Relationship between Phase Transition and Drug Release of Doxorubicin-Encapsulated Liposomes with Different Lipid Compositions Using Raman Spectroscopy

Abstract

Liposomes have a more complex structure than conventional low-molecular-weight pharmaceuticals, so there is concern that quality evaluation items will be diverse and evaluation methods will be complex. Raman spectroscopy has recently attracted attention as a Process Analytical Technology in the pharmaceutical manufacturing, and its application is expected to expand to biopharmaceuticals and other drugs with complex manufacturing in the future. We have demonstrated that the combination of probe-type Raman spectrometer and partial least squares analysis enables real-time quantification of drug inclusion rate and drug release rate from liposomes, and is useful as a new quality assessment method for liposomes. In this study, we evaluated the phase transition of drug-encapsulated liposomes and the accompanying drug release by using Raman spectroscopy. Drug-encapsulated liposomes were prepared by preparing liposomes with different cholesterol (CHOL) ratios. The phase transition and drug release of liposomes were evaluated by using Raman spectroscopy. Raman spectroscopic measurements showed that the peak intensity of the phase transition was large in systems with low CHOL ratios, while it was low in systems with high CHOL ratios. In the drug release test, a decrease in the peak intensity of the drug-derived spectra over time was observed significantly in the low CHOL ratio system compared to the high CHOL ratio system which is supposed to release drug lower due to liquid-ordered phase, suggesting that the drug release property increased in the low CHOL system. Thus, Raman spectroscopy can be used to evaluate the phase transition and the associated drug release properties of liposomes.

INTRODUCTION

Liposomes are closed vesicles with a lipid bilayer and an inner aqueous phase composed of amphiphilic lipids, such as phospholipids, and have been widely studied in many preclinical and clinical research areas.^1,2) Hydrophilic drugs can be encapsulated in the inner aqueous phase, while hydrophobic drugs can be encapsulated in the lipid bilayer.³⁾ Liposomal formulations, are already in clinical use, such as Doxil encapsulating the anticancer drug doxorubicin^4–6) and AmBisome encapsulating the antifungal drug amphotericin B.⁷⁾ Recently, lipid nanoparticles encapsulating nucleic acid drugs such as small interfering RNA have been actively studied and are attracting attention as a drug delivery system (DDS) technology that can meet unmet medical needs.^8,9) Furthermore, modification of liposome surfaces with polyethylene glycol (PEG), ligands, and antibodies is expected to improve drug efficacy and reduce side effects.¹⁰⁾ For example, PEG can be modified on the surface of liposome membranes to inhibit phagocytosis by macrophages and extend the half-life in the blood stream.

The structure and manufacturing process of such liposomes as drug delivery carriers are expected to be more complex than those of conventional formulations, making it difficult to evaluate their physicochemical properties and biological profiles.¹¹⁾

European Medicine Agency (EMA) “Data requirements for intravenous liposomal products developed with reference to an innovator liposomal product-Scientific guideline,” U.S. Food and Drug Administration (FDA) “Liposome Drug Products, Chemistry, Manufacturing, and Controls; Human Pharmacokinetics and Bioavailability; and Labeling Documentation” and the Ministry of Health, Labour and Welfare (MHLW) of Japan “Guidelines for liposomal drug products” have published documents on quality control and manufacturing of liposomal drug products.^12–14)

In the pharmaceutical process, it is necessary to take out samples during the process and use various analytical instruments to grasp the sample status during the process and determine whether the process is compliant to specification. However, most liposome drug products are sterile, and taking samples during the process is undesirable because of the risk of bacterial contamination that may occur if samples are taken during the process. Therefore, in order to achieve constant formulation quality, there is a need for analytical technology that allows operators to monitor the status of the process in real time without direct contact with samples.

Raman spectroscopy is a type of vibrational spectroscopy that offers distinct ad-vantages over other methods, such as near-IR spectroscopy (NIR) and terahertz spectroscopy. Specifically, Raman spectroscopy produces clear, well-resolved spectra and exhibits reduced sensitivity to water interference, making it a valuable analytical tool. For example, in NIR spectroscopy, strong water absorption peaks can interfere with those of other components, making accurate analysis difficult. In contrast, Raman spectra exhibit relatively weak water peaks, allowing for easier analysis of other components even in samples with high moisture content. Like NIR, Raman spectroscopy can be used for both qualitative and quantitative analysis through chemometric methods. However, whereas NIR spectra often display overlapping absorption bands, Raman spectra are generally clearer, making them more suitable for calibration curve development. Therefore, it is expected to be applied to DDS formulations such as biopharmaceuticals and liposomes, which are complex in structure and composition.¹⁵⁾

