Microfluidic Production of Lysolipid-Containing Temperature-Sensitive Liposomes

Bioengineering

Your institution must subscribe to JoVE's Bioengineering section to access this content.

Fill out the form below to receive a free trial or learn more about access:

 

Summary

The protocol presents the optimized parameters for preparing thermosensitive liposomes using the staggered herringbone micromixer microfluidics device. This also allows co-encapsulation of doxorubicin and indocyanine green into the liposomes and the photothermal-triggered release of doxorubicin for controlled/triggered drug release.

Cite this Article

Copy Citation | Download Citations | Reprints and Permissions

Cheung, C. C. L., Ma, G., Ruiz, A., Al-Jamal, W. T. Microfluidic Production of Lysolipid-Containing Temperature-Sensitive Liposomes. J. Vis. Exp. (157), e60907, doi:10.3791/60907 (2020).

Abstract

The presented protocol enables a high-throughput continuous preparation of low temperature-sensitive liposomes (LTSLs), which are capable of loading chemotherapeutic drugs, such as doxorubicin (DOX). To achieve this, an ethanolic lipid mixture and ammonium sulfate solution are injected into a staggered herringbone micromixer (SHM) microfluidic device. The solutions are rapidly mixed by the SHM, providing a homogeneous solvent environment for liposomes self-assembly. Collected liposomes are first annealed, then dialyzed to remove residual ethanol. An ammonium sulfate pH-gradient is established through buffer exchange of the external solution by using size exclusion chromatography. DOX is then remotely loaded into the liposomes with high encapsulation efficiency (> 80%). The liposomes obtained are homogenous in size with Z-average diameter of 100 nm. They are capable of temperature-triggered burst release of encapsulated DOX in the presence of mild hyperthermia (42 °C). Indocyanine green (ICG) can also be co-loaded into the liposomes for near-infrared laser-triggered DOX release. The microfluidic approach ensures high-throughput, reproducible and scalable preparation of LTSLs.

Introduction

LTSL formulation is a clinically relevant liposomal product that has been developed to deliver the chemotherapeutic drug doxorubicin (DOX) and allows efficient burst drug release at clinically attainable mild hyperthermia (T ≈ 41 °C)1. The LTSL formulation consists of 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC), the lysolipid 1-stearoyl-2-hydroxy-sn-glycero-3-phosphatidylcholine (MSPC; M stands for "mono") and PEGylated lipid 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(polyethylene glycol)-2000] (DSPE-PEG2000). Upon reaching the phase transition temperature (Tm ≈ 41 °C), the lysolipid and DSPE-PEG2000 together facilitate the formation of membrane pores, resulting in a burst release of the drug2. The preparation of LTSLs primarily uses a bulk top-down approach, namely lipid film hydration and extrusion. It remains challenging to reproducibly prepare large batches with identical properties and in sufficient quantities for clinical applications3.

Microfluidics is an emerging technique for preparing liposomes, offering tunable nanoparticle size, reproducibility, and scalability3. Once the manufacturing parameters are optimized, the throughput could be scaled-up by parallelization, with properties identical to those prepared at bench scale3,4,5. A major advantage of microfluidics over conventional bulk techniques is the ability to handle small liquid volumes with high controllability in space and time through miniaturization, allowing faster optimization, while operating in a continuous and automated manner6. Production of liposomes with microfluidic devices is achieved by a bottom-up nanoprecipitation approach, which is more time and energy efficient because homogenization processes such as extrusion and sonication are unnecessary7. Typically, an organic solution (e.g. ethanol) of lipids (and hydrophobic payload) is mixed with a miscible non-solvent (e.g. water and hydrophilic payload). As the organic solvent mixes with the non-solvent, the solubility for the lipids is reduced. The lipid concentration eventually reaches a critical concentration at which the precipitation process is triggered7. Nanoprecipitates of lipids eventually grow in size and close into a liposome. The main factors governing the size and homogeneity of the liposomes are the ratio between the non-solvent and solvent (i.e. aqueous-to-organic flow rate ratio; FRR) and the homogeneity of the solvent environment during the self-assembly of lipids into liposomes8.

Efficient fluid mixing in microfluidics is therefore essential to the preparation of homogeneous liposomes, and various designs of mixers have been employed in different applications9. Staggered herringbone micromixer (SHM) represents one of the new generations of passive mixers, which enables high throughput (in range of mL/min) with a low dilution factor. This is superior to traditional microfluidic hydrodynamic mixing devices8,10. The SHM has patterned herringbone grooves, which rapidly mix fluids by chaotic advection9,11. The short mixing timescale of SHM (< 5 ms, less than the typical aggregation time scale of 10–100 ms) allows lipid self-assembly to occur in a homogenous solvent environment, producing nanoparticles with uniform size distribution3,12.

The preparation of LTSLs with microfluidics is, however, not as straightforward compared to conventional liposomal formulations due to the lack of cholesterol8, without which lipid bilayers are susceptible to ethanol-induced interdigitation13,14,15. Until now, the effect of residual ethanol presents during the microfluidic production of liposomes has not been well understood. The majority of the reported formulations are inherently resistant to interdigitation (containing cholesterol or unsaturated lipids)16, which unlike LTSLs are both saturated and cholesterol-free.

The protocol presented herein uses SHM to prepare LTSLs for temperature triggered-release drug delivery. In the presented method, we ensured the microfluidic-prepared LTSLs are nano-sized (100 nm) and uniform (dispersity < 0.2) by dynamic light scattering (DLS). Furthermore, we encapsulated DOX using the transmembrane ammonium sulfate gradient method (also known as remote loading)17 as a validation of the integrity of the LTSL lipid bilayer. Remote loading of DOX requires the liposome to maintain a pH-gradient in order to achieve high encapsulation efficiency (EE), which is unlikely to happen without an intact lipid bilayer. In this presented method, distinctive from typical microfluidic liposome preparation protocols, an annealing step is required before the ethanol is removed to enable the remote loading capability; i.e. to restore the integrity of the lipid bilayer.

