Oleic Acid Vesicles as Model Membranes of Protocells

Nov 24, 2014 - ABSTRACT: Oleic acid vesicles represent good models of membrane protocells that could have existed in prebiotic times. Here, we report ...
0 downloads 0 Views 6MB Size
Download PDF
Article pubs.acs.org/Langmuir

Aminosilane/Oleic Acid Vesicles as Model Membranes of Protocells Jean-Paul Douliez,*,† Vanessa Zhendre,‡ Axelle Grélard,‡ and Erick J. Dufourc‡ †

UMR 1332, biologie et pathologie du fruit, INRA, centre de Bordeaux, 33883 Villenave d’Ornon, France Institute of Chemistry and Biology of Membranes and Nano-objects, UMR 5248, CNRS, Université Bordeaux, Institut Polytechnique Bordeaux, F-33600 Pessac, France

‡

S Supporting Information *

ABSTRACT: Oleic acid vesicles represent good models of membrane protocells that could have existed in prebiotic times. Here, we report the formation, growth polymorphism, and dynamics of oleic acid spherical vesicles (1−10 μm), stable elongated vesicles (>50 μm length; 1−3 μm diameter), and chains of vesicles (pearl necklaces, >50 μm length; 1−3 μm diameter) in the presence of aminopropyl triethoxysilane and guanidine hydrochloride. These vesicles exhibit a remarkable behavior with temperature: spherical vesicles only are observed when keeping the sample at 4 °C for 2 h, and self-aggregated spherical vesicles occur upon freezing/unfreezing (−20/20 °C) samples. Rather homogeneous elongated vesicles are reformed upon heating samples at 80 °C. The phenomenon is reversible through cycles of freezing/heating or cooling/heating of the same sample. Deuterium NMR evidences a chain packing rigidity similar to that of phospholipid bilayers in cellular biomembranes. We expect these bilayered vesicles to be surrounded by a layer of aminosilane oligomers, offering a variant model for membrane protocells.

■

INTRODUCTION Lipid vesicles are of interest in drug delivery and in various domains of physical-chemistry and biophysics.1−4 Fatty acid vesicles show interesting biomimetic properties and are potentially versatile substitutes to phospholipid vesicles in materials science. They are above all valuable models of membrane protocells because fatty acids are the most simple amphiphilic molecules that were present in the prebiotic times.5 Protocells are defined as the first compartments capable of encapsulating the initial life ingredients on the early Earth.5−10 Pioneering work was developed 40 years ago with vesicles made of oleic acid/sodium oleate mixtures.6,11 For the sake of clarity, we further call such vesicles as oleic acid vesicles, although they are made of oleic acid and sodium oleate mixtures. Typically, these vesicles were built by decreasing the pH of a micellar solution of sodium oleate.6 Later, short chain (e.g., decanoic acid) or other unsaturated fatty acids were also shown to form vesicles depending on the pH.7,12 pH lowering induces the formation of fatty acid mixtures under the form of sodium carboxylate/carboxylic acid that stabilize the bilayer topology via hydrogen bonding,7 and henceforth bilayered vesicles. In the same way, pouring the sodium oleate micellar solution into a slightly basic buffer (pH 8) yields the formation of vesicles that were shown to grow with time.13−15 Later, preformed vesicles were also shown to grow upon addition of fatty acid micelles that progressively insert into the initial vesicles. This yields the formation of tubular elongated vesicles that can further divide upon mild shearing.14,16 Alkanols which were also plausibly present in the prebiotic times have been shown to allow the conversion of fatty acid micelles into vesicles.7 Again, the formation of hydrogen bonds between the alcohol group of © 2014 American Chemical Society

the alkanol and the carboxylate moiety of the fatty acid helps the self-assembly into curved membranes. All these works resulted in the formation of “isolated” vesicles that may represent a constraint in dilute prebiotic times. For instance, fusion or exchange of materials between these compartments may not be facilitated in that case.17 Recently, giant vesicles made of mixtures of phospholipids and oleic acid were shown to aggregate upon addition of cationic polymers.17 Although several aspects of this finding need further investigation, the formation of aggregated vesicles allowed suggesting a novel model of primitive cell communities named “vesicle colonies”. It may then be very attractive to find other systems capable of forming such aggregated vesicles but this has not been developed up to now. Montmorillonite has been shown to accelerate the spontaneous conversion of fatty acid micelles into vesicles.18 This clay is a silica-based material (phyllosilicate) that may catalyze the formation of RNA oligomers19 and it may have been present in large amount in the prebiotic soup. There also exist derivatives of Montmorillonite called aminoclays that may be obtained either by chemical modification of that clay or directly via chemical synthesis.20 Such materials are synthesized using aminopropyl triethoxysilane (APTES), the positively charged amine group of which could interact with fatty acids but no experiments have been reported so far. Interestingly, APTES has already been used in combination with fatty acid derivatives (via ion-pairing) in concentrated systems with the purpose of building chiral solid porous silicate materials.21 Received: March 12, 2014 Revised: November 24, 2014 Published: November 24, 2014 14717

