Nuclear-Based Labeling of Cellular Immunotherapies: A Simple Protocol for Preclinical Use

Labeling and tracking existing and emerging cell-based immunotherapies using nuclear imaging is widely used to guide the preclinical phases of development and testing of existing and new emerging off-the-shelf cell-based immunotherapies. In fact, advancing our knowledge about their mechanism of action and limitations could provide preclinical support and justification for moving towards clinical experimentation of newly generated products and expedite their approval by the Food and Drug Administration (FDA). Here we provide the reader with a ready to use protocol describing the labeling methodologies and practical procedures to render different candidate cell therapies in vivo traceable by nuclear-based imaging. The protocol includes sufficient practical details to aid researchers at all career stages and from different fields in familiarizing with the described concepts and incorporating them into their work.


Introduction
Cellular immunotherapies comprise a wide and constantly expanding arsenal of different immune cell therapeutics [1].As such, they are based on living cells, either autologous (isolated from the same patient) or allogeneic (isolated from a different donor), whose role in the successful treatment of a broad range of human diseases has consolidated over the years.New emerging off-the-shelf cell-based immunotherapies are being developed worldwide by scientists and are conventionally tested in preclinical animal models without a reliable assessment of their in vivo distribution, targeting capability and long-term survival.Molecular imaging of immune cell therapeutics could provide unique insights into their mechanism of action, their success and, sometimes, their failure [2].To render them traceable, either molecular probes or contrast-generating reporters must be introduced into the cells by employing direct or indirect labeling methodologies [3,4].When doing so, aspects such as preclinical versus clinical setting, tracking time, tracking intervals, labeling strategy, modality, and probe of choice, need to be carefully considered [5].Unlike other molecular imaging modalities, preclinical nuclear-based imaging combines non-invasive whole-body tracking capabilities with exquisite sensitivity and depth penetration [6], enhanced resolution (with Positron Emission Tomography (PET) offering < 1mm resolution) [7,8], multiplexing capabilities [9], and absolute 3D quantification.
Hereby we describe a simple and comprehensive protocol for the nuclear-based labeling of immune cell therapeutics and allowing the reliable assessment of their in vivo fate and therapeutic efficacy by whole-body non-invasive preclinical imaging.The following direct and indirect methods and respective experimental workflows (Figs. 1 and 2) are tailored to the visualization and monitoring of different classes of cell-based therapies, including adoptively transferred T (Chimeric Antigen Receptor T) cells [10,11], natural killer (NK) [12,13] and CAR-NK cells [14], regulatory T cells (Treg) [15] and CAR-Tregs [16,17], gamma delta T cells [18,19] and dendritic cell vaccines [20].Detailed downstream in vitro, in vivo and ex vivo validation experiments are also included.A list of Materials was also included at the end of the protocol.Whenever applicable, we provide the reader with critical considerations and suggestion for the successful execution of this protocol.The application of these concepts should exclusively involve fully trained personnel.

