Effect of Cholesterol on Nano-Structural Alteration of Light-Activatable Liposomes via Laser Irradiation: Small Angle Neutron Scattering Study

Abstract

Although the light-activated liposomes have been extensively studied for drug delivery applications, the fundamental mechanism of the drug release based on lipid compositions has not been fully understood. Especially, despite the extensive use of cholesterol in the lipid composition, the role of cholesterol in the light-activated drug release has not been studied. In this study, the influence of cholesterol on drug release from light-responsive drug-encapsulated liposomes after activated by near infrared (NIR) laser was investigated. We prepared methotrexate (MTX)-encapsulated DSPC liposomes consisting of 0 mol% (−Chol) or 35 mol% cholesterol (+Chol), with (+Au) or without gold nanorods (−Au) on the lipid bilayer to compare drug release, morphological changes, and nanostructures after laser irradiations. Transmission electron microscopy (TEM) and small angel neutron scattering (SANS) data revealed that only +Chol +Au liposomes showed partial aggregation of the liposomes after laser irradiation. Similar trends on the drug release and structural change were observed when the liposomes were heated to above chain-transition temperature. Overall, we have found that (1) inclusion of 35 mol% cholesterol enhanced the permeability of lipid bilayers above T_c; (2) the mechanism of laser-activated liposomal drug delivery is disrupting lipid bilayer membranes by the photothermal effect in the presence of plasmonic materials. By understanding the fundamentals of the technology, precise controlled drug release at a targeted site with great stability and repeatability is anticipated.

Graphical Abstract

1.2 1. Introduction

Liposomes have been proposed as carriers for drug delivery since 1970s¹. Liposomes are one of the most widely used drug delivery materials because of inherent biocompatibility, low toxicity and ability to encapsulate both hydrophilic and hydrophobic drugs. Liposomes have been produced by various methods, including probe-sonication,² extrusion,³ reverse-phase evaporation,^{4, 5} and recently developed supercritical fluid method.^6,7 Because passive delivery of encapsulated drugs from liposomes has often limited the efficacy,⁸ liposomes have been recently engineered to release encapsulated drugs on the target of interest in a controlled fashion in response to external stimuli such as pH, light, and high temperature.⁹

In recent years, laser-activatable liposomes with gold nanoparticles as a plasmonic photothermal agent have been investigated extensively for on-demand drug release^{10 11 12 13}. This type of liposomes offered a novel method to control drug release at the targeted sites with the laser source as a precise switch.¹⁴ Although the light-activated liposomes have been extensively studied because of its advantage, few applications have been tested in clinical trials and only one is approved by FDA.¹⁵ The use of light-activated liposomal drug delivery systems will be widely applied if the amount of drug is precisely controlled at a target location with great stability. In order to achieve the goal of translating the technology to clinical settings, the fundamentals of the light-activated liposomal drug delivery need to be understood based on the lipid compositions and mechanisms. The drug release mechanism is considered as the following. When the liposomes triggered by laser, the lipid membrane is locally heated to above the chain-transition temperature via plasmon surface resonance of gold nanoparticles. During this process, the permeability of the local or whole lipid bilayer may change to allow the encapsulated drug to diffuse through the membrane. For the lipid composition, in addition of phospholipids, cholesterol has been studied and utilized in liposomes for drug release, mainly regarding its ability to stabilize the formulations.¹⁶ However, effect of cholesterol in light-activatable liposomes on controlling drug release remains unclear.

In this study, we will examine the role of cholesterol in drug release as well as morphological changes using transmission electron microscopy (TEM) and small-angle neutron scattering (SANS) analysis to identify the fundamentals. Previously we fabricated gold nanorod (AuNR)-coated liposomes containing 35 mol% cholesterol (+Chol +Au) and showed that these liposomes released drug repetitively with multiple laser cycles.^{17 18} Because cholesterol also plays a role in adjusting lipid bilayer permeability,^{19 20 21 22} we hypothesize that if the drug release mechanism of light-activatable liposomes is the permeability change by heating, cholesterol will impact drug release amount. In addition, the morphological changes in the nano-scale may indicate the reorganization of the lipids in the bilayer after “chain-melting” via the photothermal effect. Lastly, we studied the effect of bulk heating to 75 °C (above T_c) on drug release and structural changes, to compare with the irradiated liposomes, which are spatially heated. Thus, this study will advance fundamental understandings in how cholesterol affects the lipid membranes and tunes drug release of light-responsive liposomes upon laser irradiation. The results will also support drug release mechanism and promote applications of laser-controlled nanomedicine.