As mentioned previously, hydrophilic and hydrophobic payloads can also be introduced to the initial solutions for the simultaneous encapsulation of payloads during the formation of LTSLs. As a proof-of-concept, indocyanine green (ICG), an FDA-approved near-infrared fluorescent dye, which is also a promising photothermal agent, is introduced to the initial lipid mixture and successfully co-loaded into the LTSLs. An 808 nm laser is used to irradiate the DOX/ICG-loaded LTSLs and successfully induce photothermal heating-triggered burst release of DOX within 5 min.

All the instruments and materials are commercially available, ready-to-use, and without the need for customization. Since all the parameters for formulating LTSLs have been optimized, following this protocol, researchers with no prior knowledge of microfluidics could also prepare the LTSLs, which serves as the basis of a thermosensitive drug delivery system.

Subscription Required. Please recommend JoVE to your librarian.

Protocol

1. Equipment setup

  1. Assemble the syringe pumps and SHM as follows.
    1. Connect the "To Computer" port of the secondary syringe pump (Pump 02, for aqueous solution) to the "To Network" port of the master syringe pump (Pump 01, for ethanol lipid solution) using Pump to Pump network cable (Figure 1, yellow).
    2. Connect the "To Computer" port of the master pump to the "RS232 Serial" port of the computer using PC to Pump network cable (Figure 1, blue).
    3. Connect tubing to each of the inlets and outlets of the SHM using a nut and ferrule. Convert the terminal of the tubing for both inlets to female Luer using another nut and ferrule and a union assembly. Longer tubing of the inlets allows easier attachment to the syringes (Figure 2).
  2. Set up the pump control software.
    1. Assign the address of the master syringe pump and secondary syringe pump to "Ad:01" and "Ad:02", respectively, using the "Setup" button of the syringe pump. This only needs to be done for the first time.
    2. Open the pump control software on the computer. The two syringe pumps should be detected automatically, followed by a beeping sound. Otherwise, click Pumps and Search for pumps to update the connection. (Figure 3).
    3. Assign Diameter to 12.45 (mm) by choosing "HSW Norm-Ject 5 cc (Dia=12.45)".
    4. Assign Rate to 0.25 mL/min for Pump 01 (ethanol lipid solution) and 0.75 mL/min for Pump 02 (aqueous solution). The flow rates correspond to a total flow rate (TFR) of 1 mL/min and aqueous-to-ethanol flow rate ratio (FRR) of 3.
    5. Assign Volume to any values above 5 mL.
      NOTE: The targeted infusion volume is set greater than the loaded liquid volume considering the void volume of the tubing.
    6. Select INF (infusion) mode for both pumps.
    7. Press Set to confirm the settings.

2. Prepare the LTSLs

  1. Prepare a LTSL10 or LTSL10-ICG lipid mixture (see Table 1).
  2. Withdraw 1 mL of lipid mixture and at least 3 mL of (NH4)2SO4 solution using two 5 mL Luer lock syringes.
  3. Install the two syringes onto the syringe pumps in the upright position by sliding the barrel flange of the syringe to the syringe retainer of the pump, and the plunger flange of the syringe to the pusher block of the pump (Figure 4).
  4. Wrap the end of the heating tape to the syringe with the aqueous solution. Wrap the other end of the heating tape and temperature probe of the thermostat around the syringe with the lipid solution. It is helpful to practice this step with empty syringes in place in order to ease the assembly process (Figure 5A).
  5. Connect the two syringes to the female Luer adaptors of the corresponding inlets of the SHM. Make sure the syringes containing the lipid mixture and (NH4)2SO4 solutions are connected ethanol inlet and aqueous inlet, respectively. Adjust the plunger position to remove air bubbles from the syringes (Figure 5B).
    NOTE: Ensure the syringes are still securely positioned onto the syringe retainer of the pumps.
  6. Heat up the syringes to above 51 °C using the heating tape using a 10 s heating session. Allow the thermostat to update the temperature of the syringes. Repeat this step in the following steps to maintain the temperature during the infusion.
    CAUTION: Turn off the heating tape after 10 s to prevent temperature overshoot and allow the thermostat to update the actual temperature. The heating tape should also be handled with care as its temperature rises very quickly. Heating continuously may damage the equipment and syringes, due to the time delay of the thermostat for updating the measured temperature.
  7. Once the temperature is above 51 °C, run the syringe pumps by pressing Run All in the pump control software (Figure 3).
  8. Ensure the fluid flow is free of air bubbles and any leakage. Dispose the initial volume (around 0.5 mL) of liquid from the outlet as waste.
    NOTE: This initial waste volume is not definite and depends on the internal volume of the setup, which is the volume for fluid to travel from the syringes through the tubing and SHM to the outlet.
  9. Collect the rest of the liquid as liposome samples into a microcentrifuge tube or bijou vial.
  10. Pause/stop the infusion when the liquid in either of the syringes are almost empty.
    NOTE: The pumps should be stopped manually, since the pumps may not accurately detect the position when the syringes are empty.
  11. Place the collected liposome solutions in a 60 °C water bath to anneal for 1.5 h.
    NOTE: This step is essential in enabling drug loading into the liposomes.
  12. Transfer the solutions to dialysis tubes. Dialyze the solutions against 1 L of 240 mM (NH4)2SO4 at 37 °C for at least 4 h to obtain purified liposomes.
    NOTE: The protocol can be paused here. Liposomes at this step are at 5 mM of phospholipid. Purified liposomes can be stored at 4 °C.
  13. To clean the SHM for repeated use, flush the SHM sequentially with deionized water, ethanol and dry with nitrogen gas.