Experimental Protocol for Preclinical Use
CAUTION: As this protocol involves handling and manipulation of radioactive material and or animals, researchers should consult with their home institution's Radiation Safety and /or Medical Physics Department and undergo the required training prior to initiating this work.Steps should be taken to minimize exposure to ionizing radiations, including wearing whole-body and ring dosimeters.Gloves should be worn at all times.Radioactive waste MUST be disposed according to institutional radiation safety and local waste management guidelines.Existing home institution SOPs (Standard Operating Procedures) MUST be strictly followed for your own safety and that of your coworkers.
Cyclotrons, generator systems, linear accelerators (LIN-ACs) or nuclear reactors can be used as a source of radioactive isotopes for this purpose.In either of these cases, access to a radiochemistry facility is essential.Radioisotopes and radioactive probes suitable to label immune cells need to be produced with high radiochemical purity (RCP; general consensus is ≥95%, however tracer dependent) and yield (RCY; tracer dependent).To achieve adequate radio-incorporation of the radionuclide into the targeting vector molecules, several radionuclide precursor solution considerations must be taken into account, including the radionuclide specific activity as the activity per unit mass of a radionuclide (radioactivity/mg) [21], possible trace contamination with elemental impurities that may compete for radio-incorporation with the radionuclide of interest, and the chemical form of both radionuclide itself and its formulation.In most instances, this information can be obtained by reviewing the Certificate of Analysis (CoA) that has been provided by the radionuclide manufacturer along with the material.Also, because the chemical form of the radionuclide solution may change over time, as a general rule, the time between radiochemical isolation of the radionuclide by the radionuclide producer and radiolabeling should be minimized.For more detailed radioisotope considerations and radiochemical probe design, refer to Volpe et al. [22].
Cells isolation protocols and culture medium may vary depending on immune cell type (T lymphocytes [23], NK (natural killer) cells [24], Tregs [25,26] (regulatory T cells), gamma delta T cells [27], and dendritic cells [28].Some cell types can be isolated using more than one protocol (see Hoerster et al. and Sato et al.) [24,29].For further details on different isolation protocols, practical isolation steps, required reagents, materials, and equipment, we refer the reader to commercially available kits. ).To obtain optimal labeling, preliminary optimization and calculation of incorporation rates are required.
1.2.Incubate resuspended cells for 10-30 min at 37 degrees Celsius and 5% CO 2 on agitating rack (at 350 rpm) (Note 6 6 ).Incubation time may vary depending on cell type and the probe.1.3.Wash cells three times in ice-cold PBS (not containing Ca 2+ /Mg2-).Centrifuge and collect each wash solution in three separate collection tubes labeled "supernatant", (wash 1" and "wash 2", respectively.1.4.Cells are resuspended in growth medium or PBS for further in vitro and/or in vivo experiments.1.5.Radioactivity in resuspended cells and respective washes is measured using a gamma-counter measuring the radioisotope of choice (Note 77 ).1.6.Radioactive cell-labeling efficiency is calculated using Equation 1.

LE (%) = activity of cell fraction activity of cell fraction + activity of combined washes x100
Fig. 2 Schematic representation of the indirect labelling experimental workflow at 1500 rpm and supernatants are removed and collected in prelabeled tubes.Cell pellets are resuspended in their culture medium.Cell-associated radiotracer and supernatants are measured at the gamma counter and percentages are calculated using Equation 1.As cells continue to divide, expect a label/signal dilution over-time as the label will redistribute to the daughter cells.The percentage of total activity bound to the cells will decrease with time compared to the initial measurement according to radiotracer decay (the latter depending on half-life of the chose radiotracer), but cell expansion will not be detectable as the number of individually labeled cells will decrease [3], making the latter method unsuitable for a reliable long-term observation of fast-dividing cells.IMPORTANT: Any commercially available gamma counter will provide decay corrected counts.In the unlikely circumstance that is not the case, a sample of known activity from the initial radioactive stock can be re-measured alongside with experimental samples.In lieu of these options, radioactivity at any given time can be calculated using the standard formula:

Or its derivatization:
where A is the activity, A 0 is the initial activity, A t is the activity after any time t, and λ is the decay constant, which for most radioactive material is readily available through the literature, or can be otherwise calculated using the following formula: IMPORTANT: Not all radioactive probes are well suited candidates for the direct labeling of cell-based immunotherapies.One example is provided by Jacob et al., reporting on dose-dependent changes in polyclonally expanded human Tregs phenotype (downmodulation of CD4 and CD25), impaired proliferation both in vitro and in vivo, and failure to survive when labeled with 89 Zr-oxine [31].Always perform downstream assays to assess the consequences of direct radiolabeling, including viability, phenotyping, proliferation, functional and DNA damage studies.Also consider using a different PET radiotracer with a similar half-life (days) but with a potentially less harmful radioactive decay than 89 Zr for this specific immune cell type.

Cellular immunotherapies transduction
This is a general protocol for immune cells transduction.Immune cells are non-adherent cells and therefore transduction protocols will require the use of non-tissue culture treated plates and pre-coating with RetroNectin.Transduction is performed on a 6-well format with 1 x 10 6 cells per well but both plate format and number of cells can be scaled up or down if needed.

