependence of antimicrobial selectivity and potency on oligomer structure investigated using substrate supported lipid bilayers and sum frequency generation vibrational spectroscopy. Anal. Chem. 2009, 81, 8365−8372.

8499

dx.doi.org/10.1021/la501718n | Langmuir 2014, 30, 8491−8499