Analysis of the Thermotropic Phase Transitions of Liposomal Phosphatidylcholine Using Differential Scanning Fluorimetry

Abstract

Liposomes are small vesicles composed of lipid bilayers, which have been widely studied and are used in drug delivery systems (DDSs). The lipid bilayers, as two-dimensional liquid crystalline structures, show different phase states, and temperature-dependent phase transitions occur as a result of the thermotropic alteration of the physicochemical properties of the lipid bilayers, resulting in drastic changes in the morphology and dynamics of the fluctuations of the lipid bilayers. Analysis of the thermotropic phase behavior of the liposomal lipid bilayer is crucial for the development and application of functional liposomes for DDSs. We constructed a differential scanning fluorimetry (DSF) method that enabled observation and analysis of the thermotropic phase transitions and temperatures of liposomal lipid bilayers using a real-time PCR device and solvatochromic dyes, which have fluorescence characteristics that reflect alterations in the polar environment. This DSF method using Nile Red and a tandem thermal sequence enabled analysis of the phase transition temperatures of three liposomal phosphatidylcholines, and not only the T_m and T_p, but also the T_sub values, except for the T_p value of 1,2-dipalmitoyl-sn-glycero-3-phosphocholine, were clearly determined. Other solvatochromic dyes could not be used to determine the T_sub values clearly. The measured phase transition temperatures of three liposomal phosphocholines correlated well with the reported values. Our DSF method has several practical advantages over the typical thermal analytical method, differential scanning calorimetry, including reduced sample volume and analytical time, which may contribute to expanding the opportunities for the physicochemical analysis of liposomal lipid bilayers.

INTRODUCTION

Liposomes, which are formed by lipid bilayers typically containing phospholipids can be used as nanocapsules to encapsulate a wide variety of materials with various solubilities.^1–3) Liposomes have an extremely wide range of possible applications as carrier platforms that can be equipped with various functions, including reducing the rate of the metabolic elimination or degradation of the molecules of interest in vivo (stealth properties) and controlling the sites or speed of the release of the payload molecules (controlled release functions).^2,4–6) These promising features have resulted in liposomes being one of the most widely studied biomaterials in both basic research and the application of drug delivery systems (DDSs).^7,8)

The physicochemical characteristics of the liposomal lipid bilayers, including the fluidity, flexibility, and dynamics of the morphological fluctuations, are important. These properties are closely related to the phase behavior of the lipid bilayers.^9–11) Therefore, the measurement of phase transition temperatures, which are direct indicators of physicochemical properties, such as the structure, fluidity, and flexibility of lipid bilayers, is desirable in pharmaceutical research to optimize in vivo function of liposomal formulations such as the cellular uptake, the membrane fusion activity and drug release capability.^2,10)

The phase states and their transitions can be roughly classified as lyotropic and thermotropic, according to whether they occur because of changes in the components of the lipid bilayer or the temperature, respectively.^12,13) The thermotropic phases and transitions are analyzed by gradually altering the temperature of the lipid bilayer sample, and the resulting thermodynamic changes (typically endothermic reactions) that are derived from temperature-dependent structural changes in the lipid bilayer are observed using differential scanning calorimetry (DSC).^14–16) However, a drawback of thermal analysis using DSC is that the equipment is expensive and thereby it would not be necessarily available to everyone.

The most common components of liposomal formulations are phosphatidylcholines, which are neutral phospholipids that constitute the lipid bilayers of a wide range of higher organisms from plants to animals.^2–4) A diagram illustrating the thermotropic phase behavior of lipid bilayers that are formed by phosphatidylcholines is shown in Fig. 1. In phosphatidylcholines, two fatty acids and a phosphocholine group are covalently bound to the three hydroxyl groups in glycerol creating a hydrophobic acyl chain region and a hydrophilic polar head group, respectively (Supplementary Fig. S1). Typically, the two-dimensional liquid crystalline structures of lipid bilayers that are mainly formed from phosphatidylcholines have four phases: a dehydrated crystal phase (L_c); a gel phase (L_β′); a ripple phase (P_β′); and a liquid crystalline phase (L_α).^12,15,17,18) As the temperature increases, the thermal energy causes trans-gauche isomerization-like orientational changes in the fatty acid tails of the lipid molecules, and at a critical temperature that exceeds a certain thermal energetic barrier, i.e., the phase transition temperature, the topology of the entire lipid bilayer drastically changes to that of a different phase state. The temperature that causes transitions from the L_c phase to the L_β′ (S_o) phase, from the L_β′ (S_o) phase to the P_β′ phase, and from the P_β′ to the L_α (L_d) phase are the sub-, pre-, and main-transition temperatures (T_sub, T_p, and T_m), respectively¹⁹⁾ (Fig. 1). The T_m value, which has the highest thermodynamic energy, can be determined with the highest sensitivity because it is the largest signal observed via thermal analytical methods. The much less energetic T_sub and T_p values have low detection sensitivity and are difficult to observe even with DSC.^12,15,17,18)

Fig. 1. Schematic Illustration of the Thermotropic Phase Transitions of Lipid Bilayers That Consist of Typical Neutral Phospholipids, Such as Phosphatidylcholines

