Copper-64 Labeled PEGylated Exosomes for In Vivo Positron Emission Tomography and Enhanced Tumor Retention

Abstract

Exosomes have attracted tremendous attention due to their important role in physiology, pathology and oncology, and promising potential in biomedical applications. Although great efforts have been dedicated to investigating their biological properties and applications as natural cancer drug delivery systems, the systemic biodistribution of exosomes remains underexplored. In addition, exosome-based drug delivery is inevitably hindered by the robust liver clearance, leading to suboptimal tumor retention and therapeutic efficiency. In this study, we report one of the first examples using in vivo positron emission tomography (PET) for noninvasive monitoring of copper-64 (⁶⁴Cu)-radiolabeled polyethylene glycol (PEG)-modified exosomes, achieving excellent imaging quality and quantitative measurement of blood residence and tumor retention. PEGylation not only endowed exosomes with a superior pharmacokinetic profile and great accumulation in the tumor than traditionally reported native exosomes, but also reduced premature hepatic sequestration and clearance of exosomes, findings that promise enhanced therapeutic delivery efficacy and safety in future studies. More importantly, this study provides important guidelines about surface engineering, radiochemistry and molecular imaging in obtaining accurate and quantitative biodistribution information of exosomes, which may benefit future exploration in the realm of exosomes.

Graphical Abstract

INTRODUCTION

Exosomes, the naturally secreted lipid vesicles carrying myriad cellular proteins and genetic information, have attracted increasing attention since the first discovery of their biological function as intercellular messengers in 2007.^1–5 Recent efforts have revealed an important perspective on exosomes, showing their unique biogenesis and biological functions.⁶ Exosomes originate from multivesicular endosomes regulated by the endosomal sorting complexes required for transport (ESCRT) machinery.^7–9 A series of cellular membrane and cytosol proteins are contained in exosomes, including tetraspanins, heat shock proteins, lysosomal proteins, membrane transport and fusion proteins, cytoskeleton and metabolism-related proteins, and major histocompatibility complex (MHC) classes I and II molecules.^10–12 Exosomes are involved in intercellular communication, immune response, and cancer progression and metastasis, playing important roles in physiology, pathology, and oncology, and manifesting promise as novel theranostic devices.^12–15

Because of their biocompatibility and role in cancer progression, exosomes have been applied as a promising natural nanoplatform for drug delivery and immunotherapy.^{10, 11, 16, 17} Many types of anticancer drugs, including nucleotides, proteins and small molecules, have been loaded into various kinds of exosomes derived from human and murine cancer cells, lymphoma cells, embryonic fibroblasts, and immune cells.^{17, 18} Unlike traditional synthetic drug delivery systems whose applications are highly limited by vascular extravasation and cell internalization,¹⁹ exosomes are believed to be generated and also taken up by all cells, and therefore can potentially reach every therapeutic area,^{5, 20} emphasizing their promise for effective drug delivery. Significant presence of surface proteins on exosomes not only improves their stability but also endows them with organ tropism without necessitating external targeting moieties.²¹ In addition, exosomes play important roles in cellular immunity, angiogenesis and regeneration, which have inspired a novel therapeutic approach (i.e. exosome therapy) without the need to load drugs, showing great potential in applications from oncology to regenerative medicine.²² However, the majority of the exosomes injected systemically are cleared by the liver with suboptimal tumor retention,^{5, 14, 23} impeding their potential applications and clinical translation in cancer treatment. Considering that exosome production still suffers from remarkably low yield,¹⁶ an effective delivery means to bypass liver clearance and increase tumor retention is urgently needed to guarantee sufficient anticancer efficacy. Inspired by PEGylation of synthetic nanoparticles that greatly improves nanoparticle behavior in vivo,^24–29 we introduce PEGylated exosomes in this study, which exhibit enhanced in vivo pharmacokinetic profile and tumor retention.