dx.doi.org/10.1021/la503908z | Langmuir 2014, 30, 14717−14724

Langmuir

Article

quadrupolar splitting of the entire lipid chain is calculated, and a Fourier transformation is performed, leading to simulated spectra. Adjustable parameters are the individual quadrupolar splittings and their intrinsic line width that can be measured at best on experimental spectra. All measurable quadrupolar splittings are taken as input parameters, with the weight of each splitting being set according to the molecular structure (i.e., 1 for each of the CD group of the double bond, 2 for each CD2 and 3 for the terminal CD3 group). Comparison between experimental spectra and calculated spectra is made until a satisfactory superimposition is obtained.33 As a starting set of quadrupolar splitting for oleic acid chains, literature values reported for specific labeling on the oleic chain of palmitoyl-oleoylphosphatidylcholine (POPC) were used.34,35 Growing of vesicles was assessed by turbidity measurements using a Helios UV spectrometer (Thermo Electron Corp., Fischer Bioblock Scientific, Illkirch, France), the wavelength of which being set to 500 nm. Samples were poured in 1 mL cells having a 2 mm path length. The reaction was initiated by addition of GuHCl to a solution of OlNa/APTES (see above), and an aliquot of 100 μL was taken every 2−3 min and diluted 10 times prior performing measurements. Microscopy. Samples were observed by dark-field microscopy on an Eclipse E800 microscope (Nikon, Paris, France), with either a dry 40× or a water immersion 63× objective. Images were acquired with a Coolsnap HQ2 digital camera (Photometrics, Paris, France) using the Metavue 7.7.1 software (Roper Scientific, Paris, France). Phase contrast microscopy images were obtained at room temperature at a 20× magnification using an optical microscope in the phase contrast mode (Nikon Eclipse E-400, Tokyo, Japan) equipped with a 3-CCD JVC camera allowing digital images (768 × 512 pixels) to be collected. A drop of the lipid dispersion (about 20 μL) was deposited on the glass slide surface (76 × 26 × 1.1 mm3, RS, France) and covered with a cover slide (22 × 22 mm2, Menzel-Glaser, Germany). The glass slides were previously cleaned with ethanol. Epifluorescence images were acquired (at 40× magnification) with a Nikon Eclipse 800 microscope equipped with a Coolsnap HQ2 digital camera (Photometrics) using the Metavue 7.7.1 software (Roper Scientific). Fluorescence was recovered with a long pass 590 nm filter under an excitation wavelength from 510 to 560 nm.

Recently, phospholipids bearing a silicate headgroup were synthesized and shown to self-assemble into vesicles and polymerize yielding the formation of artificial-cell like systems with a ceramic surface called “cerasome”.22−25 Ion pairing, that is, mixing of molecules bearing opposite charges, has led to the well-known catanionic systems26−28 that can form vesicles under specific conditions.1−4,29 Altogether, this suggested that ion-pairing of APTES with fatty acids in diluted systems could also yield vesicles that would be good models of protocells, which is the aim of the present work. Here, we show that APTES accelerates the conversion of sodium oleate micelles into vesicles in the presence of guanidine hydrochloride (GuHCl) which is already known to strongly interact with saturated fatty acids.30,31 We also found that this system exhibits a peculiar reversible thermal behavior. We observed the formation of aggregated vesicles (colonies) after freezing/unfreezing samples that convert into isolated vesicles when samples are kept at 4 °C and transit to elongated tubular vesicles upon heating.

■

MATERIALS AND METHODS

Chemicals were obtained from Sigma-Aldrich (Paris, France) and were of the highest purity available. Oleate micelle solutions were prepared by weighting 50 mg of oleic acid in a sample tube and adding 5 mL of pure water (yielding an oleic acid concentration of 34 mM) and the appropriate volume of a 1 M stock solution of NaOH to reach equimolarity, which was obtained at a pH of about 9.5. Samples were heated to 40 °C and vigorously shaken in a vortex mixer for 15 min to ensure homogeneity. Sodium oleate (OlNa)/aminopropyl triethoxysilane (APTES) solutions were prepared by addition of a desired volume of pure APTES to a previously prepared sodium oleate micellar solution. Samples were vigorously shaken and kept at least 2 h at room temperature (20 °C) before use. Then, a desired volume of GuHCl solution (1M) was added and samples were again shaken in a vortex mixer. For epifluorescence experiments, red Nile from SigmaAldrich (Paris, France) was dispersed at a concentration of 1 mg/mL in ethanol and 50 μL of that dye dispersion was poured in 5 mL of vesicle dispersion under vigorous stirring. A sample was produced using calcein (200 μL of a 2 mg/mL calcein in water in 5 mL of OlNa micellar solution) to study the entrapment of that dye in vesicles. Once APTES added, the vesicles were allowed to form as commented above. Then, 1 mL of that dispersion was eluted on a PD10 desalting column (GE Helthcare). Eluted volumes were recovered manually in nine tubes. Vesicles were mainly eluted between 2 and 3 mL as attested by a marked increase of turbidity (visual observation). Each sample was diluted three times with ethanol to break vesicles (to prevent any interference due to the turbidity during absorption measurements). The absorption spectrum for each of the nine tubes was recorded using a Helios UV spectrometer (Thermo Electron corp., Fischer Bioblock Scientific, Illkirch, France) between 300 and 600 nm, and the intensity at 500 nm was noted. For NMR experiments, perdeuterated oleic acid (Eurisotop, Gif-surYvette, France) was used and samples were prepared in the same way as previously mentioned except that only 1 mL was produced. Samples were placed into a 100 μL ZrO2 rotor (Cortecnet, Paris, France). Deuterium NMR experiments were carried out in the static mode at 76.8 MHz for deuterium on a Bruker II Avance 500 WB spectrometer (Bruker, Wissembourg, France) with a CP-MAS dual 4 mm 1H/2H probe. Quadrupolar-echo sequences32 were performed to record time dependent signals that were Fourier transformed with 100−200 Hz Lorentzian filtering to yield wide line spectra. Pulse durations and echo time delay were respectively 2.75 and 40 μs, spectral window was 500 kHz and 4k acquisitions were accumulated with a repetition time of 1.5 s. Experiments were performed at 25 °C. Experimental (powder) NMR spectra were simulated using a Fortran routine (EJ Dufourc, unpublished): the time domain trace composed of the weighed sum of signals corresponding to each

