Intracellular labeling properties of zirconium-89 oxine in comparison to indium-111 oxine labeling

Introduction: A zirconium-89 (⁸⁹Zr)-oxine ex vivo cell labeling method has been developed recently for in vivo tracking of cells by positron emission tomography (PET). It has successfully visualized many cell types with high sensitivity and resolution in various animal models and its clinical application has been started. In this study, we analyzed nature of ⁸⁹Zr-oxine intracellular labeling.

Methods: Mouse naïve T cells, natural killer cells, dendritic cells, monocytes, as well as EL4 (mouse lymphoma), DC2.4 (mouse dendritic cell), and Kit225K6 (human T cell leukemia) cell lines, were incubated with ⁸⁹Zr-oxine (18.5-592 kBq/10⁶ cells) synthesized in-house. Cellular incorporation, incorporation threshold, protein binding, intracellular distribution, and retention of ⁸⁹Zr were examined. Protein binding and intracellular distribution were also analyzed for indium-111 (¹¹¹In)-oxine cell labeling for comparison. Protein mass of the cells was measured using labeled cell lysates. SDS-PAGE of sub-fractionated labeled cell lysates were transferred to membranes followed by autoradiography.

Results: Different cell types that were labeled using the same condition incorporated different amounts of ⁸⁹Zr-oxine, which positively correlated with the protein mass of the cells. Almost all ⁸⁹Zr was protein-bound (>97%) in EL4, DC2.4 and Kit225K6 cell lines. Intracellularly, ⁸⁹Zr primarily localized in the cytoplasm, plasma/intracellular membranes, and soluble nuclear fraction, consisting of >75% activity, followed by chromatin-bound protein fraction. By contrast, in ¹¹¹In-oxine labeled EL4 and DC2.4 cells (370 kBq/10⁶ cells), >35% of ¹¹¹In was non-protein-bound (p<0.0001, vs. ⁸⁹Zr-oxine). ¹¹¹In mainly localized in the cytoplasm, followed by membrane fraction (>86% in the two fractions). Autoradiography indicated binding of ⁸⁹Zr to various proteins of different sizes in all sub-cellular fractions, whereas no band was shown in ¹¹¹In-oxine labeled cell samples. Labeling with increasing doses of ⁸⁹Zr-oxine resulted in plateauing of incorporated activity. Labeling a cell above its saturation point seemed to impact its ability to retain ⁸⁹Zr.

Conclusions: ⁸⁹Zr-oxine cell labeling results in virtually all activity bound to various intracellular proteins, showing much higher and stable protein binding than ¹¹¹In-oxine. Intra-cellular protein binding of ⁸⁹Zr could contribute to stability of the label within the cells.