Furthermore, using a probe-type Raman spectrometer makes it possible to evaluate various physical properties in real time and in situ.^16,17) In fact, there are examples of in-line real-time monitoring using a probe Raman spectrometer as a Process Analytical Technology (PAT) tool in cell culture is in practical use. Therefore, we focused on the drug inclusion rate and drug release of the active ingredient among the quality evaluation items of liposome formulations. In general, ultracentrifugation and gel filtration chromatography followed by photometric measurement with HPLC are used to evaluate the inclusion rate and in vitro drug release characteristics, respectively, of active ingredients contained in liposomes.^18,19) However, these analytical methods make it difficult to separate the encapsulated and un-encapsulated drug components, and there are concerns about changes in drug concentration.^20,21) Although the HPLC measurement has an advantage of high sensitivity for separating and measuring other components at a time, it is impossible to evaluate liposomes in real time during the manufacturing process because of sample pretreatment. Therefore, we proposed real-time monitoring by Raman spectroscopy using a probe-type Raman spectrometer and demonstrated its usefulness as a new quality evaluation method for drug inclusion rate and drug release of liposome formulations.²²⁾

Thermodynamic properties of liposome membranes are listed as one of the quality characteristics to be considered in the “Guidelines for Liposome Preparation” established by MHLW and is useful as an indicator of the fluidity and uniformity of the lipid bilayer in liposomes, especially when the phase transition temperature is not clearly determined by the inclusion of CHOL. Thermodynamic properties of liposome membranes are a quality characteristic related to drug release, stability, and other factors.²³⁾ For liposomes that release active ingredients upon external stimuli such as temperature change, an in vitro release test method must be developed to identify the release of active ingredients from liposomes upon assumed temperature change.¹⁴⁾

Differential scanning calorimetry (DSC) is a typical method for thermodynamic characterization of liposomes; by measuring liposomes with DSC, the transition temperature from the gel phase to the liquid crystal phase and the heat of transition (transition enthalpy) are directly obtained.²⁴⁾ Phospholipids, which constitute liposomes, have a specific phase transition temperature and the fluidity of liposome membranes changes after the phase transition.²⁵⁾ Temperature-sensitive liposomes containing anticancer drugs are expected to improve tumor specificity when used in combination with hyperthermia therapy.²⁶⁾ However, DSC measurement has problems such as the necessity of sampling and low detection sensitivity for liposome analysis. In this context, the development of a new analytical method that can simultaneously and sensitively evaluate the thermodynamics and in vitro drug release properties of the active ingredient contained in liposomes is expected to facilitate the development of liposome drug formulations.

In this study, we evaluated the thermodynamic properties of the active ingredient and drug release properties of liposomes, which are considered to be particularly important quality evaluation items for various liposomes.

MATERIALS AND METHODS

Materials

The anthracycline anticancer drug doxorubicin hydrochloride (DOX) was purchased from Combi Blocks as an encapsulated drug in liposomes. Liposomal lipid dipal-mitoyl-sn-glycero-3-phosphatidylcholine (DPPC), whose phase transition temperature is reported to be 41–42 °C,²⁷⁾ and distearoyl-sn-glycero-3-phosphatidylcholine (PEG) as a membrane modifying agent for liposomes were purchased from NOF CORPORATION (Tokyo, Japan). CHOL as a membrane stabilizer was purchased from FUJIFILM Wako Pure Chemical Corp. (Osaka, Japan). Ammonium Chloride (AC) as a concentration gradient forming agent in liposome preparation, sodium chloride as a buffer for dialysis, and chloroform and methanol for thin film preparation were purchased from FUJIFILM Wako Pure Chemical Corp. 2-[4-(2-Hydroxyethyl)-1-piperazinyl]ethanesulfonic acid (HEPES) used as a buffer solution for dialysis was purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). The reagents used were those of special reagent grade. The structures of the main reagents are shown in Fig. 1.

Fig. 1. Molecular Structure of Reagents Used

(a) Dipalmitoyl phosphatidyl choline (DPPC). (b) Cholesterol (CHOL). (c) Doxorubicin hydrochloride (DOX). (d) 1,2-Distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(polyethylene glycol)-2000] (DSPE-PEG2000).

Preparation of Doxorubicin-Encapsulated Liposomes

Doxorubicin hydrochloride-encapsulated liposomes (DOX-Lp) were prepared by a remote loading method using an AC gradient^28,29) with DOX as the encapsulated drug. DPPC : CHOL = 10 : 0, 9 : 1, 8 : 2, 7 : 3, 6 : 4 (mol ratio) as liposome constituents, and PEG was added to the eggplant flask at a composition of 1 mol% of DPPC. After dissolving the liposomes in chloroform:methanol = 2 : 1 organic solvent, a thin film was formed by solvent distillation, and liposomes were formed by sonication with 350 mM AC solution. The outer phase was replaced by dialysis with 20 mM HEPES buffer at pH 7.5 for 5 h to form a pH gradient between the inner and outer phases. Then, 0.5 mg/mL of DOX was added and incubated at 50 °C for 90 min. Finally, DOX-Lp was prepared by dialysis again to remove any unincorporated drug.