■

RESULTS Mixtures were prepared using chemicals, the structures of which are shown Scheme 1. Experiments started with “stock” Scheme 1. Main Chemicals Used in the Studya

a

Top: oleic acid. Bottom left: guanidine hydrochloride, GuHCl. Bottom right: aminopropyl triethoxysilane, APTES. In water, depending on the pH, GuHCl can dissociate, forming guanidium and chloride ions, and the amine group of APTES can be positively charged.

micellar solutions of sodium oleate (OlNa) at a concentration of 10 mg/mL (34 mM) oleic acid in pure water. The pH of these solutions was about 10, and is determined by the nature of the mixtures since no buffer was employed. Using buffer could interfere via electrostatic interactions with either fatty acids or APTES. Besides, the pH was measured for all samples and was not shown to vary with time. At such a high pH of 10, only micelles can form, with oleic acid being fully deprotonated under the form of sodium oleate. We further used APTES since it may interact with the carboxylate group of oleic acid upon ion-pairing. Pure APTES when poured in water at concen14718

dx.doi.org/10.1021/la503908z | Langmuir 2014, 30, 14717−14724

Langmuir

Article

trations equal to those used here (50μL APTES in 5 mL water) had a pH of 10.5. We also used guanidine hydrochloride (GuHCl, see Scheme 1) that can dissociate in water into a guanidium moiety (positively charged) and Cl− and at 1 M concentration in water; it had a pH of 6.5. It is already known to strongly interact with saturated fatty acid, preventing their crystallization.30,31 First, in a “control” experiment, addition of GuHCl to a sodium oleate micellar solution yielded a slight and slow increase of turbidity that suggests the formation of vesicles. The pH of the OlNa mixture only decreased from 10 to 9.5 upon addition of GuHCl. Interestingly, if the pH of the micellar solution is decreased to 9.5 with HCl, we do not observe any formation of vesicles. For that, the pH must be decreased to a value lower than 9.6 That control experiment allows a comparative study when one further includes APTES. This aminosilane was incorporated in “stock” OlNa micellar solutions for molar ratios, R APTES/OlNa, ranging from 0.2 to 1.6. Samples were vigorously shaken for 2 min and further gently shaken for 1 h. These samples remained transparent without any increase of turbidity, but the viscosity slightly increased with time (during the first 30 min) and further decreased as seen by visual inspection, upon tube manipulation. Once APTES was added, the pH of the solutions remained equal to 10. Then, again at such a high pH, vesicles do not form. Then, GuHCl was added to the OlNa micellar solution containing various amounts of APTES and the turbidity of the samples was followed as a function of time. Again, the viscosity of the samples was seen to increase (again, during the first 30 min) and decrease with time, with a concomitant and marked increase of turbidity (see SI1, photo SI1a in the Supporting Information), suggesting the formation of vesicles. The pH of these solutions was about 10, a value slightly higher than that in the absence of APTES. This was a rather high value for the formation of vesicles in such solutions. Indeed, at that pH, pure oleic acid is known to form only micelles, being under the form of sodium oleate. Results in Figure 1 clearly show that the

Information). Because 1-decanol is not soluble in water, forming separated “oil-like” droplets, vigorous shaking was necessary to obtain a homogeneous sample. As a consequence, the kinetic of conversion between micelles and vesicles cannot be measured accurately. Decanol could also be added with pure oleic acid prior to hydration, but since the presence of the alkanol yields the formation of oleic acid vesicles without APTES, this neither allowed performing a kinetic study. Note that here GuHCl was not added showing, that the presence of APTES alone has a strong effect on the sample turbidity. Our results are then very similar to those obtained with montmorillonite.18 Doping sodium oleate micellar solution with that clay or with APTES yields a strong increase in the rate of vesicle formation. The sample turbidity indeed resulted from the conversion of OlNa micelles to vesicles as further demonstrated by both microscopy and solid state NMR. Optical microscopy showed circles (Figure 2A and B) and threadlike elongated structures (Figure 2C and D) that are very

Figure 1. Turbidity curves (obtained at room temperature) in relation with vesicle formation (see text) as a function of time after addition at time t = 0 of GuHCl to sodium oleate micelle solutions upon increasing the amounts of APTES (R = APTES/OlNa molar ratio; R = 0, R = 0.4, R = 0.8, and R = 1.3 (the R value is indicated on the curves)).