Although systemic biodistribution is one of the most critical parameters affecting the efficacy of a drug delivery system, this information remains missing for exosomes. Previous studies attempting to delineate the biodistribution profile of exosomes relied heavily on optical imaging,³⁰, ³¹ which is neither quantitative nor accurate due to the limited tissue penetration of the light. Exosome-like vesicles were radiolabeled for single-photon emission computed tomography (SPECT) in few reported studies.^32–34 However, more quantitative and reliable elucidation of exosome biodistribution in the body remains an underexplored and crucial knowledge gap. In this study, radioisotope copper-64 (⁶⁴Cu) was labeled on the surface of exosomes via stable chelation and noninvasively monitored by longitudinal positron emission tomography (PET), an imaging approach with enhanced sensitivity, spatiotemporal resolution and quantifiability than SPECT.³⁵ Since PET is widely used in clinical oncology,^36–39 the incorporation of a PET isotope with exosomes can facilitate future clinical and translational studies of exosome-based drug delivery systems.

RESULTS AND DISCUSSION

Exosomes were isolated from culture media of 4T1 breast cancer cells using the widely employed “gold standard” differential ultracentrifugation method.⁴⁰ In brief, the culture media was sequentially centrifuged at speeds of 1800 × g and 2800 × g, passed through a 0.22 μm filter to remove the floating cells and cell debris, and subsequently ultracentrifuged at 100,000 × g to precipitate the exosomes. The isolation of the exosomes was confirmed by ultraviolet-visible (UV-VIS) spectrometry (Figure S1), and the protein concentration was measured by spectrophotometry. Compared with other isolation approaches including ultrafiltration and immunoisolation, which may deform the structure or alter the biological function of exosomes, ultracentrifugation is a cheap, easy and convenient method to retain the structure and function of exosomes, although the production yield is low.¹⁷, ⁴¹ The isolated exosomes were conjugated with bifunctional chelator 1,4,7-triazacyclononane-1,4,7-triacetic acid (NOTA) for ⁶⁴Cu radiolabeling and polyethylene glycol (PEG) for enhanced in vivo properties (Figure 1). The exosomes before surface engineering and the final products, NOTA-Exosome and NOTA-Exosome-PEG, were then characterized by transmission electron microscopy (TEM) (Figure 2a–f), exhibiting a plate shape due to dehydration during sample preparation without any visible aggregation. Notably, no significant change in morphology was observed after conjugating NOTA and PEG on the surface of the exosomes. In addition, there was no significant change of the hydrodynamic size before or after NOTA conjugation (Exosome: number weighted: 64.0 ± 12.1 nm; volume weighted: 106.3 ± 0.3 nm; intensity weighted: 183.6 ± 3.7 nm; n = 3. NOTA-Exosome: number weighted: 63.5 ± 5.5 nm; volume weighted: 112.7 ± 5.2 nm; intensity weighted: 182.0 ± 2.6 nm; n = 3). The average hydrodynamic size of NOTA-Exosome-PEG (number weighted: 63.6 ± 10.0 nm; volume weighted: 127.1 ± 14.4 nm; intensity weighted: 226.6 ± 17.9 nm; n = 3) was slightly larger than that of NOTA-Exosome (Figure 2g, Figure S2), due to the conjugation of PEG. The zeta potentials of Exosome, NOTA-Exosomes, and NOTA-Exosome-PEG were – 33.4 ± 2.2 mV, – 26.6 ± 2.5 mV, and – 3.3 ± 3.2 mV, respectively (Figure 2h). The changes in surface charge suggest successful conjugation of NOTA and PEG. Of note, no dramatic differences of morphology, hydrodynamic size and surface charge were observed between Exosome and NOTA-Exosome, indicating that NOTA conjugation and following radiolabeling should not significantly alter the properties of exosomes. However, since PEGylation greatly neutralized surface charge of Exosome and NOTA-Exosome, NOTA-Exosome-PEG may exhibit completely different in vivo profile.

Figure 1.