Characterization of Doxorubicin-Encapsulated Liposomes

Observation of Liposome Morphology by Cryo-Transmission Electron Microscopes (Cryo-TEM)

Morphologies and internal structures of the liposomes were examined by cryo measurements using a field-emission transmission electron microscope (JEM-2100F, JEOL Ltd., Tokyo, Japan). The samples were placed on a copper − carbon grid (Cu200 mesh) hydrophilized by glow-discharge irradiation with 3 µL of liposome suspension, and the excess suspension was removed using filter paper. The samples were immediately flash-frozen in liquid ethane and cooled in liquid nitrogen. Acceleration voltage and observation magnification during the measurement were 120 kV and ×120 k, respectively. All sample manipulations up to the observation stage were conducted in liquid nitrogen.

Particle Size Evaluation of Liposomes by Dynamic Light Scattering (DLS)

The particle size distributions of the DOX-Lp used in each monitoring process were measured using a DLS instrument (Zetasizer ZS, Malvern Panalytical Ltd., Malvern, U.K.). The refractive index used was equivalent to that of water (1.33). The samples were diluted 200-fold by adding purified water to 10 µL of DOX-Lp acquired via each preparation method and then added to a 12 mm square polystyrene cell followed by measurement. All size analyses of liposome dispersions were performed using a He/Ne laser emitting at 633 nm with a laser power of 4.0 mW to evaluate the number-average particle size (APS) and dispersibility of the liposome suspension. Number-average particle sizes were acquired based on the results of three experiments. Size distribution was examined by using the polydispersity index (PDI). Liposome size and dispersion in water were analyzed at 25 °C.

Thermodynamic Evaluation of Liposome Membranes by DSC

DSC equipment (Thermo plus EVO2 DSC vesta Rigaku, Rigaku, Ltd., Tokyo, Japan) was used to detect the phase transition temperature of DOX-Lp. Approximately 15 mg of the sample was weighed in an aluminum pan (sealed system), and the temperature was measured in the range of 20–60 °C at a rate of 2 °C/min under a nitrogen flow of 50 mL/min. Thermo plus EVO (Rigaku Corp., version 2.086-1) was used for measurement and analysis.

Evaluation of DOX Inclusion Rate in Liposomes by UV-Visible Spectrophotometer

The DOX incorporation rate into liposomes was evaluated by quantifying the concentration of unincorporated DOX in the outer aqueous phase. The DOX-Lp suspension was added to a cellulose tube and dialyzed at 4 °C for 5 h to separate the unincorporated DOX from the DOX-Lp. The concentration of unincorporated DOX was calculated by measuring the absorbance of the dialysate at a wavelength of 490 nm using a UV-visible spectrophotometer (V-730, JASCO Ltd., Tokyo, Japan). The inclusion rate of DOX was calculated by dividing the administered dose of DOX by the number of doses minus the amount of unincorporated DOX.

Evaluation of Phase Transition and Drug Release Properties of Liposome Membranes of Probe-Type Raman Spectrometer

Evaluation of Phase Transitions in Liposome Membranes

Phase transition behavior was measured by Raman spectroscopy using a probe Raman spectrometer (MarqMetrix ALL-IN-ONE Raman System, MarqMetrix Ltd., Seattle, WA, U.S.A.) and subjected to liposome suspensions in dialysis tubes. The peaks used in the analysis are at 2850 and 2880 cm⁻¹, where the phase transition of lipids is observed. Conditions of the evaluation are as follow: sample: 10 mL, Exposure time: 10 s, Integration Count: 6 times, Laser Power: 450 mW, Laser Wavelength: 785 nm, Measuring temperature: 25, 30, 35, 40, 45, 50, 70 °C

Evaluation of Drug Release Properties

The drug release process of liposomes was monitored by Raman spectroscopy using the dialysis principle for DOX-Lp in each of the preparations described in “Preparation of Doxorubicin-Encapsulated Liposomes.” The Raman spectra used in the analysis were 1290 cm⁻¹ from DOX and 1299 cm⁻¹ from DPPC. Conditions of the evaluation are as follow: sample: 10 mL, Exposure time: 10 s, Integration Count: 6 times, Measurement Interval: 15 min, Measurement Time: 24 h, Laser Power: 450 mW, Laser Wavelength: 785 nm, Measuring temperature: 30, 37, 44, 50 °C

Prediction of Drug Quantitation Using Multivariate Analysis

The partial least squares (PLS) analysis was used to predict the concentration of DOX in the liposome suspension in the dialysis tubes and the pretreatment of Raman spectra during the dialysis process obtained during each of the test processes in “Evaluation of Phase Transition and Drug Release Properties of Liposome Membranes of Probe-Type Raman Spectrometer” was performed using Solo + MIA (ver. 8.6.2, Eigenvector Research Inc, Manson, WA, U.S.A.). DOX-Lp samples were fixed at a phospholipid concentration of 11.775 mg/mL and prepared according to each preparation method at DOX concentrations of 0, 0.1, 0.2, 0.3, 0.4 and 0.5 mg/mL. The Raman spectral range of 1170 to 1330 cm⁻¹ was used, as it allows for the measurement of peaks from both DPPC and DOX. Standard normal variate (SNV) was applied to the Raman spectra, followed by mean centering.³⁰⁾ The predictive performance indexes used for the model were the coefficient of determination R², the root mean square error of calibration (RMSEC), the root-mean-square error of cross-validation (RMSECV), and the root-mean-square error of prediction (RMSEP).³¹⁾ The cross-validation method (leave-one-out) was used to evaluate the regression model obtained.³²⁾ After evaluating the predictive accuracy of the regression model by calculating these values, the DOX concentration in the dialysis tubing during each dialysis monitoring process was quantified. In addition, the drug release rate was calculated using the following equation based on the quantified DOX concentrations.