Figure 2. (A−D) Dark field images (obtained at room temperature) of dispersions made of Ol/Na/APTES/GuHCL (molar ratio 1/1/1/1). Scale bars: 10 μm. (E−G) epifluorescence images of similar vesicles labeled with red Nile. Scale bar is the same as above.

similar to vesicles observed in other systems.14 Elongated vesicles are believed to be intermediate steps in vesicle division.14,36 In the case of elongated structures, some reached a length of more than 50 μm with a diameter of about 1−3 μm. In addition, pearling of such elongated structures was also observed (Figure 2D). Micrographs recorded from a sample after a period of rest of 1 month (kept at room temperature) showed similar images, suggesting that these structures were very stable, even when observations were made after vigorous shaking of the sample. Images were identical for samples

higher the amount of APTES, the higher the increase of turbidity. Since alkanols are known to allow the conversion of fatty acid micelles into vesicles, 7 1-decanol was also incorporated in the mixtures. First, in the control experiment, addition of 1-decanol to OlNa micelles yielded a slight increase of turbidity (see SI1, photo SI1b in the ). By comparison, addition of 1-decanol to OlNa micelles containing APTES yielded a very turbid sample (SI1, photo 2 Supporting 14719

dx.doi.org/10.1021/la503908z | Langmuir 2014, 30, 14717−14724

Langmuir

Article

Figure 3. (A) Characterization by solid state 2H NMR of a 1 day rested sample made of sodium oleate (perdeuterated) in the presence of APTES and GuHCl at T = 25 °C (sample concentration: 1 mL sodium oleate (10 mg/mL Ol, 34 mM) in the presence of 10 μL APTES and 36 μL GuHCl (1 M)). (B) Simulated spectrum for palmitic acid (a saturated chain) embedded in fluid bilayer phases; data used for the simulation were those obtained from experimental data previously published39 (see also Figure 4). Double arrow stands for the quadrupolar splitting that can be measured from the distance between Pake doublets. (C) Simulated spectrum of POPC with a perdeuterated oleic chain; data used for the simulation were those obtained in the literature (see text and SI3 Table S1 in the Supporting Information). (D) Simulated spectrum of the experimental one represented in (A) (see text).

prepared with OlNa mixed with APTES first and then GuHCl or for OlNa mixed with GuHCl first and then APTES. Any attempt to prepare similar elongated structures by using HCl or even arginine (which also bears a guanidine moiety) failed. In these cases, only smaller spherical vesicles were obtained (not shown). The vesicular structure was further confirmed by epifluorescence microscopy. Red Nile is a hydrophobic fluorescent dye that can insert into the fatty acid bilayers. Figure 2E−G shows the images acquired for solutions made of Ol/Na/APTES/ GuHCL (molar ratio 1/1/1/1). Circles that were initially observed by dark field microscopy are also visible on the images, together with the pearl necklaces and elongated structures. The fluorescent dye is clearly located on the surface of the objects confirming that these dispersions are composed of vesicles. Finally, we also prepared a sample using calcein and showed that vesicles entrapped that dye (see SI2 in the Supporting Information). These vesicular solutions were further studied by solid-state NMR using perdeuterated oleic acid. This technique is powerfull for studying biomembranes.32,33,37,38 Figure 3A shows the spectrum of a given sample of OlNa/APTES/ GuHCl (see composition in the legend). One can note the presence of a small isotropic peak that accounts for the deuterated water present in natural abundance. First of all, the spectrum was broad, confirming that oleic acids are embedded in a solidlike structure as bilayers but not as oily droplets or micelles. The spectrum exhibited a shape which was very different to that observed for saturated fatty acid chains

embedded in membranes, an example of which is shown Figure 3B. It was however very similar (see Figure 3C) to that obtained for a vesicle-forming phospholipid, POPC, that bears a perdeuterated oleic chains.34 From these spectra, each doublet allows measuring the well-known quadrupolar splitting (see arrow Figure 3B) that accounts for the dynamic of a methylene group.32 The experimental spectrum for OlNa/APTES/GuHCl mixture (Figure 3A) was simulated with a homemade program that returns the quadrupolar splittings vs the carbon labeled position (Figure 3D). The simulation used literature34,35,38 data as starting parameters for the assignment of labeled carbon positions to quadrupolar splittings. They were obtained from the above-mentioned vesicle-forming phospholipid, POPC. Since it bears a perdeuterated oleic chain and that spectra are similar (Figure 3A and C), it seemed to be a good starting point. Quadrupolar splittings were adjusted until a satisfactory agreement was obtained between simulated and experimental spectrum (Figure 3A and D; SI3 Table S1 in the Supporting Information). Plotting the quadrupolar splitting as a function of labeled carbon position along the fatty acyl chain (see Figure 4) allows a minute description of the orientational ordering of CD bonds (orientation and fluidity). For comparison, we also plotted the quadrupolar splittings extracted from a spectrum of palmitic acid/lysine mixtures39 and those of POPC with the oleic chain specifically deuterated,33,34,37 both embedded in bilayer vesicles in the fluid state. The main difference between saturated and unsaturated chains was the strong decrease of splittings around positions 9−10 due to the presence of the double bond that 14720

