In vivo tracking of [89Zr]Zr-labeled engineered extracellular vesicles by PET reveals organ-specific biodistribution based upon the route of administration

Abstract

Extracellular vesicles (EVs) have garnered increasing interest as delivery vehicles for multiple classes of therapeutics based on their role as mediators in an important, natural intercellular communication system. We recently described a platform to allow the design, production and in vivo study of human EVs with specific properties (drug or tropism modifiers). This article seeks to compare and expand upon historical biodistribution and kinetic data by comparing systemically and compartmentally administered labeled engineered EVs using in vivo and ex vivo techniques. MethodsEVs were surface-labeled to high radiochemical purity and specific activity with 89Zirconium deferoxamine ([89Zr]Zr-DFO) and/or cy7-scrambled antisense oligonucleotide (Cy7-ExoASOscr), or luminally loaded with GFP for in vivo tracking in rodents and non-human primates (NHPs). Positron Emission Tomography (PET) and subsequent immunohistochemistry (IHC) and autoradiography (ARG) cross-validation enabled assessment of the anatomical and cellular distribution of labeled EVs both spatially and temporally. ResultsOver time, systemic administration of engineered EVs distributed preferentially to the liver and spleen (Intravenous, IV), gastrointestinal tract and lymph nodes (Intraperitoneal, IP) and local/regional lymph nodes (Subcutaneous, SC). Immunostaining of dissected organs displaying PET signal revealed co-localization of an EV marker (PTGFRN) with a subset of macrophage markers (CD206, F4/80, IBA1). Compartmental dosing into NHP cerebrospinal fluid (CSF) resulted in a heterogenous distribution of labeled EVs depending upon whether the route was intrathecal (ITH), intracisterna magna (ICM) or intracerebroventricular (ICV), compared to the homogeneous distribution observed in rodents. Thus anatomically, ITH administration in NHP revealed meningeal distribution along the neuraxis to the base of the skull. In contrast ICM and ICV dosing resulted in meningeal distribution around the skull and to the cervical and thoracic spinal column. Further characterization using IHC shows uptake in a subset of meningeal macrophages. ConclusionsThe present studies provide a comprehensive assessment of the fate of robustly and reproducibly labeled engineered EVs across several mammalian species. The in vivo distribution was observed to be both spatially and temporally dependent upon the route of administration providing insight into potential targeting opportunities for engineered EVs carrying a therapeutic payload.