  

RESULTS AND DISCUSSION

Characterization of DOX-Encapsulated Liposomes

Observation of the Morphology and Internal Structure of Liposomes by Cryo-TEM

Figure 2 shows Cryo-TEM images of DOX-Lp at various CHOL ratios prepared according to each preparation method. The DPPC : CHOL = 10 : 0 liposomes were 2 to 3 times larger in diameter than the other liposomes prepared. This was considered to be due to the fusion of liposomes caused by the lack of stability due to the absence of CHOL. In the case of DPPC : CHOL = 9 : 1 and 8 : 2, polyhedral liposomes and flat bicelles were observed. These were considered to be due to the low ratio of CHOL, as reported in previous studies.³³⁾ In the DPPC : CHOL = 7 : 3 and 6 : 4, stable spherical liposomes were obtained due to the high content of CHOL.

Fig. 2. Cryo-TEM Images of DOX-Lp with Different Lipid Compositions

(a) 10 : 0. (b) 9 : 1. (c) 8 : 2. (d) 7 : 3. (e) 6 : 4. All ratios are DPPC : CHOL (mol ratio).

Particle Size Evaluation of Liposomes by DLS

The results of particle size measurements for each preparation method are shown in Table 1. As liposome formulations, an average particle size of 50–100 nm and a PDI of 0.3 or less, which is considered monodisperse in the pharmaceutical field, are desirable from survey of commercialized liposomal products.³⁴⁾ In the DLS measurements after liposome preparation, the particle size was large for DPPC : CHOL = 10 : 0 to 8 : 2 with low CHOL ratios, and close to 50 to 100 nm for 7 : 3 and 6 : 4 ratios. The PDI values were greater than 0.5 for 10 : 0, greater than 0.3 for 9 : 1 and 8 : 2, and less than 0.3 for 7 : 3 and 6 : 4. These results correlate with the results of Cryo-TEM measurements. The results of the particle size measurements after the drug release test are shown in Table 2. For the liposomes prepared with DPPC : CHOL = 7 : 3 and 6 : 4 used in the drug release test, no significant change in particle size was observed compared to immediately after preparation. This indicates that the drug release was not due to the collapse of the liposomes, but rather due to the permeation of the drug through the liposomal membrane.

Table 1. Results of Particle Size Measurement of DOX-Lp for Each Lipid Composition by DLS

Molar ratio of DPPC : CHOL	10 : 0	9 : 1	8 : 2	7 : 3	6 : 4
Particle size (nm)	235.6	88.7	93.9	96.3	81.6
PDI	0.58	0.33	0.33	0.24	0.25

Table 2. Results of Particle Size Measurement of DOX-Lp after the Drug Release Test by DLS

Molar ratio of DPPC : CHOL	Measurement items	30 °C	37 °C	44 °C	50 °C
6 : 4	Particle size (nm)	86.36	81.18	90.04	92.31
	PDI	0.249	0.232	0.242	0.253
7 : 3	Particle size (nm)	87	80.33	95.81	85.82
	PDI	0.208	0.228	0.23	0.171

Thermodynamic Evaluation of Liposome Membranes by DSC

Figure 3 shows the DSC results of DOX-Lp at various CHOL ratios prepared according to each preparation method. An endothermic peak due to the gel-liquid crystal phase transition³⁵⁾ was observed at 43 °C in the DSC curves of DPPC : CHOL = 10 : 0 and 9 : 1. On the other hand, no endothermic peak was observed in the temperature range of 25–60 °C in the DSC curves of DPPC : CHOL = 8 : 2, 7 : 3, and 6 : 4. The phase transition phenomenon targeted in this study is the melting of hydrophobic chains of phospholipids, which cause a large endothermic peak to appear. However, as the amount of CHOL increases, the fluidity of the gel state increases and the fluidity of the liquid crystal state decreases, obscuring the phase transition phenomenon according to the amount of CHOL.³⁶⁾ This suggests that the phase transition temperature exists around 43 °C in DOX-Lp, but is difficult to observe in stable liposomes due to the influence of CHOL.

Fig. 3. All Ratios in DSC Measurements of DOX-Lp with Different Lipid Compositions Show DPPC : CHOL (Mol Ratio)

Evaluation of DOX Inclusion Rate in Liposomes by UV-Visible Spectrophotometer

The inclusion rate was calculated from the absorbance of the dialysate collected during the preparation of DOX-Lp. The results are shown in Table 3. The results were 58.5% for 10 : 0 of DPPC : CHOL; 50.3%, 9 : 1; 54.4%, 8 : 2; 97.7%, 7 : 3; and 97.8%, 6 : 4, indicating that liposomes having a globular shape formulated with 7 : 3 and 6 : 4 of DPPC : CHOL achieved the high inclusion rate close to 100%. Therefore, quantitative evaluation of drug release studies using multivariate analysis was performed toward the liposomes with 7 : 3 and 6 : 4 of DPPC : CHOL from the viewpoint of the stability of the inclusion rate.