dx.doi.org/10.1021/la503908z | Langmuir 2014, 30, 14717−14724

Langmuir

Article

indicates that either a larger amount of vesicles formed in the presence of that aminosilane or that the size and/or surface of vesicles scatters light more efficiently. Such an increase of turbidity was also observed in the presence of montmorillonite.18 The most striking result is the change in the rate of vesicle formation. Again, as previously observed for montmorillonite,18 APTES markedly increases with time the spontaneous conversion of micelles into vesicles. The reason remains difficult to explain but it is obvious that the amine moiety of APTES may interact with the carboxylate group of oleic acid. In these mixtures, one may suggest that APTES is hydrolyzed and can further condense to form Si−O−Si bonds20 yielding aminosilane oligomers. The final structure of the vesicles is probably very similar to that of cerasome.22−25 These later vesicles are formed by using a synthetic lipid on which is anchored an inorganic, silica precursor. In water, condensation of silica groups occurs forming a ceramic layer at the surface of the bilayers. We can then propose a model of vesicles based on cerasomes and our present results as shown Figure 6. Both oligomers of APTES and guanidinium groups could electrostatically be bound to the fatty acid bilayers. In addition, the guanidium groups could stabilize Si−O− if present. It is noteworthy that vesicles form upon addition of GuHCl alone (without APTES), that is, probably interacting with fatty acids, and then have a role in the formation of the bilayer structure. Note that both sides of the bilayer are concerned; that is, APTES oligomers are present inside and outside the vesicles. Our vesicles form at relatively high pH compared to previous vesicles obtained by decreasing the pH.6,7 Vesicles at such a high pH have been observed only upon mixing fatty acid micelles with alkanols,7 that act as a cosurfactant and allow formation of hydrogen bonds with the carboxylate groups of fatty acids. In other words, alkanols replace the fatty acids under their carboxylic form (COOH) that are present when the pH is decreased. Here, we suspect that guanidium groups and cationic oligomers of APTES interacting with the carboxylate groups of fatty acids stabilize the bilayer structure instead of micelles at high pH. This suggests that cationic polymers such as polylysine could also stabilize vesicles at such a high pH. NMR spectroscopy provides information on the fatty acid chain dynamics and has already been widely used to characterize lipid fluidity in biomembranes.32,33,37,38 However, most of the studies published in the literature report on saturated alkyl chains, and only few of them deal with unsaturated ones.34 The main difference between saturated and unsaturated chains is the strong decrease of quadrupolar splittings around positions 9−10 due to the presence of the double bond that shows a peculiar topology with respect to the lipid long axis (C−D bonds do not have an average 90° orientation as observed for saturated chains).34 This must not be mistaken with disorder; the cis double bond has indeed lesser degrees of freedom than methylene groups, a character that prevents close packing of chains, leading to very low temperature melting transitions. Of great interest is the fact that the order profile obtained for oleic acid/APTES/GuHCl parallels that of the oleoyl chain in POPC bilayered vesicles, in their fluid state (see Figure 4). This indicates that the physical packing of oleic chains in our vesicles is almost the same as that of the oleoyl chain bound to the glycerol moiety of phospholipids in biomembranes. In other words, oleic acid is in its fluid state in our vesicles. An additional point of interest is that some vesicles exhibit elongated shapes and pearl-like structures (Figure 2). This is

Figure 4. Orientational ordering profile: quadrupolar splittings versus the labeled carbon position from simulation of our experimental data (●), from palmitic acid lysine mixtures (○, ref 39), and from POPC multilamellar vesicles with the oleyl chain specifically 2H-labeled, at T = 25/27 °C (+, refs 34, 35, and 38). See text for details.

may account for a local increase of disorder or peculiar orientation of that deuterons group. It is strongly suspected that variations of temperature may have occurred in prebiotic times on the early Earth, and we thus investigated the thermal behavior of these systems. After 1 day resting at room temperature, one of the samples was frozen at −20 °C for 5 h and subsequently warmed up at room temperature. Microscope inspection revealed that the sample was now only constituted by spherical and aggregated vesicles; threadlike elongated vesicles were no longer observed (Figure 5). These aggregates were stable for long periods (>weeks) when samples were further kept at room temperature. We suggest calling such aggregates “protocell colonies” by analogy with bacteria and a previous work reporting the aggregation of phospholipid/fatty acid vesicles induced by cationic polymers.17 Similarly, another sample prepared at room temperature but further kept at 4 °C for 2 h showed only spherical vesicles; elongated vesicles were no longer observed (Figure 5). After heating these samples (either “fresh” or frozen/unfrozen or kept at 4 °C for 1 day) at 80 °C for 30 min and cooling back to room temperature, one observed spherical and elongated vesicles (Figure 5). Cooling or freezing/unfreezing such a sample as previously done yielded the similar observations as before, that is, only spherical and aggregated vesicles, respectively. In other words, the modification of vesicular shape, from spherical to elongated, and the formation of aggregates can be reversibly triggered with the temperature.