The L_c phase corresponds to the sub-gel phase and is also called the ‘pseudo-crystalline phase’.²⁰⁾ The L_c phase is mainly formed under low temperature conditions, where the packing between the lipid molecules is tight, the degree of hydration is relatively low, and the mobility of individual lipid molecules and the fluidity or flexibility of the lipid bilayers are restricted. When the temperature is increased, the L_c phase shows a transition to the lamellar gel phase (L_β′ phase) at a certain phase transition temperature. In the L_β′ phase, also called the ‘S_o phase’, the intermolecular packing because of hydrophobic and Van der Waals interactions between the lipid molecules is highly ordered, but the lipid bilayers are more hydrated, fluid, and flexible than in the L_c phase. If the temperature is further increased, the L_β′ (S_o) phase undergoes a transition to the ripple gel phase (P_β′ phase).^21–24) In the P_β′ phase, relatively short thin and thick regions are generated in the lipid bilayer, forming major and minor arms, respectively. The P_β′ phase is classified as a gel phase, the same as the L_β′ (S_o) phase. Some phosphatidylcholines, including 1,2-dimyristoyl-sn-glycero-3-phosphocholine (DMPC), 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC), and 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC), are known as ripple-forming lipids and can form the P_β′ phase.^12,15,18,21) The L_α phase, which is formed by a phase transition caused by further increasing the temperature, is also known as the ‘liquid disordered (L_d) phase’. In the L_α (L_d) phase of lipid bilayers, the hydrophobic or Van der Waals interactions between the lipid molecules are weaker than those in the gel phases, and therefore the thermal mobility of individual lipid molecule is high.

In experiments to observe these phase behavior of lipid bilayers, it is necessary to control the sample temperature accurately and precisely. The use of real-time PCR equipment not only enables highly accurate control of the sample temperature but also enables real-time observations of temperature-dependent changes in the fluorescence intensity of the samples. Because of the versatility of real-time PCR systems, these systems are widely used and are potentially less expensive compared with DSC. Furthermore, recent advances have been made in the development of solvatochromic fluorescent dyes, which can sensitively detect slight changes in polarity.^25–27) The solvatochromic fluorescent dye, Nile Red, is a powerful probe that has been reported to be able to sensitively reflect changes in the polarity of the hydrophobic environment within lipid bilayers^19,28) (Supplementary Fig. S2). Recently, determination of the thermotropic phase behavior of lipid bilayers has been reported to be possible using temperature-dependent fluorescence analysis with solvatochromic dyes, such as 9,10-bis(phenylethynyl)-anthracene (BPEA) and Cl-BPEA²⁹⁾ (Supplementary Fig. S2). Unlike conventional fluorescent dyes, such as Laurdan (6-dodecanoyl-2-dimethylaminonaphthalene), the fluorescence of next-generation dyes that show solvatochromism characteristics can be detected with real-time PCR equipment because their excitation and emission wavelengths can match the wavelengths used in real-time PCR equipment. Although examination of the thermal denaturing curves of samples using real-time PCR and fluorescent dyes (DSF) is relatively widely performed to measure the thermal denaturing temperature of biomolecules, such as proteins or nucleic acids, there have been no reports to date using DSF approaches for investigation of the thermotropic phase behavior and transitions of lipid bilayers.

Here, we investigated the thermotropic phase behavior and transition temperatures of lipid bilayers of liposomal phosphatidylcholine using a DSF approach using real-time PCR equipment and solvatochromic fluorescent dyes in a hybrid manner, and conducted a detailed examination and assessment of the practical aspects and remaining challenges involved in using this system.

MATERIALS AND METHODS

Materials

DMPC, DPPC, DSPC, and 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) were purchased from Avanti Polar Lipids (Birmingham, AL, U.S.A.). Nile Red, 1-chloro-9,10-bis(phenylethynyl)anthracene (Cl-BPEA), and naphtho[2,3-a]pyrene were purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). BPEA was purchased from Angene International Ltd. (Nanjing, China). Unless stated otherwise, all other materials used for this study were purchased from Sigma-Aldrich (St. Louis, MO, U.S.A.) and Nacalai Tesque, Inc. (Kyoto, Japan).

Liposome Preparation

Liposomes were prepared by an established procedure based on the Bangham method and as previously reported^29–31) in the analysis of temperature-dependent phase transitions of liposomal membranes using fluorescent dyes, and the molar ratio between the phospholipids and the fluorescent dyes was fixed at 98 : 2 mol%, according to a previous report.²⁹⁾ In brief, the lipids and a fluorescent dye were dissolved in chloroform at a molar ratio of 98 : 2 mol% in a 5 mL round-bottom glass flask and a homogeneous thin lipid cake was created on the inner-wall by removing the solvent using a rotatory evaporator. Residual chloroform was thoroughly eliminated by leaving the flask in an evacuated container overnight in the dark. The lipid cake was then completely dissolved in 0.5 mL of 5% (w/v) D-glucose solution to give concentrations of the lipids and the fluorescence dye of 1.96 and 0.04 mM, respectively. The lipid/dye mixture was hydrated by repeating cycles of heating in an 80 °C water bath followed by mechanical agitation by vortexing five times. The resulting spontaneously generated large multilamellar lipid vesicles were passed through a polycarbonate filter with a pore size of 100 nm 21 times using a mini-extruder system (Avanti Polar Lipids, AL, U.S.A.) to create small multilamellar lipid vesicles with a uniform size distribution. The total lipid concentration of the prepared liposomes was determined by the choline oxidase N-ethyl-N-(2-hydroxy-3-sulfopropyl)-3,5-dimethoxyaniline sodium salt (DAOS) method using a Phospholipids C-test Wako assay kit (FUJIFILM Corp., Tokyo, Japan). Characterization of the prepared liposomes was performed by measuring the hydrodynamic diameter, polydispersity index (PDI), and surface zeta-potential values of the liposomal particles via dynamic light scattering analysis using a nanoPartica SZ-100V2 and SZ-100 software (HORIBA, Kyoto, Japan)³²⁾ and these data are presented in Supplementary Table S1.

