Simplified 89Zr-Labeling Protocol of Oxine (8-Hydroxyquinoline) Enabling Prolonged Tracking of Liposome-Based Nanomedicines and Cells

In this work, a method for the preparation of the highly lipophilic labeling synthon [89Zr]Zr(oxinate)4 was optimized for the radiolabeling of liposomes and human induced pluripotent stem cells (hiPSCs). The aim was to establish a robust and reliable labeling protocol for enabling up to one week positron emission tomography (PET) tracing of lipid-based nanomedicines and transplanted or injected cells, respectively. [89Zr]Zr(oxinate)4 was prepared from oxine (8-hydroxyquinoline) and [89Zr]Zr(OH)2(C2O4). Earlier introduced liquid–liquid extraction methods were simplified by the optimization of buffering, pH, temperature and reaction times. For quality control, thin-layer chromatography (TLC), size-exclusion chromatography (SEC) and centrifugation were employed. Subsequently, the 89Zr-complex was incorporated into liposome formulations. PET/CT imaging of 89Zr-labeled liposomes was performed in healthy mice. Cell labeling was accomplished in PBS using suspensions of 3 × 106 hiPSCs, each. [89Zr]Zr(oxinate)4 was synthesized in very high radiochemical yields of 98.7% (96.8% ± 2.8%). Similarly, high internalization rates (≥90%) of [89Zr]Zr(oxinate)4 into liposomes were obtained over an 18 h incubation period. MicroPET and biodistribution studies confirmed the labeled nanocarriers’ in vivo stability. Human iPSCs incorporated the labeling agent within 30 min with ~50% efficiency. Prolonged PET imaging is an ideal tool in the development of lipid-based nanocarriers for drug delivery and cell therapies. To this end, a reliable and reproducible 89Zr radiolabeling method was developed and tested successfully in a model liposome system and in hiPSCs alike.


Introduction
The development of nuclear-imaging-guided, liposome-based nanomedicines or nanotheranostics occasionally requires prolonged tracing time due to the slower kinetics, enhanced circulation and excretion time of the candidate compounds. Zirconium-89 ( 89 Zr)labeled oxine (8-hydroxyquinoline) has recently emerged as a favorable positron emission tomography (PET) alternative to single photon emission computed tomography (SPECT) indium-111-labeled oxine [1]. The positron emitter 89 Zr (T 1 2 = 78.4 h) offers the opportunity of tracking cells or lipid-based nanomedicines by PET for up to one week.
Based upon these findings, the present study focused on simplifying and optimizing the [ 89 Zr]Zr(oxinate) 4 production. Liquid-liquid extraction or solvent extraction methods were applied in which the 89 Zr radioisotope was transferred from the aqueous raffinate to the chloroform solution of chelating 8-hydroxyquinoline. We tested different buffer environments, temperatures, reaction times and stirring methods. The present study was primarily aimed at identifying the key factors to allow for a reproducible and robust synthesis protocol at the highest isotope incorporation yield, while minimizing time and effort. The validity of the established method was verified by PET on a model liposome system and via labeling trials of human induced pluripotent stem cells (hiPSCs). As of more recently, human iPSCs and their progenies can be generated with high efficiency by advanced protocols [8,9] and have great potential for regenerative medicine. However, progress towards the clinical translation of hiPSCs requires efficient labeling technologies to monitor aspects of transplantation safety and efficiency [10]. The established protocols were employed in advanced nanomedicinal and stem cell transplantation studies to be presented elsewhere.

Liposome Formulation and Labeling
The model liposome for the radiolabeling studies was TargoSphere ® [11,12], an umbrella term coined for various lipid-based nanocarriers developed by Rodos Biotarget. The thin-film hydration method followed by extrusion [13] was used for preparing liposomes. In brief, phospholipids were dissolved, the stock solutions were combined in round-bottomed flasks, and lipid films were subsequently formulated by removing the

Characterization of Liposomes
Particle size distributions were determined by dynamic light scattering (DLS) using a Malvern Zetasizer Nano ZS device (Malvern Panalytical), thus gaining the polydispersity index (PDI) and ζ potential. The labeling yield of the [ 89 Zr]Zr(oxinate) 4 -liposome complex was checked after 5, 10, 15, 20, 30 and 60 min as well as after 6, 18 and 24 h; the reactions were performed at RT or 60 • C, respectively. To this end, we employed thin-layer chromatography (TLC), size-exclusion chromatography (SEC) and separation by centrifugation. As a thin layer, ITLC-SG strips (Merck, Darmstadt, Germany) were developed by 0.1 M citrate buffer, chloroform, chloroform-MeOH 5% and 20 mmol EDTA solutions [14].

