Article pubs.acs.org/Langmuir

α-Tocopherol’s Location in Membranes Is Not Affected by Their Composition Drew Marquardt,*,† Norbert Kučerka,‡,§,∥ John Katsaras,†,⊥,#,∇ and Thad A. Harroun*,† †

Department of Physics, Brock University, St. Catharines, Ontario L2S 3A1, Canada National Research Council, Canadian Neutron Beam Centre, Chalk River, Ontario K0J 1J0, Canada § Department of Physical Chemistry of Drugs, Comenius University, 832 32 Bratislava, Slovakia ∥ Frank Laboratory of Neutron Physics, Joint Institute for Nuclear Research, 141980 Dubna - Moscow Region, Russia ⊥ Neutron Sciences Directorate, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831-6393, United States, # Joint Institute for Neutron Sciences, Oak Ridge, Tennessee37831-6453, United States ∇ Department of Physics and Astronomy, University of Tennessee, Knoxville, Tennessee 37996, United States ‡

ABSTRACT: To this day, α-tocopherol’s (aToc) role in humans is not well known. In previous studies, we have tried to connect aToc’s biological function with its location in a lipid bilayer. In the present study, we have determined, by means of small-angle neutron diffraction, that not only is aToc’s hydroxyl group located high in the membrane but its tail also resides far from the center of 1-palmitoyl-2-oleoyl-snglycero-3-phosphocholine (POPC) bilayers. In addition, we located aToc’s hydroxyl group above the lipid backbone in 1palmitoyl-2-oleoyl-sn-glycero-3-phosphoethanolamine (POPE), 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-L-serine (POPS), and sphingomyelin bilayers, suggesting that aToc’s location near the lipid−water interface may be a universal property of vitamin E. In light of these data, how aToc efficiently terminates lipid hydroperoxy radicals at the membrane center remains an open question.

■

INTRODUCTION

abstracted from the phenolic hydroxyl group by a lipid peroxyl radical, producing a tocopheroxyl radical.6,7 Cellular membranes are barriers that separate the inside of a cell from its environment. The main structural component of the cell membrane is lipids, a diverse group of molecules that play a crucial role in the structure−function relationship of the cell membrane. Phospholipids are a major group of membrane lipids and come with a variety of headgroup moieties, such as the zwitterionic phosphotidylcholine (PC) or phosphotidylethanolamine (PE) moiety, or possess a formal charge, such as in the case of phosphotidylserine (PS).8 These different headgroup lipids are not always colocated. For instance, PC is the most common phospholipid headgroup and is found in most eukaryotic cells. PE is found in membranes of the nervous system (e.g., brain cells) and accounts for approximately half of the phospholipid content.9 The negatively charged PS is found as a component of cell and blood platelet membranes.9 The highest concentration of PS lipids is found in the plasma membrane, typically in the cytoplasmic part of the membrane.9,10 In different cell types or organelles, lipid headgroups exist, in part, to modulate various biophysical properties such as the area

1

Discovered in 1922 by Evans and Bishop, the biological role of vitamin E, of which α-tocopherol (aToc) is the human active component, remains unknown. Importantly, there is no credible scientific evidence for the health benefits of vitamin E supplements. There are, however, several known conditions associated with vitamin E deficiency, including profound muscle weakness2 and hemolytic anemia3 to name a few. Despite the extended list of ill effects associated with vitamin E deficiency, a molecular picture describing how these conditions arise has yet to be formulated. Vitamin E is composed of two families of molecules known as tocopherols and tocotrienols, each with four members (i.e., α, β, δ, and γ). Although all eight molecules share many similarities and are consumed by humans, only aToc is taken up by the human body.4 The various vitamin E compounds are known to be fat-soluble antioxidants and are commonly used as preservatives in cosmetics and foods. When combined with water-soluble ascorbic acid, they make for one of the most powerful commercial preservative. In vivo, it is hypothesized that tocopherol protects highly vulnerable polyunsaturated phospholipids (PUPL) and polyunsaturated fatty acids (PUFA), although this has been challenged by other hypotheses and thus remains an open question.5 As the oxidation of tocopherol takes place, the hydroxyl hydrogen is © XXXX American Chemical Society

Received: July 2, 2014 Revised: September 1, 2014

A

dx.doi.org/10.1021/la502605c | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

per lipid of the membrane and, by extension, its curvature. For example, PE lipids are known to form regions of negative curvature in membranes because of the small neutral structure of their headgroup. In addition, the umbrella model predicts that the small PE headgroup is unable to shield the different guest molecules (e.g., cholesterol) from water, unlike its larger PC counterpart, thus leading to a lower solubilty of cholesterol in PE-rich portions of the membrane.11,12 The charged PS lipid can facilitate the binding of metals to the membrane’s surface, typically through electrostatic interactions between the negatively charged PS and metal cations. Beyond the fatty acid tail and headgroup compositions, the third significant phospholipid moiety is the glycerol backbone, which is ubiquitous in most phospholipids. However, the ceramide backbone is found in a class of phospholipids called sphingolipids, with the most common, sphingomyelin, commonly found in the myelin sheath of nerve cells, acting as an insulator. Sphingomyelin may be best known for its ability to solubilize cholesterol. In fact, cholesterol’s affinity for sphingomyelin is so high that they pack tightly together in the so-called liquid-ordered phase, along with other saturated acyl chain phospholipids, which is associated with lipid rafts.13−15 Recently, new structural data has been obtained in the hope that one can explain and predict the role of aToc in biological membranes. For example, recent neutron diffraction studies have located aToc’s sacrificial hydroxyl group in bilayers with different degrees of acyl chain unsaturation. It was noted that regardless of acyl chain unsaturation aToc resides at the lipid− water interface.16 On the basis of that data, we proposed a mechanism in which aToc’s antioxidant action takes place at the lipid−water interface as a “first line of defense” against water-borne oxidants; the termination of lipid peroxyl radicals relies on the intrinsic disorder of polyunsaturated fatty acid (PUFA) chains to bring lipid peroxides to the surface, where they can be terminated by aToc.17−22 The one exception to aToc’s high location in the bilayer was found in the prototypical lipid, dimyristoyl-phosphatidylcholine (DMPC).23 By means of small-angle neutron diffraction, oxidation assays, and 2H NMR, we unambiguously determined that aToc resides in the center of a DMPC bilayer. However, we could not determine the reasons for aToc’s unusual behavior in DMPC bilayers. In this study, we examined the location of aToc in bilayers made up of lipids with several different headgroup species, with a focus on POPC bilayers. First, we measured the tail distribution of aToc in a POPC bilayer in order to determine aToc’s tail structure. Figure 1 shows the 2H labeling of aToc’s 5′ and 9′ methylene positions. Second, we kept the chemical composition of the fatty acid chains the same for all different headgroup bilayers so that aToc’s location in the different bilayers and whether there is an umbrella or charge effect could be determined. Finally, we looked for a lipid backbone effect by measuring aToc’s hydroxyl position in sphingomyelin, again using the 2H3-labeled C5 methyl as a proxy for the hydroxyl group.