DSF Experiments

A 20 µL aliquot of the liposome suspension was added to the well of a hard-shell 96-well PCR plate (Bio-Rad, CA, U.S.A.). After sealing the wells of the plate with a clear optical film Microseal ‘B’ (Bio-Rad), DSF experiments were performed with CFX Opus 96 real-time PCR equipment (Bio-Rad) with the lid temperature set to off. Simple DSF measurements were performed in a similar manner to a standard thermal shift assay, a single thermal gradient from 10 to 70 °C was applied to the sample at a ramp rate of 0.5 °C/5 s and the resulting fluorescence emission of the dye in the sample was monitored in real time (Fig. 2). The tandem thermal-gradient program, as shown in Fig. 3A, was prepared by tandemly connecting two single thermal-gradient sequences for each sample. For observation of the fluorescence of Nile Red, the data were collected with the ALL-channel of the CFX Opus 96 equipment, and the resulting fluorescence emission data were obtained in the Texas Red channel. For observation of the other dyes, the data were collected and analyzed using the fluorescein (FAM) channel. All the measurements were performed in triplicate, and the resulting melting temperature data for each liposomal sample were determined as the average ± standard deviation values as shown in Table 1.

Fig. 2. Representative DSF Spectra of DMPC/Nile Red (98 : 2 mol%) Liposomes Obtained from a Single Thermal-Gradient Program

(A) A schematic diagram of the single thermal-gradient program, which was based on a typical thermal shift assay. (B) The melt curve and the melt peak spectra (the left and right panels, respectively) of DSF that were obtained using the single thermal-gradient program. The melt peak spectrum (the right panel), in which the vertical axis is the temperature-derivative values of the relative fluorescence units (−dRFU/dT), was generated by analyzing the melt curve data (the left panel). (C) The DSF spectral data as shown in panel D were collected using the same sample used for the experiment shown in panels A and B and by performing the same DSF experiment again as soon as possible after the DSF experiment shown in panels A and B. (D) The melt curve and the melt peak spectra (the left and right panels, respectively) of the DSF were obtained from the second data of the two consecutive single thermal-gradient DSF experiments (shown in panel C).

Fig. 3. Representative DSF Spectra of DMPC/Nile Red (98 : 2 mol%) Liposomes Measured Using the Tandem Thermal-Gradient Sequence

(A) Schematic diagram of the tandem thermal-gradient sequence for DSF. (B) The DSF melt curve and melt peak spectra (the left and right panels, respectively), of the liposomes measured from the first section of the tandem temperature-gradient sequence (denoted by the dashed lined open box in panel A). (C) The DSF melt curve and melt peak spectra (the left and right panels, respectively), of the liposomes measured from the second section of the tandem temperature-gradient sequence (denoted by the dotted lined open box in panel A). The asterisk in the right panel in C indicates the artifact peak that appeared regardless of the type of lipids or fluorescent dyes used.

Table 1. Values of the Phase Transition Temperatures of Liposomal Phospholipid Bilayers Measured by DSF Experiments Using Nile Red

Composition of liposomea)	Tsubb,c)	Tpb,c)	Tmb,c)
DMC	15.3 ± 0.50	14.7 ± 0.61	22.4 ± 0.32
DPPC	20.9 ± 0.83	(Not determined)d)	39.7 ± 0.81
DSPC	28.8 ± 1.06	52.7 ± 1.30	54.9 ± 1.29

a) The molar ratio between each lipid and Nile Red was 98 : 2 mol%. b) The units are °C. c) The average ± standard deviation of these data was determined by performing three independent DSF experiments using the tandem thermal-gradient program. d) The T_p values of the DPPC liposomes could not be determined because the signal was overlapped with an artifact peak.

RESULTS

DSF Analysis of Phosphatidylcholines/Nile Red Liposomes Using a Single-Gradient Thermal Program

As the acyl chain lengths of the fatty acids in a phospholipid increase, the hydrophobicity and Van der Waals interactions between the fatty acids become stronger, leading to an increase in the phase transition temperature of a lipid bilayer formed by the phospholipids.^12,18) To investigate the utility of using solvatochromic fluorescent dyes and DSF for measuring the phase transition temperatures of lipid bilayers, DSF experiments were performed using various liposomes containing Nile Red, prepared using different phosphatidylcholines, DMPC (14 : 0), DPPC (16 : 0), DSPC (18 : 0), DOPC [18 : 1 (Δ9-Cis)], which have varying acyl chain lengths and known lipid bilayer phase transition temperatures (Supplementary Fig. S1). Prior to starting the DSF experiments, we confirmed that the excitation and detection of the fluorescence emission of Nile Red were possible with an CFX Opus 96 (Bio-Rad), which is a typical real-time PCR device, by performing fluorescence spectroscopy (Supplementary Table S2).

The DSF spectrum of DMPC/Nile Red liposomes was measured by applying a single-gradient thermal program, which is a typical method for standard DSF experiments (Fig. 2A). A weak signal was observed at approximately 17 °C along with a large drift in the baseline (Fig. 2B). Then, the same DSF experiment was performed again using the same sample with as little time as possible between the two measurements (Fig. 2C). The spectral profile from the second measurement was completely different from that of the first run, and signals were observed at approximately 15 and 22 °C (Fig. 2D). From the reported T_sub, T_p, and T_m values for DMPC membranes, which are approximately 17, 15, and 24 °C, respectively,^{12,14,15,18,29,33)} it was thought that the signal corresponding to T_sub was detected in the first DSF run and the signals corresponding to T_p and T_m were detected in the second DSF run. In general, the rate of change in fluorescence intensity observed in a DSF experiment is expressed as the temperature derivative of the fluorescence intensity, dRFU/dT. In the thermal denaturing process of a target molecule, the fluorescence intensity of the dye increases as the hydrophobic interactions between the fluorescent dye and the target molecule become stronger. In a typical DSF experiment, the magnitude of the temperature derivative of the fluorescence intensity is expressed in the negative direction as −dRFU/dT. In a similar manner, the phase transition temperatures for the liposomal samples were observed as negative signals in the temperature-derivative fluorescence intensity spectra (Fig. 2).