■

DISCUSSION For the sake of clarity, we first summarize our results as follows: (i) upon addition of GuHCl to oleate micelles, it yields vesicles but with a very low turbidity; (ii) upon addition of APTES to oleate micelles, the fatty acid remains dispersed in solution as micelles; and (iii) upon addition of GuHCl to oleate micelles that contain APTES, vesicles formed with a strong turbidity. (iv) The kinetics of vesicle formation upon addition of GuHCl is markedly different when APTES is present. Then, our results clearly show that APTES and GuHCl have a strong effect on the conversion of oleate micelles into vesicles. The marked increase of sample turbidity compared to those lacking APTES 14721

dx.doi.org/10.1021/la503908z | Langmuir 2014, 30, 14717−14724

Langmuir

Article

Figure 5. Phase contrast (left) and dark field images (right) showing the evolution of vesicle shape upon (top) freezing/unfreezing (−20 °C (2 h)/ 20 °C), (middle) cooling (4 °C, 1 h), or (bottom) heating (80 °C, 1 h). The sample is made of 5 mL of sodium oleate (10 mg/mL Ol, 34 mM) in the presence of 50 μL of APTES and 180 μL of GuHCl (1 M). Bar size: 10 μm.

already known to occur in oleic acid vesicles, upon addition of oleate micelles to pre-existing vesicles.15 It must be noted however that upon mild shearing of the samples, pearl like structures and elongated vesicles were shown to vanish yielding only spherical vesicles.15 This feature is supposed to illustrate the growing and division mechanisms of protocells. In our case, the elongated and even pearl-like structures could be observed for long time and even after vigorous shaking of the samples. The other striking result is the reversible thermal dependence of the shape of vesicles and the formation of “ultrastructures” of aggregated vesicles. The reason for which this morphological change occurs as a function of temperature remains difficult to explain. First, it is not related to the melting point of oleic acid which is about 14 °C. Moreover, decreasing the temperature below the fluid to gel phase is not known to induce aggregation. Finally, at 4 °C, vesicles are not aggregated but rather mostly

Figure 6. Schematic representation of a vesicle (left) formed by a fatty acid bilayer (middle). At high pH, the fatty acids are deprotonated forming a negatively charged surface. On the right is shown a cationic oligomer of APTES electrostatically interacting with the negatively charged surface of the fatty acid bilayer. Guanidine which is shown in its “triarm” representation (see Scheme 1) also interacts with fatty acids but also possibly with Si−O− groups. Both guanidinum groups and the cationic oligomers are expected to stabilize the bilayer structure versus that of micelles.

14722

dx.doi.org/10.1021/la503908z | Langmuir 2014, 30, 14717−14724

Langmuir

Article

Notes

spherical; that is, elongated vesicles or pearl-like structures are no longer observed. Self-aggregation of fatty acid containing vesicles has already been reported and was induced by cationic polymers.17 It had been suggested that formation of such protocell colonies could exhibit various advantages compared to isolated vesicular protocells. For instance, self-reproduction of fatty acid vesicles is a key issue in the context of protocells for the development of life and may be induced by addition of fatty acid micelles to pre-existent vesicles.5,13,15,16 However, in diluted prebiotic times, the probability that a vesicle meets fatty acid micelles could have been very low. Horizontal transfer (via fusion of vesicles) of prebiotic materials between vesicles could be a complementary mechanism of self-reproduction or for exchange of nutrients between vesicles.17 Our finding of aggregated vesicles when samples are frozen is in favor of such a mechanism because vesicles would have been in close contact for exchanging materials. Moreover, some vesicles present in aggregates may have served as a lipid reservoir (instead of free micelles) for other vesicles to grow up and further divide. Heating would have then allowed mixing of materials or maybe self-division. In summary, we have demonstrated the formation of vesicles and threadlike vesicles in mixtures of oleic acid and APTES, an aminosilane, upon addition of GuHCl. The vesicles exhibit internal bilayer packing akin to that found for phospholipid in cellular membranes. Above all, this finding may be of interest with respect to protocell models and represent one possible model for the emergence of cells. Aminosilane derivatives, but also guanidine or some analogues, may have been present in the prebiotic times and helped for the formation of such robust and polymorphic vesicles. The peculiar behavior we observed as a function of temperature is a first step to a fascinating scenario initially proposed17 for the aggregation of similar vesicles in which colonies could break and reassociate, grow, or revert to isolated vesicles. Several aspects of our findings require further investigations such as encapsulation and release of ingredients and how they could be exchanged between or inside colonies when thermal events occur and we are exploring this avenue in the future. These findings could also be of interest in materials chemistry. These vesicles could be used as primers and templates for the synthesis of silica-based capsules or hollow spheres.40 Finally, although it may be speculative, the formation of such vesicles could offer a novel membrane model for diatoms. These “cells” are known to be contained within a unique silica capping wall formed by the intracellular polymerization of silicic acid monomers which are further extruded out of the cell.41,42 Although diatoms are very complex structures, our present vesicles could also be valuable models of such cells.

■

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS J.-P.D. would like to thank Brigitte Batailler for her help during the dark field microscopy experiments and Bérénice Houinsou Houssou for the phase contrast microscopy images. Financial support from the TGIR-RMN-THC Fr3050 CNRS for conducting the research is gratefully acknowledged.

