Thermosensitive Exosome–Liposome Hybrid Nanoparticle‐Mediated Chemoimmunotherapy for Improved Treatment of Metastatic Peritoneal Cancer

Abstract Metastatic peritoneal carcinoma (mPC) is a deadly disease without effective treatment. To improve treatment of this disease, a recently developed hyperthermic intraperitoneal chemotherapy (HIPEC) has emerged as the standard of care. However, the efficacy of this approach is limited by inefficient drug penetration and rapidly developed drug resistance. Herein, a nanotechnology approach is reported that is designed to improve drug delivery to mPC and to augment the efficacy of HIPEC through delivery of chemoimmunotherapy. First, the drug delivery efficiency of HIPEC is determined and it is found that chemotherapy agents cannot be efficiently delivered to large tumors nodules. To overcome the delivery hurdle, genetically engineered exosomes‐thermosensitive liposomes hybrid NPs, or gETL NPs, are then synthesized, and it is demonstrated that the NPs after intravenous administration efficiently penetrates into mPC tumors and releases payloads at the hypothermia condition of HIPEC. Last, it is shown that, when granulocyte‐macrophage colony‐stimulating factor (GM‐CSF) and docetaxel are co‐delivered, gETL NPs effectively inhibit tumor development and the efficacy is enhanced when HIPEC is co‐administered. The study provides a strategy to improve drug delivery to mPCs and offers a promising approach to improve treatment of the disease through combination of locoregional delivery of HIPEC and systemic delivery of chemoimmunotherapy via gETL NPs.


Results
The experiments were carried out in triplicate.

Figure S6. Characterization of cytotoxicity of the indicated nanoparticles CT26 cells (a) or HCoEpic cells (b).
Exos, wild type exosomes; gExos, exosomes with CD47&GM-CSF overexpression; Blank lips, blank thermosensitive liposomes; Blank NPs, engineered exosomes hybridized with blank liposomes. The data was presented as the mean ± SD from three repeated experiments, One-way ANOVA was used to determine statistical differences. (* p < 0.05). Figure S7. M2 to M1 repolarization of RAW264.7 macrophages induced by gETL NPs was evaluated by flow cytometry. The representative percentages of M1 macrophages (CD68 + CD86 + ) and M2 macrophages (CD68 + CD206 + ) were displayed here, the quantification results were analyzed and showed in Figure 5c and 5d, three independent experiments were repeated. Figure S8. Synergistic therapeutic benefits of G/D-gETL NPs and HIPEC in CT26-derived mPC xenografts (n=5). Abdominal circumference and ascites at pre-treatment (day 10) and at endpoint (day 30) were recorded, scale bar: 1 cm. Figure S9. gETL NPs combined with HIPEC displayed synergistic anti-tumor effect (n=5). Mice were killed at the endpoint of observation (day30), the histological morphology of tumor was analyzed by hematoxylin-eosin staining of tissue sections, and the apoptosis of tumor cells was analyzed by TUNEL staining. Scale bar: 200 μm.

Human tissues, cell lines, and animals
The tumor tissues which were used to construct PDX animal models were obtained from mPC patients from general surgery department of the Third Affiliated Hospital Committee, Sun Yat-Sen University.

Construction of mPC animal models
To construct patient derived xenografts (PDX) animal models of mPC [1] , tumor tissues that were obtained from mPC patients were cut into tissue fragments at size of 20 mm 3 and transplanted into the peritoneal cavity of BALB/c nude mice as soon as possible. For animal surgery, a 5 mm incision was performed on the middle abdomen of each anesthetized mouse. Then the tumor fragments were seeded into the lower peritoneal cavity, and the incision was closed by 6-0 nylon surgical sutures (Ethicon Inc., USA). The passage and amplification of tumors between mice were implemented through surgery as previously described, and the third generation of PDX model mice were used for experiment. The expression of HER2 and Ki-67 in tumors that obtained from animals was detected by human antibodies and compared with primary patient-derived tumors to confirm the homology.
To construct cells-derived xenografts (CDX) animal model of mPC [2] , each BALB/c mouse was intraperitoneally injected with 1×10 6 CT26 cells in logarithmic growth phase of.