2.2.1
Pre-coat a culture plate with RetroNectin (add at least 1mL per well in a 6-well plate to ensure complete surface coverage; scale quantities up or down depending on the chosen plate format) 2.2.2 Preserve plate sterility and prevent RetroNectin evaporation by sealing it with parafilm.Then store overnight at 4 °C or for 2h at RT 2.2.3 Harvest viral containing supernatant from virus producing cells and filter them through a 0.45 μm syringe filter (Note 12 12 ) 2.2.4 Replace media in the virus producing cells for subsequent transduction rounds 2.2.5 Supplement viral containing media with interleukin-2 (IL-2).Final concentrations used may vary depending on immune cell type (e.g., for T cells, 20 IU/mL; for Tregs 1000 IU/mL; for gamma delta T cells 300-1000 IU/mL) (Note 13 13 ) 2.2.6 Remove RetroNectin coating by aspiration and wash wells with 1 mL PBS supplemented with 10% FCS and leave at RT for 30 min 2.2.7 Collect immune cells (Note 14 14 ) and proceed counting them with using either an hematocytometer or an automated cell counter 2.2.8Take the desired cell number for transduction and centrifuge at 500 x g for 5 min at 4 °C 2.2.9 Resuspend pellet in filtered viral containing supernatant at 1 x 10 6 cells/3 mL and incubate for 15 min at RT (Note 15 15  ity controls" can be performed depending on the reporter (e.g., sodium perchlorate is used as a competitive substrate for the sodium iodide symporter NIS and resulted in uptake depletion demonstrating radioactive uptake specificity) [48].In that case, pre-treat cells with blocking or competitive agent, then add it again with the radioactivity (keep concentration constant throughout the assay).We encourage to also add a negative control (parental non-transduced cells) to determine basal uptake levels 2.3.2.5.Centrifuge samples at 1500 rpm at RT, collect 100 μL of supernatant and transfer it into a pre-labeled tube named "supernatant".Safely remove the remaining 900 μL and dispose of it according to institutional radiation safety guidelines 2.3.2.6.Wash the cells twice with ice-cold PBS alternating centrifugation steps.Collect 100 μL from each washing solution and transfer them into pre-labeled tubes named "wash 1" and "wash 2", respectively 2.3.2.7.Follow steps 1.4 to 1.5.2.3.2.8.Calculate radiotracer uptake using the following Equation 6, where CPM represents the decay-corrected radioactivity counts per minute (Note 19 19 ) (6) %Uptake = CPM(cells) CPM (cells) + counts of combined washes (supernatant + wash 1 + wash 2) x 100 New emerging cell therapeutics (including human cardiac progenitor cells and human-induced pluripotent stem cells (iPSCs)) are also amenable to indirect labeling, and step-by-step protocols describing their engineering using the sodium iodide symporter NIS reporter gene are available [49,50].
IMPORTANT: When the chosen reporter gene is a transporter or a surface protein, localization studies must be performed (confocal microscopy) prior to the radioactive uptake assay (as described in 2.3.2) to prove its correct localization as a fundamental pre-requisite for its correct function (imaging capability shown by the successful uptake of desired radioactive probe).Staining controls (e.g., membrane marker) and negative controls (parental non-transduced cells) are required (see Volpe et al. as an example) [47].
IMPORTANT: Whether Direct or Indirect Labeling was performed, we strongly recommend the reader to perform all application-dependent downstream experiments, including assessing the effects of cell labeling on their long-term viability, proliferation, phenotype, functional status (including activation markers and/or CAR expression), cytotoxic capability and potential radiation-induced DNA damage as compared to the parental cells (see referenced examples) [24,31,47].These are crucial steps to determine whether the labeling procedure had detrimental effects on the cellular function and are a sole responsibility of each user.Perform these experiments alongside a negative control (parental non-transduced cells).Other controls may be needed depending on the type of experiment and its complexity and will therefore not be discussed in this protocol.