■

Figure 1. Chemical structures of α-tocopherol (aToc), N-palmitoyl-Derythro-sphingosylphosphorylcholine (sphingomyelin, SM), 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), 1-palmitoyl-2oleoyl-sn-glycero-3-phosphoethanolamine (POPE), and 1-palmitoyl2-oleoyl-sn-glycero-3-phospho-L-serine (POPS). 5′ and 9′ indicate the locations of the 2H labels. The other label is located on the methyl group of the 5 carbon on the chromanol ring. nolamine (POPE), and 1-palmitoyl-2-oleoyl-sn-glycero-3-phospho-Lserine (POPS). We purified 2H-labeled α-[5-2H3]tocopherol (headgroup-labeled aToc-C 5d3), α-[5-2H2]tocopherol (tail-labeled aToc-C5′d2), and α[9′-2H2]tocopherol (tail-labeled aToc-C9′d2) from previously synthesized preparations following published protocols,16,24 and protonated α-tocopherol (aToc) was purchased and used as received from ColeParmer (Vernon Hills, IL). Small-Angle Neutron Diffraction. Mixtures of approximately 12 mg of lipid containing 10 mol % aToc (labeled or unlabeled) were codissolved in a 1:1 chloroform/trifluoroethanol solution. Care was taken to prevent exposure of the lipids to oxygen and to ensure minimal exposure to light. Each sample was then deposited on the surface of a 1-mm-thick silicon wafer (25 × 60 mm2) in a glovebox filled with nitrogen. The wafer was rocked during the evaporation of the organic solvent.25 The samples were then dried under vacuum for several hours (no less than 6 h). Neutron diffraction data were collected at the Canadian Neutron Beam Centre’s N5 and D3 beamlines located at the National Research Universal (NRU) reactor (Chalk River, Ontario, Canada), using 2.37Å-wavelength neutrons. The appropriate wavelength neutrons were selected by the (002) reflection of a pyrolytic graphite (PG) monochromator, and a PG filter was used to eliminate higher-order reflections (i.e., λ/2, λ/3, etc.). Samples were placed in an airtight sample cell purged with argon and hydrated to 94% relative humidity (RH) using a series of D2O/H2O mixtures (i.e., 100, 70, 40, and 8% D2O). RH was controlled by saturating the D2O/H2O solutions with KNO3 salt. Experimental stability over the course of the data collection was determined by the reproducibility of the lamellar repeat spacing and the intensity of the Bragg peaks over time. The reproducibility indicated that the sample chemistry and conditions remained static throughout the measurement. Data from samples that were underwent change were discarded, and new samples were prepared. Data analysis was carried out in a manner similar to that in studies determining the location of cholesterol.26,27 The diffraction peaks were

MATERIALS AND METHODS

Phospholipids were purchased from Avanti Polar Lipids (Alabaster, AL) and after experimentation were evaluated for degradation by thin layer chromatography. The lipids studied were sphingomyelin (from chicken eggs), which is predominately N-palmitoyl-D-erythro-sphingosylphosphorylcholine (SM), 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoethaB

dx.doi.org/10.1021/la502605c | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

obtained from θ − 2θ scans when the Bragg condition was satisfied, where θ and 2θ are the angles of the sample and the detector, respectively. Corrected form factors were then applied in a cosine Fourier sum to produce a real-space map of the neutron scattering length density (NSLD, ρ), analogous to an electron density map determined by X-ray scattering. ρ=

F0 2 + d d

h

⎛ 2πhz ⎞ ⎟ d ⎠

∑ Fh cos⎜⎝ h=1

Table 1. Measured and Calculated Structural Parameters for Different Membranes Containing 10 mol % aToc lipid/aToc label POPC/ aToc-C 5d3a POPC/ aToc-C 5′d2 POPC/ aToc-C 9′d2 POPE/ aToc-C 5d3 POPS/ aToc-C 5d3 SM/aToc-C 5d3

(1)

d represents the repeat spacing of the unit cell, h is the diffraction order, and F0 is the total scattering length of the unit cell (calculated). The difference in scattering length density between samples (Δρ) is determined by cosine reconstruction using the difference between labeled and unlabeled corrected form factors, FLh and FUh , respectively. Δρ =

F0L − F0U 2 + d d

h

⎛ 2πhz ⎞ ⎟ d ⎠

∑ (FhL − FhU)cos⎜⎝ h=1

(2)