DSF Analysis of Phosphatidylcholine/Nile Red Liposomes Using a Tandem-Gradient Thermal Sequence

On the basis of the data shown in Fig. 2, we constructed an original thermal sequence in which two single-gradient thermal programs were tandemly combined, termed a tandem-gradient thermal sequence (Fig. 3A), and DSF experiments using liposomal samples were performed. In the DSF spectra, a signal at 17 °C was observed from the first gradient section, and signals at 15 and 23 °C were observed from the second gradient section (dashed- and dotted-line sequences in Fig. 3A, respectively). From the reported T_sub, T_p, and T_m values of DMPC membranes mentioned above,^{12,14,15,18,29,33)} it was thought that the signal corresponding to T_sub was detected in the first thermal-gradient section and the signals corresponding to T_p and T_m were detected in the second thermal-gradient section in this DSF spectra (Fig. 3, Table 1). In the DSF spectrum that was obtained from the second thermal gradient section in the tandem-gradient thermal sequence, a relatively strong signal at approximately 35 °C was also observed (denoted as an asterisk in the right panel of Fig. 3C).

The DSF spectrum of DPPC/Nile Red liposomes was measured using the tandem-gradient thermal sequence. A signal at 20 °C was observed from the first gradient section, and a signal at 40 °C was observed from the second gradient section (Fig. 4). Because the reported T_sub, T_p, and T_m values for DPPC membranes are approximately 22, 35, and 41 °C, respectively,^{12,14,15,18,29,33)} it was thought that the signal corresponding to T_sub was detected in the first thermal-gradient section and the signal corresponding to T_m was detected in the second thermal-gradient section in this DSF spectrum (Fig. 4, Table 1). In the DSF spectrum that was obtained from the second thermal-gradient section in the tandem-gradient thermal sequence, similar to the results for DMPC/Nile Red shown in Fig. 3, a relatively strong signal at approximately 35 °C was observed (denoted by the asterisk in the right panel of Fig. 4B). Although it has been reported that the T_p value of DPPC membranes was approximately 35 °C as mentioned above, the unknown signal at 35 °C also appeared in other samples, not only in DMPC but also in DSPC and DOPC membranes, as will be discussed later (Figs. 2–5, Supplementary Figs. S3–S15). Therefore, we did not assign that signal as the T_p of the DPPC membrane because the T_p signal may have been overlapped with the unknown signal (Fig. 4, Table 1). Determination of the T_m values was possible because the signal intensity was clearly stronger than the signal at 35 °C although each signal was partially overlapped (Fig. 4B, Table 1).

Fig. 4. Representative DSF Spectra of the DPPC/Nile Red Liposomes Measured Using the Tandem Thermal-Gradient Sequence

(A) The DSF melt curve and melt peak spectra (the left and the right panels, respectively) of the liposomes measured from the first section of the tandem temperature-gradient sequence (denoted by the dashed lined open box in Fig. 3A). (B) The DSF melt curve and melt peak spectra (the left and the right panels, respectively) of the liposomes measured from the second section of the tandem temperature-gradient sequence (denoted by the dotted lined open box in Fig. 3A). The asterisk shown in the right panel indicates the artifact peak that appeared regardless of the type of lipids or fluorescent dyes used.

Fig. 5. Representative DSF Spectra of the DSPC/Nile Red Liposomes Measured Using the Tandem Thermal-Gradient Sequence

(A) The DSF melt curve and melt peak spectra (the left and the right panels, respectively) of the liposomes measured from the first section of the tandem temperature-gradient sequence (denoted by the dashed lined open box in Fig. 3A). (B) The DSF melt curve and melt peak spectra (the left and the right panels, respectively) of the liposomes measured from the second section of the tandem temperature-gradient sequence (denoted by the dotted lined open box in Fig. 3A). The asterisk shown in the right panel indicates the artifact peak that appeared regardless of the type of lipids or fluorescent dyes used.

Next, the DSF spectra of DSPC/Nile Red liposomes were measured using the tandem-gradient thermal sequence. In the DSF spectra, a signal at 29 °C was observed from the first gradient section, and signals at 52 and 55 °C were observed from the second gradient section (Fig. 5). The reported T_sub, T_p, and T_m values for DSPC membranes are approximately 28, 50, and 55 °C, respectively,^{12,14,15,18,29,33)} therefore, the signal corresponding to T_sub was thought to be in the first thermal-gradient section and the signals corresponding to T_p and T_m were thought to be in the second thermal-gradient section in this DSF spectrum (Fig. 5, Table 1). The unknown signal at 35 °C (denoted by the asterisk in the right panel of Fig. 5B) did not interfere with the analysis of the DSF data because it was sufficiently separated from the T_p and T_m signals of the DSPC/Nile Red liposomes.

Subsequently, the DSF spectra of DOPC/Nile Red liposomes were measured using the tandem-gradient thermal sequence. In both the first and second thermal-gradient sections, the observed DSF spectra was noisy and no reproducible signals appeared within the 10–30 °C and 40–70 °C regions (Supplementary Fig. S3). The phase transition temperatures of DOPC membranes are extremely low (the reported T_m is approximately −17 °C) and therefore DOPC membranes show high fluidity even though the fatty acid chains are the same length as those of DSPC. However, DOPC has a fatty acid tail that contains an unsaturated bond that causes a gauche conformation, resulting in steric effects that prevent the Van der Waals interactions between the lipid molecules.^10,20) The results that the signals derived from some phase transitions were not observed in the DSF spectra were consistent with the fact that the phase transition temperatures of DOPC membranes are extremely low. A relatively strong signal at approximately 35 °C was observed (denoted by the asterisk in the right panel of Supplementary Fig. S3B) similar to the results for the DMPC/Nile Red and DSPC/Nile Red liposomes shown in Figs. 3 and 5. This finding indicated that the unknown signal that appeared at approximately 35 °C was an artifact that did not reflect a phase transition of the lipid bilayers.