3. Remote loading of DOX into LTSLs by transmembrane pH gradient

  1. Exchange external buffer of LTSLs to HEPES-buffered saline (HBS) by using size exclusion chromatography (SEC) to establish a transmembrane pH gradient.
    1. Add a total of 25 mL of HBS to the top of a SEC column to prepare the column. Allow all eluent to elute through the column and dispose the eluate.
    2. Add 1 mL of dialyzed liposomes, prepared from step 2.12, to the column and dispose the elute.
    3. Add 1.5 mL of HBS to the column and dispose the elute.
    4. Add 3 mL of HBS to the column and collect the 3 mL of elute.
      NOTE: The protocol can be paused here. Liposomes are collected at this step and are at 1.67 mM of phospholipid. Buffer exchanged liposomes can be stored at 4 °C.
  2. Incubate LTSLs with doxorubicin (DOX) and purify LTSLs.
    1. Add DOX solution in 1:20 DOX-to-phospholipid molar ratio into 1 mL of buffer-exchanged liposomes solution (1.67 mmol) contained in a bijou vial. This can be achieved by adding 48.4 µL of 1 mg/mL DOX solution (83.4 µmol).
    2. Place the bijou vial in a 37 °C water bath for 1.5 h to allow DOX loading into the liposomes.
    3. Mix 10 µL of the liposomes with 170 µL of HBS and 20 µL of 1% (v/v) Triton X-100 solution in a black 96-well plate. Repeat for three wells. These wells correspond to the "before purification" DOX content.
    4. In case of preparing LTSL10-ICG, mix 40 µL of the liposomes with 160 µL of DMSO in a clear 96-well plate. Repeat for three wells. These wells correspond to the "before purification" ICG content.
    5. Purify the liposome solution as described in step 3.1.
      NOTE: To reuse the column for future purification, clean the column from free DOX by first adding 1 mL of diluted 0.5 M NaOH solution before performing step 3.1.1. Free DOX in red will turn violet-blue and elute through the column quickly.
    6. Mix 30 µL of the purified liposomes solution with 150 µL of HBS and 20 µL of 1% (v/v) Triton X-100 solution in a black 96-well plate. Repeat for three wells. These wells correspond to the "after purification" DOX content.
    7. In case of LTSL10-ICG, mix 40 µL of the purified liposomes solution with 160 µL of DMSO in a clear 96-well plate. Repeat for three wells. These wells correspond to the "after purification" ICG content.
    8. Measure the DOX fluorescence intensity of the wells before (step 3.2.3) and after (step 3.2.5) purification, using a microplate reader (λex = 485 nm, λem = 590 nm).
    9. Calculate the encapsulation efficiency of DOX (DOX EE) by taking the ratio of the fluorescence intensities before and after purification.
      Equation 1
    10. Measure the ICG absorbance of the wells before and after purification, using a microplate reader (600 to 1000 nm).
    11. Calculate the encapsulation efficiency of ICG (ICG EE) by taking the ratio of the absorbance at 792 nm before and after purification, taking into account the dilution factor (3 times) during the purification.
      Equation 2

4. Dynamic Light Scattering (DLS)

  1. Add 50 µL of liposomes solution (step 2.12) to 450 µL of deionized water.
  2. Place the cuvette inside the DLS instrument and perform the measurement according to the manufacturer's instructions.
  3. Record the mean Z-average diameter and dispersity of three measurements for each sample.

5. Differential scanning calorimetry (DSC)

  1. Concentrate 1 mL of the liposomes samples (step 2.12) with a centrifugal filter unit to 0.5 mL (final lipid concentration of 10 mM). Using a fixed-angle rotor, spin at 7500 x g for approximately 15 min.
  2. Transfer 20 µL of (NH4)2SO4 solution and liposomes samples to two respective DSC pans. Seal the pans with DSC hermetic lids using the DSC sample press kit.
  3. Measure the sample from 30 °C to 60 °C at a heating rate of 1 °C/min using a differential scanning calorimeter.
  4. Analyze the data with appropriate software. Take the phase transition temperature (Tm) as the onset of the phase transition (melting peak), which is measured by the x-intercept of the tangent of the point of maximum slope.

6. Doxorubicin release

  1. Preheat HBS at designated temperature (37 or 42 °C) using a water bath. Prepare an ice water bath for quenching the samples.
  2. Add 100 µL of purified DOX-loaded liposomes (step 3.2.5) into 1.9 mL of HBS in a microcentrifuge tube. Place the tube into the water bath of the designated temperature.
  3. Withdraw immediately 200 µL of samples from the tube and quickly place it in the ice water bath to quench any subsequent drug release. This sample corresponds to the initial (t = 0) time point.
  4. Withdraw 200 µL of samples at subsequent time points (t = 5, 10, 15, 30, 60 min) and quickly place it in the ice water bath to quench any drug release.
  5. Mix 50 µL of sample of each time point with 150 µL of HBS in a black 96-well plate. Measure the DOX fluorescence intensity using a plate reader.
  6. Add 20 µL of 1% (v/v) Triton X-100 into random selected wells prepared in step 6.5. Measure the DOX fluorescence intensity of these wells using a plate reader. These values correspond to the fully released (t = ∞; 100% release) time point.
  7. Calculate and plot the percentage of DOX released by interpolating the fluorescence intensity of each time points (I(t)), compared to the initial (I(0)), and fully released (I(∞)), value.
    Equation 3