1.3 2. Materials and Methods

1.3.1 2.1. Materials

Cholesterol was purchased from Sigma-Aldrich (St. Louis, MO). 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC) and 1,2-Distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(polyethylene glycol)-5000] (DSPE-PEG 5000) were obtained from Avanti Lipids ((Alabaster, AL). Potassium chloride and deuterium oxide (D₂O, 99.8 atom %D) were purchased from Fisher Scientific (Waltham, MA). The rest of the materials used in this study was mentioned in our previous publication.¹⁷ An aqueous citrate-capped gold nanorod dispersion with average size of 10 nm × 59 nm was purchased from Nanopartz, Inc. (Loveland, CO).

1.3.2 2.2. Synthesis of light-activatable liposomes

Briefly, the AuNR-attached liposomes were fabricated by a probe-sonication method as previously described¹⁷. For preparing liposomal formulations with and without cholesterol, DSPC, cholesterol, stearylamine and DSPE-PEG 5000 were dissolved in chloroform and mixed at two molar ratios, 50:35:10:5 (+Chol) and 85:0:10:5 (−Chol), respectively. 10 μmol of total lipids amount were kept constant for each mixture. After lipids were completely dried overnight in a fume hood at room temperature, 2 mL of D₂O was added and sonicated for 10 min × 3 cycles to form liposomes. D₂O was used to have contrast variation between the solvent/core and the liposomes bilayer in the EQ-SANS experiments. The liposome suspension was then centrifuged twice, at 4000 rpm × 30 min followed by 6000 rpm × 20 min, to increase the final lipid concentration to 84.6 mg/mL. AuNRs (4.7×10¹³/mL) were tethered on the liposomal membranes via electrostatic attractions by mixing the AuNR suspension with the liposome suspension.

To prepare drug-loaded liposomes for laser-triggered release tests, liposomes were prepared via the same method with methotrexate (MTX) solution (2 mL, 25 mg/mL dissolved in DI water) substituted for D₂O. Unencapsulated MTX was removed by a dialysis device prior to drug releasing tests.¹⁷

2.3. Dynamic light scattering (DLS) and Zeta potential

The average liposome/nanodroplet size and size distribution was determined using dynamic light scattering (NanoBrook Omni, Brookhaven Instruments, Holtsville, NY, USA). 20 μL of gold nanorod-coated liposomes was diluted 50 times in DI water and measurements were repeated three times for a period of 60 seconds each at room temperature at a reading angle of 90°. The zeta potential was measured using the same instrument with an electrode provided in the same condition as the DLS measurement.

1.3.3 2.4. Transmission Electron Microscope (TEM)

The TEM images of the liposomes before and after laser irradiations were obtained as described previously^{17, 18}. Briefly, liposomal samples were treated with 0 sec or 20 sec laser irradiations. Subsequently, 10× diluted samples were applied on a formvar/carbon-coated grid and negatively stained with 2% uranyl acetate. Then the samples on the TEM grid were lyophilized in a FreeZone Freeze Dry system (Kansas City, MO) and immediately imaged.

1.3.4 2.5. ImageJ Analysis

TEM images were analyzed by ImageJ software²³ to estimate the sizes and polydispersity of AuNR, Au spheres and liposomes. For AuNR or Au spheres, the sizes are measured manually. After the scale was set for each TEM image, each Au nanoparticle was selected and measured three times to get an average value of the sizes. For liposomes, the area measurements were done by both automatic counting and manual counting. First, the TEM images were converted to binary images after adjusting the threshold. After selecting a proper size cutoff and circularity (above 0.2), each disjoint liposome in the image can be automatically outlined and measured using the command of Analyze Particles. The results of sphere areas were exported, and the mean radii and polydispersity (PD) were calculated correspondingly. Meanwhile, tightly packed liposomal clusters or deformed liposomes were considered as aggregates, and they can’t be accurately recognized or measured by the Analyze particles command. Thus the individual liposomes in a cluster were manually measured three times to obtain the average radius (R_liposome). The radius of an aggregate (R_aggregate) was calculated via:

$$V_{\mathit{\text{shell}},\mathit{\text{individual liposome}}}=\frac{4}{3}\pi R_{\mathit{\text{liposome}}}{}^3-\frac{4}{3}\pi {\left(R_{\mathit{\text{liposome}}}-d_t\right)}^3$$ (1)