■

(1) Bhattacharya, S.; Bajaj, A. Recent Advances in Lipid Molecular Design. Curr. Opin. Chem. Biol. 2005, 9, 647−655. (2) Bhattacharya, S.; De, S. Vesicle Formation from Dimeric IonPaired Amphiphiles. Control over Vesicular Thermotropic and IonTransport Properties as a Function of Intra-amphiphilic Headgroup Separation. Langmuir 1999, 15, 3400−3410. (3) Bhattacharya, S.; De, S.; Subramanian, M. Synthesis and Vesicle Formation from Hybrid Bolaphile/Amphiphile Ion-Pairs. Evidence of Membrane Property Modulation by Molecular Design. J. Org. Chem. 1998, 63, 7640−7651. (4) Bhattacharya, S.; Haldar, S. Synthesis, Thermotropic Behavior, and Permeability Properties of Vesicular Membranes Composed of Cationic Mixed-Chain Surfactants. Langmuir 1995, 11, 4748−4757. (5) Morigaki, K.; Walde, P. Fatty Acid Vesicles. Curr. Opin. Colloid Interface Sci. 2007, 12, 75−80. (6) Gebicki, J. M.; Hicks, M. Ufasomes are Stable Particles surrounded by Unsaturated Fatty Acid Membranes. Nature 1973, 243, 232−234. (7) Apel, C. L.; Deamer, D. W.; Mautner, M. N. Self-Assembled Vesicles of Monocarboxylic Acids and Alcohols: Conditions for Stability and for the Encapsulation of Biopolymers. Biochim. Biophys. Acta 2002, 1559, 1−9. (8) Chen, I. A.; Roberts, R. W.; Szostak, J. W. The Emergence of Competition Between Model Protocells. Science 2004, 305, 1474− 1476. (9) Mann, S. The Origins of Life: Old Problems, New Chemistries. Angew. Chem., Int. Ed. 2013, 52, 155−162. (10) Meierhenrich, U. J.; Filippi, J.-J.; Meinert, C.; Vierling, P.; Dworkin, J. P. On the Origin of Primitive Cells: From Nutrient Intake to Elongation of Encapsulated Nucleotides. Angew. Chem., Int. Ed. 2010, 49, 3738−3750. (11) Hicks, M.; Gebicki, J. M. Microscopic Studies of Fatty Acid Vesicles. Chem. Phys. Lipids 1977, 20, 243−252. (12) Namani, T.; Ishikawa, T.; Morigaki, K.; Walde, P. Vesicles from Docosahexaenoic Acid. Colloids Surf., B 2007, 54, 118−123. (13) Walde, P.; Wick, R.; Fresta, M.; Mangone, A.; Luisi, P. L. Autopoietic Self-Reproduction of Fatty Acid Vesicles. J. Am. Chem. Soc. 1994, 116, 11649−11654. (14) Budin, I.; Debnath, A.; Szostak, J. W. Concentration-Driven Growth of Model Protocell Membranes. J. Am. Chem. Soc. 2013, 134, 20812−20819. (15) Chen, I. A.; Szostak, J. W. A Kinetic Study of the Growth of Fatty Acid Vesicles. Biophys. J. 2004, 87, 988−998. (16) Zhu, T. F.; Szostak, J. W. Coupled Growth and Division of Model Protocell Membranes. J. Am. Chem. Soc. 2009, 131, 5705−5713. (17) Carrara, P.; Stano, P.; Luisi, P. L. Giant Vesicles “Colonies”: A Model for Primitive Cell Communities. ChemBioChem 2012, 13, 1497−1502. (18) Hanczyc, M. M.; Fujikawa, S. M.; Szostak, J. W. Experimental Models of Primitive Cellular Compartments: Encapsulation, Growth, and Division. Science 2003, 302, 618−622. (19) Ferris, J. P.; Ertem, G. Oligomerization of Ribonucleotides on Montmorillonite: Reaction of the 5′-Phosphorimidazolide of Adenosine. Science 1992, 257, 1387−1389. (20) Mann, S.; Burkett, S. L.; Davis, S. A.; Fowler, C. E.; Mendelson, N. H.; Sims, S. D.; Walsh, D.; Whilton, N. T. Sol−Gel Synthesis of Organized Matter. Chem. Mater. 1997, 9, 2300−2310.

ASSOCIATED CONTENT

S Supporting Information *

Images showing turbidity of samples in the presence and absence of APTES, and the encapsulation of calcein. Table showing quadrapolar splittings of samples. Additional references. This material is available free of charge via the Internet at http://pubs.acs.org.

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. 14723

dx.doi.org/10.1021/la503908z | Langmuir 2014, 30, 14717−14724

Langmuir

Article

Properties of Fatty Acid/Lysine Aqueous Dispersions. Langmuir 2010, 26, 5329−5334. (40) Caruso, F.; Caruso, R. A.; Möhwald, H. Nanoengineering of Inorganic and Hybrid Hollow Spheres by Colloidal Templating. Science 1998, 282, 1111−1114. (41) Nassif, N.; Livage, J. From Diatoms to Silica-Based Biohybrids. Chem. Soc. Rev. 2011, 40, 849−859. (42) Round, F. E.; Crawford, R. M. The Diatoms. Biology and Morphology of the Genera; Cambridge University Press: Cambridge, 1990.