The correlations between the values of the phase transition temperatures that were measured in the DSF experiments and those reported from DSC experiments were analyzed (Fig. 6). The results demonstrated that the T_sub, T_p, and T_m values determined from the DSF experiments showed high agreement with the reported values for all the examined DMPC/Nile Red, DPPC/Nile Red, and DSPC/Nile Red liposomes (Fig. 6).

Fig. 6. Correlation Plots between the Measured Phase Transition Temperatures That Were Obtained from DSF Experiments Using Nile Red as a Fluorescent Dye and the Reported Values^{12,14,15,18,29,33)}

The asterisk shown on the plot for T_p (the middle panel) indicates that the T_p values of the DPPC liposomes could not be determined from DSF experiments because the signal was overlapped with an artifact peak. The error bars of the measured data were standard deviations that were calculated from three independent experiments.

Investigation of Other Solvatochromic Fluorescence Dyes for DSF

In addition to Nile Red, other potential solvatochromic fluorescent dyes, BPEA, Cl-BPEA,³¹⁾ and naphtho[2,3-a]pyrene,^26,34) which have been used for fluorescence analysis of the physicochemical properties of lipid bilayers (Supplementary Fig. S2), were also investigated as detection probes for observing lipid bilayer phase behavior in our DSF experiments. Prior to performing the DSF experiments, we confirmed that the excitation and detection of the fluorescence emission of these dyes were possible with CFX Opus 96 equipment in a similar manner to that for Nile Red as described above (Supplementary Table S2).

DSF experiments using the tandem-gradient thermal sequence were performed using PC liposomes containing Cl-BPEA. No clear signals were observed from the first thermal-gradient section for DMPC, DPPC, or DSPC liposomes (Supplementary Figs. S4A–S6A). Signals at 16 and 22 °C (DMPC liposomes), 41 °C (DPPC liposomes), and 51 and 54 °C (DSPC liposomes) were observed from the second thermal-gradient section (Supplementary Figs. S4B–S6B). A comparison of these results with the reported values for T_sub, T_p, and T_m for DMPC, DPPC, and DSPC lipid bilayers mentioned above^{12,14,15,18,29,33)} indicated that Cl-BPEA could be used to determine the T_p and T_m values, although determination of the T_sub value was difficult in the DSF experiments (Supplementary Figs. S4–S6, Supplementary Table S3).

Next, DSF experiments using the tandem-gradient thermal sequence were performed using phosphatidylcholine liposomes containing BPEA. Similar to the results for Cl-BPEA, no clear signals were observed from the first thermal-gradient section for DMPC, DPPC, or DSPC liposomes (Supplementary Figs. S8A–S10A). Signals at 17 and 22 °C (DMPC liposomes) and 43 °C (DPPC liposomes) were observed from the second thermal-gradient section (Supplementary Figs. S8B–S10B). These results demonstrated that BPEA can be used to determine the T_p and T_m values for DMPC and DPPC liposomes but not all the phase transition temperatures could be determined for DSPC liposomes, and determination of the T_sub value was difficult regardless of the type of phosphatidylcholine in the DSF experiments (Supplementary Figs. S8–S10, Supplementary Table S3).

DSF experiments using the tandem-gradient thermal sequence were also performed using phosphatidylcholine liposomes containing naphtho[2,3-a]pyrene. No clear signals were observed from the first thermal-gradient section for DMPC, DPPC, or DSPC liposomes (Supplementary Figs. S12A–S14A) similar to the results using Cl-BPEA. Signals at 40 °C (DPPC liposomes) and 50 and 54 °C (DSPC liposomes) were observed from the second thermal-gradient section (Supplementary Figs. S12B–S14B). These results indicated that naphtho[2,3-a]pyrene can be used to determine the T_p and T_m values of DPPC and DSPC liposomes but not all the phase transition temperatures could be detected using DMPC liposomes, and determination of the T_sub value was difficult regardless of the type of phosphatidylcholine in the DSF experiments (Supplementary Figs. S12–S14, Supplementary Table S3).

The correlations between the values of the phase transition temperatures that were measured in the DSF experiments using Cl-BPEA, BPEA, and naphtho[2,3-a]pyrene and the values reported from DSC experiments were analyzed (Supplementary Fig. S16). The correlation plots for T_sub are not shown because the T_sub signals could not be detected regardless of the type of phosphatidylcholine when Cl-BPEA, BPEA, or naphtho[2,3-a]pyrene were used. In particular, the T_p and T_m values determined from the DSF experiments using Cl-BPEA showed high agreement with the reported values, although the signal intensities were lower than those of Nile Red. The T_p and T_m values of DMPC and DPPC liposomes determined in the DSF experiments using BPEA showed high agreement with the reported values, although the signal intensities were further attenuated compared with those of Cl-BPEA. The T_p and T_m values of DPPC and DSPC liposomes determined in the DSF experiments using naphtho[2,3-a]pyrene showed high agreement with the reported values and the signal intensities were similar to those of BPEA.

For the DOPC liposomes analysis using BPEA, Cl-BPEA, and naphtho[2,3-a]pyrene, no reproducible signals appeared in the 10–30 °C and 40–70 °C regions in the DSF spectra from either the first or second thermal-gradient sections (Supplementary Figs. S7, S11, S15).

A relatively strong signal at approximately 35 °C (denoted by an asterisk in the right panel of Supplementary Figs. S3B–S15B) was observed in the spectra of all the examined liposomes regardless of the type of phosphatidylcholines, similar to the results for Nile Red.