PET Imaging of Liposomes
PET imaging was used to evaluate the in vivo integrity of the [ 89 Zr]Zr(oxinate) 4liposome complex. Labeled liposomes were injected intravenously (IV) in healthy C57BL/6 mice (n = 5). As a control, PBS-buffered [ 89 Zr]Zr(OH) 2 (C 2 O 4 ) was applied in two animals. Injected activities and volumes were 5.0 MBq in 100 µL per animal. The final lipid concentration of the solutions applied in vivo was 0.91-0.94 µg/µL. Mice were then subjected to serial PET imaging using a small animal microPET/CT system (Inveon DPET and CT120; Siemens Healthineers, Erlangen, Germany). Right after IV application, dynamic PET images were acquired over a 60 min period, and a static PET acquisition was performed 24 h later. Values of the I.D./g tissue (i.e., injected activity per gram unit of tissue) were determined from the region of interest per volume of interest (ROI/VOI) selections from spatial PET images using PMOD software (PMOD Technologies, Zürich, Switzerland).

Cell Culture and Labeling
The hiPSC lines hHSC_1285i_iPS2 (MHHi006-A [15]; or MHHi001-A-5 [16]) were cultured by conventional surface-adherent 2D culture and in 3D suspension culture, as previously described [9]. In brief, cryopreserved hiPSCs were thawed and cultured over 2-3 passages on Geltrex ® -coated T-flasks in Essential 8 medium (E8) with Rhokinase inhibitor (RI). Subsequently, hiPSCs were dissociated using Accutase TM treatment; 10 million single cells were inoculated in 20 mL E8 + RI in a 125 mL Erlenmeyer flask and placed on a horizontal shaker rotating at 70 rpm, placed in a conventional incubator at 37 • C, 5% CO 2 and 95% RH for 2-3 days to allow hiPSC aggregation and expansion in suspension. The prelabeled [ 89 Zr]Zr(oxinate) 4 was redissolved in DMSO and 30, 60 or 90 µL of this solution was added to 3 mL aliquots of~3 × 10 6 cells in suspension (in individual wells in a 6-well plate), reaching DMSO concentrations of 1%, 2% or 3%, respectively. Cells were incubated for 5, 10, 15, 30 and 60 min as well as 6 and 24 h while applying stirring at 100 rpm. Cells were collected by pelleting at 100× g for 5 min and resuspended in PBS. Labeling yields were determined by assessing both percentages of radioactivity remaining in the supernatant and cell-bound radioactivity.

Characterization of Liposomes
The mean diameter of the TargoSphere ® formulation employed proved to be 91.51 nm (Zaverage: 89.12 nm), with a low 0.209 PDI. Measured ζ potential value was mbox−0.0612 ± SD 0.0950. For a particle size distribution histogram, see Figure S3.

MicroPET Studies of Liposomes
PET studies confirmed the radiochemical stability and colloidal integrity of the labeled complex after IV administration. Most of the injected radioactivity quickly accumulated in organs of the reticuloendothelial system within the first 15 min, while only moderate radioactivity was detected in other organs (Figure 2). The fast kinetics and early biodistribution of the [ 89 Zr]Zr(oxinate) 4 liposomes compared to the [ 89 Zr]Zr(OH) 2 (C 2 O 4 ) over 60 min p.i. are depicted in the time-activity curves generated from the dynamic PET imaging data ( Figure 2C). Different uptake kinetics and organ distributions are clearly visible, predominately demonstrating hepatic, splenic and renal uptake. Specifically, high uptake by liver and spleen was observed over the entire 24 h tracing period. 60 min p.i. are depicted in the time-activity curves generated from the dynamic PET imaging data ( Figure 2C). Different uptake kinetics and organ distributions are clearly visible, predominately demonstrating hepatic, splenic and renal uptake. Specifically, high uptake by liver and spleen was observed over the entire 24 h tracing period.
This pharmacokinetic profile and the 24 h organ distribution characteristics unequivocally corresponded to a nanosized labeled compound that was slowly released from the hepatic and splenic macrophages and excreted renally: the 24 h kidney uptake was 8.77% I.D./g (± SD 10.75%) compared to the control group with 1.24% I.D./g (± SD 0.01%). The 6.80% I.D./g (±SD 3.07%) lung activity indicated a low ratio of an aggregated particle fraction compared to the control group with 1.36% I.D./g (± SD 0.03%) corresponding values [18]. PET images of the [ 89 Zr]Zr(OH)2(C2O4) control group showed skeletal uptake by the spine, extremities and skull.
This pharmacokinetic profile and the 24 h organ distribution characteristics unequivocally corresponded to a nanosized labeled compound that was slowly released from the hepatic and splenic macrophages and excreted renally: the 24 h kidney uptake was 8.77% I.D./g (±SD 10.75%) compared to the control group with 1.24% I.D./g (±SD 0.01%). The 6.80% I.D./g (±SD 3.07%) lung activity indicated a low ratio of an aggregated particle fraction compared to the control group with 1.36% I.D./g (±SD 0.03%) corresponding values [18]. PET images of the [ 89 Zr]Zr(OH) 2 (C 2 O 4 ) control group showed skeletal uptake by the spine, extremities and skull.