(21) Che, S.; Liu, Z.; Ohsuna, T.; Sakamoto, K.; Terasaki, O.; Tatsumi, T. Synthesis and Characterization of Chiral Mesoporous Silica. Nature 2004, 429, 281−284. (22) Katagiri, K.; Hashizume, M.; Ariga, K.; Terashima, T.; Kikuchi, J.-i. Preparation and Characterization of a Novel Organic−Inorganic Nanohybrid “Cerasome” Formed with a Liposomal Membrane and Silicate Surface. Chem.Eur. J. 2007, 13, 5272−5281. (23) Kikuchi, J.-i.; Yasuhara, K. Cerasomes: A New Family of Artificial Cell Membranes with Ceramic Surface. In Advances in Biomimetics; Cavrak, M., Ed.; InTech: Rijeka, Croatia, 2011; Chapter 11, pp 231−250. (24) Liang, X.; Yue, X.; Dai, Z.; Kikuchi, J.-i. Photoresponsive Liposomal Nanohybrid Cerasomes. Chem. Commun. 2011, 47, 4751− 4753. (25) Yue, X.; Dai, Z. Recent Advances in Liposomal Nanohybrid Cerasomes As Promising Drug Nanocarriers. Adv. Colloid Interface Sci. 2013, 207, 32−42. (26) Dubois, M.; Deme, B.; Gulik-Krzywicki, T.; Dedieu, J.-C.; Vautrin, C.; Desert, S.; Perez, E.; Zemb, T. Self-Assembly of Regular Hollow Icosahedra in Salt-Free Catanionic Solutions. Nature 2001, 411, 672−675. (27) Zemb, T.; Dubois, M.; Demé, B.; Gulik-Krzywicki, T. Self Assembly of Flat Nanodiscs in Salt Free Catanionic Surfactant Solutions. Science 1999, 283, 816−819. (28) Blanzat, M.; Perez, E.; Rico-Lattes, I.; Prome, D.; Lattes, A. New catanionic Glycolipids. 1. Synthesis, Characterisation and Biological Activity of Double-Chain and Gemini Catanionic Analogues of Galactosylceramide (Galβ1cer). Langmuir 1999, 15, 6163−6169. (29) Soussan, E.; Cassel, S.; Blanzat, M.; Rico-Lattes, I. Drug Delivery by Soft Matter: Matrix and Vesicular Carriers. Angew. Chem., Int. Ed. 2009, 48, 274−288. (30) Douliez, J.-P.; Houinsou-Houssou, B.; Fameau, A.-L.; Novales, B.; Gaillard, C. Self assembly of anastomosis-like superstructures in fatty acid/guanidine hydrochloride aqueous dispersions. J. Colloid Interface Sci. 2010, 341, 386−389. (31) Fameau, A.-L.; Houinsou-Houssou, B.; Ventureira, J.; Novales, B.; Navailles, L.; Nallet, F.; Douliez, J.-P. Self Assembly, Foaming, and Emulsifying Properties of Sodium Alkyl Carboxylate/Guanidine Hydrochloride Aqueous Mixtures. Langmuir 2011, 27, 4505−4513. (32) Davis, J. The Description of Membrane Lipid Conformation, Order and Dynamics by 2H-NMR. Biochim. Biophys. Acta 1983, 737, 117−171. (33) Grélard, A.; Guichard, P.; Bonnafous, P.; Marco, S.; Lambert, O.; Manin, C.; Ronzon, F.; Dufourc, E. J. Hepatitis B Subvirus Particles Display Both a Fluid Bilayer Membrane and a Strong Resistance to Freeze Drying: A Study by Solid-State NMR, Light Scattering, and Cryo-Electron Microscopy/Tomography. FASEB J. 2013, 27, 4316−4326. (34) Seelig, J.; Waespe-Sarcevic, N. Molecular Order in cis and trans Unsaturated Phospholipid Bilayers. Biochemistry 1978, 17, 3310−3315. (35) Dufourc, E. J.; Smith, I. C. P.; Jarrell, H. C. The Role of Cyclopropane Moieties in the Lipid Properties of Biological Membranes: A Deuterium NMR Structural and Dynamical Approach. Biochemistry 1984, 23, 2300−2309. (36) Chen, I. A.; Salehi-Ashtiani, K.; Szostak, J. W. RNA Catalysis in Model Protocell Vesicles. J. Am. Chem. Soc. 2005, 127, 13213−13219. (37) Uzureau, P.; Uzureau, S.; Lecordier, L.; Fontaine, F.; Tebabi, P.; Homble, F.; Grelard, A.; Zhendre, V.; Nolan, D. P.; Lins, L.; Crowet, J.-M.; Pays, A.; Felu, C.; Poelvoorde, P.; Vanhollebeke, B.; Moestrup, S. K.; Lyngso, J.; Pedersen, J. S.; Mottram, J. C.; Dufourc, E. J.; PerezMorga, D.; Pays, E. Mechanism of Trypanosoma brucei gambiense resistance to human serum. Nature 2013, 501, 430−434. (38) Dufourc, E. J.; Smith, I. C. P.; Jarrell, H. C. A 2H-NMR Analysis of Dihydrosterculoyl-Containing Lipids in Model Membranes: Structural Effects of a Cyclopropane Ring. Chem. Phys. Lipids 1983, 33, 153−177. (39) Novales, B.; Riaublanc, A.; Navailles, L.; Houinsou Houssou, B.; Gaillard, C.; Nallet, F.; Douliez, J.-P. Self-Assembly and Foaming 14724

dx.doi.org/10.1021/la503908z | Langmuir 2014, 30, 14717−14724