Table 3. Drug Inclusion Rate in DOX-Lp for Each Lipid Composition by UV-Visible Absorbance Spectrophotometry

Molar ratio of DPPC : CHOL	10: 0	9 : 1	8 : 2	7 : 3	6 : 4
Drug encapsulation efficiency (%)	58.5	50.3	54.4	97.7	97.8

Evaluation of Phase Transition and Drug Release Properties of Liposome Membranes of Probe-Type Raman Spectrometer

Evaluation of Phase Transitions in Liposome Membranes

Liposomes undergo a phenomenon called phase transition, in which membrane fluidity changes with temperature variations. This phase transition is classified into two types: the pre-transition and the main transition. During the pre-transition, the solid state of the hydrophobic chains is maintained, but changes in interactions between hydrophilic head groups cause the lipid bilayer to adopt a ripple structure. The main transition, in which the hydrophobic chains melt, is observed at around 41–42 °C. In this main transition, the alkyl chains of phospholipids undergo isomerization from the trans conformation, which predominantly exists in the gel phase, to the gauche conformation, which is predominant in the liquid crystalline phase, upon heating. There are two main regions where phase transitions can be observed by Raman spectroscopy. The first is the C–H stretching vibration in the range of 2800–3000 cm⁻¹ and the second is the C–C stretching vibration in the range of 1000–1200 cm⁻¹, both of which are based on trans-gauche isomerization of alkyl chains.³⁷⁾ It is known that the 2880 cm⁻¹ vibration is due to the C–H asymmetric stretching vibration, and when the temperature is raised above the phase transition temperature, the symmetric vibration of the methylene group at 2850 cm⁻¹ and the symmetric vibration of the methyl group at 2935 cm⁻¹ increase. In this study, we focused on 2880 and 2850 cm⁻¹, where the changes are most clearly visible. To compare the fluidity of liposome membranes, we evaluated membrane fluidity by plotting the ratio of the peak intensity at 2850 cm⁻¹ derived from the Gauche conformation to that at 2880 cm⁻¹ derived from the trans conformation (I_2850/2880) versus each temperature.³⁵⁾ Figure 4 shows the results of membrane fluidity evaluation. As shown in Fig. 4, I_2850/2880 increased rapidly with increasing temperature at low CHOL ratios of 10 : 0 to 8 : 2, while at high CHOL ratios of 6 : 4 and 7 : 3, I_2850/2880 increased slowly with increasing temperature. In other words, the trans-gauche isomerization associated with this phase transition may have caused a change in the C–H stretching vibration mode in the gel and liquid crystal phases due to the rotation of the C–C bond in the hydrophobic hydrocarbon chain of the phospholipid. It has been reported that the addition of high concentrations of CHOL to liposome membranes composed of saturated lipids generates a new domain called the liquid-ordered (Lo) phase, which exhibits fluidity intermediate between the gel and liquid crystal phases.³⁸⁾ Below the phase transition temperature, the gel phase and the liquid-ordered phase coexist, while above the phase transition temperature, the liquid crystalline phase and the liquid-ordered phase coexist. In other words, the liquid-ordered phase increases fluidity in the gel state and decreases it in the liquid crystalline state, resulting in a more gradual phase transition. Therefore, the gradual phase transition observed at higher CHOL ratios (6 : 4 and 7 : 3) is considered to be due to the presence of the liquid-ordered phase. As described above, Raman spectroscopy can observe the phase transition phenomena of phospholipids even when CHOL is abundant and reduction of the extreme physical property changes in membrane fluidity associated with the gel-liquid crystal phase transition due to increase in CHOL ratio.

Fig. 4. Flowability Evaluation of Lipid Bilayers

All measurement ratios are DPPC : CHOL (mol ratio).

Evaluation of Drug Release Properties

Figure 5 shows the changes in Raman spectra of drug release over time for each of the DPPC : CHOL = 7 : 3 and 6 : 4 at various temperatures by Raman spectroscopy. The spectra used in the release tests were evaluated and quantified using for the spectra used in the emission test, the peak of CH₂ torsional vibration³⁷⁾ that appeared at 1299 cm⁻¹ was used for DPPC. For doxorubicin, the O–H in-plane angular vibration peak at 1209 cm⁻¹ and the C–OH and NH₂ bending vibration-derived³⁹⁾ peaks were used. These peaks were used for Raman spectral evaluation and quantitative evaluation of the drug in the drug release test. The measurement results showed that the intensity of the doxorubicin-derived peak at 1209 cm⁻¹ decreased at 44 °C, above the phase transition temperature as the temperature increased for both 7 : 3 and 6 : 4, as compared to that of the lipid-derived peak at 1299 cm⁻¹. It was also observed that the peak intensity was half as strong beyond 24 h as compared to the 0 h. In other words, differences in drug release were observed around 41 °C, the phase transition temperature of DPPC.