7. Laser Heating and Triggered Release

  1. Set water bath temperature to 37 °C and allow the temperature to stabilize.
  2. Add 200 µL of DOX loaded LTSL10-ICG ([ICG] = 10 µg/mL) to a clear 96-well plate, then place it in the water bath, keep the bottom immersed in water.
  3. Set the current of the laser system to 2.27 A. Place the collimator of the laser system at 5 cm vertically above the surface of the 96-well plate, which corresponds to an energy flux of 0.5 W/cm2 [Figure 6].
    CAUTION: The laser system should be operated in compliance with relevant laser safety measures.
  4. Switch on the laser and monitor the temperature every minute using a fiber optic temperature probe.
  5. At 5 and 10 min, withdraw 10 µL of laser-irradiated liposomes from the clear 96-well plate at and mix with 190 µL of HBS for three wells in a black 96-well plate.
  6. Mix 10 µL of the liposomes with 170 µL of HBS and 20 µL of 1% (v/v) Triton X-100 solution for three wells in a black 96-well plate. These wells correspond to the "100% released" DOX content. Measure the DOX fluorescent intensity and calculate the DOX release as described in step 6.7.

Subscription Required. Please recommend JoVE to your librarian.

Representative Results

The preparation of LTSLs by microfluidics requires the lipid composition of DPPC/MSPC/DSPE-PEG2000 (80/10/10, molar ratio; LTSL10). Figure 7A (left) shows the appearance of as-prepared LTSL10 from step 2.9, as a clear and non-viscous liquid. LTSL10 formulation is developed from the conventional formulation, LTSL4 (DPPC/MSPC/DSPE-PEG2000, 86/10/4, molar ratio) since LTSL4 forms a gel-like viscous sample, as indicated by the large amount of air bubbles trapped in the sample (Figure 7A; right).

DLS measurement of LTSL10 (Figure 7B, red) showed that the Z-Average diameter and dispersity of LTSL10 were 95.28 ± 7.32 nm and 0.100 ± 0.022, respectively, indicating the success of the experiment. Figure 7B (gray) also shows a suboptimal sample, which was prepared at 20 °C, where larger and more dispersed liposomes were obtained.

Figure 7C shows that the DOX EE of LTSL10. DOX EE should usually be around 80%. LTSLs prepared by the conventional method of lipid film hydration with extrusion (LF) are included for comparison, prepared as described elsewhere18. DOX EE of LTSL4 (LF) and LTSL10 (LF) showed decent DOX loading of around 70% and 50%, respectively. Annealing of as-prepared LTSL10 (step 2.11) is essential to enable DOX loading. In the absence of the annealing step, low DOX EE (< 20%) was persistent, regardless of incubation temperature (20 °C to 42 °C) and duration (1 to 24 h). This indicated the failure of LTSL10 to maintain a transmembrane pH gradient, where DOX was instead loaded passively or by adsorption. By annealing the as-prepared LTSL10, DOX EE increased significantly to a mean of 85%, indicating the success of the remote loading of DOX and the presence of the transmembrane pH gradient.

Figure 7D shows the DOX release profile of LTSL10. At 37 °C, the release of encapsulated DOX over 60 min was about 10%. In contrast, at 42 °C, all of the encapsulated DOX was released within 5 minutes, demonstrating the temperature-sensitivity of LTSL10. Similar results were observed with LTSL10 (LF) as a control.

Figure 8 shows the phase transition temperature (Tm) of LTSL10 characterized using differential scanning calorimetry (DSC). Dotted lines as tangent of the point of maximum slope, are added as a visual aid of the onset phase transition temperature (x-intercept of the tangent line). LTSL10 has a relatively broad phase transition with onset at 41.6 °C and peak at 42.6 °C. Similar results were observed with LTSL10 (LF), suggesting a minor difference between the preparation techniques. As a comparison, LTSL4 (LF) has a lower and sharper phase transition, in agreement with the literature1.

Figure 9 shows the characterization of LTSL10-ICG. Effect of initial ICG concentration on size (Figure 9A) and loading efficiencies of DOX and ICG (Figure 9B) are categorized into three concentration ranges. At low ICG concentration (ICG-to-lipid molar ratio of 0.003; initial concentration of 60 µM ICG and 20 mM lipid), Z-average, dispersity and DOX EE were similar to LTSL10 without ICG loading; ICG EE was around 75%. The efficient co-loading of DOX and ICG into LTSL10 can be achieved at this ICG concentration. At intermediate ICG concentrations, while the size and dispersity of the samples were satisfactory, both DOX and ICG EE were reduced. In particular, the decrease in DOX EE indicated the disruption of the liposomal membrane and thus, the pH-gradient. At high ICG concentrations, samples were again gelled; DOX and ICG EE were both significantly decreased.

LTSL10-ICG was irradiated with near-infrared laser (section 7) to induce photothermal heating and triggered the release of DOX (Figure 10). Upon laser irradiation, the sample first heated up to 49.7 °C with a gradual reduction of temperature. Subsequent laser irradiation increased the temperature to 36.7 °C. Quantification of the released DOX indicated that a complete burst release of encapsulated DOX was achieved after the first heating cycle. This was as expected since the temperature reached above 42 °C, in agreement with the DOX release profile shown in Figure 7D. In contrast, LTSL10 without ICG cannot provide photothermal heating, and thus did not release DOX upon laser irradiation.

Figure 1
Figure 1: Photograph of the syringe pumps setup. The "To Network" port of the master pump (Pump 01) is attached to the "To Computer" of the secondary pump (Pump 02; yellow); the "To Computer" port of the master pump is attached to the RS-232 port of the computer (blue). Please click here to view a larger version of this figure.