Schematic illustration of NOTA conjugation, PEGlyation and ⁶⁴Cu radiolabeling of exosomes. Surface engineering was facilitated by amine-reactive reagents (p-SCN-Bn-NOTA or mPEG_5k-NHS) and the protein-NH₂ on the surface of the exosomes.

Figure 2.

Characterization of exosome conjugates. The morphology and size of Exosome (a), NOTA-Exosome (b) and NOTA-Exosome-PEG (c) at magnification of 100k were characterized by TEM after negative staining. Enlarged TEM of Exosome (d), NOTA-Exosome (e) and NOTA-Exosome-PEG (f) were acquired at 200k. (g) Their hydrodynamic sizes were compared based on number weighted, volume weighted and intensity weighted distributions. (h) Their zeta potentials were also compared.

NOTA-Exosome and NOTA-Exosome-PEG were radiolabeled with ⁶⁴CuCl₂ in sodium acetate buffer (pH 5). The radiolabeling was so rapid that the chelation was complete in 1 minute, with 91.2 ± 0.2 % (n = 3) ⁶⁴Cu²⁺ successfully labeled onto NOTA-Exosome, and 85.7 ± 0.7 % (n = 3) ⁶⁴Cu²⁺ successfully labeled onto NOTA-Exosome-PEG (Figure 3a and 3b, Figure S3). The radiolabeling yield of NOTA-Exosome-PEG was slightly lower than that of NOTA-Exosome at 1 min after incubation, probably due to the physical barrier presented by the PEG layer. The radiolabeling reaction was essentially complete after 15 min of incubation (NOTA-Exosome: 95.6 ± 0.1 % at 15 min and 95.6 ± 0.3 % at 30 min; NOTA-Exosome-PEG: 97.9 ± 0.2 % at 15 min and 97.7 ± 0.2 % at 30 min; Figure 3a and 3b, Figure S3; n = 3). To confirm their radiostability, both ⁶⁴Cu-NOTA-Exosome and ⁶⁴Cu-NOTA-Exosome-PEG were purified with PD-10 desalting columns and incubated with phosphate-buffered saline solution (PBS) or mouse serum (25 % concentration) for 24 h, showing excellent stability (> 80 %; n = 3; Figure 3c and 3d, Figure S4, Figure S5) that was sufficient for accurate in vivo studies. Importantly, although ⁶⁴Cu-NOTA-Exosome and ⁶⁴Cu-NOTA-Exosome-PEG exhibited nearly identical radiostability in PBS at 24 h post incubation (⁶⁴Cu-NOTA-Exosome: 94.5 ± 0.1 %; ⁶⁴Cu-NOTA-Exosome-PEG: 92.6 ± 0.5 %; n = 3), ⁶⁴Cu-NOTA-Exosome-PEG showed significantly greater stability in mouse serum at 24 h (95.7 ± 0.9 %) than that of ⁶⁴Cu-NOTA-Exosome (80.4 ± 1.3 %) without PEGylation. The enhancement of radiostability may be attributed to the protection of the isotope by the PEG layer from competitive non-specific chelation by serum proteins, further manifesting the importance of PEGylation.

Figure 3.

Radiolabeling efficiency and stability. The radiolabeling efficiencies of ⁶⁴Cu-NOTA-Exosome (a) and ⁶⁴Cu-NOTA-Exosome-PEG (b) were measured by TLC after reacting for 1 min, 15 min and 30 min. The radiolabeling stabilities of ⁶⁴Cu-NOTA-Exosome (c) and ⁶⁴Cu-NOTA-Exosome-PEG (d) were measured by ITLC after incubating in mouse serum for 3 h, 6 h and 24 h. The ITLC was performed using 50 mM EDTA (pH 5.5) as the mobile phase.