Fig. 5. Raman Spectra of Drug Release at Various Measurement Temperatures

(a) 6 : 4 at 30 °C. (b) 6 : 4 at 37 °C. (c) 6 : 4 at 44 °C. (d) 6 : 4 at 50 °C. (e) 7 : 3 at 30 °C. (f) 7 : 3 at 37 °C. (g) 7 : 3 at 44 °C. (h) 7 : 3 at 50 °C. All ratios are DPPC : CHOL (mol ratio).

Prediction of Drug Quantitation Using Multivariate Analysis

Regression Modeling Using PLS

To quantitatively evaluate the DOX concentration from the Raman measurements in each monitoring study, a PLS regression model was developed using the method shown in Prediction of Drug Quantitation Using Multivariate Analysis (Fig. 6). Figure 6 shows the R² values, RMSECV and RMSEP values for each regression model. When the latent variable was 2, the regression models with DPPC : CHOL = 7 : 3 and 6 : 4 both showed R² > 0.995, RMSECV <0.0196 mg/mL and RMSEP <0.0160 mg/mL, indicating the best accuracy from the RMSEP values. Additionally, we examined the limit of detection (LOD) and limit of quantification (LOQ). For DPPC : CHOL = 6 : 4, the LOD was 0.0223 mg/mL and the LOQ was 0.0670 mg/mL. For DPPC : CHOL = 7 : 3, the LOD was 0.0254 mg/mL and the LOQ was 0.0760 mg/mL. These results demonstrate that the minimum detectable and quantifiable amounts using the probe-type Raman spectrometer are both sufficiently accurate. The regression models with very high prediction accuracy were successfully generated from these results for each preparation method. The loading plots were shown in the Fig. 7. Factor 1 showed high similarity compared to the Raman spectrum of DOX, suggesting that Factor 1 reflects DOX. Furthermore, the cumulative contribution of Factor 1 exceeded 85% in both cases. These results indicate that the PLS regression model developed in this study was sufficiently accurate for quantitative evaluation of DOX.

Fig. 6. PLS Regression Model Used to Quantify Drug Release Rates

(a) 7 : 3. (b) 6 : 4. All ratios are DPPC : CHOL (mol ratio).

Fig. 7. Loading Plots for Each Regression Model

(a) 7 : 3. (b) 6 : 4. All ratios are DPPC : CHOL (mol ratio).

Determination of Drug Release Rate by PLS

We have previously shown that a combination of Raman spectroscopy and PLS analysis, a multivariate analysis method, can provide quantitative results comparable to those of conventional evaluation methods.³⁴⁾ Therefore, we calculated the drug release rate up to 24 h after the drug release test at DPPC : CHOL = 7 : 3 and 6 : 4. The results are shown in Fig. 8. As in the Raman spectral evaluation, the first point shows that the drug is released due to the phase transition of the lipid membrane. In other words, the drug was released at 44 °C and 50 °C in both systems. Second, above the phase transition temperature, the drug is released faster in the 7 : 3 system, which has a low CHOL ratio. The increase in drug release can be attributed to changes in the membrane due to phase transitions. Normally, in the gel phase, the alkyl chains of liposome membranes are regularly aligned, resulting in low fluidity and reduced drug release. However, the fluidity of liposomes increased due to the disorderly arrangement of lipid alkyl chains in the liposome membrane following the gel-liquid crystal phase transition, which was presumably responsible for the increased release of DOX. We also considered that the release of drug was reduced in the 6 : 4 phase compared to the 7 : 3 phase because the fluidity of the membrane decreases as the CHOL ratio increases and the Lo phase is reached. These results allow us to observe the release behavior of different liposomes with different lipid compositions and different phase transitions. These results show a high correlation with the detection of the phase transition temperature of liposomes by DSC, and also show a change in drug release depending on the lipid composition. Therefore, Raman spectroscopy can be used to evaluate the thermodynamic properties and drug release of liposome formulations more easily.

Fig. 8. Quantitative Evaluation of Drug Release Rate Using PLS Regression Model

(a) 7 : 3. (b) 6 : 4. All ratios are DPPC : CHOL (mol ratio).

CONCLUSION

This study shows that Raman spectroscopy can be used to comprehensively evaluate the quality characteristics of liposome formulations. Specifically, by monitoring the liposome drug release test process and calculating the intensity ratio of lipid alkyl chains at each temperature, it was found that the phase transition of liposomes and the drug release of the active ingredient can be evaluated simultaneously. The Raman spectroscopic method can also be used to indirectly evaluate the changes in vibrational modes associated with the phase transition of liposome membranes and is expected to be a useful tool that can overcome the problems of detection sensitivity and scale-up that have been issues with DSC, a method for evaluating thermodynamic properties. In addition, the use of a probe Raman spectrometer eliminates the need for sampling during measurement and is expected to be a PAT tool that can be used in continuous production. This is expected to be a new characterization method that can replace conventional analytical methods and will contribute to the advancement of liposome formulation development.