DISCUSSION

In the present study, we investigated the phase transition temperatures of liposomal lipid bilayers using a DSF method with a real-time PCR device and solvatochromic fluorescent dyes. To objectively assess the validity and potential of the DSF approach, we used liposomes that had a phosphatidylcholine as the main component. Phosphatidylcholines are the most common lipid components in liposomal formulations and the phase transition temperatures of lipid bilayers formed from formulations containing phosphatidylcholines has been previously studied. As the fluorescent probes, we used Nile Red, BPEA, and Cl-BPEA which display solvatochromism and have fluorescence emissions that are detectable using a real-time PCR device, and naphtho[2,3-a]pyrene, which is a representative fluorescent dye with a pyrene structure, because this type of dye has been used to analyze the thermotropic behavior of lipid bilayers in previous studies.^26,34) We demonstrated that it was possible to determine the three phase transition temperatures of phosphatidylcholine membranes, T_sub, T_p, and T_m, by performing two consecutive DSF runs. In addition, we developed a thermal sequence that allowed the analysis of all the phase transition temperatures in one DSF run. In the first gradient section of the tandem-gradient thermal sequence, the temperature-derivative fluorescence intensity spectra showed a considerably meandering baseline with increasing sample temperatures. This result indicated that the mobility and dynamics of the lipid bilayer were increased and therefore the hydration of the membrane was increased with increasing temperatures, resulting in a decrease in the area of the hydrophobic environment surrounding the fluorescent dyes in the lipid bilayers.³⁵⁾ While the thermodynamic energy, which causes the large distortion of the baseline is high, the energy causing the phase transition from pseudo-crystalline (sub-gel) to gel phases is weak, which makes it difficult to detect the T_sub signals even with a sensitive DSC method. The present study is the first to report that T_sub signals can be clearly observed using Nile Red with DSF.

After the lipid bilayer has been hydrated during the process of the phase transition from the sub-gel to gel phase, the lipid bilayer becomes stable in the gel phase.^10,36,37) Therefore, the phase transition temperatures from the gel to the liquid disordered phases, T_p and T_m, could be measured using DSF by re-increasing the sample temperature in the second thermal-gradient section. The results of the present study suggested that the thermal history of the lipid bilayer samples is important in the measurement of the thermotropic phase transitions, which has been suggested in previous reports.^18,35,37) The tandem thermal-gradient sequence developed in the present study can be used to give lipid bilayer samples an appropriate thermal history and provide reliable measurement of thermotropic phase transitions.

The T_m signal, which was observed in the DSF spectra of the second thermal-gradient section, was also weakly present in the first thermal-gradient section of the tandem thermal-gradient sequence of the DMPC and DPPC liposomes (Figs. 3–4). Similar results were obtained to those using Nile Red with other dyes, including Cl-BPEA (Supplementary Fig. S5). It was assumed that the T_m signal could be weakly observed even in the first gradient section because of its strong signal intensity compared with that of the T_sub and T_p signals. Although the DSF spectra of liposomal samples are often noisy and cause situations in which the signal assignments are difficult, signals that could be observed in the spectra of both the first and second thermal-gradient sections would be useful to confirm the assignment of the peak of interest. In addition, it is useful to perform DSF measurements multiple times to increase the confidence of the signal assignments because effects due to noise appear randomly but genuine signals appear reproducibly at the same temperatures. The DSF spectra shown in Figs. 2–5 and Supplementary Figs. S1–S15 were representative spectra that showed peaks that appeared reproducibly in experiments repeated multiple times.

The present study revealed that the characteristics required for the detection of lipid bilayer phase behavior varied depending on the type of fluorescent dye used, and it was shown that Nile Red could be used to determine all three phase transition temperatures, T_sub, T_p, and T_m, of phosphatidylcholines with the highest signal intensity of all the dyes investigated. The possibility that Nile Red can be used to determine T_p and T_m has been previously reported in a study using DPPC multilamellar vesicles,¹⁹⁾ which supports our results. Nile Red has a structure containing dimethylamino and carbonyl groups attached to aromatic rings (Supplementary Fig. S2). This chemical structure leads to a phenomenon known as ‘intramolecular charge transfer’ (ICT), in which an electron from the lone pair of the dimethylamino group is transferred to the carbonyl group via the conjugated bonds of the aromatic rings.³⁸⁾ The efficiency of ICT in Nile Red can be sensitively altered in response to changes in the polar environment surrounding the Nile Red molecules, and that the fluorescence properties of Nile Red, such as the fluorescence wavelength and intensity, are changed accordingly.³⁸⁾ Changes in the phase state or higher-order structure of the lipid bilayer lead to changes in the polar environment of the lipid bilayer, and therefore it is thought that Nile Red can be used to detect those environmental changes more sensitively than non-ICT type dyes, such as BPEA, Cl-BPEA, and naphtho[2,3-a]pyrene. Although the structural difference between Cl-BPEA and BPEA is the presence or absence of one chloride group, Cl-BPEA resulted in a higher signal intensity than BPEA when used to determine phase transition temperatures. In addition, Cl-BPEA could be used to determine the T_p and T_m signals of DMPC, DPPC, and DSPC liposomes, whereas the T_p and T_m signals of DSPC liposomes were unable to be detected using BPEA. These results indicated that Cl-BPEA was superior to BPEA as a probe for observation of the phase transition temperatures of lipid bilayers, and this result was consistent with a previous report.²⁹⁾ Unlike BPEA, naphtho[2,3-a]pyrene was unable to be used to detect the T_p and T_m signals of DMPC liposomes. Naphtho[2,3-a]pyrene has a bulky structure and is highly hydrophobic; therefore, naphtho[2,3-a]pyrene should be relatively suitable for observing the phase behavior of DSPC liposomes, in which the hydrophobic region of the lipid bilayer is larger than that of DMPC liposomes. Thus, it was suggested that the potential for detecting the phase behavior of lipid bilayers would vary depending, not only on the structure and type of the lipid, but also on the structure and physicochemical characteristics of the fluorescent dyes. It is expected that the insights obtained from the present study will lead to the development of new functional fluorescent dyes.