As a test of the quality of our model, we introduce the commonly used idea of a crystallography R factor.28 However, our model is constructed from Δρ (eq 1), which relies on the difference in labeled and unlabeled form factors, so we modify the standard R factor to what we call the modified R factor, Rmod, R mod =

a

d spacing (Å)

bilayer thickness (Å)

label depth (Å)

label width (Å)

53.2

35.72(0.7)

24.2(0.03)

4.22(0.04)

0.29

53.3

36.4(0.3)

20.8(0.2)

5.3(0.2)

0.11

53.3

36.4(0.3)

18.7(0.2)

8.7(0.2)

0.005

51.5

40.3(0.3)

20.7(0.1)

6.4(0.10)

0.19

53.5

43.7(0.3)

18.7(0.1)

5.6(0.10)

0.0965

60.7

44.2(0.1)

20.8(0.1)

3.1(0.10)

0.6

Rmod

Data is reproduced from Marquardt et al.16

∑ ||(FhL − FhU)| − |ΔFhcalc|| ∑ |(FhL − FhU)|

(3)

Fcalc h

represents the form factors generated from our model. where Thus, the quality of the models can be compared between samples. Differential Scanning Calorimetry. Lipid/aToc mixtures were prepared by mixing appropriate amounts of lipid/chloroform solution with aToc/chloroform solution to yield lipid mixtures with 0.3 and 10 mol % aToc. A pure lipid sample was also prepared to create a consistent point of reference. The mixtures were placed under vacuum overnight to ensure that all residual chloroform was removed. The samples were then hydrated with ultrapure water to a concentration of 0.05 g/mL and subjected to five freeze/thaw cycles with vortex mixing. Thirty microliters of solution was loaded into Shimadzu aluminum hermetic pans and crimp sealed. The thermograms were collected at a scan rate of 2 °C/min and a sampling rate of 1 s using a Shimadzu DSC-60 and a TA-60WS thermal analyzer.

■

Figure 2. Raw (unnormalized) data of SM bilayers containing 10 mol % aToc-5d3 and hydrated with 8% 2H2O, resulting in five lamellar Bragg reflections. The inset to the figure is a schematic diagram of Bragg scattering from an oriented sample.

RESULTS Oriented membrane multilayers were adsorbed onto silicon single-crystal substrates and hydrated in a 94% relative humidity environment. In monochromatic neutron beam, these samples produce four to six quasi-Bragg peaks, with lamellar repeat spacings of d ≈ 50−60 Å for the different membranes studied; the results are summarized in Table 1. Figure 2 shows a representative data set for SM bilayers containing 10 mol % aToc-C 5d3 and hydrated with 70% 2H2O. Figure 3 shows rocking curves collected for samples containing protiated aToc. These data are a measure of the sample’s orientational quality and are collected by rotating (θ) the sample in the incident neutron beam while keeping the detector angle fixed. We see a sharp, narrow peak at the Bragg angle (centered at 0° for ease of presentation) accounting for 76−96% of the total scattering intensity (indicative of wellordered bilayer stacks), sitting on top of a broad shoulder. The broad shoulder is the product of misaligned bilayers within the sample and contributes little to the reconstruction of the scattering-length density profile. The width of the central peak also happens to be narrower than the instrumental resolution. The central peaks of POPC, POPE, and POPS account for 96, 92, and 85% of the scattering intensity, respectively. SM has the broadest central peak of all samples (i.e., 76% of the scattered

intensity), which we attribute to the use of egg sphingomyelin extract with up to 15% of the sphingmyelin having acyl chains other than 16:0, thus making these bilayers more difficult to align. aToc’s Tail Distribution in POPC. The dashed line in Figure 4 is the 1D neutron scattering length density (NSLD) profile along the direction of the bilayer normal for POPC bilayers containing 10 mol % protiated aToc and hydrated with 8% 2H2O. Here, the water layer is “invisible” as it is contrast matched, and the NSLD is solely the result of the lipid and aToc. The lipid phosphates and backbone are the main contributors to the NSLD peaks, and the trough in the middle of the bilayer is attributed to the CH3 terminal methyls. The blue curve in Figure 4 indicates the water distribution between bilayers, which is determined by subtracting the diffraction data at 8% 2H2O from that at 70% 2H2O. We see that the 1/e level of water penetration occurs before the peak in the NSLD profile of the 8% 2H2O data, indicating that water does not penetrate past the glycerol backbone, which is typical for POPC bilayers.8 These parameters (d spacing, bilayer C

dx.doi.org/10.1021/la502605c | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

thickness, and water distribution) are also in agreement with the protiated aToc in POPC measured in Marquardt et al.16 See Tables 1 and 2 for a complete list of the structural parameters. Table 2. Water Distribution for Membranes Containing 10 mol % aToc at 94% RH

a

lipid

depth (Å)

width (Å)

Rmod

POPCa POPE POPS SM

26.6(1.1) 25.9(0.1) 26.76(0.03) 30.3(1.1)

7.71(0.4) 8.7(0.1) 10.5(0.8) 7.3(0.4)

