Mechanisms of enhanced antiglioma efficacy of polysorbate 80‐modified paclitaxel‐loaded PLGA nanoparticles by focused ultrasound

Abstract The presence of blood‐brain barrier (BBB) greatly limits the availability of drugs and their efficacy against glioma. Focused ultrasound (FUS) can induce transient and local BBB opening for enhanced drug delivery. Here, we developed polysorbate 80‐modified paclitaxel‐loaded PLGA nanoparticles (PS‐80‐PTX‐NPs, PPNP) and examined the enhanced local delivery into the brain for glioma treatment by combining with FUS. Our result showed PPNP had good stability, fast drug release rate and significant toxicity to glioma cells. Combined with FUS, PPNP showed a stronger BBB permeation efficiency both in the in vitro and in vivo BBB models. Mechanism studies revealed the disrupted tight junction, reduced P‐glycoprotein expression and ApoE‐dependent PS‐80 permeation collectively contribute to the enhanced drug delivery, resulting in significantly stronger antitumour efficacy and longer survival time in the tumour‐bearing mice. Our study provided a new strategy to efficiently and locally deliver drugs into the brain to treat glioma.

potential neurotoxins by the way of an active transport mechanism mediated by P-glycoprotein. In some cases, drugs have to be administered directly into the cerebrospinal fluid (CSF), where it can enter the brain by crossing the blood-cerebrospinal fluid barrier. 8,9 However, not all drugs that are delivered directly to the CSF can effectively penetrate the CSF barrier and enter the brain, severely restricting the availability for most anticancer agents. 10,11 Therefore, it is desirable to develop new techniques to selectively deliver a chemotherapeutic agent into brain tumour for glioma therapy. 12 To date, many strategies have been applied to deliver drug across the BBB, including the intra-arterial injection of hyperosmotic fluids (such as mannitol) to induce osmotic blood-brain barrier disruption, the design of drug delivery systems targeted to transport receptors highly expressed at the BBB (such as transferrin receptor) or using a naturally occurring compound as adjuvant (such as borneol) to improve drug delivery to the brain. [13][14][15][16][17] Recently, polysorbate 80 (PS-80), known as a surfactant, was identified to be able to enhance the BBB transport by modifying nanoparticles. [18][19][20] Researchers demonstrated the possible mechanism is that PS-80 can selectively adsorb apolipoprotein E (ApoE) on the brain capillary endothelium and induce low-density lipoprotein receptor-related protein (LRP) mediated transcytosis across the BBB. [21][22][23] However, drug delivery system mediated with PS-80 often produces whole brain but not local drug delivery, resulting in undesired drug accumulation at health sites and inadequate drug concentrations in the tumour region. In addition, PS-80modified nanoparticles would be ineffective for these glioma patients who do not express or only express low level of ApoE proteins.
Recently, focused ultrasound (FUS) combined with microbubbles (MBs) provides a promising approach to achieve local delivery into the brain. 12,24 Taking advantage of MB cavitation induced by FUS, transient and reversible BBB opening can be induced, without obvious damage to the cellular ultrastructure. 25,26 Also, US transducer is portable and US beams can be easily focused on the brain tumour site, giving the approach great advantages in the local drug delivery into brain. So far, FUS combined with MBs have successfully delivered drugs, genes, proteins and cells for brain tumour treatment. 13,[27][28][29][30][31] In this study, paclitaxel (PTX), a chemotherapeutic agent that can hardly cross the BBB, was chosen as a model drug.
Herein, we fabricated PS-80-modified PTX-loaded PLGA nanoparticles (PS-80-PTX-PLGA, PPNP) and locally delivered it into brain for the enhanced antitumour efficacy through combining with FUS irradiation. The preparation, characterization, in vitro and in vivo drug delivery efficiency across BBB, and antitumour effect of PPNP were investigated. Importantly, the mechanisms of penetration through BBB were also explored in the study. All other reagents were of analytical grade.

| Preparation of PPNP
PTX-loaded PLGA nanoparticles (PTX-PLGA-NPs, PNP) were prepared following the emulsion solvent evaporation method with slight modifications. Briefly, PTX and PLGA, dissolved in 1 mL of dichloromethane, were added dropwise to 5% w/v aqueous PVA solution (5 mL), followed by sonication for 5 minutes (30% amplitude, 50 ms on, 10 ms off) to generate water/oil emulsion. This emulsion was stirred at 700 rpm for 6-8 hours until complete removal of organic solvent, followed by centrifugation at 9 500 g for 10 minutes at 4°C.
The pellet obtained was resuspended in water and washed twice. The nanoparticles were freeze-dried and stored at 4°C under anhydrous conditions. To obtain the PS-80-coated PTX-loaded PLGA, PNP formulation was treated as the procedure described by Wilson et al. 32 In brief, PNP were resuspended in water and PS-80 was added to 1% PS-80 final concentration; then, the mixture was incubated for 30 minutes. The resulting PPNP were collected by centrifugation and finally lyophilized. For visualization, PS-80-coated IR-780-loaded PLGA NPs (PS-80-IR-780-NPs, PINP) and PS-80-coated Coumarin-6loaded PLGA NPs (PS-80-Cou-6-NPs, PCNP) were also fabricated using IR-780 or Coumarin-6 instead of PTX drugs.

