Abstract
Drop-coating deposition Raman (DCDR) spectroscopy was employed to study liposome suspensions. The method is based on a specific drying process on the hydrophobic surface that efficiently accumulates the macromolecular sample in a ring of the edge of the dried drop. We studied liposome suspensions purchased from two sources (Avanti Polar Lipids, Inc. and Sigma-Aldrich, Co.) and prepared under different conditions. Structure of the dried drop substantially depends on the lipid concentration, lipid composition of the sample, and used solvent. Optimal lipid concentration is about 0.3 mg/ml in all cases, asolectin and DSPC suspensions form compact dried drops when dissolved in water and phosphate buffer, respectively. Drying process of the sample drop does not influence the initial phase state (gel or liquid-crystalline) of the studied liposomes excepting DSPC from Sigma-Aldrich, Co.
1. Introduction
Liposomes as spherical sacs formed by curved lipid bilayer are often used as a model system of biological membranes. Raman spectroscopy, suitable noninvasive technique for their studies, requires high concentration and often also large volume of sample. Low sensitivity of classical Raman technique can be overcome by drop-coating deposition Raman (DCDR) spectroscopy based on deposition of a small volume of the sample solution (units of μL) on a special hydrophobic surface [1]. Deposited sample evaporates with the “coffee ring effect” leading to formation of a ring in edge part of the droplet containing majority of the sample [2]. The sample in this ring is highly concentrated and gives reproducible Raman signal. DCDR technique allowed measuring Raman spectra of low-concentrated biomolecules (e.g., peptides about 0.01 mg/mL [3]). In our previous studies [4, 5] we demonstrated potential of DCDR technique for liposome studies. Raman spectra of liposomes in the initial solution concentration about 0.3 mg/mL can be obtained, and drying process does not change their initial phase state [5]. Moreover, separation of different part of the sample (free and in complex) can serve to study interaction of liposomes with bimolecular complexes [4]. On the other hand, we found out that various preparation parameters can influence the drying process and consequently acquired DCDR spectra.
In this paper we studied two different liposomes: first formed by a synthetic lipid (1,2-distearoyl-sn-glycero-3-phosphocholine, DSPC) and second by a lipid extract (asolectin from soybean). We focused our attention to the influence of liposome preparation procedure, namely lipid concentration, used solvent (water versus buffer) and lipid sources (Avanti Polar Lipids, Inc. versus Sigma-Aldrich, Co.) on obtained DCDR spectra.
2. Experimental
Synthetic phospholipid, 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC), and one natural extract asolectin, from soybean, a mix of different lipids (lecithin, cephalin, phosphatidylinositol and other phospholipids) were purchased from two suppliers: Avanti Polar Lipids, Inc. and Sigma-Aldrich, Co.. Samples of liposome suspensions were prepared upon basic protocol from [6]. Details of the preparation will be discussed furthermore. A phosphate buffer (5 mM, pH 7.4) and deionized water were used as solvents. Original suspensions (1 mg/mL) were threefold diluted in a solvent for the DCDR spectroscopy and concentrated at ≈50 mg/mL for classical Raman spectroscopy from solution. In a case of DSPC dissolved in phosphate buffer, the diluted suspension was dialyzed by the Millipore “drop dialysis” technique for about 15 minutes. Millipore floating filter disc (Millipore “V” series membranes, 0.025 μm) was put on water surface and about 5 μL of sample was deposited on it. This way salts from buffer can diffuse through the membrane. 2 μL drops of the sample were subsequently taken and deposited on a DCDR plate SpectRIM (Tienta Sciences, slide with surface consisting of a polished stainless steel base coated with a thin layer of Teflon). Samples were dried at room temperature at about 40 min.
A confocal Raman microspectrometer LabRam HR800 (Horiba Jobin and Yvon) equipped with nitrogen-cooled CCD detector, He-Ne excitation laser (632.8 nm, ~2 mW), 600 grooves/mm grating, objectives 100x with and without water immersion for accumulation of spectra from suspensions and dried drops, respectively, were used. Spectra were measured at room temperature and collected in 300–3600 cm−1 spectral region with s acquisition time.
3. Results and Discussion
Our study comprised two selected lipids: one synthetic pure DSPC, often used as a model system of biological membranes and one lipid extract asolectin, mixture of different lipids, as an example of sample that is close to real biological membranes.