Regardless of the type of lipid or dye that was used in the present study, a relatively strong signal appeared at 35 °C in the DSF spectra of the second thermal-gradient section (Figs. 2–5, Supplementary Figs. S3–S15). In many cases, this unidentifiable signal was overlapped with the T_p signal of the DPPC liposomes. It is possible that this signal appeared because of some interaction between the fluorescent dyes and the surrounding molecules. In an effort to investigate this possibility, liposomes containing no fluorescent dyes were prepared, and then the fluorescent dye was mixed with the liposomes, and this mixture was used for the DSF measurements. A weak signal was observed at 35 °C, and this signal was also weakly present when the fluorescent dyes were dissolved in a solvent that did not contain lipids (Supplementary Fig. S17). These results suggested that the signal at 35 °C was non-specifically generated from the interaction of the fluorescent dyes with solvent molecules or by the hydration of the dyes. However, the signal at 35 °C was remarkably weak compared with the situation when the dye was pre-incorporated into the liposomes as shown in Figs. 2–5. This result indicated that the signal at 35 °C was mainly generated by the hydration of the fluorescent dyes that occurred in the lipid bilayer environments, and by the complex interactions between the dyes, lipids, and water molecules in the lipid bilayer. Because DPPC is a representative neutral phospholipid that is widely used for a variety of pharmaceutical purposes and basic research, finding a way to avoid this generation of the artificial and non-specific signal at 35 °C is important for the further development of the DSF method. In the future, it is expected that the cause of the generation of the signal at 35 °C will be further elucidated, leading to the development of new dyes that do not generate artificial signals.

This study demonstrated that measurement of the thermotropic phase transitions and the phase transition temperatures of the lipid bilayers was possible by monitoring the temperature-dependent behavior of solvatochromic fluorescent dyes incorporated in the lipid bilayers. However, the molecular mechanisms by which the solvatochromic fluorescent dyes reflect the morphological changes in the lipid bilayers, and what relationship exists between the lipid bilayer phase transitions and the fluorescence fluctuation, largely remain unknown. In the future, it is expected that unraveling the molecular mechanisms will further expand the applicability of the DSF method.

In the DSF method developed in the present study, only 20 µL of a liposomal formulation was required per measurement and the data collection could be completed in approximately 1 h. By contrast, with typical single-cell type DSC equipment, approximately 300-µL samples are required, and the total machine time may be 16–20 h including the pre-warming (typically overnight) time to stabilize the device and the actual measurement time (approximately 3 h per sample). Compared with the DSC method, the DSF approach that only requires 1/15 of the amount of the sample and 1/20 of the machine time is advantageous in terms of throughput.

The DSF method demonstrated in the present study that uses a versatile real-time PCR device will expand the practicality of the investigation of the thermotropic phase behavior and transitions of lipid bilayers, and has the potential to become a useful application as an alternative or complementary approach to DSC. Further research will be needed to expand the application of this method to lipids other than phosphatidylcholine.

Acknowledgments

This work was financially supported in part by JSPS KAKENHI (Grant Numbers: 23K24017 and 24K22394 for K.S-K).

Conflict of Interest

The authors declare no conflict of interest.

Data Availability

All data collected and analyzed in this research are available within the published article.

Supplementary Materials

This article contains supplementary materials.