0.15 0.056 0.11 0.078

POPC data is reproduced from Marquardt et al.16

The solid black line in Figure 4 is an example of a POPC bilayer containing 2H-labeled aToc, specifically the NSLD profile for POPC bilayers with 10 mol % aToc-C9′d2 hydrated with 8% 2H2O. The d spacing and bilayer thickness for both aToc-C9′d2 and aToc-C5′d2 (data not shown) agree with the sample containing protiated aToc, which allows for a direct subtraction of the 8% 2H2O hydration data, yielding the 2H aToc label distribution. A Gaussian distribution is assumed for the distribution of the 2H label, and its Fourier transform is fitted to the corrected data in reciprocal space.16 We perform the subtraction in reciprocal space because in real space such subtractions are subject to fluctuations as a result of having only four or five terms (quasi-Bragg reflections) sampling the Fourier series. In addition, the effects of the repeating centrosymmetric unit cell are not present in reciprocal space, and thus we can draw the label distributions near the edges (or center) of the unit cell and they are not distorted by those from adjacent unit cells. Figure 4 shows Gaussian fits for aToc-C 5d3 (from Marquardt et al.16), aToc-C5′d2, and aToc-9′d2 (this work). The label distributions appear to become more localized as the label position gets closer to the lipid headgroups; we will discuss this observation below. The fit parameters for the label distributions are also listed in Figure 1. Location of aToc’s Hydroxyl in Bilayers with Different Lipid Headgroups. In a previous report we examined aToc’s location in PC bilayers with different acyl chain lengths, degrees of unsaturation, and acyl chain combinations.16 It was determined that aToc’s overall location in different bilayers was invariant, the exception being DMPC.23 Here we present the approximate location of aToc’s hydroxyl in phospholipid bilayers whose chain composition is 1-palmitoyl-2-oleoyl but with different headgroups. We do so by labeling the hydrogen of the adjacent C5 methyl group and measuring its location. The dashed curves in Figure 5 show bilayers containing 10 mol % protonated aToc. The NSLD profiles have a shape that is similar to that of POPC; however, the height of the headgroup peaks differ because of the differing headgroup compositions that, as expected, possess different scattering length densities. There is also a difference in water distribution (outermost, blue lines in Figure 5) compared to that in PC. The water penetrates more deeply past the lipid headgroup peak, especially in the case of POPS bilayers. PS is a charged headgroup, rich in oxygen, and thus more hydrophilic than the others studied. The repeat spacing of POPE and POPS bilayers is comparable to that of POPC (±2 Å); however, the peak-topeak distance is significantly greater for POPS and POPE bilayers. The thickness of POPE bilayers with 10 mol % aToc is

Figure 3. Rocking curves are collected by rotating the sample, in the incident neutron beam, through an angle (θ) while keeping the detector fixed at the Bragg scattering condition. The narrow peak at 0° indicates highly aligned multilayers. Dips in the intensity occur when the sample is aligned parallel to the incident or diffracted neutron beam and undergoes maximum absorption.

Figure 4. Dashed black line is the NSLD profile for POPC bilayers containing 10 mol % protiated aToc and hydrated with 8% 2H2O. The solid black line is an example of an NSLD profile containing labeled aToc, specifically, POPC containing 10 mol % aToc-C9′d2 hydrated with 8% 2H2O (water distribution indicated by the blue curve). The 2 H distributions for the three labels examined in POPC bilayers are shown by the colored lines: aToc-C 5d3 (red), aToc-C5′d2 (magenta), and aToc-9′d2 (green). The aToc-C 5d3 data is from ref 16. The aToc molecule is not drawn to scale but is enlarged to show which labels on the molecule correspond to which distribution.

D

dx.doi.org/10.1021/la502605c | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

backbone, other than glycerol, is ceramide found in sphingolipids. Here we examined egg sphingomyelin, which contains predominately 16:0 chains (SM in Figure 5). Again we see an NSLD profile similar in shape to POPC bilayers. However, the repeat spacing is over 10% greater (Table 1), something commonly seen with sphingomyelin. The ceramide backbone is significantly smaller than the glycerol analogue, which results in the ordering of the hydrocarbon chains, which has also been observed by Bartels et al.31 This effect is also reflected in the bilayer thickness. The egg SM bilayer thickness and d spacing are also in good agreement with the value of 44.4 Å determined by Maulik and Shipley32 for pure 16:0 SM bilayers. The difference between the solid and the dashed black lines of SM bilayers containing aToc results in a red line, which is the Gaussian distribution of the 2H label of aToc-C 5d3. Like the other lipids examined, the aToc-C 5d3 label sits high in the bilayer; however, in the case of SM bilayers, the distribution is very narrow. The localized distribution of aToc can be attributed to the rigidity and ordering effect that is provided by the ceramide backbone. Furthermore, compared to PC, PE, and PS bilayers, there is less overlap of the water with the aToc headgroup when in SM bilayers. Differential Scanning Calorimetry. Figure 6 and Table 3 show the effect that aToc has on the main chain melting phase transition of the examined phospholipids over temperatures allowable by the calorimeter. The chain melting transitions of pure bilayers, with the exception of sphingomyelin, are in excellent agreement with previously reported values.33−36 Note that the value for the sphingomyelin main transition is 4 °C lower than the reported value for pure 16:0 SM (41 °C,32), which we can attribute to the use of egg sphingomyelin, which is only ∼86% 16:0. Upon the addition of naturally abundant (0.3 mol %) aToc, the main transition for POPE, POPS, and egg SM bilayers are significantly broadened, but with only a modest shift in transition temperature and enthalpy (Table 3). Moreover it is known that POPE undergoes a transition to a hexagonal phase at ∼75 °C. We observe this transition in both the pure lipid sample and the sample containing a natural abundance aToc. However, this transition is eliminated upon the addition of 10 mol % aToc (data not shown). Interestingly, we were able to observe the main chain melting transition for POPC (∼−5 °C) at a total lipid concentration of 50 mg/mL. However, the main transition is diminished in the cooling thermogram with the natural abundance of aToc and is completely obscured in the presence of 10 mol % aToc. For DPPC, the addition of a natural concentration of aToc (0.3 mol %) suppresses the main chain melting transition (Tm) as well as significantly decreasing the enthalpy of melting (ΔH). This suppression is observed in both the heating and cooling thermograms (Table 3). When the concentration of aToc is increased to 10 mol %, further decreases in Tm and ΔH are observed. This trend in DPPC is in qualitative agreement with the observations by Quinn.37 Curiously, DMPC does not follow this same trend; upon the addition of 0.3 mol % aToc, there is no observable effect on Tm and ΔH in the heating thermograms and only a modest decrease in the cooling scans. However, when 10 mol % aToc is incorporated into DMPC bilayers, an effect similar to what is seen in other lipid bilayers is observed. This result is associated with aToc’s unusual location in DMPC, as we will discuss below.