$$V_{\mathit{\text{shell}},\mathit{\text{aggregate}}}=\sum V_{\mathit{\text{shell}},\mathit{\text{ individual liposome}}}=\frac{4}{3}\pi R_{\mathit{\text{aggregate}}}{}^3$$ (2)

where R_liposome is the average radius of liposome measured by ImageJ, d_t is the shell thickness (4.3 nm for non-irradiated sample and 4.5 nm for laser-irradiated sample), V_shell represents the volume of the lipid shell for liposome or aggregate. The ImageJ analysis was performed based on approximately 500 liposomes in multiple TEM images.

1.3.5 2.6. Laser-triggered repetitive drug release tests

The liposomes were irradiated following a procedure mentioned previously.^{17, 18} Briefly, the liposome suspension was loaded in a glass capillary and irradiated with 1 to 4 cycles (5 sec per cycle) at 1.1 Watts by a near infrared pulse laser beam (wavelength: 1064 nm; pulse repetition frequency: 10kHz). The glass capillary is fully covered by the beam upon irradiation experiments to ensure all the liposomes are irradiated upon each cycle. The released amount of MTX from irradiated liposomes was determined using a UV-vis spectrometer as described previously.¹⁷

1.3.6 2.7. Heating-triggered drug release tests

The MTX-encapsulated liposome suspensions were placed in a water bath at 75°C for 15 minutes. After heating, the samples were cooled to room temperature and diluted 20 × using DI water. Samples were then stored at room temperature for 18 hours. The released amount of MTX was determined using a UV-vis spectrometer as described in previous work.¹⁷

1.3.7 2.8. Small angle neutron scattering (SANS)

SANS measurements were performed at the extended q-range small angle neutron scattering diffractometer (EQ-SANS) of the Spallation Neutron Source (SNS) at Oak Ridge National Laboratory (ORNL). A single instrument configuration with 4 m sample to detector distance was used for all experiments. Frame-skipping (30 Hz) was used to have two bandwidths of 2 Å–5.7 Å and 9.3 Å–13.0 Å, resulting a q-range of 0.004 Å⁻¹– 0.44 Å^{−1 24}. The neutron scattering data was reduced by a standard procedure implemented in Mantid Plot²⁵. During this procedure the collected data was corrected for open beam neutron flux and background scattering. The data was then circularly and annularly averaged to produce scattering intensity (I) profile as a function of scattering vector q,

$$q=\frac{4\pi }{\lambda }\text{ sin}(\theta )$$ (3)

where θ is the scattering angle and λ is the neutron wavelength.

The measured scattering intensity is modeled by

$$I(q)=n\Delta {\rho }^2{V}^{2}P(q)S(q)+I_{\mathit{\text{inc}}}$$ (4)

where n is the number density of liposomes, Δρ is the difference in the scattering length density between the solvent and the liposomal shells, V is the volume of the liposome shells, P(q) is the form factor corresponding to the internal structures of liposomes, and S(q) is the structure factor that described the inter-particle interactions. For a diluted system such as this study, S(q) is considered as 1. I_inc is the incoherent background.

Total 90 μL of liposomes solution that was prepared in D₂O was irradiated by above mentioned laser irradiation method in 3 capillaries. The irradiated liposomes solution was collected and was diluted 4 times using D₂O. The liposomes solution was placed inside a banjo-style cuvette with 1 mm path length and 350 μl total liquid volume. All the EQ-SANS measurements were then done at room temperature (~25°C).

1.3.8 3. Results and Discussion

1.3.9 3.1. Effect of cholesterol on laser-triggered drug release

The characterization of the liposomes is described in our previous publication.¹⁷ Briefly, the size was 149.53 ± 6.18 nm with a polydispersity index (PDI) of 0.18 ± 0.04 for −Chol −Au (no cholesterol and no gold nanorods) liposomes, and 155.34 ± 2.57 nm with a PDI of 0.15 ± 0.08 for +Chol −Au liposomes, according to the dynamic light scattering (DLS) data. The zeta potential value was 3.08 ± 1.82 mV for the −Chol −Au liposomes and 3.06 ± 1.52 mV for the +Chol −Au liposomes. The positive values were designed for electrostatic binding of negatively charged citrate-coated AuNRs on the liposome surface. The p-values for the size and zeta potential between −Chol −Au and +Chol −Au were 0.90 and 0.99, respectively, implying no statistically significant difference. TEM data analysis provided additional morphological information, including location of AuNR on the liposomes and the size distribution, which will be discussed in section 3.2 in details.