REFERENCES

• 1) Akbarzadeh A, Rezaei-Sadabady R, Davaran S, Joo SW, Zarghami N, Hanifehpour Y, Samiei M, Kouhi M, Nejati-Koshki K. Liposome: classification, preparation, and applications. Nanoscale Res. Lett., 8, 102 (2013).
• 2) Guimarães D, Cavaco-Paulo A, Nogueira E. Design of liposomes as drug delivery system for therapeutic applications. Int. J. Pharm., 601, 120571 (2021).
• 3) Giordani S, Marassi V, Zattoni A, Roda B, Reschiglian P. Liposomes characterization for market approval as pharmaceutical products: analytical methods, guidelines and standardized protocols. J. Pharm. Biomed. Anal., 236, 115751 (2023).
• 4) Li M, Du C, Guo N, Teng Y, Meng X, Sun H, Li S, Yu P, Galons H. Composition design and medical application of liposomes. Eur. J. Med. Chem., 164, 640–653 (2019).
• 5) Liu P, Chen G, Zhang J. A review of liposomes as a drug delivery system: current status of approved products, regulatory environments, and future perspectives. Molecules, 27, 1372 (2022).
• 6) Wen P, Ke W, Dirisala A, Toh K, Tanaka M, Li J. Stealth and pseudo-stealth nanocarriers. Adv. Drug Deliv. Rev., 198, 114895 (2023).
• 7) Sercombe L, Veerati T, Moheimani F, Wu SY, Sood AK, Hua S. Advances and challenges of liposome assisted drug delivery. Front. Pharmacol., 6, 286 (2015).
• 8) Yuba E. Development of functional liposomes by modification of stimuli-responsive materials and their biomedical applications. J. Mater. Chem. B, 8, 1093–1107 (2020).
• 9) Kraft JC, Freeling JP, Wang Z, Ho RJ. Emerging research and clinical development trends of liposome and lipid nanoparticle drug delivery systems. J. Pharm. Sci., 103, 29–52 (2014).
• 10) Danaei M, Kalantari M, Raji M, Samareh Fekri H, Saber R, Asnani GP, Mortazavi SM, Mozafari MR, Rasti B, Taheriazam A. Probing nanoliposomes using single particle analytical techniques: effect of excipients, solvents, phase transition and zeta potential. Heliyon, 4, e01088 (2018).
• 11) Chen C, Chen C, Li Y, Gu R, Yan X. Characterization of lipid-based nanomedicines at the single-particle level. Fundam. Res., 3, 488–504 (2023).
• 12) Matsuki H, Goto M, Tada K, Tamai N. Thermotropic and barotropic phase behavior of phosphatidylcholine bilayers. Int. J. Mol. Sci., 14, 2282–2302 (2013).
• 13) Chavda VP, Dyawanapelly S, Dawre S, Ferreira-Faria I, Bezbaruah R, Rani Gogoi N, Kolimi P, Dave DJ, Paiva-Santos AC, Vora LK. Lyotropic liquid crystalline phases: drug delivery and biomedical applications. Int. J. Pharm., 647, 123546 (2023).
• 14) Mabrey S, Sturtevant JM. Investigation of phase transitions of lipids and lipid mixtures by sensitivity differential scanning calorimetry. Proc. Natl. Acad. Sci. U.S.A., 73, 3862–3866 (1976).
• 15) Lewis RN, Mak N, McElhaney RN. A differential scanning calorimetric study of the thermotropic phase behavior of model membranes composed of phosphatidylcholines containing linear saturated fatty acyl chains. Biochemistry, 26, 6118–6126 (1987).
• 16) Demetzos C. Differential scanning calorimetry (DSC): a tool to study the thermal behavior of lipid bilayers and liposomal stability. J. Liposome Res., 18, 159–173 (2008).
• 17) de Vries AH, Yefimov S, Mark AE, Marrink SJ. Molecular structure of the lecithin ripple phase. Proc. Natl. Acad. Sci. U.S.A., 102, 5392–5396 (2005).
• 18) Matsuki H, Goto M, Tamai N. Membrane states of saturated glycerophospholipids: a thermodynamic study of bilayer phase transitions. Chem. Pharm. Bull., 67, 300–307 (2019).
• 19) Sot J, Gartzia-Rivero L, Banuelos J, Goni F, Alonso A. Liquid-crystalline, liquid-ordered, rippled and gel lipid bilayer phases as observed with Nile Red fluorescence. J. Mol. Liq., 363, 119874 (2022).
• 20) Robertson JD. The structure of biological membranes. Current status. Arch. Intern. Med., 129, 202–228 (1972).
• 21) Heimburg T. A model for the lipid pretransition: coupling of ripple formation with the chain-melting transition. Biophys. J., 78, 1154–1165 (2000).
• 22) Nagle JF, Tristram-Nagle S. Lipid bilayer structure. Curr. Opin. Struct. Biol., 10, 474–480 (2000).
• 23) Smith EA, Dea PK. The interdigitated gel phase in mixtures of cationic and zwitterionic phospholipids. Biophys. Chem., 196, 86–91 (2015).
• 24) Khakbaz P, Klauda JB. Investigation of phase transitions of saturated phosphocholine lipid bilayers via molecular dynamics simulations. Biochim. Biophys. Acta Biomembr., 1860, 1489–1501 (2018).
• 25) Klymchenko AS. Solvatochromic and fluorogenic dyes as environment-sensitive probes: design and biological applications. Acc. Chem. Res., 50, 366–375 (2017).
• 26) Sot J, Esnal I, Monasterio BG, León-Irra R, Niko Y, Goñi FM, Klymchenko A, Alonso A. Phase-selective staining of model and cell membranes, lipid droplets and lipoproteins with fluorescent solvatochromic pyrene probes. Biochim. Biophys. Acta Biomembr., 1863, 183470 (2021).
• 27) Klymchenko AS. Fluorescent probes for lipid membranes: from the cell surface to organelles. Acc. Chem. Res., 56, 1–12 (2023).
• 28) Teo W, Caprariello AV, Morgan ML, Luchicchi A, Schenk GJ, Joseph JT, Geurts JJG, Stys PK. Nile Red fluorescence spectroscopy reports early physicochemical changes in myelin with high sensitivity. Proc. Natl. Acad. Sci. U.S.A., 118, e2016897118 (2021).
• 29) Brkljača Z, Ćehić M, Portada T, Butumović M, Bakarić D, Crnolatac I. Monitoring lipid phase transition temperatures using fluorescent probes and temperature-dependent fluorescence spectroscopy. Dyes Pigm., 206, 110621 (2022).
• 30) Bangham AD, Horne RW. Negative staining of phospholipids and their structural modification by surface-active agents as observed in the electron microscope. J. Mol. Biol., 8, 660–668 (1964).
• 31) Takeda A, Tachibana A, Nagumo H, Sakai-Kato K. An in vitro lipid-mixing assay to investigate the fusion between small extracellular vesicles and endosome. Anal. Biochem., 669, 115130 (2023).
• 32) Endo N, Aoki C, Sugiki T, Sakai-Kato K. Quantitative lipid composition characterization of intact liposomes via ³¹P nuclear magnetic resonance spectroscopy. Anal. Sci., 40, 871–879 (2024).
• 33) Koynova R, Caffrey M. Phases and phase transitions of the phosphatidylcholines. Biochim. Biophys. Acta Rev. Biomembr., 1376, 91–145 (1998).
• 34) Niko Y, Didier P, Mely Y, Konishi G, Klymchenko AS. Bright and photostable push-pull pyrene dye visualizes lipid order variation between plasma and intracellular membranes. Sci. Rep., 6, 18870 (2016).
• 35) Chen C, Han D, Cai C, Tang X. An overview of liposome lyophilization and its future potential. J. Control. Release, 142, 299–311 (2010).
• 36) Veksli Z, Salsbury NJ, Chapman D. Physical studies of phospholipids: XII. Nuclear magnetic resonance studies of molecular motion in some pure lecithin-water systems. Biochim. Biophys. Acta Biomembr., 183, 434–446 (1969).
• 37) Lewis RN, McElhaney RN. Subgel phases of n-saturated diacylphosphatidylcholines: a Fourier-transform infrared spectroscopic study. Biochemistry, 29, 7946–7953 (1990).
• 38) Gajo C, Shchepanovska D, Jones JF, Karras G, Malakar P, Greetham GM, Hawkins OA, Jordan CJC, Curchod BFE, Oliver TAA. Nile Red fluorescence: where’s the twist? J. Phys. Chem. B, 128, 11768–11775 (2024).