Discussion
More recent studies disseminated [ 89 Zr]Zr(oxinate) 4 radiolabeling by applying variable labeling protocols. For cell labeling, Charoenphun et al. first presented a protocol that employed a different buffering method, which resulted in~60% labeling efficiency of the prelabeled oxinate [2]. Sato and colleagues published a more complex process that obtained 89 ZrCl 4 from 89 Zr-oxalate, including an ion exchange step prior to solvent extraction [3]. Weist 4 -labeling yields similar to those achieved in the present study by applying the formulation of a more complex kit prior to cell labeling studies [6]. For liposome tracking, Li et al. [7] combined oxine prelabeling with a second liposome-incorporated bifunctional chelator deferoxamine [19], and then introduced a post-purification step for obtaining the final product. In contrast to all these approaches, here we provide a more simple, reproducible [ 89 Zr]Zr(oxinate) 4 -labeling protocol for prospective liposome and cell-imaging investigations that requires no further post-purification process.
Besides the importance of appropriate buffering, our study illustrated the importance of proper mixing during the solvent extraction. Specifically, maximum 89 Zr chelation requires a maximum distribution ratio, which cannot be reached by laminar mixing. Turbulent vortexing must be maintained, thus maximizing the liquid-liquid interface, preferably not in a "v-shape" microreaction vessel, but in an asymmetrically shaped liquid container at high-speed vortexing. For comparison, non-turbulent mixing resulted in either low or highly variable labeling yields. By considering these crucial process parameters, we obtained a reliably reproducible 89 Zr-labeling protocol requiring a 15-30 min effort that involved simple and cost-effective quality control. The incorporation efficacy of the prelabeled agent was proven by the successful efficient labeling of liposomes and of human iPSCs. Due to the poor water solubility of oxine and oxinates, the use of DMSO and ethanol as a cosolvent was required. Thus, the short-term DMSO tolerance of the cells [20] and the ethanol's impact on liposome size and stability [21] must be considered. However, DMSO-induced cell toxicity effects at concentrations of around ≥ 1% only became apparent upon much longer incubation times [22]. 89 Zr-Labeled liposomes were injected into healthy mice to evaluate their stability and basic pharmacokinetic characteristics by PET. The pharmacokinetic profile and the 24 h organ distribution matched the characteristics of a nanosized labeled compound that was slowly excreted renally.

Conclusions
Prolonged PET imaging is an ideal tool in the development of lipid-based nanocarriers for drug delivery and cell therapies. For this objective, a reliable, reproducible and simplified 89 Zr-radiolabeling method was developed and tested successfully in model liposomes and hiPCSs alike. This method may be adapted to other lipid-based nanomedicines or nanotheranostics, and to other cell species of interest.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/pharmaceutics13071097/s1, Table S1: Radiolabeling yields of [89Zr]Zr(oxinate)4 until the extraction (60 min) and in the following stability samplings (2 h to 24 h). Table S2: Radiolabeling yields of [89Zr]Zr(oxinate)4-liposome complex until extraction (18-24 h) and in the following stability samplings (+ 24 h). Table S3: Radiolabeling yields of cells at different times of incubation and different DMSO concentrations. Table S4: 24 h ex vivo biodistribution results of [89Zr]Zr(oxinate)4-liposomeinjected animals and the 89Zr control group (I.D./g organ). Table S5:   Data Availability Statement: The datasets and PET images used and/or analyzed during the current study are available from the corresponding author on reasonable request.