Figure 1 shows the effect of cholesterol and AuNR on total MTX release as a function of the number of laser irradiation cycles (5 sec each cycle). +Chol +Au liposomes showed that the amounts of total MTX released after the 1st, 2nd, 3rd, and 4th cycles were 15.8±5.0, 28.3±6.0, 31.6±7.0 and 53.3±7.0 μg/mL respectively. The difference between these MTX releases with respect to irradiation cycles were statistically significant (p < 0.05), except the 2nd and 3rd cycles (p=0.253) (Supplementary Table S1). The results indicate that the MTX release was proportional to the number of laser irradiation cycle. The MTX released of +Chol +Au liposomes at each cycle was significantly higher than other groups.

Figure 1.

Laser-triggered MTX release measurement for −Chol −Au liposomes (hatched green), −Chol +Au liposomes (hatched yellow), +Chol −Au liposomes (solid blue) and +Chol +Au liposomes (solid orange) after 1 to 4 laser cycles (5 sec laser irradiation per cycle, 3 min cooling time between each cycle) at 1.1 W.

The amount of total MTX released from +Chol −Au liposomes after the 1st, 2nd, 3rd, and 4th cycles were 0.7±5.4, 3.9±2.3, 2.0±2.3 and 2.0±2.3 μg/mL respectively. The MTX releases with respect to irradiation cycles were not significantly different (p > 0.05 for all pairs. Supplementary Table S2). The results suggest strong surface plasmon heating of AuNR on MTX release from +Chol liposomes after laser irradiation cycles.

In the case of −Chol −Au liposomes, the total amount of MTX released after the 1st, 2nd, 3rd, and 4th irradiation cycles were 1.3±4.0, 3.9±3.0, 6.5±7.0, 6.5±0.8 μg/mL, respectively. The difference of these MTX releases between irradiation cycles were not statistically significant for all (p > 0.05 for all pairs. Supplementary Table S3). −Chol +Au liposomes released 0.0±7.0, 1.0±3.0, 3.0±11.0 and 7.0±7.0 μg/mL after the 1st, 2nd, 3rd, and 4th irradiation cycles, respectively. The p values between all pairs were greater than 0.05 (Supplementary Table S4), demonstrating that the MTX release was not significantly affected by laser irradiation without cholesterol in the liposome composition. The differences between total MTX released between −Chol −Au and −Chol +Au liposomes at each cycle were also not statistically significant, according to the t-test (p > 0.05 for all pairs. Supplementary Table S5).

By comparing the MTX releases between −Chol +Au and +Chol +Au group at each cycle (p < 0.05 for all pairs. Supplementary Table S6), we found that the presence of cholesterol was of crucial importance for the drug release of these liposomes when activated by laser. This is in agreement with other study²⁶, in which García et al. found that AuNP-anchored liposomes containing 40 mol% cholesterol reached higher release amount than those with only 3.35 mol% cholesterol at temperature higher than T_c. They proved that increasing temperature from 37°C to above T_c enhanced the drug release ability for AuNP-liposomes with 40 mol% of cholesterol²⁶. Note that the opposite phenomenon was also reported previously. Jeong Eun Shin et al.²⁷ has reported that hollow gold nanoshell conjugated-DPPC liposomes with 40% Chol possessed a higher threshold of light power to release payload than cholesterol-free liposomes. One reason that possibly leads to the contradictory findings is the difference in triggering mechanism. Shin et al. has attributed the payload release to the transient nanobubbles formed by the hollow gold nanoshells upon laser exposure, which caused mechanical lysing of liposomes^{27 28}.

1.3.10 3.2. Effect of Cholesterol on morphological changes upon irradiation

The morphological changes of the light-activated liposomes, −Chol +Au and +Chol +Au liposomes, upon laser irradiation were examined by TEM. Some liposomes appeared to be an ellipsoidal shape due to potential collapse during the freeze-drying process of the TEM sample preparation. Figure 2A and Figure 2B show the TEM images of −Chol +Au liposomes before irradiation and after laser irradiation, respectively. No visible changes were observed. In contrast to the −Chol +Au liposomes, laser-irradiated +Chol +Au liposomes showed notable changes (Figure 2D). Some +Chol +Au liposomes aggregated, deformed, and seemed fused, especially near the AuNR. Figure S1 in Supplementary also showed the trend. In addition, some gold nanospheres were also observed, indicative of the melting of AuNR (Figure S1). DLS measurements were also conducted to characterize the size change of +Chol +Au liposomes before and after laser. However, no significant changes of hydrodynamic diameter were found (Supplementary Figure S2). Interestingly, the size distribution shifted to the right (regime of larger size) after the laser irradiation, but the average diameter was unchanged.