Figure 2
Figure 2: Photograph of the SHM setup. (A) Assembled view of the SHM setup. (B) Exploded view of the SHM setup. Inlets and outlets of the SHM are connected to tubing using a nut and ferrule. The tubing of both inlets is extended by a longer tubing with nut and ferrule on each end, terminated by a female Luer adapter using a union assembly. Please click here to view a larger version of this figure.

Figure 3
Figure 3: Interface of the pump control software. The two syringe pumps should be detected automatically upon initiating the software; otherwise, click Pumps on the top left corner and Search for pumps. Parameters to be configured are highlighted in red boxes. Please click here to view a larger version of this figure.

Figure 4
Figure 4: Photograph of the syringe pumps and installation of a syringe. (A) Syringe retainer bracket and syringe retainer thumbscrew (2, one on each side) of the syringe pump (yellow box). Pusher block, adjustment thumbscrew, and drive-nut button (white button on the left) of the syringe pump (green box). (B) Position of an installed syringe on the syringe pump. Please click here to view a larger version of this figure.

Figure 5
Figure 5: Photograph of the assembly of syringes, heating element and SHM. (A) Assembly of syringes, heating tape, and thermostat. (B) Assembly of syringes, heating tape, thermostat, and SHM. Please click here to view a larger version of this figure.

Figure 6
Figure 6: Photograph of the laser setup. (A) Photograph of the fiber coupled laser system during operation. (B) Position of the collimator 5 cm above the 96-well plate. Please click here to view a larger version of this figure.

Figure 7
Figure 7: Characterization of LTSL10. (A) Photograph of (left) LTSL10 and (right) LTSL4, before dialysis. LTSL10 appeared as clear and non-viscous liquid, while LTSL4 was gel-like and viscous. (B) Z-average diameter and dispersity of 10 mM of LTSL10 prepared at 20 and 51 °C. Solid bars and open circles (○) indicate Z-average diameter and dispersity, respectively. (C) DOX EE of LTSL4 (LF; white), LTSL10 (LF; white), and LTSL10 before (gray) and after annealing (red). (D) DOX release of LTSL10 (circle) and LTSL10 (LF; cross) at 37 °C (black) and 42 °C (red). Data represent mean ± SD of at least three independent experiments. ***p < 0.001; two-tailed unpaired t-tests. Please click here to view a larger version of this figure.

Figure 8
Figure 8: Thermal properties of LTSL10. Thermographs of LTSL4 (LF), LTSL10 (LF), and LTSL10 characterized by DSC. Dotted lines are added as a visual aid of the onset phase transition temperature. Data represent the mean of at least three independent experiments. Please click here to view a larger version of this figure.

Figure 9
Figure 9: Characterization of LTSL10-ICG. (A) Z-average diameter (red) and dispersity (blue) of ICG-loaded LTSL10. (B) DOX EE (red) and ICG EE (green) of ICG-loaded LTSL10. Data represent the mean ± SD of at least three independent experiments. Please click here to view a larger version of this figure.

Figure 10
Figure 10: Laser-induced photothermal heating triggered the release of DOX -loaded LTSL10 and LTSL10-ICG. (A) The temperature of irradiated samples and (B) DOX release of DOX-loaded LTSL10 (blue) and LTSL10-ICG (red) during the laser-induced photothermal heating. Data represent the mean ± SD of at least three independent experiments. Please click here to view a larger version of this figure.

Table 1: Lipid mixtures, buffers, and stock solutions.

20 mM LTSL10 lipid mixture
(DPPC/MSPC/DSPE-PEG2000; 80/10/10, molar ratio)
For 1 mL of 20 mM LTSL10 lipid mixture:
Dissolve 11.75 mg of DPPC with 210 µL of MSPC ethanol stock (5 mg/mL), 558 µL DSPE-PEG2000 ethanol stock (10 mg/mL), and 232 µL of ethanol.
Alternatively, equivalent amount of MSPC (1.05 mg) and DSPE-PEG2000 (5.58 mg) can be added as powder.
20 mM LTSL10-ICG lipid mixture
(DPPC/MSPC/DSPE-PEG2000; 80/10/10, molar ratio;
ICG-to-lipid molar ratio = 0.003)
For 1 mL of 20 mM ICG-loaded LTSL10 lipid mixture:
Dissolve 11.75 mg of DPPC with 210 µL of MSPC ethanol stock (5 mg/mL), 558 µL DSPE-PEG2000 ethanol stock (10 mg/mL), and 46.5 µL of ICG ethanol stock (1 mg/mL) and 185.5 µL of ethanol.
Alternatively, equivalent amount of MSPC (1.05 mg) and DSPE-PEG2000 (5.58 mg) can be added as powder.
Contains 60 µM of ICG and 20 mM of lipid.
240 mM Ammonium sulfate solution (NH4)2SO4, pH 5.4 Dissolve 31.71 g of (NH4)2SO4 per L of deionized water.
The pH of the solution is natively 5.4, additional pH adjustment is not required.
Doxorubicin solution (DOX), 1 mg/mL Dissolve 1 mg of DOX per mL of deionized water.
DSPE-PEG2000 ethanol stock, 10 mg/mL Dissolve 10 mg of DSPE-PEG2000 per mL of ethanol.
HEPES-buffered saline (HBS), pH 7.4 Dissolve 8.0 g of NaCl and 4.766 g of HEPES per L of deionized water.
Adjust pH to 7.4 with 2.5 M NaOH solution.
ICG ethanol stock, 1 mg/mL Dissolve 1 mg of ICG per mL of ethanol.
MSPC ethanol stock, 5 mg/mL Dissolve 5 mg of DPPC per mL of ethanol.