To explore the in vivo biodistribution of the exosomes, ⁶⁴Cu-NOTA-Exosome and ⁶⁴Cu-NOTA-Exosome-PEG were intravenously (i.v.) injected into 4T1 tumor-bearing mice for serial PET scans at different time points post injection (p.i.). Without the protection of PEG, ⁶⁴Cu-NOTA-Exosome exhibited very short blood circulation time and robust hepatic clearance (Figure 4a). The signal in blood significantly reduced to 1.2 ± 0.2 %ID/g at 1 h p.i. and remained at similar level at later time points (1.0 ± 0.2 %ID/g and 0.9 ± 0.2 %ID/g at 4 h and 24 h p.i.; n = 3; Figure 4b, Table S1). The signal in liver was high in the earlier scan and gradually decreased over time due to degradation and clearance through hepatic pathway (44.3 ± 2.5 %ID/g, 38.0 ± 2.1 %ID/g, and 17.0 ± 0.8 %ID/g at 1 h, 4 h, and 24 h p.i.; n = 3; Figure 4b and 4e, Table S1). Low tumor uptake (0.7 ± 0.2 %ID/g, 1.2 ± 0.4 %ID/g, and 0.9 ± 0.2 %ID/g at 1 h, 4 h, and 24 h p.i.; n = 3; Figure 4b, Table S1) and poor tumor contrast (represented by tumor/muscle ratio; 2.3 ± 0.9, 4.0 ± 1.2, and 3.5 ± 1.4 at 1 h, 4 h, and 24 h p.i.; n = 3; Figure 4d) were observed, probably due to the short blood circulation and strong hepatic clearance, suggesting that native exosomes without appropriate surface coating are not desirable for in vivo biomedical applications, especially tumor-oriented drug delivery.

Figure 4.

In vivo PET imaging and ex vivo biodistribution. (a) Serial coronal PET images of 4T1 tumor-bearing mice at different time points post-injection of ⁶⁴Cu-NOTA-Exosome and ⁶⁴Cu-NOTA-Exosome-PEG. Tumors are pointed by red arrows. Time-activity curves of the tumor, blood, liver and muscle upon intravenous injection of ⁶⁴Cu-NOTA-Exosome (b) and ⁶⁴Cu-NOTA-Exosome-PEG (c) were obtained by ROI analysis and presented as %ID/g. Tumor/muscle ratio (d) and liver uptake (e) were compared between ⁶⁴Cu-NOTA-Exosome and ⁶⁴Cu-NOTA-Exosome-PEG (n = 3). (f) The biodistribution results were obtained by gamma counting of wet-weighted different tissues after 24 h p.i. of ⁶⁴Cu-NOTA-Exosome and ⁶⁴Cu-NOTA-Exosome-PEG (n = 3). Data are expressed as mean ± standard deviation. * P< 0.05; ** P< 0.01.

On the contrary, after appropriate PEGylation, ⁶⁴Cu-NOTA-Exosome-PEG exhibited prolonged blood circulation and reduced hepatic clearance (Figure 4a). After intravenous injection, the blood uptake was as high as 7.9 ± 0.4 %ID/g at 1 h p.i. and slowly decreased over time (5.6 ± 1.1 %ID/g, and 1.8 ± 0.3 %ID/g at 4 h and 24 h p.i., respectively; n = 3; Figure 4c, Table S2), while the liver uptake (25.5 ± 4.0 %ID/g, 28.6 ± 3.7 %ID/g, and 13.8 ± 0.9 %ID/g at 1 h, 4 h and 24 h p.i.; n = 3; Figure 4c and 4e, Table S2) was significantly lower than that of ⁶⁴Cu-NOTA-Exosome at the corresponding time points, manifesting the improved in vivo pharmacokinetic properties of exosomes after PEGylation. ⁶⁴Cu-NOTA-Exosome-PEG gradually accumulated in the tumor, which exhibited 3-fold higher tumor uptake after 24 h p.i. (2.7 ± 0.3 %ID/g; n = 3; Figure 3c, Table S2) in comparison to that of ⁶⁴Cu-NOTA-Exosome (0.9 ± 0.2 %ID/g), suggesting that PEGylation can be a promising approach to increasing the accessibility of exosome-based nanoplatforms in the tumor. The tumor contrast was thereby enhanced by PEGylation (represented by tumor/muscle ratio; 3.5 ± 1.1, 6.1 ± 1.1, and 7.4 ± 0.5 at 1 h, 4, and 24 h p.i.; n = 3; Figure 4d). The enhanced tumor retention is due to improved EPR effect of exosomes after appropriate PEGylation, resulting from reduced hepatic clearance and improved in vivo pharmacokinetics. The accumulation of radiolabeled exosomes in tumor and other important organs was verified by ex vivo biodistribution studies after the last PET scan at 24 h p.i., confirming the accuracy of PET imaging and analysis (Figure 4f).