Effect of solvent on formation of the ring upon drying is seen on Figure 1. DSPC liposome suspension prepared in water according to standard procedure originates in formation of nonhomogenous and noncompact structure without a distinct ring. On the other hand, dried drop of liposome suspension prepared in phosphate buffer formed round structure with clearly distinguishable ring. Salt crystals of buffer remain in the centre and/or in the ring of droplets and can interfere with Raman spectra of lipids. To avoid this, the sample can be dialyzed before deposition on DCDR plate. Dialysis for about 15 minutes seems to be optimal. The dialyzed sample gives rise to dried droplets with round shapes (~1400 μm diameter) and nicely formed edge rings (about 180 μm width). Asolectin liposome suspension in contrast to DSPC one makes compact drop after drying when dissolved in water. Thus, we suggest that a condition for optimal dried drop structure strongly depends on the lipid composition.
(a)
(b)
To optimize the preparation process of sample for DCDR spectroscopy, the appropriate concentration of liposome suspension for deposition is necessary to be assessed. Optimal concentration appears to be about 0.3 mg/mL for all samples. In this case, reproducible Raman spectra of lipids from edge rings can be obtained. Higher concentrations than 0.5 mg/mL led to a thicker dried layer with Raman spectra varied according to the excitation laser beam focusing. If lower initial concentrations were used, compact dried drops were not formed. It is worth to mention that asolectin concentration of 0.4 mg/mL is only about one hundredth of that necessary to obtain good Raman spectra of the bulk liposome suspension.
Figure 2 shows typical Raman spectra of asolectin (A) and DSPC (B) purchased from Sigma-Aldrich, Co. (b) and Avanti Polar Lipids, Inc. (c) suppliers. Comparison of the spectra from liposome aqueous solution (a) and from the ring of the dried drop is demonstrated (b), (c). The concentration of samples in aqueous solutions had to be at least about 50x higher than initial ones. Spectra measured in aqueous suspension (a) contain two broad bands at ~3400 cm−1 corresponding to OH stretching vibrations of water. These bands do not occur in the spectra from dried drops so the drying processes of the samples were completed. Other spectral features show typical C–C stretching skeletal modes in 1000–1150 cm−1 and C–H stretching modes in 2800–3100 cm−1 [5, 7]. The and intensity ratios characterize phase transitions from the ordered gel phase at low temperatures to the disordered liquid-crystalline phase at high temperatures [7, 8]. Raman spectra from aqueous solutions are the same for samples from both suppliers: DSPC is in the gel phase and asolectin in the liquid-crystalline phase, respectively (see spectra (a), lipids from Sigma-Aldrich, Co.). Intensity ratios of C–H stretching bands in Raman spectra measured from dried drops (spectra (b) and (c)) clearly demonstrate that sample of DSPC coming from Avanti Polar Lipids, Inc. keeps a gel phase after droplet evaporating while the sample from Sigma-Aldrich, Co. undergoes the phase change from gel to liquid-crystalline phase. In contrast to that, dried asolectin sample from both companies keeps its liquid-crystalline phase from solution. Thus, we suggest that lipid phase of dried samples depends on lipid source (supplier, purity).
4. Conclusions
DCDR technique is efficient tool in liposome study. Samples appropriately prepared with optimal lipid concentration form round drop with a distinct ring after deposition on a DCDR plate. Very low sample concentrations are sufficient for the DCDR spectroscopy. Lipid concentrations of about 0.3–0.5 mg/mL were found optimal for DCDR method, while for the Raman spectroscopy from bulk suspension 30–80 mg/mL are often necessary.
Formation of ring substantially depends also on used solvent. Asolectin liposome suspension makes compact round drop after drying when dissolved in water. DSPC drop is not compact when prepared at the same conditions. Lipid dissolution in phosphate buffer and subsequent dialysis before deposition on DCDR plate are necessary in this case. During the evaporating process the liposome does not change the initial phase in all samples excepting DSPC from Sigma-Aldrich, Co. Thus, we suggest that the lipid phase of dried samples depends on lipid source (supplier, purity).
Spectra measured from the inner part of the ring show excellent reproducibility. Drops dried on a DCDR plate having the same properties as the samples in aqueous solutions can be used for biological membrane study. The DCDR technique is about two orders of magnitude more sensitive than the Raman spectroscopy from solution.
Acknowledgment
The support by the Czech Science Foundation (no. 208/10/0376) is gratefully acknowledged.