Subscription Required. Please recommend JoVE to your librarian.

Discussion

The presented protocol describes the preparation of low temperature-sensitive liposomes (LTSLs) using a staggered herringbone micromixer (SHM). The LTSL10 formulation enables temperature-triggered burst release of doxorubicin within 5 minutes at a clinically attainable hyperthermic temperature of 42 °C. Indocyanine green (ICG) can also be co-loaded for photothermal heating triggered the release of DOX. The method relies on: (i) self-assembly of phospholipids into liposomes under a homogenized solvent environment provided by the rapid, chaotic mixing of ethanol and ammonium sulfate solution in the SHM11; (ii) annealing of the liposomes to preserve the integrity of the lipid bilayer essential for DOX loading; and (iii) remote loading of DOX into LTSLs by an ammonium sulfate pH-gradient17. Since the equipment utilized in this protocol is commercially available off-the-shelf and the parameters are optimized, this approach is manageable for users without prior knowledge or microfluidics experience.

One of the most critical steps within the protocol is to ensure the whole assembly is properly secured and fluid can be properly infused (step 2.5). Since the reproducibility of the self-assembly of liposomes relies on a homogenized solvent environment, any instability, such as dislodgement of syringes or introduction of air bubbles, will disturb the stability of the fluid flow and result in suboptimal liposome size and dispersity. This is also the rationale behind step 2.8, where the volume, consisting of the fluid initially occupying the channel and before a stable flow is reached, should be disposed of.

A second critical step for a successful experiment is the annealing step, to enable high DOX EE (step 2.11). In cholesterol-free liposomes, micelle-forming membrane components (i.e. MSPC and DSPE-PEG2000) will accumulate at grain boundaries with a high degree of defects to accommodate a high membrane curvature2. These arrangements thermodynamically favor the formation and stabilize membrane pores, opened liposomes, or bilayer discs. The low DOX EE of LTSL10 without annealing suggested that porous structures existed even below Tm, resulting in the absence of the pH gradient required for DOX loading (Figure 7C). The premature formation of pores below Tm was not observed for LTSLs prepared by lipid film (LTSL4 (LF) and LTSL10 (LF)), where annealing is not required. Furthermore, cholesterol-containing formulations prepared by microfluidics also do not require annealing8. It is, therefore, speculated that the premature formation of pores is a combined effect of the presence of ethanol during the preparation and the lack of cholesterol in the lipid bilayer. Structural defects within the bilayer membrane have been reported to be eliminated by annealing the liposomes above Tm, allowing lipid molecules to redistribute homogeneously and defects to be corrected19. In addition, the annealing process is an irreversible process where annealed liposomes returning to a temperature below Tm do not recreate leaky vesicles19, in agreement with the annealed LTSL10.

The nature of microfluidic preparation of liposome is a nanoprecipitation process, which requires two miscible solvents with distinctive solubility for the lipids: typically, ethanol (as a lipid solvent) and aqueous solution (as a lipid non-solvent). Thus, the presence of ethanol is unavoidable. Therefore, formulations that are sensitive or prone to alcohol-induced interdigitation13, such as cholesterol-free liposomes20, may require modification of the formulation or re-optimization of the protocol. As demonstrated with the preparation of LTSL4, the highly viscous gel was obtained, (Figure 7A), which was likely due to the formation of an interdigitated gel phase15. On the other hand, LTSL10, with its higher polymer concentration that prevents interdigitation21, was prepared successfully. Consequently, an ethanol removal procedure also must be performed; here, it was removed simply by dialysis. While on-chip continuous purification techniques such as tangential flow filtration (capable of both ethanol removal and buffer exchange) have been developed22,23, their implementation (as one-chip or modular) are beyond the aim of this protocol. Nonetheless, in the future, we expect these modular or standardized designs to be optimized and increased in availability, streamlining the microfluidic production process.

Another limitation of the protocol is the sample loss due to the travelling distance of the liquids, namely the initial waste volume (step 2.8) as well as the last fraction of solutions that would be injected but wouldn't reach the outlet. These sample losses are almost unavoidable and may contribute to a significant portion of the preparation volume at bench-scale production, especially when a small volume or precious samples are to be prepared. When necessary, the lipid recovery could be quantified by high performance liquid chromatography-evaporative light scattering detector method that enables rapid quantification of lipid concentrations24. However, once the process is optimized and scaled up, such as by using a larger syringe or fluid reservoir, the throughput could be further scaled up and sample losses would be less significant.

The main difference between this method and the existing preparation method is that liposomes are self-assembled in a controllable solvent environment in a high-throughput, continuous manner. The lipid film method is a batch manufacturing process and requires size homogenization. While very feasible at a bench-scale, it remains challenging to scale up for clinical production. Within existing microfluidic techniques, for instance, microfluidic hydrodynamic focusing, SHM offers a shorter mixing timescale11 and a greater throughput (in the range of mL/min) with lower dilution factor; notwithstanding the preparation of LTSLs has not been reported using other microfluidic devices so far. The major advantage of our approach is the high-throughput, scalable production of thermosensitive liposomes.

Thus far, our microfluidic protocol offers continuous production of LTSL10 with drug loading capability. Payloads other than DOX and ICG are also viable. However, ethanol removal by dialysis, drug remote loading, and purification by SEC column remain as batch processes and are the bottlenecks of the overall formulation process. Future development could focus on utilizing microfluidic approaches (such as tangential flow filtration) to enhance the throughput of these downstream processes and increase the scalability of the protocol.