Figure 2.

TEM images of (A)−Chol +Au liposomes before laser irradiation and (B) after 20 sec (5 sec × 4 cycles) laser, (C) +Chol +Au liposomes before laser irradiation and (D) after 20 sec (5 sec × 4 cycles) laser.

1.3.11 3.3. Effect of cholesterol on nanostructures

1.3.12 3.3.1. Effect of laser

The EQ-SANS curves for all liposome formulations with different laser irradiation durations are presented in Figure 3. Besides +Chol +Au liposomes (Figure 3D), all the other formulations (−Chol −Au, +Chol −Au, −Chol +Au) showed no remarkable changes between no laser and with laser. +Chol +Au exhibited a curve shift from high q to low q with increased laser irradiation durations, which was revealed by the slope change at q range of 0.04 Å⁻¹ ~0.1 Å⁻¹. It indicates a size increase after laser irradiation, which will be discussed later in details.

Figure 3.

EQ-SANS scattering curves for (A) −Chol −Au liposomes with 0 sec or 20 sec laser, (B) +Chol −Au liposomes with 0 sec or 20 sec laser, (C) −Chol +Au liposomes with 0 sec, 5 sec or 20sec laser, (D) +Chol +Au liposomes with 0 sec, 5 sec or 20 sec laser.

We first evaluated the bilayer thicknesses through the Kratky-Porod approximation based on the range from 0.001 Ų < q² <0.006 Å^{2 29, 30}. The equations are:

$$I(q)=I(0)q^{-2}\text{ exp}\left(-R_t^2{q}^2\right)$$ (5)

$$I_{\mathit{\text{exp}}}(q)q^2=I(q)q^2+I_{\mathit{\text{inc}}}{q}^2$$ (6)

where R_t is the radius of gyration of the bilayer, I_exp(q) is the experimental scattering intensity, and the I_inc is the incoherent scattering. Hence, by plotting ln[I(q) · q²] as a function of q² (Figure 4), the lamellar radius of gyration R_t was obtained from the slope of a linear regression. Note that the scattering background I_inc has been subtracted from the scattering intensities prior to plotting. The bilayer thickness can be estimated via:

$$d_t^2\cong 12R_t^2$$ (7)

The bilayer thicknesses obtained from the Kratky-Porod plots are summarized in Table 1. For −Chol −Au liposomes, the bilayer thickness was determined to be 5.0 ± 0.1 nm, which is which is in agreement with 5.1 nm that was previously reported elsewhere³¹. Addition of 35 mol% cholesterol decreased the bilayer thickness to 4.4 ± 0.1 nm. This is likely caused by the kink or deformation of the hydrocarbon chains of the DSPC lipids to accommodate the cholesterol molecules³². It is also noteworthy that the effect of adding cholesterol on bilayer thickness depend on the acyl chain lengths of the lipids. For saturated phosphatidylcholines with shorter chains, e.g., 12 to 16 carbons, inclusion of certain amount of cholesterol was found to increase the bilayer thickness^{29, 32}. After 5 sec or 20 sec laser irradiation, +Chol +Au liposomes showed a thickness of ~4.5 nm.

Figure 4.

The Kratky-Porod plots of the EQ-SANS scattering data for (A) −Chol −Au, (B)+Chol −Au, (C) +Chol +Au after 5 sec laser irradiation, and (D) +Chol +Au after 20 sec laser irradiation. Dashed lines are the

Table 1.

Parameters of the lipid bilayers for different liposomes obtained from the Kratky-Porod plots.