Acknowledgments

The authors gratefully acknowledge Dr. Kenjirou Higashi and Mr. Taiki Fujimoto of the Graduate School of Pharmaceutical Sciences, Chiba University, for technical assistance with the Cryo-TEM experiments and helpful discussions.

Conflict of Interest

The authors declare no conflict of interest.

REFERENCES

• 1) Bangham AD, Standish MM, Watkins JC. Diffusion of univalent ions across the lamellae of swollen phospholipids. J. Mol. Biol., 13, 238–252 (1965).
• 2) Sessa G, Weissmann G. Incorporation of lysozyme into liposomes. J. Biol. Chem., 245, 3295–3301 (1970).
• 3) Pavelić Ž, Škalko-Basnet N, Jalšenjak I. Characterisation and in vitro evaluation of bioadhesive liposome gels for local therapy of vaginitis. Int. J. Pharm., 301, 140–148 (2005).
• 4) Symon Z, Peyser A, Tzemach D, Lyass O, Sucher E, Shezen E, Gabizon A. Selective delivery of doxorubicin to patients with breast carcinoma metastases by stealth liposomes. Cancer, 86, 72–78 (1999).
• 5) Perez AT, Domenech GH, Frankel C, Vogel CL. Pegylated liposomal doxorubicin (Doxil^®) for metastatic breast cancer: the Cancer Research Network, Inc., experience. Cancer Invest., 20 (Suppl. 2), 22–29 (2002).
• 6) O’Shaughnessy JA. Pegylated liposomal doxorubicin in the treatment of breast cancer. Clin. Breast Cancer, 4, 318–328 (2003).
• 7) Adler-Moore J. AmBisome targeting to fungal infections. Bone Marrow Transplant., 14 (Suppl. 5), S3–S7 (1994).
• 8) Niu NK, Yin JJ, Yang YX, Wang ZL, Zhou ZW, He ZX, Chen XW, Zhang X, Duan W, Yang T, Zhou SF. Novel targeting of PEGylated liposomes for codelivery of TGF-β1 siRNA and four antitubercular drugs to human macrophages for the treatment of mycobacterial infection: a quantitative proteomic study. Drug Des. Devel. Ther., 9, 4441–4470 (2015).
• 9) Tamaru M, Akita H, Nakatani T, Kajimoto K, Sato Y, Hatakeyama H, Harashima H. Application of apolipoprotein E-modified liposomal nanoparticles as a carrier for delivering DNA and nucleic acid in the brain. Int. J. Nanomedicine, 9, 4267–4276 (2014).
• 10) Lin C, Zhang X, Chen H, Bian Z, Zhang G, Riaz MK, Tyagi D, Lin G, Zhang Y, Wang J, Lu A, Yang Z. Dual-ligand modified liposomes provide effective local targeted delivery of lung-cancer drug by antibody and tumor lineage-homing cell-penetrating peptide. Drug Deliv., 25, 256–266 (2018).
• 11) Parot J, Caputo F, Mehn D, Hackley VA, Calzolai L. Physical characterization of liposomal drug formulations using multi-detector asymmetrical-flow field flow fractionation. J. Control. Release, 320, 495–510 (2020).
• 12) European Medicines Agency. “Reflection paper on the data requirements for intravenous liposomal products developed with reference to an innovator liposomal product. No. 806058, February 21, 2013.”: ‹https://www.ema.europa.eu/en/documents/scientific-guideline/draft-reflection-paper-data-requirements-intravenous-liposomal-products-developed-reference-innovator-liposomal-product_en.pdf›, accessed 29 January, 2025.
• 13) Zhou Y, Gong XJ, Yang JB. Introduction to the guidance for industry on liposome drug products: chemistry, manufacturing, and controls; human pharmacokinetics and bioavailability; and labeling documentation issued by FDA. Chinese J. New Drugs, 27, 1835–1840 (2018).
• 14) MHLW. Guideline for the Development of Liposome Drug Products. March 2016, MHLW, Japan, pp. 1–30 (2016).
• 15) Weber D, Hubbuch J. Raman spectroscopy as a process analytical technology to investigate biopharmaceutical freeze concentration processes. Biotechnol. Bioeng., 118, 4708–4719 (2021).
• 16) Inoue M, Hisada H, Koide T, Carriere J, Heyler R, Fukami T. Real-time formation monitoring of cocrystals with different stoichiometries using probe-type low-frequency Raman spectroscopy. Ind. Eng. Chem. Res., 56, 12693–12697 (2017).
• 17) Nomura K, Titapiwatanakun V, Hisada H, Koide T, Fukami T. In situ monitoring of the crystalline state of active pharmaceutical ingredients during high-shear wet granulation using a low-frequency Raman probe. Eur. J. Pharm. Biopharm., 147, 1–9 (2020).
• 18) Mehn D, Capomaccio R, Gioria S, Gilliland D, Calzolai L. Analytical ultracentrifugation for measuring drug distribution of doxorubicin loaded liposomes in human serum. J. Nanopart. Res., 22, 158–164 (2020).