Histological studies were conducted to further confirm the in vivo and ex vivo biodistribution observations, by i.v. administration of Alexa Fluor 488-conjugated Exosome (Alexa Fluor 488-Exosome) or Exosome-PEG (Alexa Fluor 488-Exosome-PEG) in 4T1 tumor-bearing mice. As shown in Figure 5, Alexa Fluor 488-Exosome-PEG exhibited significantly stronger fluorescence (green channel) in the tumor than that of Alexa Fluor 488-Exosome, demonstrating the enhanced tumor retention of PEGylated exosomes. More importantly, a mismatch between the signals from Exosome-PEG and tumor vasculature (CD31, which is specifically expressed on vascular endothelial cells; orange channel) was observed, indicating that exosomes after appropriate PEGylation can successfully extravasate from the vessels into the tumor microenvironment. Of note, high-level extravasation from vasculature has been achieved only rarely with conventional synthetic nanoplatforms, because of their relatively large sizes and suboptimal surface attributes, strictly limiting their applications to vasculature targeting and peri-vessel delivery of therapeutic agents. As naturally secreted vehicles, however, exosomes have the potential to reach every therapeutic area beyond the barrier of vasculature,⁵ and thus may be an effective drug delivery approach for tumor cells. In addition, liver was selected as the positive control, since it is the major clearance organ of nanoparticles. Both Alexa Fluor 488-Exosome and Alexa Fluor 488-Exosome-PEG showed strong accumulation in the liver after 24 h p.i., and no correlation was found between the signal from exosomes and vasculature. In addition, minimal retention was observed in the muscle (negative control) in both Alexa Fluor 488-Exosome and Alexa Fluor 488-Exosome-PEG groups, which is consistent with the results from PET images and ex vivo biodistribution studies.

Figure 5.

Histology. Cryo-sectioned tumor, muscle and liver were analyzed by fluorescent microscopy for the distribution of exosomes (green channel; with Alexa Fluor 488-conjugated exosome or exosome-PEG), vasculature (orange channel, with anti-CD31 primary antibody and Alexa Fluor 555-conjugated secondary antibody) and nuclei (blue channel, with Hoechst 33342). Scale bar: 200 μm.

CONCLUSION

Herein, we reported one of the first examples for radiolabeling and in vivo PET imaging of exosomes. We demonstrated that appropriate engineering approaches are required for improving the biomedical applications of exosomes. ⁶⁴Cu-labeled PEGylated exosomes exhibited enhanced tumor uptake and imaging capacity. PEGylation also reduced hepatic clearance of exosomes, which promises enhanced delivery efficacy and safety for potential applications in future studies. In addition, this study illustrated the application of molecular imaging in obtaining accurate and quantitative information about in vivo biodistribution of exosomes, and provided important guidelines about the surface engineering, radiolabeling and PET imaging of exosomes, which may benefit future exploration in the realm of exosomes.

METHODS

Reagents.