Composition	R2t (nm2)	dt (nm)
−Chol−Au	2.1 ± 0.1	5.0 ± 0.1
+Chol−Au	1.6 ± 0.1	4.4 ± 0.1
+Chol +Au, 5 sec laser	1.7 ± 0.1	4.5 ± 0.1
+Chol +Au, 20 sec laser	1.7 ± 0.1	4.5 ± 0.2

Before proceeding to model fitting, we examined the structures of AuNR in the scattering curves by intensity subtraction. As shown in Figure 5A, by subtracting the scattering curve I _{(+Chol −Au)} from the curve I _{(+Chol +Au)}, the residual intensity theoretically only contained the structure information of AuNR. By fitting the residual intensity ΔI with a cylinder model, we found that the fitting yields a 11 nm (radius of cross section) × 132 nm (length) rod structure (Table 2). This is in agreement with the average size of AuNRs observed in TEM images (shown in Figure 5A inset). The polydispersity index used in fitting were obtained from ImageJ analysis of TEM images (Supporting Information, Table S8).

Figure 5.

The subtracted scattering curves and the corresponding fittings. (A) AuNR residual intensity obtained by subtracting I_{(+Chol −Au)} from I_{(+Chol +Au)}. (B) AuNR and Au sphere residual intensity obtained by subtracting I_{(−Chol −Au)} from I_{(−Chol +Au, 20 sec laser)}; the ΔI curve has been shifted for better illustration. Insets showed the corresponding TEM images of AuNR or Au sphere structures.

Table 2.

Fitting parameters for the ΔI curves in Figure 5 (A) and (B), respectively.

Figure	5(A)	5(B)
Fitting Model (component)	Cylinder (AuNR)	Cylinder (AuNR) + Sphere (Au sphere)
Size (nm)	AuNR: 11 (radius) × 132 (length)	AuNR: 11 (radius) × 132 (length) Au sphere: 22.8 (radius)
PDa	0.14 (radius, AuNR) 0.20 (length, AuNR) 0.12 (radius, Au sphere)

^aPD represents polydispersity. PD values were obtained from TEM image analysis.

Next, we examined AuNR structures solely after laser irradiation. By subtracting I _{(−Chol −Au)} from I_{(−Chol +Au, 20 sec)}, we can extract the AuNR structures that were exposed to 20sec-laser irradiation (Figure 5B). According to TEM images (Figure 5B inset and Figure S1), a portion of AuNRs melted and formed Au spheres after 20 sec-laser. Thus, the ΔI curve was fitted with a customized model consisting of a sphere-model and a cylinder-model, representing the Au spheres and AuNRs respectively. The fitting curve slightly deviated at the q range of 0.04 Å⁻¹ ~0.1 Å⁻¹, probably because of liposomes in the subtracted curve. The fitting parameters of the combined model are presented in Table 2, suggesting the co-existence of Au spheres with a radius of 22.8 nm and AuNRs with a size of 11 nm ×132 nm. This result is consistent with the sizes of AuNR and Au spheres measured in TEM images (Figure 5B inset).

Combination of a core-shell-sphere model and AuNR/Au spheres was used to fit the liposomes, based on the volume fraction of the particles (Figure 6A and Table 3). For simplicity, only fitting parameters for the +Au groups are shown. The fittings on −Chol −Au and +Chol −Au liposomes shared the same parameters with −Chol +Au and +Chol +Au on the part of core-shell-sphere model, respectively. The polydispersity obtained from TEM image analysis (Table S8) was used for fitting. For the +Chol +Au data with 5 sec or 20 sec laser, as discussed in section 3.2, there was no overall size increase observed by DLS. Thus, a second core-shell-sphere model with bigger size was used for the laser irradiated aggregated/fused liposomes samples.

Figure 6.

EQ-SANS scattering curves for (A) +Chol +Au liposomes and −Chol +Au liposomes, showing the difference caused by inclusion of cholesterol. (B) +Chol +Au liposomes with 0 sec, 5 sec or 20sec laser, respectively, showing the changes caused by laser irradiation. The black solid lines are corresponding fitting curves. Some scattering curves are shifted vertically for better illustration.

Table 3.

Fitting models and parameters for the −Chol +Au liposomes, +Chol +Au liposomes with 0 sec, 5 sec or 20 sec laser irradiation, respectively.