• 19) Gven A, Ortiz M, Constanti M, O’Sullivan CK. Rapid and efficient method for the size separation of homogeneous fluorescein-encapsulating liposomes. J. Liposome Res., 19, 148–154 (2009).
• 20) Faya M, Hazzah HA, Omolo CA, Agrawal N, Maji R, Walvekar P, Mocktar C, Nkambule B, Rambharose S, Albericio F, de la Torre BG, Govender T. Novel formulation of antimicrobial peptides enhances antimicrobial activity against methicillin-resistant Staphylococcus aureus (MRSA). Amino Acids, 52, 1439–1457 (2020).
• 21) Yang Z, Tian L, Liu J, Huang G. Construction and evaluation in vitro and in vivo of tedizolid phosphate loaded cationic liposomes. J. Liposome Res., 28, 322–330 (2018).
• 22) Sato T, Haneishi K, Hisada H, Fujii MY, Koide T, Fukami T. Real-time quantitative evaluation of a drug during liposome preparation using a probe-type Raman spectrometer. Langmuir, 40, 7962–7973 (2024).
• 23) Sakai-Kato K. Overview of analytical techniques for evaluation of nano-DDS formulations. Yakugaku Zasshi, 139, 255–260 (2019).
• 24) Janiak MJ, Small DM, Shipley GG. Nature of the thermal pretransition of synthetic phospholipids: dimyristoyl- and dipalmitoyllecithin. Biochemistry, 15, 4575–4580 (1976).
• 25) Lentz BR, Barenholz Y, Thompson TE. Fluorescence depolarization studies of phase transitions and fluidity in phospholipid bilayers. 1. Single component phosphatidylcholine liposomes. Biochemistry, 15, 4521–4528 (1976).
• 26) Deng Z, Xiao Y, Pan M, Li F, Duan W, Meng L, Liu X, Yan F, Zheng H. Hyperthermia-triggered drug delivery from iRGD-modified temperature-sensitive liposomes enhances the anti-tumor efficacy using high intensity focused ultrasound. J. Control. Release, 243, 333–341 (2016).
• 27) Gardikis K, Hatziantoniou S, Viras K, Wagner M, Demetzos CA. DSC and Raman spectroscopy study on the effect of PAMAM dendrimer on DPPC model lipid membranes. Int. J. Pharm., 318, 118–123 (2006).
• 28) Liu C, Shi J, Dai Q, Yin X, Zhang X, Zheng A. In-vitro and in-vivo evaluation of ciprofloxacin liposomes for pulmonary administration. Drug Dev. Ind. Pharm., 41, 272–278 (2015).
• 29) Du Y, He W, Xia Q, Zhou W, Yao C, Li X. Thioether Phosphatidylcholine liposomes: a novel ROS-responsive platform for drug delivery. ACS Appl. Mater. Interfaces, 11, 37411–37420 (2019).
• 30) Rinnan Å, van den Berg F, Engelsen SB. Review of the most common pre-processing techniques for near-infrared spectra. TrAC-Trends Analyt. Chem., 28, 1201–1222 (2009).
• 31) Westad F, Marini F. Validation of chemometric models-A tutorial. Anal. Chim. Acta, 893, 14–24 (2015).
• 32) Stone M. Cross-validatory choice and assessment of statistical predictions. J. R. Stat. Soc. Series B Stat. Methodol., 36, 111–133 (1974).
• 33) Cheng CY, Lai YF, Hsieh YL, Wu CH, Chiu CC, Yang YM. Chiu C cheng, Yang YM. Divergent effects of cholesterol on the structure and fluidity of liposome and catanionic vesicle membranes. FEBS Lett., 596, 1827–1838 (2022).
• 34) Danaei M, Dehghankhold M, Ataei S, Hasanzadeh Davarani F, Javanmard R, Dokhani A, Khorasani S, Mozafari MR. Impact of particle size and polydispersity index on the clinical applications of lipidic nanocarrier systems. Pharmaceutics, 10, 1–17 (2018).
• 35) Gardikis K, Hatziantoniou S, Viras K, Demetzos C. Effect of a bioactive curcumin derivative on DPPC membrane: A DSC and Raman spectroscopy study. Thermochim. Acta, 447, 1–4 (2006).
• 36) Kawakami K. Molecular assemblies composed of phospholipids for pharmaceutical formulations. J. Jpn. Soc. Colour Mater., 89, 238–243 (2016).
• 37) Kitt JP, Bryce DA, Harris JM. Calorimetry-derived composition vectors to resolve component Raman spectra in phospholipid phase transitions. Appl. Spectrosc., 70, 1165–1175 (2016).
• 38) Kitayama H, Takechi Y, Tamai N, Matsuki H, Yomota C, Saito H. Thermotropic phase behavior of hydrogenated soybean phosphatidylcholine- cholesterol binary liposome membrane. Chem. Pharm. Bull., 62, 58–63 (2014).
• 39) Das G, Nicastri A, Coluccio ML, Gentile F, Candeloro P, Cojoc G, Liberale C, De Angelis F, Di Fabrizio EFT-IR. Raman, RRS measurements and DFT calculation for doxorubicin. Microsc. Res. Tech., 73, 991–995 (2010).