Figure 5. Neutron scattering length data of aToc-C 5d3 in different liquid-crystalline phospholipid membranes. Solid black lines: lipid and labeled aToc in 8% 2H2O. Dashed black lines: lipid and protiated aToc in 8% 2H2O. Red lines: The mass distribution of aToc’s C5-methyl label calculated from the difference between sol-5id and the dashed black lines and scaled by a factor of 2. Blue lines: Difference between the 8% 2H2O data and the same sample in 70% 2H2O, showing the mass distribution of the interbilayer water.

slightly less than a previously reported value of 41.8 Å for pure POPE bilayers,29 and our POPS with 10 mol % aToc bilayer thickness is slightly thicker than a recently published value of 42.2 Å reported by Pan et al.30 for pure POPS. This can be partially explained by the lower RH in our study, resulting in some steric constraints in the lateral direction. In other words, the acyl chains cannot “curl up” as much. The solid black lines in Figure 5 are the NSLD profiles for the phospholipid bilayer containing 10 mol % aToc-C 5d3. The d spacing and bilayer thickness for aToc-C 5d3 agree with samples containing protiated aToc, which allow for a direct subtraction of the 8% 2H2O data to yield the C 5d3 label distribution. The red lines in Figure 5 are the Gaussian distributions of the label, which are determined by performing the subtraction and fits to the data in reciprocal space as was discussed above. In the case of POPE and POPS bilayers, we observe aToc sitting high in the bilayer with slightly broader distributions than in POPC bilayers. The location and width of these distributions are summarized in Table 1. However, compared to POPC bilayers, aToc resides lower in POPE and POPS bilayers. For completeness, the water distribution in these lipid bilayers is shown in Table 2. There is considerable overlap of aToc’s hydroxyl group and water, with the exception being SM bilayers, indicating that, for the most part, aToc is anchored at the lipid/water interface in PC, PE, and PS bilayers. Compared to headgroups, there are significantly fewer different backbone moieties found in nature. A common E

dx.doi.org/10.1021/la502605c | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

Figure 6. Heating (bottom curves) and cooling (top curves) scans of POPE, POPS, and egg SM MLVs prepared in the presence of (black) 0 mol % aToc, (red) 0.3 mol % aToc, and (green) 10 mol % aToc.

■

DISCUSSION That the distribution of the three different tail labels on aToc in POPC are not found at the bilayer center is reasonable considering the disordered nature of aToc’s phytanyl tail. A fully stretched aToc has a distance from the C5-methyl on the chromane to the C9′ tail position of ∼15 Å, which is more than twice the distance between the centers of their measured distributions. It is likely then that aToc spends little time fully extended. The conformational freedom of the labels is reflected in the 1/e distribution width of each label. The distribution becomes larger as the label is placed further down the phytanyl side chain and consequently deeper into the membrane. If the aToc-C 5d3 label distribution represents the local motion of aToc in and out of the bilayer, then the excess width of the tail labels represents the extra motion from their conformational degrees of freedom. This increased motional freedom taking place deeper in the bilayer is how we previously interpreted the 2 H NMR order parameters for aToc-C 5d3, aToc-C5′d2, and aToc-9′d2.16 In addition, we previously reported the 2H NMR order parameter of aToc labeled at the 5′ and 9′ positions16 and for d31-POPC bilayers containing protiated aToc.23 In both cases, we made the point that aToc resides very high in the bilayer and that the majority of aToc’s impact on POPC’s acyl chains is

near the lipid’s glycerol backbone. In this work, the location of the aToc-C5′d2 and aToc-C9′d2 labels near the lipid−water interface, as determined by neutron diffraction, implies that aToc’s tail is rarely fully extended. This curling of the tail near the water interface could act as an additional barrier for oxidants diffusing into the bilayer through a volume-filling mechanism, as shown in the schematic in Figure 7. The location of aToc in different membranes will provide simulators with a reference to fine-tune their force fields. An MD simulation study of aToc in different membranes, including POPC and POPE, reported that aToc resides much deeper in the bilayer than our neutron results.38 In fact, we locate the tail labels closer to the lipid−water interface of POPC bilayers than do simulations. Our data clearly show that aToc’s hydroxyl group sits near the lipid−water interface in POPE, POPS, and SM bilayers, above the lipid backbone. This result supports the previous hypothesis that aToc’s biological function is exercised at the bilayer surface. The fact that the bilayer’s structural properties remain relatively unchanged in the presence or absence of aToc suggests a structural role similar to that of cholesterol.29 Similar to cholesterol, aToc has a significant effect on the phase behavior of the phospholipid bilayers in which it resides in. In addition, it is known that adding material to a lipid bilayer F

dx.doi.org/10.1021/la502605c | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

the choline methyls that remain unfrozen down to −70 °C.42 The phosphate water associated cluster is, however, not large enough to serve as a point of nucleation, and the use of ultrapure water enabled us to supercool the water in the sample. Our measured Tm for POPC was consistent with the value determined by 2H NMR.41,42 The thermotropic behavior of all of the lipids studied appears to trend with the location of the C5-methyl of aToc in the bilayer. In bilayers where C5-methyl is above the glycerol/ phosphate group (i.e., DPPC, POPC, and POPE; Figure 5), there is a distinct decrease in Tm (5%) in the presence of physiological concentrations of aToc (0.3 mol %).43 In contrast, in those bilayers where aToc’s C5-methyl sits lower, such as DMPC, POPS, and egg SM, there is either no decrease in Tm or even a small increase. This observation is consistent with categories put forward by Papahadjopoulos et al.44 for protein−membrane interactions where the proteins are adsorbed. The disparate effect that aToc has on the phase behavior of DMPC could be attributed to its unexpected location in those bilayers. The heating phase behavior observed when the natural abundance aToc was incorporated into DMPC bilayers is consistent with its location and other relationships put forward in ref 44. Furthermore, the other phospholipids exhibit the phase behavior expected for a category 2 molecular interaction, where there is partial penetration of aToc. A category 2 interaction is described as typically no deeper than the glycerol backbone of the lipids,44 which is in excellent agreement with the location of the C5-methyl label determined by the neutron diffraction, both in this study and in our past studies of DMPC and DPPC.16,23 The phase behavior of DMPC with the natural abundance aToc is consistent with the behavior exhibited by DMPC and DPPC lipids in the presence of squalane.45,46 It is known that squalane is located in the midplane of the bilayer,47 and studies by Castelli et al.45 and Simon et al.46 demonstrate that squalane has no effect on the phase behavior of DMPC and DPPC multilamellar vesicles (MLV).