| In vitro PTX release
In vitro PTX release from PPNP was studied in 0.02 mol/L phosphatebuffered saline (PBS, pH 7.4) with 0.5% PS-80. In a dialysis tubing, approximately PPNP equivalent to 1 mg PTX were suspended in 2 mL distilled water. Then, the dialysis tubing was placed into 250 mL medium at 37°C, followed by continuous shaking at 100 rpm and 37°C in a constant temperature shaking bath. Samples (1 mL) were withdrawn at predetermined time intervals for HPLC analysis and replaced with an equal volume of fresh PBS. The release rate (RR) was calculated according to the equation: RR% = (W i /W total ) 9 100%, where W i is the measured amount of PTX at the indicated time point and W total is the total PTX amount in the same volume of NPs suspensions.

| Cell viability
In order to investigate toxicity to tumour cells, we treated U87 cells with PPNP or free PTX at 0, 3, 6, 9, 12 and 24 lg/mL of PTX concentrations. After 3 days, the media were replaced with 110 lL of fresh solution (100 lL medium plus 10 lL of CCK-8 solution). After 45 min of incubation, the CCK-8 assay was performed. Cell viability in each group was expressed as a percentage relative to that of the untreated control.

| In vitro BBB model
bEnd.3 cells were seeded into transwell to establish the in vitro bloodbrain barrier model. Briefly, transwell permeable supports with 0.4-lm porous membranes (Corning Inc., NY, USA) were used. bEnd.3 cells were seeded into the support at 5 9 10 4 cells per support and incubated for 48 hours. The culture media was replaced every 3 days. After 1 week, sodium fluorescein was used as probe to measure the paracellular or junctional permeability. To calculate endothelial permeability coefficient (Pe), flux of the sodium fluorescein across monolayers and cell-free inserts is measured 33

| Ultrasound equipment
The sonications were performed using a single-element focused

| In vitro BBB penetration
A given amount of CNP or PCNP (equal Coumarin-6) were added into transwell insert which had been developed the in vitro BBB model. To ensure a close contact between MBs and cells, the transwells were sealed with sealing film and inverted. The transwell was exposed under 1.26 W/cm 2 for 60 seconds. The microbubbles that have good stability were obtained according to our previous study. 35

| In vivo BBB opening
Before sonication, the mice were anaesthetized with 1.5% isoflurane (RWD Life Science, Co., Ltd., Shenzhen, China) and placed on a heat-

| Immunofluorescence assays
Immunofluorescence staining was performed to detect expression and localization of ZO-1 and P-glycoprotein (P-gp). The cells or brain slices were fixed with 4% paraformaldehyde solution and permeabilized with 0.1% Triton X-100, followed by blocking with PBS containing 1% BSA for 1 hour at room temperature. After that, the cells or slices were incubated with goat polyclonal anti-ZO-  bars. The skull was exposed through a 1-cm midline incision, and a burr hole was made 2 mm to the right of the bregma and 1.3 mm posterior to the coronal suture using a syringe with a 1-mm tip.

| Animal models
Using a microsyringe attached to the stereotaxic frame, 5 lL of U87-Luc cells (1 9 10 5 cells) was injected over 5 min at a depth of 3 mm. After the injection, the syringe was kept in place for 1 minute prior to withdrawal and the incision was closed with Vetbond skin glue. The animals were monitored for tumour growth using an in vivo imaging system (IVIS Lumina II, Caliper, USA).

| In vivo antitumour efficacy
Ten days after intracranial inoculation of U87-Luc cells, the in vivo bio-

| Histological analysis
At the end of the treatment, the mice were sacrificed and tumours were excised for histological analysis. Organ tissues including liver, heart, lung, spleen and kidney were collected for haematoxylin and eosin (H&E) staining. All tissues were prepared and sectioned to standard procedures. For the H&E staining, tissues were then embedded in paraffin wax, cut into 5-lm-thick sections and stained with haematoxylin and eosin dyes. Microscopic images of the tissues were acquired using an optical microscope. For immunofluorescence staining, tissues were embedded in OCT, cut into 8-lm-thick sections. Then slides were stained with anti-ki67 antibodies to assess tumour proliferation. Tumour apoptosis was also assessed by TUNEL assay according to the product instruction.

