(G-245) Enabling extracellular vesicle isolation and analysis via temperature-responsive polymeric reagents

Introduction:: Extracellular vesicles (EVs) are secreted by cells serve as vital mediators of intercellular communication between immune, parenchymal, and stromal cells in tissues. EVs are also secreted by the disease cells (e.g., tumor cells).[1] Therefore, they can potentially be utilized for delivering therapeutics, diagnosing diseases, and facilitating life science research. Characterization and isolation methods serve as the foundation for the EV technology development. For example, isolating mesenchymal stem cell-derive EVs for drug delivery applications.[1] Current mainstream separation methods include ultracentrifugation, size exclusion chromatography, ultrafiltration, precipitation, and immunoaffinity separation. However, these approaches might involve complex procedure, are time consuming, exhibit limited capacity, and result in moderate purity as well as yield.[1] To address the need, we have developed a temperature-responsive polymeric reagent system to enable efficient EV isolation to facilitate the downstream analyses.

Materials and Methods:: The magnetic nanoparticles were synthesized by following our previous publication.[2] In brief, the synthesis was templated using a diblock copolymer, poly(N-isopropylacrylamide)-block-ploy(acrylic acid)/pNIPAAm-b-pAAc, synthesized via reversible addition-fragmentation chaintransfer (RAFT) polymerization. The polymer was mixed with FeCl₃, so the pAAc segment could coordinated the particle core formation by complexing with Fe³⁺. Then, NH₄OH, a reducing reagent, was added to form 𝛾-Fe₂O₃ cores with pNIPAAm displayed on the surfaces. The resulting particles were purified via size-exclusion chromatography, and characterized for their particle sizes (e.g., DLS) as well as temperature-responsiveness.

The polymer-antibody conjugates were synthesized by end-grafting pNIPAAm to antibodies via click chemistry. In brief, the anti-mouse IgG antibody’s lysine residues were modified with dibenzocyclooctyne (DBCO) using a linker, DBCO-PEO₄-TFP. After removing the excess linker, the DBCO-antibodies were mixed with azido-pNIPAAm (Sigma) overnight for the conjugation.

The EV isolation started by mixing specimens and antibodies that target specific surface markers, including CD63, CD81, CD9, and phosphatidylserine (PS). After the addition of the polymer-antibody conjugates and the magnetic particles, the solution was heated to ca. 40°C, which facilitate the magnetic separation by aggregating the reagents as well as the bound EVs by hydrophobizing pNIPAAm. After the isolation, the captured EVs were further processed for the downstream analysis such as RT-PCR.


Results, Conclusions, and Discussions:: EVs from human semen were used as a model analyte. The separation targeted common EV tetraspanin markers CD81, CD9, and CD63. We assessed the efficiency of EV purification by quantifying microRNAs (let-7b and miR-29b) known to be carried by EVs from semen by RT-qPCR and normalized to the levels of spike-in control. Compared to an equivalent volume of untreated EVs or EVs subjected to the isolation protocol using control antibodies which do not specifically select for EV, let-7b and miR-29b were detected at 32 and 178 fold higher concentrations, respectively, in the specifically selected sample. Taken together, these results indicate that the binary reagent system with anti-tetraspanin antibodies selectively isolated EVs. Compared to the Dynabeads®, which requires a 2-day separation process, the temperature-responsive binary reagent system isolated EV expressing one or more common tetraspanin markers rapidly in 90 minutes. The reagent system was also utilized in conjunction with transmembrane immunoglobin and mucin domain-4 (TIM-4) to isolate EVs from bovine milk by target PS. To evaluate the isolation process, microRNA let-7b, found in high abundance in bovine milk EV, was used. An equivalent volume of unprocessed milk was also analyzed to estimate capture efficiency. Ct values for let7-b were normalized to the miR-cel-39 spike-in control, added before RNA extraction of the captured EVs. Compared to the control group, mouse IgG1 Κ isotype, TIM-4 construct isolation led to more let-7b, with 4 – 13 folds increase. The EV isolation efficiency was estimated to be 51 – 77%. The total process duration is 75 min and consists only of 4 steps.

The temperature-responsive polymeric reagents demonstrated the isolation by targeting EV surface markers. These nanoscale reagents achieved rapid and efficient separation via rapid diffusion and there were many “particles” of polymer–antibody conjugate diffusing in solution at any antibody concentration. Our previous publication also demonstrated significantly faster biomarker separation when compared to the mainstream magnetic microbeads.[3] Therefore, we conclude the temperature-responsive polymeric reagents can be employed for efficient EV affinity isolation to enable high quality analytical assays (e.g., PCR).


Acknowledgements (Optional): : Financial support was provided by NIH HD089679 and CA174581.

References (Optional): : 1. Lai JJ, Chau ZL, Chen S-Y, Hill JJ, Korpany KV, Liang N-W, Lin L-H, Lin Y-H, Liu JK, Liu Y-C, Lunde R, Shen W-T. Exosome Processing and Characterization Approaches for Research and Technology Development. Advanced Science. 2022;9(15):2103222 .

2. Jauregui R, Srinivasan S, Vojtech LN, Gammill HS, Chiu DT, Hladik F, Stayton PS, Lai JJ. Temperature-Responsive Magnetic Nanoparticles for Enabling Affinity Separation of Extracellular Vesicles. ACS Appl Mater Interfaces. 2018;10(40):33847-56.

3. Nehilla BJ, Hill JJ, Srinivasan S, Chen Y-C, Schulte TH, Stayton PS, Lai JJ. A Stimuli-Responsive, Binary Reagent System for Rapid Isolation of Protein Biomarkers. Analytical Chemistry. 2016;88(21):10404-10.