Table 3. Chain-Melting Temperatures and Enthalpies from DSCa heating

cooling

ΔH (J/g)

ΔH (J/g)

sample

Tm (°C)

POPC + natural aToc + 10 mol % aToc POPE + natural aToc + 10 mol % aToc POPS + natural aToc + 10 mol % aToc SM + natural aToc + 10 mol % aToc DMPC + natural aToc + 10 mol % aToc DPPC + natural aToc + 10 mol % aToc

−4.8(3) −5.2(3)

17(6) 13(9)

−7.8(3) −5(8)

16(0) 4(6)

25.6(8) 23.7(7) 22.2(4) 10.3(2) 10.9(2) 6(9) 37.1(2) 37.5(1)

28(4) 24(5) 16(7) 37(1) 37(7) 7(6) 36(9) 15(8)

22.2(1) 19.6(6) 18.(5) 7.1(1) 7.8(2) 5(2) 34.7(3) 34.7(9)

28(7) 20(7) 10(4) 29(9) 28(3) 4(6) 36(0) 24(8)

23.32(2) 23.32(1) 20.64(3) 40.94(4) 39.14(8) 36.98(9)

35.3(2) 27.5(4) 18(8) 20.5(3) 10.1(3) 10(7)

21.64(6) 21.64(5) 18.13(3) 39.04(4) 37.2(3) 33.99(7)

29.0(2) 29(0) 21(7) 21.4(1) 11.1(0) 11(5)

Tm (°C)

a

Errors were determined through the average of multiple DSC experiments. The DSC heating and cooling rate was 2 °C/min.

Figure 7. Schematic of the aToc location in a POPC membrane as determined by three different 2H labels. The zones of aToc antioxidant action were determined in ref 16.

■

CONCLUSIONS Recently, Li et al. have discussed the rational design of naphthyridinol-based lipophilic antioxidants.48 One conclusion from their data is that an improved antioxidant must be “more reactive to peroxyl radicals than” aToc while at the same time “at least as regenerable by phase-separated (i.e., water-soluble) reductants as aToc”. Thus, the reducing group must be available to both hydrophilic and lipophilic peroxyl radicals at the same time. There is, however, no explanation on the structural level as to how these phase-separated radicals can be accessible to the same reducing group. Our results demonstrate that 10 mol % aToc does not significantly affect the structural parameters of the different bilayers studied, although it has a significant effect on their chain melting temperature. Moreover, the relative location of aToc in POPE, POPS, and egg SM bilayers is similar to that of previously reported phospholipid bilayers.16 aToc’s invariant location in bilayers with different headgroup, backbone, and tail compositions is a unique result that bolsters the notion that aToc’s biological role is manifested at the lipid−water interface of a bilayer.

lowers the cooperativity of the chain melting, causing the transition to be broadened, an effect that we also observe. The loss of cooperativity and broadening of the transition peak is often attributed to the disruption of the chain packing or headgroup water interactions. On the basis of the present neutron data, we attribute this loss of cooperativity to the latter.39 Our results, however, can still be adequately descibed by the alternative roles of aToc put forward by Traber4 and Azzi.5 The location of aToc at the lipid−water interface in different membranes has implications related to the transfer to and from the α-tocopherol transfer protein (TTP), the fatty acid binding protein for tocopherol. It is reported that such proteins bind directly to the membrane’s surface where they extract or deliver their payload. Zhang et al.40 have demonstrated that the rate of aToc transfer from TTP to different model membranes was unaltered with headgroup composition and charge. Although POPC’s Tm transition is below zero, we were able to study the main chain melting transition of the membrane with and without the presence of aToc. Past 2H NMR work by Lee and coworkers determined that water in lipid emulsions remains unfrozen well below 0 °C.41 Specifically, Hsieh and Wu believe that there are between five and six water molecules near

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. G

dx.doi.org/10.1021/la502605c | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

*E-mail: [email protected].