Polyethylene glycol (PEG; MW: 5000 Da) linkers mPEG_5k-NHS and Chelex 100 resin (50–100 mesh) were purchased from Sigma-Aldrich (St. Louis, MO). Water and all buffers were of Millipore grade and pretreated with Chelex 100 resin to ensure that the aqueous solutions were free of heavy metal. ⁶⁴Cu was produced by a GE PETtrace cyclotron using the ⁶⁴Ni(p,n)⁶⁴Cu reaction at the institutional Cyclotron Radiochemistry Facility at The University of Texas MD Anderson Cancer Center (Houston, TX). 1,4,7-Triazacyclononane-1,4,7-triacetic acid (NOTA) was purchased from Macrocyclics, Inc. (Dallas, TX). PD SpinTrap G-25 and PD-10 desalting columns were purchased from GE Healthcare (Piscataway, NJ). Complete mouse serum was purchased from Jackson Immuno Research Laboratories (West Grove, PA). Exosome-depleted fetal bovine serum (FBS) and ‎other chemicals and buffers were obtained from Thermo Fisher Scientific (Waltham, MA).

Cell lines and animal models.

4T1 murine breast cancer cells were obtained from ATCC (Manassas, VA) and cultured according to the supplier’s instructions. When they reached ~80% confluence, the cells were harvested for tumor implantation. Six-to-eight-week-old female Balb/c mice (Taconic Biosciences, Rensselaer, NY) were each subcutaneously injected with 1.5 × 10⁶ 4T1 cells in the flank to generate subcutaneous 4T1 breast cancer model. In vivo experiments were performed when the tumor diameter reached 6–8 mm. All animal studies were conducted under a protocol approved by the Institutional Animal Care and Use Committee at The University of Texas MD Anderson Cancer Center.

Exosome isolation, purification and characterization.

4T1 murine breast cancer cells were cultured in RPMI 1640 (high glucose) cell culture medium supplemented with 10 % (v/v %) exosome-depleted FBS and 1 % (v/v %) penicillin-streptomycin for 2 to 3 days until reaching ~90 % confluence. Of note, using exosome-depleted FBS is critical in cell cultures for exosome isolation, since the normal FBS also contains exosomes which will introduce impurity. The cell culture medium was collected and centrifuged at speeds 1800 × g and 2800 × g followed by passing through a 0.22 μm filter to remove the floating cells and cell debris. The supernatant was subsequently ultracentrifuged (Optima XE-90 Ultracentrifuge, Beckman Coulter, Brea, CA) twice at 100,000 × g to precipitate the exosomes. The resulting exosomes were re-suspended in sterile PBS (without calcium and magnesium) and stored at 4 ºC.

The isolation of the exosomes was confirmed by UV-VIS spectrometry (DU 800, Beckman Coulter, Brea, CA), and the protein concentration was measured on a spectrophotometer (NanoDrop 1000 Spectrophotometer, Thermo Fisher Scientific, Waltham, MA), serving as the concentration standard in later experiments. The morphology and size of the exosomes were characterized by transmission electron microscopy (TEM), immediately after exosome isolation and purification without fixation. In brief, TEM samples were placed on 100 mesh carbon-coated and formvar-coated copper grids treated with poly-l-lysine for approximately 1 h and then negatively stained with Millipore-filtered aqueous 1% uranyl acetate for 1 min. The stain was blotted dry from the grids with filter paper and samples were allowed to dry. The samples were then examined with a JEM-1010 TEM (JOEL USA, Inc., Peabody, MA) at an accelerating voltage of 80 Kv. Digital images were obtained using the AMT Imaging System (Advanced Microscopy Techniques Corp., Danvers, MA). The hydrodynamic size distribution was measured by dynamic light scattering (DLS; Brookhaven Instruments Corporation, Holtsville, NY). The zeta potential was measured with ZetaPlus Zeta Potential Analyzer (Brookhaven Instruments Corporation, Holtsville, NY).

NOTA conjugation and PEGylation.