HIPEC procedure
HIPEC procedure was performed in animals as reported [2] . Before operation, animals received inhalation anesthesia by 3% isoflurane with oxygen. Then, to build inflow and outflow channels, puncture needles were inserted into the left upper and the right lower abdomen, respectively. After connecting all pipes and setting the HIPEC equipment correctly, oxaliplatin was mixed with 100 mL 5% glucose solution and then added into the micro-perfusion system for HIPEC treatment for 1 hour, and the temperature of perfusion solution was controlled at 42 ℃ during the whole process.
In order to prevent perioperative infection, the needle was removed after treatment and the puncture wounds were sutured. The operations were completed in the super clean platform at specific pathogen free (SPF) animal laboratory center.

Preparation of thermosensitive liposomes
The thermosensitive liposomes were prepared by membrane hydration method as reported [3] .

Gene transfection
Three kinds of transfection expression vectors to achieve over-expression of CD47, GM-CSF and CD47&GM-CSF were acquired from VectorBuilder Inc. (Guangzhou, China). The plasmid identities were validated by sequencing and the different plasmids were packaged by lentiviruses. Then mouse embryonic fibroblasts (BALB/c 3T3) were transfected with lentiviruses [4] . Specifically, 3T3 cells with a cell density of 1×10 4 cells per well were planted in 12-well culture plate for 12 h, then lentiviruses were added into fresh mediums with dose at multiplicity of infection (MOI) of 10, and 10 μg/ml polybrene was also added into the transfection system. 12 hours later, culture medium was refreshed and cells were cultured for 48 hours. Transfected cells were screened out by streptomycin and cultured in the medium that was prepared by exosome-free FBS.

Purification and characterization of engineered exosomes
The purification and characterization of engineered exosomes were performed according to reported [5] . About 1× 10 7 fibroblasts was cultured by 30 ml exosome-free medium for 72 h, and the medium was collected and centrifuged at Samples were diluted by PBS to the appropriate concentration, the particle size and concentration of exosomes were detected by Flow Nano Analyzer (N30) (Nano FCM Inc., Xiamen, China).

Western blot analysis
The samples were separated on 10% Bis-Tris gel by electrophoretic (Bio-Rad) electrophoresis and transferred to 0.45 μm PVDF membrane. The membranes were sealed in 10% (w/v) skimmed milk powder (NC9022655; Thermo Fisher) at 37 °C for 2 h and incubated with primary antibodies overnight at 4 °C. Then membranes were washed by 0.2% Tween-20 containing PBS for 3 times. After incubated with secondary antibodies for 2 hours and washed 3 times, the membranes were developed with chemiluminescent reagents.

ELISA
The concentrations of GM-CSF in the fibroblasts supernatants, fibroblasts derived exosomes and NPs were determined by murine GM-CSF ELISA kit [4] . The concentrations of TNF-α and IFN-γ in tumor tissue homogenate and serum compared with control group were determined by murine TNF-α and IFN-γ ELISA kit.

Preparation and Characterization of NPs
NPs were prepared based on thermosensitive liposomes and engineered exosomes by membrane fusion technology using freeze-thaw method as reported. [6,7] Briefly, docetaxel-loaded liposomes and engineered exosomes were rapidly mixed at ratio of 1:1 (mol / mol) and incubated at 37 °C for 30 min, then the samples were frozen by liquid nitrogen for 5 min and thawed at room temperature for 15 min. After 3 freezethaw cycles were repeated, the hybrid NPs were formed via the lipid membrane fusion between the liposomes and the exosomes. In order to remove the free drugs, the samples were added into 3500 kDa dialysis bags and dialyzed in PBS for 8 h. To determine the presence of CD47 on the surface of NPs, sample was incubated with FITC-CD47 antibody at 37 °C for 30 min, 10-fold volume of PBS was added into the sample, then 100 kDa ultrafiltration tube was used to filter and concentrate the sample to original volume, 3 cycles of ultrafiltration were repeated to remove the free antibodies, then the sample was detected by Flow Nano Analyzer.