Subscription Required. Please recommend JoVE to your librarian.

Disclosures

The authors have nothing to disclose.

Acknowledgments

We thank Prostate Cancer UK (CDF-12-002 Fellowship), and the Engineering and Physical Sciences Research Council (EPSRC) (EP/M008657/1) for funding.

Materials

Name Company Catalog Number Comments
1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) Lipoid PC 16:0/16:0 (DPPC)
1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(polyethylene glycol)-2000] (DSPE-PEG2000) Lipoid PE 18:0/18:0-PEG 2000
(MPEG 2000-DSPE)
1-stearoyl-2-hydroxy-sn-glycero-3-phosphocholine (MSPC) Avanti Polar Lipid 855775P-500MG Distributed by Sigma-Adrich; also known as Lyso 16:0 PC
(Not to be confused with 14:0/18:0 PC, which is also termed MSPC)
4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) Sigma-Aldrich H3375-100G
Adapters, Female Luer Lock to 1/4"-28UNF IDEX Health & Science P-624 Requires 2 units. For the inlets
Adapters, Union Assembly, 1/4"-28UNF IDEX Health & Science P-630 Requires 2 units. (One unit included 2 nuts and 2 ferrules)
Ammonium Sulfate ((NH4)2SO4) Sigma-Aldrich 31119-1KG-M
Bijou vial VWR 216-0980 7 mL, clear, polystyrene vial
Centrifugal Filter Unit Sigma-Aldrich UFC801008 10 kDa MWCO, Amicon Ultra-4 Centrifugal Filter Unit
Centrifuge ThermoFisher Scientific Heraeus Megafuge 8R With HIGHConic III Fixed Angle Rotor
Cuvette Fisher Scientific 11602609 Disposable polystyrene cuvette, low volume, for DLS measurement
Dialysis Kit - Pur-A-Lyzer Maxi Sigma-Aldrich PURX12015-1KT 12-14 kDa MWCO
Dimethyl Sulfoxide (DMSO) Sigma-Aldrich 34943-1L-M
DLS Instrument Malvern Panalytical Zetasizer Nano ZS90
Doxorubicin Hydrochloride (DOX) Apollo Scientific BID0120
DSC Instrument TA Instruments TA Q200 DSC
DSC Tzero Hermetic Lids TA Instruments 901684.901 For DSC measurement
DSC Tzero Pans TA Instruments 901683.901 For DSC measurement
DSC Tzero Sample Press Kit TA Instruments 901600.901 For DSC measurement
Ethanol VWR 20821.330 Absolute, ≥99.8%
FC-808 Fibre Coupled Laser System CNI Optoelectronics Tech FC-808-8W-181315 FOC-01-B Fiber Collimator included.
Ferrule, 1/4"-28UNF to 1/16" OD IDEX Health & Science P-200 For the outlet
Fibre Optic Temperature Probe Osensa PRB-G40
Glass Staggered Herringbone Micromixer (SHM) Darwin Microfluidics Herringbone Mixer - Glass Chip
Heating Tape Omega DHT052020LD Can be replaced by other syringe heater such as "HTC" or "SRT series" for slower heating. Manual wiring to a 3-pin plug required for 240V models
Indocyanine Green Adooq A10473-100 Distributed by Bioquote Limited (U.K.)
Luer-lock Syringe, 5 mL VWR 613-2043 Hanke Sass Wolf SOFT-JECT 3-piece syringes, O.D. 12.45 mm
Microplate Reader BMG Labtech FLUOstar Omega Installed with 485 nm (exictation) and 590 nm (emission) filters
Microplate, 96-well, Black, Flat-bottom ThermoFisher Scientific 611F96BK For fluorescence measurement in microplate reader
Microplate, 96-well, Clear, Flat-bottom Grenier 655101 For absorbance measurement microplate reader
Nut, 1/4"-28UNF to 1/16" OD IDEX Health & Science P-245 For the outlet
PC to Pump Network Cable for Aladdin, 7ft World Precision Instruments NE-PC7 Optional: Syringe pumps can be operated manually
Pump control software - SyringePumpPro Software License for 2 World Precision Instruments SYRINGE-PUMP-PRO-02 Optional: Syringe pumps can be operated manually
Pump to Pump Network Cable for Aladdin, 7 ft World Precision Instruments NE-NET7 Optional: Syringe pumps can be operated manually
Size exclusion chromatography (SEC) column GE Life Science 17085101 Sephadex G-25 resin in PD-10 Desalting Columns
Sodium chloride (NaCl) Sigma-Aldrich 31434-1KG-M
Sodium hydroxide (NaOH) Sigma-Aldrich S5881-500G
Syringe Pumps & Cable (DUAL-PUMP-NE-1000) World Precision Instruments ALADDIN2-220/AL1000-220
Thermostat Temperature Controller Inkbird ITC-308 Can be replaced by other syringe heater kit/thermostat
Triton X-100 Sigma-Aldrich X100-100ML
Tubing, ETFE (1/16" OD) IDEX Health & Science 1516
USB To RS-232 Converter World Precision Instruments CBL-USB-232 Optional: For computer without RS-232 port
Water Bath Grant Instruments Ltd. JB Nova 12