| Statistical analysis
Statistical analysis was carried out with SPSS version 23.0 (SPSS Inc., Chicago, IL). All values were expressed as means AE SD. Data were analysed by ANOVA, and then, differences among the means were analysed using the Tukey-Kramer multiple comparison test. Differences were considered significant at *P < .05 and very significant at **P < .001. Figure 1A illustrates the schematic diagram to fabricate PPNP through an emulsion/solvent evaporation method, followed by coating with PS-80. Dynamic light scanning (DLS) measurement showed that the average size diameter of PPNP was 170.5 AE 7.1 nm with a 0.11 AE 0.03 polydiserpersity index (PDI) ( Figure 1B). The zeta potential of PPNP was À54.7 AE 0.46 mV. The TEM and AFM images revealed that PPNP appeared a well-defined spherical morphology and comparative particle size ( Figure 1C, D). The stability over time at 4°C were investigated for PPNP, revealing there were not significant changes on the particle size, PDI and zeta potential for 4 weeks (Figure 2A Figure 2D). Obviously, higher the PTX encapsulation efficiency and drug loading rate can be achieved when the ratio of PTX to PLGA was at 1:5 or 1:10. Based on the above results, the final ratio of PTX and PLGA at 1:10 was chose for the further studies. Figure 2E demonstrates the release rate of PTX from PPNP was faster than that from PNP which was not coated with PS-80. About 50% of drug was released from PPNP after 24 hours and nearly 80% after 48 hours, indicating that PS-80 as a cosolvent/surfactant could favour drug release from NPs. The cell viability revealed PPNP had dose-dependent cytotoxicity to U87 glioma cells, having a similar toxicity with free PTX at same concentration (P > .05; Figure 2F).

| Drug delivery across BBB
To investigate the drug delivery across BBB, we first developed the in vitro BBB model ( Figure

| Mechanisms of FUS-assisted BBB opening
Evans blue (EB) staining for the in vivo mouse brain treated with FUS is presented in the Figure 4B, confirming the successful BBB opening ( Figure 4A). H&E staining of the FUS-irradiated brain tissue did not observe the apparent pathological damages such as RBC extravasations, demonstrating the safe and feasible ultrasound parameters used for BBB opening in the study. As tight junction through BBB into the brain of ApoEÀ/À mice when FUS was utilized ( Figure 5C,D), revealing ApoE deficiency did not significantly disturb FUS-induced BBB penetration for nanoparticles.

| In vivo antiglioma efficacy
Next, the in vivo antiglioma efficacy was evaluated using of PNP or PPNP combined with or without FUS. U87-Luc glioma orthotopictransplantation mouse model was constructed and the tumour growth was recorded by an in vivo bioluminescence imaging machine. The experimental scheme over time is shown in Figure 6A.    Figure 1A). The resulting PPNP not only had good stability and fast drug release rate, but also had significant toxicity to U87 cells (Figure 2). ApoE is involved in the delivery of lipids to tumour cells and in the recycling of lipids by macrophages, raising the possibility that ApoEmediated transport in brain tumours. 45 Mechanisms study demonstrated ApoE proteins possess extraordinary features which is readily assembled with hydrophobic compounds via its compact hydrophobic units. These assemblies can then be converted to stable particles by protein-protein interactions via coiled coil regions which exist in ApoE structure. 46 To improve the drug local delivery and enhance chemotherapeutic efficacy against brain tumour, in this study, we utilized FUS-combined MB cavitation to induce the BBB opening, giving the PS-80 modified nanoparticles higher efficiency to cross the BBB. Data from the in vivo antitumour experiment demonstrated that PPNP + FUS group has the stronger antitumour efficacy when compared to the FUS + PNP or only PPNP groups (Figure 6). Longer survival time was also observed in tumour-bearing mice treated with PPNP + FUS, suggesting that FUS combined with PS-80 modification is significantly more efficient to deliver drugs into the brain tumours than either of them. Indeed, the enhanced antitumour effect was confirmed by histological analysis, revealing an increased apoptosis and decreased proliferation in tumour xenografts treated with PPNP + FUS (Figure 7). Thus, our study provides a new strategy to efficiently and locally deliver drugs into the brain to treat glioma through combining FUS with PS-80-modification of nanoparticles. As schematically shown in Figure S4, FUS can activate the MBs to produce cavitation, resulting in the disruption of TJs of BBB and enlargement of the endothelial cell gaps, by which PPNPs can enter into the brain.
Also, the down-regulated expression of P-gp from FUS cavitation would cause reduced drug efflux. In addition, PPNP can attach with ApoE receptor and then activate receptor-mediated endocytosis. All of the factors collectively contribute to the enhanced BBB permeation and antitumour efficacy.

| CONCLUSION S
In this study, we successfully fabricated PPNP and locally deliver these nano-drugs into the brain tumour. Also, we demonstrated the enhanced drug delivery mechanisms which can contribute to the temporary disruption TJs of BBB and reduction in P-gp expression from FUS-induced BBB opening and the PS-80-meadiated ApoEdependent permeation. In conclusion, our study provided a new strategy to efficiently and locally deliver drugs into the brain to treat glioma.