Sample	Model (component)	dt (nm)	Rcore (nm)	PD	Scale	Percentage of liposome/aggregate
−Chol +Au	Core-shell-sphere (liposome)	5.0	47	0.30	0.060	100% liposome
	Cylinder (AuNR)	Radius × Length: 13 × 132		0.14 (radius) 0.20 (length)	0.005
+Chol +Au	Core-shell-sphere (liposome)	4.3	51	0.27	0.060	100% liposome
0 sec laser	Cylinder (AuNR)	Radius × Length: 13 × 132		0.14 (radius) 0.20 (length)	0.005
+Chol +Au	Core-shell-sphere (liposome)	4.5	51	0.30	0.049	81% liposome 19% aggregate
5 sec laser	Core-shell-sphere (aggregate)	4.5	230	0.50	0.011
	Cylinder (AuNR)	Radius × Length: 13 × 132		0.14 (radius) 0.20 (length)	0.0045
	Sphere (Au Sphere)	Radius: 23		0.12	0.0005
	Core-shell-sphere (liposome)	4.5	51	0.50	0.027
+Chol +Au	Core-shell-sphere (aggregate)	4.5	230	0.50	0.033	44% liposome
20 sec laser	Cylinder (AuNR)	Radius × Length: 13 × 132		0.14 (radius) 0.20 (length)	0.004	56% aggregate
	Sphere (Au Sphere)	Radius: 23		0.12	0.001

The fittings of the scattering curve of the +Chol +Au liposomes with 5 sec laser (Table 3) showed liposomal aggregates with a radius of ~230 nm. The volume fraction (represented by scale) of the original population of liposomes decreased from 0.06 to 0.049, suggesting about 19% of original liposomes aggregated/fused. Moreover, the fitting parameters of the +Chol +Au with 20 sec laser suggested that the size of liposomal aggregate remained at 230 nm while the fraction of original liposomes further decreased from 0.049 to 0.027. This indicates approximately 56% of liposomes aggregated with longer laser duration. Note that the volume fraction percentage is based on lipid volume, excluding water core in the liposomes.

Overall, the fitting results suggested that a portion of liposomes aggregated after laser irradiation accompanied with melting of some AuNRs and with no significant changes in the lipid bilayer thickness. This result is in good agreement with the morphological changes observed by TEM and the Kratky Plot results. Despite of some deviations in the fitting curves, the combined custom model described the shift of the scattering curves with different irradiation durations.

The discrepancy between DLS and the SANS data or the TEM data have been reported elsewhere.³³ We suspect that it is because of the sample preparation step for the DLS measurement, which requires dilution of the sample for a sufficient volume, resulting in the breakdown of the aggregates. It implies that the aggregation is likely to be reversible. This hypothesis also supports our previous finding that the number density of liposomes didn’t change before and after laser irradiation when being measured in a highly diluted condition by an ultramicroelectrode.¹⁷

1.3.13 3.3.2. Effect of heating

The effect of heating on the drug release and structural changes of liposomes were examined. Figure 7 shows that −Chol +Au and +Chol +Au liposomes released 7.9 ± 5.5 μg/mL and −7.6 ± 4.6 μg/mL MTX, respectively (p = 0.006), during 18 hours at room temperature. The results suggest that cholesterol does not play a role in drug release at temperature below T_c. Similarly, many previous studies have shown that cholesterol-rich liposomes have higher drug retention rate compared to cholesterol-free liposomes at temperature below T_c.^{34 35 36} This is probably because the high content of cholesterol can suppress the permeability as rigid “barriers” in the bilayer membrane.³⁷

Figure 7.

Effect of temperature on the concentration of released MTX of the −Chol +Au and +Chol +Au liposomes.

On the other hand, after heated to 75°C (above T_c) for 15 minutes, −Chol +Au liposomes released 137.8 ± 2.6 μg/mL MTX, while +Chol +Au liposomes released 155.3 ± 5.1 μg/mL MTX (p = 0.001). The result indicated the effect of cholesterol on facilitating drug release above T_c.

However, the effect of cholesterol on the drug release when the whole sample was heated seemed less compared to the 20 s laser-irradiated drug release (Figure 1). The drug release amount from +Chol +Au liposomes was ~20 μg/mL more than −Chol +Au by the 15 min heating whereas the difference was ~50 μg/mL by 15 second laser irradiation. The results suggest that bulk heating and the laser irradiation shares the drug release mechanism, but not in the exactly the same manner. The effect of heating versus laser can be comparable to the effect of a continuous wave (CW) laser versus pulsed laser on drug release. One study has shown that a femtosecond pulsed laser resulted in higher percentage of drug release than CW at the same laser power and duration.³⁸ The overall temperature increase for the pulsed laser group was lower than the CW laser group. The peak power from the pulsed laser is in fact several orders higher than the CW, leading to effective “melting” of the lipid bilayer locally and specifically only to the liposomes irradiated. Therefore, shorter laser duration with a pulsed laser yields the same amount of drug release compared to CW. In addition, because long exposure to heat can damage healthy tissue by hyperthermia, drug release by the pulsed laser would be advantageous for practical reasons.