DOWNLOAD MATERIALS LIST

References

  1. Needham, D., Park, J., Wright, A. M., Tong, J. Materials characterization of the low temperature sensitive liposome (LTSL): effects of the lipid composition (lysolipid and DSPE-PEG2000) on the thermal transition and release of doxorubicin. Faraday Discussions. 161, 515-534 (2013).
  2. Ickenstein, L. M., Arfvidsson, M. C., Needham, D., Mayer, L. D., Edwards, K. Disc formation in cholesterol-free liposomes during phase transition. Biochimica et Biophysica Acta - Biomembranes. 1614, (2), 135-138 (2003).
  3. Valencia, P. M., Farokhzad, O. C., Karnik, R., Langer, R. Microfluidic technologies for accelerating the clinical translation of nanoparticles. Nature Nanotechnology. 7, (10), 623-629 (2012).
  4. Chen, D., et al. Rapid discovery of potent siRNA-containing lipid nanoparticles enabled by controlled microfluidic formulation. Journal of the American Chemical Society. 134, (16), 6948-6951 (2012).
  5. Forbes, N., et al. Rapid and scale-independent microfluidic manufacture of liposomes entrapping protein incorporating in-line purification and at-line size monitoring. International Journal of Pharmaceutics. 556, 68-81 (2019).
  6. Dittrich, P. S., Manz, A. Lab-on-a-chip: microfluidics in drug discovery. Nature Reviews Drug Discovery. 5, (3), 210-218 (2006).
  7. Capretto, L., Carugo, D., Mazzitelli, S., Nastruzzi, C., Zhang, X. Microfluidic and lab-on-a-chip preparation routes for organic nanoparticles and vesicular systems for nanomedicine applications. Advanced Drug Delivery Reviews. 65, (11-12), 1496-1532 (2013).
  8. Cheung, C. C. L., Al-Jamal, W. T. Sterically stabilized liposomes production using staggered herringbone micromixer: Effect of lipid composition and PEG-lipid content. International Journal of Pharmaceutics. 566, 687-696 (2019).
  9. Suh, Y. K., Kang, S. A. Review on Mixing in Microfluidics. Micromachines. 1, (3), 82-111 (2010).
  10. Jahn, A., Vreeland, W. N., Gaitan, M., Locascio, L. E. Controlled Vesicle Self-Assembly in Microfluidic Channels with Hydrodynamic Focusing. Journal of the American Chemical Society. 126, (9), 2674-2675 (2004).
  11. Stroock, A. D. Chaotic Mixer for Microchannels. Science. 295, (5555), 647-651 (2002).
  12. Belliveau, N. M., et al. Microfluidic Synthesis of Highly Potent Limit-size Lipid Nanoparticles for In Vivo Delivery of siRNA. Molecular Therapy - Nucleic Acids. 1, (8), 37 (2012).
  13. Patra, M., et al. Under the influence of alcohol: The effect of ethanol and methanol on lipid bilayers. Biophysical Journal. 90, (4), 1121-1135 (2006).
  14. Komatsu, H., Rowe, E. S., Rowe, E. S. Effect of Cholesterol on the Ethanol-Induced Interdigitated Gel Phase in Phosphatidylcholine: Use of Fluorophore Pyrene-Labeled Phosphatidylcholine. Biochemistry. 30, (9), 2463-2470 (1991).
  15. Lu, J., Hao, Y., Chen, J. Effect of Cholesterol on the in Lysophosphatidylcholine Formation of an Interdigitated Gel Phase and Phosphatidylcholine Binary. Journal of Biochemistry. 129, (6), 891-898 (2001).
  16. Vanegas, J. M., Contreras, M. F., Faller, R., Longo, M. L. Role of unsaturated lipid and ergosterol in ethanol tolerance of model yeast biomembranes. Biophysical Journal. 102, (3), 507-516 (2012).
  17. Haran, G., Cohen, R., Bar, L. K., Barenholz, Y. Transmembrane ammonium sulfate gradients in liposomes produce efficient and stable entrapment of amphipathic weak bases. Biochimica et Biophysica Acta (BBA) - Biomembranes. 1151, (2), 201-215 (1993).
  18. Sadeghi, N., et al. Influence of cholesterol inclusion on the doxorubicin release characteristics of lysolipid-based thermosensitive liposomes. International Journal of Pharmaceutics. 548, (2), 778-782 (2018).
  19. Lawaczeck, R., Kainosho, M., Chan, S. I. The formation and annealing of structural defects in lipid bilayer vesicles. Biochimica et Biophysica Acta (BBA) - Biomembranes. 443, (3), 313-330 (1976).
  20. Komatsu, H., Okada, S. Ethanol-induced aggregation and fusion of small phosphatidylcholine liposome: participation of interdigitated membrane formation in their processes. BBA - Biomembranes. 1235, (2), 270-280 (1995).
  21. Marsh, D., Bartucci, R., Sportelli, L. Lipid membranes with grafted polymers: physicochemical aspects. Biochimica et Biophysica Acta (BBA) - Biomembranes. 1615, (1-2), 33-59 (2003).
  22. Hood, R. R., Vreeland, W. N., DeVoe, D. L. Microfluidic remote loading for rapid single-step liposomal drug preparation. Lab on a Chip. 14, (17), 3359-3367 (2014).
  23. Dalwadi, G., Benson, H. A. E., Chen, Y. Comparison of diafiltration and tangential flow filtration for purification of nanoparticle suspensions. Pharmaceutical Research. (2005).
  24. Roces, C., Kastner, E., Stone, P., Lowry, D., Perrie, Y. Rapid Quantification and Validation of Lipid Concentrations within Liposomes. Pharmaceutics. 8, (3), 29 (2016).
  25. Kim, S. -H., Kim, J. W., Kim, D. -H., Han, S. -H., Weitz, D. A. Enhanced-throughput production of polymersomes using a parallelized capillary microfluidic device. Microfluidics and Nanofluidics. 14, (3-4), 509-514 (2013).

Comments

0 Comments


    Post a Question / Comment / Request

    You must be signed in to post a comment. Please or create an account.

    Usage Statistics