Evaluation of membrane fusion between exosomes and liposomes
The membrane fusion between liposomes and exosomes was confirmed by a two-way verification. Firstly, NBD-labeled liposomes were prepared by NBD-DSPE-PEG2000, and exosomes were labeled by CD9 immunomagnetic beads, then liposomes and exosomes were mixed with each other at ratio of 1:1(mol/mol).
Then the mixture was extracted in magnetic field and unfused liposomes were removed. Finally, samples including exosomes, liposomes, and mixture were observed by TEM. Samples were also used to treat HCT116 cells for 2 h and the fluorescence was observed by CLSM.
In order to quantitatively evaluate the fusion efficiency, the methods of immunomagnetic separation and FITC-labeled lipids for nanoflow analysis was adopted. In short, similar to previous two-way validation experiment, after fusion process, the mixtures containing hybrid vesicles and unfused exosomes were sorted by the immunomagnetic beads, then the fluorescence signal and particle counts of fused hybrid NPs was detected by Flow NanoAnalyzer. According to the accurate count results of nanoparticles, the efficiency of membrane fusion between exosomes and liposomes was evaluated.

Thermosensitivity of gETL NPs in vitro
The drug release behavior of NPs at different temperature in PBS with 10% FBS was determined by dialysis method. Briefly, 6 mg NPs were suspended in 5 mL

Cytotoxicity of gETL NPs
To test the cytotoxicity of blank vectors, CT26 cells or HCoEpic cells were seeded in 96-well plates and incubated with PBS, wild type exosomes, engineered exosomes, blank liposomes, and blank gETL NPs at various concentrations at 37 °C for 12 h.

In vivo biodistribution of gETL NPs
In order to study the biodistribution of gETL NPs, PDX model of mPC in BALB/c nude mice was constructed by using patient-derived xenograft, and Dir was loaded

Anti-tumor activity of gETL NPs combined with HIPEC in animal models
The anti-tumor effect of combination therapeutic that based on HIPEC and NPs After grouped, the ascites of each mouse was removed by abdominal puncture, and the volume of ascites was measured as ascites baseline. Then free docetaxel or NPs was injected through tail vein at an identical docetaxel dose of 1.5 mg/kg immediately.
And 12 hours later, 75 mg/m 2 oxaliplatin was dissolved in 100 mL 5% glucose solution and added into the micro

Evaluation of tumor burden by Peritoneal carcinosis index (PCI)
To evaluate the tumor burden of animals, PCI evaluation system that based on sizes and distributions of tumor, and ascites was adapted from literature report [2] . Briefly, the abdominal cavity was divided into four regions anatomically: region I for sub-diaphragm; region II for the surface of liver, spleen, stomach; region III for the surface of small intestine, colon, urogenital system, rectum and mesenterium; region IV for the surface of abdominal wall. The tumor burden of each region was scored according to the greatest diameter of the nodules. The scoring criteria was set as follows: score 0 for no tumor nodules; score 1 for nodule size ≥ 2 mm; score 2 for nodule size ≥ 2 mm and up to 5 mm; score 3 for nodule size ≥ 5 mm and up to 10 mm; score 5 for nodule size ≥ 10 mm. And another score 1 for appearance of ascites, the sum of all score was the PCI score for each animal (ranging from 0 to 21).

Analysis of hematological and histological
To explore hematological and histological analysis, blood samples were collected at day 30 and centrifuged at 2000 g for 15 min, then serum samples were tested for biochemical markers. All tumors and major organs were fixed by 4% paraformaldehyde, then tumors were dissected and sectioned for TUNEL staining, H&E staining, immunohistochemical staining, and immunofluorescence staining.
Organs including heart, lung, liver, spleen, and kidney were dissected and sectioned for H&E staining. Hematological and histological analysis were carried out by Servicebio (Wuhan, China).