Figure 8 shows SANS data for the liposomes heated above T_c and cooled to room temperature. Similar to the scattering curves of laser-irradiated liposomes (Figure 3), the +Chol +Au curve was shifted to left, whereas −Chol +Au had no changes between before heating and after heating. According to the Kratky-Porod plots shown in Figure 9, the bilayer thickness for heated −Chol +Au and heated +Chol +Au liposomes were determined to be 4.9 ± 0.1 nm and 4.5 ± 0.2 (Table 4), respectively. It is close to the original thickness prior to heating. The heating temperature of 75°C was not able to induce melting or deformation of AuNRs. Therefore, the differences in scattering curves only represents a structural change of liposomes.

Figure 8.

EQ-SANS scattering curves for (A) −Chol +Au liposomes after heated to 75°C for 20 min and cooled to room temperature, (B) +Chol +Au liposomes after heated to 75°C for 20 min and cooled to room temperature.

Figure 9.

The Kratky-Porod plots of the SANS scattering data for (A) −Chol +Au and (B)+Chol +Au liposomes after heated to 75°C and cooled to room temperature. Dashed lines are the trendlines of linear fittings.

Table 4.

Parameters of the lipid bilayers for liposomes with heating obtained from the Kratky-Porod plots.

Composition	Condition	R2t (nm2)	dt (nm)
−Chol +Au	Heated to 75C	2.0 ± 0.1	4.9 ± 0.1
+Chol +Au	Heated to 75C	1.7 ± 0.1	4.5 ± 0.2

As validated in the previous section, the customized model was used to fit the data for +Chol +Au heating to 75°C (Figure 10). Similarly, we used a second core-shell-sphere model representing the aggregated liposomes to quantify the changes. Table 5 summarizes the parameters associated with the fittings shown in Figure 10. The fitting results show that, after heating, about 29% of liposomes transformed to aggregates with a radius of ~155 nm.

Figure 10.

EQ-SANS scattering curves for +Chol +Au liposomes after heated to 75°C for 20 min and cooled to room temperature. Solid black lines represent the fitting curve. The scattering curve has been shifted for better illustration.

Table 5.

Fitting models and parameters for the +Chol +Au liposomes without heating (25°C), +Chol +Au liposomes after heating to 75°C and cooled down, respectively.

Sample	Component (model)	dt (nm)	Rcore (nm)	PD	Scale	Percentage of liposome/aggregate
+Chol+Au,	Core-shell-sphere (liposome)	4.3	51	0.27	0.060	100%
25°C (no heating)	Cylinder (AuNR)	Radius × Length: 13 × 132		0.14 (radius) 0.20 (length)	0.005
+Chol +Au,	Core-shell-sphere (liposome)	4.5	51	0.4	0.042	71% liposome
heated to 75°C then cooled down	Core-shell-sphere (aggregate)	4.5	155	0.4	0.018	29% aggregate
	Cylinder (AuNR)	Radius × Length: 13 × 132		0.14 (radius) 0.20 (length)	0.005

1.3.14 4. Conclusions

In this study, we found that inclusion of 35 mol% cholesterol (+Chol +Au) remarkably facilitated the drug release of laser-activated liposomes. In contrast, the cholesterol-free liposomes (−Chol +Au) barely released drug at the same laser conditions. The TEM results and the SANS data showed an aggregation or fusion only for the +Chol +Au liposomes after irradiation, indicating cholesterol played an important role in changing lipid bilayer membrane affecting aggregation during and after laser irradiation, resulting in drug release through the membrane. However, the Kratky-Porod plot analysis suggested the lipid bilayer thickness did not change. In addition, heating experiments confirmed more drug release with cholesterol in the lipid composition than in the absence of cholesterol, suggesting cholesterol enhances fluidity at temperature above chain-transition temperature T_c. Overall, the study revealed two important findings: (1) inclusion of 35 mol% cholesterol enhanced the permeability of lipid bilayers above T_c; (2) the mechanism of laser-activated liposomal drug delivery is disrupting lipid bilayer membranes by the photothermal effect in the presence of plasmonic materials. By understanding the fundamentals of the technology, more precise controlled drug release at a targeted site with great stability and repeatability is expected.