Proteins expressed on the surface of the exosomes were utilized for conjugation with NOTA for ⁶⁴Cu labeling. Exosomes were reacted with amine-reactive NOTA (p-SCN-Bn-NOTA) with the weight ratio of 10:1 (weight of proteins in the exosomes/weight of NOTA) at pH 8.5 for 2 h. The resulting NOTA-Exosome was purified with PD SpinTrap G-25 desalting column using PBS (without calcium and magnesium) as the mobile phase to remove free NOTA that was not bound to the exosomes. As-prepared NOTA-Exosome was then reacted with PEG (mPEG_5k-NHS) with the weight ratio of 1:10 (weight of proteins in the exosomes/weight of PEG) at pH 7.4 for 2 h to form NOTA-Exosome-PEG. NOTA-Exosome-PEG was not further purified since free PEG will not hamper the in vivo tumor imaging and delivery. A fluorescent dye, amine-reactive Alexa Fluor 488 (NHS-Alexa Fluor 488), was conjugated to exosomes to form Alexa Fluor 488-Exosome and Alexa Fluor 488-Exosome-PEG using the same protocol as described above.

Radiolabeling and in vivo PET imaging.

⁶⁴CuCl₂ (74 MBq) was diluted in 300 μL of 0.1 M sodium acetate buffer (pH 5.5) and mixed with NOTA-Exosome and NOTA-Exosome-PEG (0.3 mg based on protein concentration). The reactions were conducted at 37 ºC for 30 min with constant shaking. The resulting ⁶⁴Cu-NOTA-Exosome and ⁶⁴Cu-NOTA-Exosome-PEG were purified by PD-10 desalting columns using PBS (without calcium and magnesium) as the mobile phase. The radiostability of ⁶⁴Cu-NOTA-Exosome and ⁶⁴Cu-NOTA-Exosome-PEG was determined by incubating each with 25 % mouse serum for 24 h. The radiolabeling yield and radiostability were tested with thin-layer chromatography (TLC) using 50 mM EDTA (pH 5.5) as the mobile phase (n = 3),⁴² measured by Bioscan AR-2000 Radio-TLC Imaging Scanner (Eckert & Ziegler, Valencia, CA).

4T1 tumor-bearing mice were intravenously injected with ⁶⁴Cu-NOTA-Exosome (50 μCi/mouse, 200 μL) or ⁶⁴Cu-NOTA-Exosome-PEG (50 μCi/mouse, 200 μL) and then underwent serial PET scans on an Albira PET/SPECT/CT scanner (Bruker, Billerica, MA) at different time points (1 h, 4 h and 24 h; n = 3) post-injection (p.i.). Quantitative data from region-of-interest (ROI) analysis on tumor and other organs was presented as percentage injected dose per gram of tissue (%ID/g). After the final scan at 24 h p.i., mice were euthanized under anesthesia with isoflurane for ex vivo biodistribution studies. Tumor, blood and major organs/tissues were collected and weighted. The radioactivity in the tissue was measured using Perkin-Elmer (Packard) Cobra II Gamma Counter (Perkin-Elmer, Waltham, MA).

Histology.

Two groups of 4T1 tumor-bearing mice (n = 3) were each intravenously injected with Alexa Fluor 488-Exosome or Alexa Fluor 488-Exosome-PEG. At 24 h p.i., the injected mice were euthanized, and the tumor, liver (positive control) and muscle (negative control) were frozen and cryo-sectioned for histological analysis. Frozen tissue slices with 6 μm thickness were fixed with cold acetone and stained for endothelial marker CD31, using rabbit anti-mouse CD31 antibody as the primary antibody and Alexa Fluor 555-conjugated goat anti-rabbit antibody as the secondary antibody. The frozen tissues were further incubated with Hoechst 33342 solution for nuclei staining. All images from Alexa Fluor 488 (representing the location of exosomes in the tissues), Alexa Fluor 555 (representing the location of vasculature) and Hoechst 33342 (representing the nuclei of cells) channels were acquired with Zeiss Axio Observer Z1 Inverted Motorized Microscope (Carl Zeiss Microscopy, LLC, Thornwood, NY) and analyzed with AxioVision 4.8 software (Zeiss).