Acid - Differences in Lipid Matrix Properties from the Loss of One Double Bond. J. Am. Chem. Soc. 2003, 125, 6409−6421. (18) Soubias, O.; Gawrisch, K. Docosahexaenoyl chains isomerize on the subnanosecond time scale. J. Am. Chem. Soc. 2007, 129, 6678− 6679. (19) Niki, E. Role of vitamin E as a lipid-soluble peroxyl radical scavenger: in vitro and in vivo evidence. Free Radical Biol. Med. 2014, 66, 3−12. (20) Rajamoorthi, K.; Petrache, H. I.; McIntosh, T. J.; Brown, M. F. Packing and viscoelasticity of polyunsaturated ω-3 and ω-6 lipid bilayers as seen by 2H NMR and X-ray diffraction. J. Am. Chem. Soc. 2005, 27, 1576−1588. (21) Salmon, A.; Steven, W.; Dodd, G. D. W.; Beach, J. M.; Brown, M. F. Configurational Statistics of Acyl Chains in Polyunsaturated Lipid Bilayers from 2H NMR. J. Am. Chem. Soc. 1987, 109, 2600− 2609. (22) Petrache, H. I.; Salmon, A.; Brown, M. F. Structural Properties of Docosahexaenoyl Phospholipid Bilayers Investigated by Solid-State 2 H NMR Spectroscopy. J. Am. Chem. Soc. 2001, 123, 12611−12622. (23) Marquardt, D.; Williams, J. A.; Kinnun, J. J.; Kučerka, N.; Atkinson, J.; Wassall, S. R.; Katsaras, J.; Harroun, T. A. Dimyristoyl phosphatidylcholine: A remarkable exception to α-tocopherol’s membrane presence. J. Am. Chem. Soc. 2014, 136, 203−210. (24) Ekiel, I. H.; Hughes, L.; Burton, G. W.; Jovall, P. A.; Ingold, K. U.; Smith, I. C. P. Structure and Dynamics of α-tocopherol in Model Membranes and in Solution: A Broad-Line and High-Resolution NMR Study. Biochemistry 1988, 27, 1432−1440. (25) Tristram-Nagle, S. A. In Methods in Membrane Lipids; Dopico, A. M., Ed.; Methods in Molecular Biology; Humana Press: Totowa, NJ, 2007; Vol. 400; Chapter 5, pp 63−75. (26) Harroun, T. A.; Katsaras, J.; Wassall, S. R. Cholesterol Is Found To Reside in the Center of a Polyunsaturated Lipid Membrane. Biochemistry 2008, 47, 7090−7096. (27) Kucerka, N.; Marquardt, D.; Harroun, T. A.; Nieh, M.-P.; Wassall, S. R.; de Jong, D. H.; Schafer, L. V.; Marrink, S. J.; Katsaras, J. Cholesterol in Bilayers with PUFA Chains: Doping with DMPC or POPC Results in Sterol Reorientation and Membrane-Domain Formation. Biochemistry 2010, 49, 7485−7493. (28) Hamilton, W. C. Significance tests on the crystallographic R factor. Acta Crystallogr. 1965, 18, 502−510. (29) Rand, R. P.; Parsegian, V. A. Hydration forces between phospholipid bilayers. Biochim. Biophys. Acta 1989, 988, 351−376. (30) Pan, J.; Cheng, X.; Monticelli, L.; Heberle, F. A.; Kucerka, N.; Tieleman, D. P.; Katsaras, J. The molecular structure of a phosphatidylserine bilayer determined by scattering and molecular dynamics simulations. Soft Matter 2014, 10, 3716−3725. (31) Bartels, T.; Lankalapalli, R. S.; Bittman, R.; Beyer, K.; Brown, M. F. Raftlike Mixtures of Sphingomyelin and Cholesterol Investigated by Solid-State 2H NMR Spectroscopy. J. Am. Chem. Soc. 2008, 130, 14521−14532. (32) Maulik, P. R.; Shipley, G. G. N-Palmitoyl Sphingomyelin Bilayers: Structure and Interactions with Cholesterol and Dipalmitoylphosphatidylcholine. Biochemistry 1996, 35, 8025−8034. (33) Caffrey, M.; Hogan, J. LIPIDAT: A database of lipid phase transition temperatures and enthalpy changes. {DMPC} data subset analysis. Chem. Phys. Lipids 1992, 61, 1−109. (34) Pili, B.; Bourgaux, C.; Meneau, F.; Couvreur, P.; Ollivon, M. Interaction of an anticancer drug, gemcitabine, with phospholipid bilayers. J. Therm. Anal. Calorim. 2009, 98, 19−28. (35) Epand, R. M.; Bottega, R. Determination of the phase behaviour of phosphatidylethanolamine admixed with other lipids and the effects of calcium chloride: implications for protein kinase C regulation. Biochim. Biophys. Acta, Biomembr. 1988, 944, 144−154. (36) Jimenez-Monreal, A. M.; Villalain, J.; Aranda, F. J.; GomezFernandez, J. C. The phase behavior of aqueous dispersions of unsaturated mixtures of diacylglycerols and phospholipids. Biochim. Biophys. Acta, Biomembr. 1998, 1373, 209−219.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS D.M. thanks Mark Frampton for many fruitful discussions. We acknowledge the Canadian Neutron Beam Centre (CNBC, Chalk River, ON) for providing generous amounts of neutron beam time. D.M. is supported by an NSERC Vanier Canada Graduate Scholarship. T.A.H. is partially supported by the National Science and Engineering Research Council of Canada (NSERC). J.K. is supported through the Department of Energy (DOE) Scientific User Facilities Division, Office of Basic Energy Sciences (contract no. DEAC05-00OR2275).

■

REFERENCES

(1) Evans, H. M.; Bishop, K. S. On The Existence of a Hitherto Unrecognized Dietary Factor Essential for Reproduction. Science 1922, 56, 650−651. (2) Gohil, K.; Vash, V. T.; Cross, C. E. Dietary α-tocopherol and neuromuscular health: Search for optimal dose and molecular mechanisms continues! Mol. Nutr. Food Res. 2010, 54, 693−709. (3) Traber, M. G.; Stevens, J. F. Vitamins C and E: Beneficial effects from a mechanistic perspective. Free Radical Biol. Med. 2011, 51, 1000−1013. (4) Traber, M. G. Vitamin E Regulatory Mechanism. Annu. Rev. Nutr. 2007, 27, 347−362. (5) Azzi, A. Molecular mechanism of α-tocopherol action. Free Radical Biol. Med. 2007, 43, 16−21. (6) Tappel, A. Vitamin-E as the biological lipid antioxidant. Vitam. Horm. 1962, 20, 493−510. (7) Niki, E.; Traber, M. G. A History of Vitamin E. Ann. Nutr. Metab. 2012, 61, 207−12. (8) Kucerka, N.; Nieh, M.-P.; Katsaras, J. Fluid phase lipid areas and bilayer thicknesses of commonly used phosphatidylcholines as a function of temperature. Biochim. Biophys. Acta 2011, 1808, 2761− 2771. (9) Vance, J. E.; Tasseva, G. Formation and function of phosphatidylserine and phosphatidylethanolamine in mammalian cells. Biochim. Biophys. Acta, Mol. Cell Biol. Lipids 2013, 1831, 543− 554. (10) Vance, J. E.; Steenbergen, R. Metabolism and functions of phosphatidylserine. Prog. Lipid Res. 2005, 44, 207−234. (11) Huang, J.; Feigenson, G. A Microscopic Interaction Model of Maximum Solubility of Cholesterol in Lipid Bilayers. Biophys. J. 1999, 76, 2142−2157. (12) Huang, J.; Buboltz, J. T.; Feigenson, G. W. Maximum solubility of cholesterol in phosphatidylcholine and phosphatidylethanolamine bilayers. Biochim. Biophys. Acta, Biomembr. 1999, 1417, 89−100. (13) Milhas, D.; Clarke, C. J.; Hannun, Y. A. Sphingomyelin metabolism at the plasma membrane: Implications for bioactive sphingolipids. FEBS Lett. 2010, 584, 1887−1894. (14) Heberle, F. A.; Petruzielo, R. S.; Pan, J.; Drazba, P.; Kučerka, N.; Standaert, R. F.; Feigenson, G. W.; Katsaras, J. Bilayer Thickness Mismatch Controls Domain Size in Model Membranes. J. Am. Chem. Soc. 2013, 135, 6853−6859. (15) Armstrong, C. L.; Marquardt, D.; Dies, H.; Kucerka, N.; Yamani, Z.; Harroun, T. A.; Katsaras, J.; Shi, A.-C.; Rheinstädter, M. C. The Observation of Highly Ordered Domains in Membranes with Cholesterol. PLoS One 2013, 8. (16) Marquardt, D.; Williams, J. A.; Kučerka, N.; Atkinson, J.; Wassall, S. R.; Katsaras, J.; Harroun, T. A. Tocopherol Activity Correlates with Its Location in a Membrane: A New Perspective on the Antioxidant Vitamin E. J. Am. Chem. Soc. 2013, 135, 7523−7533. (17) Eldho, N. V.; Feller, S. E.; Tristram-Nagle, S.; Polozov, I. V.; Gawrisch, K. Polyunsaturated Docosahexaenoic vs Docosapentaenoic H

dx.doi.org/10.1021/la502605c | Langmuir XXXX, XXX, XXX−XXX

Langmuir

Article

(37) Quinn, P. Is the distribution of α-tocopherol in membranes consistent with its putative functions? Biochemistry (Moscow) 2004, 69, 58−66. (38) Qin, S.-S.; Yu, Z.-W. Structural and Kinetic Properties of αTocopherol in Phospholipid Bilayers, a Molecular Dynamics Simulation Study. J. Phys. Chem. B 2009, 113, 16537−16546. (39) Rozycka-Roszak, B.; Przyczyna, A. Interaction between Ndodecyl-N,N-dimethyl-N-benzylammonium halides and phosphatidylcholine bilayers-the effect of counterions. Chem. Phys. Lipids 2003, 123, 209−221. (40) Zhang, W. X.; Frahm, G.; Danny Manor, S. M.; Atkinson, J. Effect of Bilayer Phospholipid Composition and Curvature on Ligand Transfer by the alpha-Tocopherol Transfer Protein. Lipids 2009, 44, 631−641. (41) Lee, D.-K.; Kwon, B. S.; Ramamoorthy, A. Freezing Point Depression of Water in Phospholipid Membranes: A Solid-State NMR Study. Langmuir 2008, 24, 13598−13604. (42) Hsieh, C.-H.; Wu, W.-G. Structure and dynamics of primary hydration shell of phosphatidylcholine bilayers at subzero temperatures. Biophys. J. 1996, 71, 3278−3287. (43) Atkinson, J.; Epand, R. F.; Epand, R. M. Tocopherols and tocotrienols in membranes: A critical review. Free Radical Biol. Med. 2008, 44, 739?764. (44) Papahadjopoulos, D.; Moscarello, M.; adn, T.; Isac, E. E. Effects of proteins on thermotropic phase transitions of phospholipid membranes. Biochim. Biophys. Acta 1975, 401, 317−335. (45) Castelli, F.; Sarpietro, M. G.; Micieli, D.; Stella, B.; Rocco, F.; Cattel, L. Enhancement of gemcitabine affinity for biomembranes by conjugation with squalene: Differential scanning calorimetry and Langmuir-Blodgett studies using biomembrane models. J. Colloid Interface Sci. 2007, 316, 43−52. (46) Simon, S.; Lis, L.; MacDonald, R.; Kauffman, J. The noneffect of a large linear hydrocarbon, squalene, on the phosphatidylcholine packing structure. Biophys. J. 1977, 19, 83−90. (47) Hauβ, T.; Dante, S.; Dencher, N. A.; Haines, T. H. Squalane is in the midplane of the lipid bilayer: implications for its function as a proton permeability barrier. Biochim. Biophys. Acta, Bioenerg. 2002, 1556, 149−154. (48) Li, B.; Harjani, J. R.; Cormier, N. S.; Madarati, H.; Atkinson, J.; Cosa, G.; Pratt, D. A. Besting Vitamin E: Sidechain Substitution is Key to the Reactivity of Naphthyridinol Antioxidants in Lipid Bilayers. J. Am. Chem. Soc. 2013, 135, 1394−1405.

I

dx.doi.org/10.1021/la502605c | Langmuir XXXX, XXX, XXX−XXX