In vivo tracking of cell-based immunotherapies in relevant preclinical models:
IMPORTANT: Prior to animal testing, ensure that the in vivo protocol is approved by the local Ethical Review Panel and is compliant with the institutional ethical and safety guidelines and regulations.Laboratory personnel is required to undergo training before performing invasive procedures in living animals.Access to an imaging facility is essential to perform the following protocol.
Relevant therapeutic models will have to be established prior to starting the labeling of therapeutic cells.) 3.2.5.Follow steps 3.1.11to 3.1.133.2.6.To observe the early stages of cell immunotherapies distribution, inject the radioactive probe immediately after the cells.If prolonged monitoring is needed, the radioactive probe can be administered at the desired time point and repeated imaging can be performed indefinitely (Note 24 24 )

Nuclear imaging by microSPECT-CT or microPET-CT and data analysis:
Parameters such as in vivo distribution, metabolism and clearance from the body may vary depending on the radioactive probe used (refer to Volpe et al. for preferential distribution and properties of available radioactive probes) [22].Therefore, it is important to add some waiting time prior to imaging in order to allow radiotracer in vivo biodistribution and avoid high blood signal.
Depending on the radiotracer characteristics (gamma rays or positron emitter), either SPECT (Single-photon emission computed tomography) or PET (Positron emission tomography) will be used.Then acquire PET image (Note 25 25 and Note 26 26 ) 4.4.If repeated animal imaging is needed, allow animals to fully recover from anesthesia and transfer them into a maintenance unit 4.5.If this is a terminal imaging procedure, euthanize animals according to approved protocol (see Shomer et al. [55]) 4.6.Reconstruct the SPECT-or PET-CT images using a 3D fully iterative Monte Carlo-based algorithm 4.7.Ensure that images are calibrated to the injected radioactivity and corrected for tissue attenuation, dead time and radioisotope decay (Note 27 27 ) 4.8.Save the data in the standard exchange format for medical images "DICOM" (Digital Imaging Communication in Medicine) and load them in the selected image analysis software (Note 28 28 ) 4.9.Before proceeding with image-based analysis, ensure that the CT and SPECT or PET are correctly co-registered 4.10.Delineate regions of interest (ROIs) and quantify corresponding radioactive signal (Note 29 29 and Note 30 30 ) 4.11.Non normalized results can be expressed as percent injected dose (%ID).Normalized results can be expressed as (i) percent injected dose per volume (%ID/mL), when considering the individual organs/tissues volumes, or (ii) standardized uptake value (SUV), with the latter accounting for the average radioactivity across the whole animal (Note 31 31 ).4.12.Express image data as either %ID (Equation 7) or %ID/mL (Equation 8).To best convey the quantitative information and to greatly facilitate the comparison of images across different studies, it is important to apply the same intensity scale (Note 32 32 ).The corresponding intensity scale bar should be always present for each figure.For more guidelines on radionuclide-based image analysis, refer to Weber et al. [58] For those with little to no experience in nuclear-based image analysis, refer to a recently published protocol from Cawthorne et al. [59] 5. Ex vivo biodistribution: 5.1.Following to step 4.5, measure the whole animal radioactivity and note the value and time of measurement 5.2.Proceed with animal dissection and harvest relevant organs and tissues 5.3.Measure the radioactivity in the tail and in the remaining carcass post dissection (remember to take note of values and time of measurement) (Note 31) 5.4.Selected organs and tissues are collected in preweighted tubes and weighted 5.5.Prepare radioactive calibration standards with a known activity to simplify decay correction: 5.5.1.On the day of the experiment, prepare standards from the same stock solution used for radiotracer injection 5.5.2.Prepare duplicate samples in 1.5 mL Eppendorf tubes 5.5.3.Each duplicate will be read at the gamma counter before and after your experimental samples using the same settings (i.e., exposure time and energy window) 5.5.4.At the end of the measurements, a calibration curve (activity vs counts per second) can be generated by linear regression 5.6.Measure radioactivity in all selected organs and tissues alongside with radioactive calibration standards using an automated gamma counter (note the time of measurement) (Note 33 33 ) 5.7.Express values as %ID (Equation 7), %ID/g (Equation 8 Equation 8) or SUV (Equation 9), where the latter are calculated using the following formulas: Additional downstream analyses include (i) autoradiography, (ii) tissue histology (the latter preceded by organ/ tissue embedding (sample preparation may differ depending on embedding procedure selected), (iii) tissue homogenization followed by staining for immune cell markers and flow cytometry.

Conclusions
When rendering immune cell therapeutics in vivo traceable is of interest, one should familiarize with concepts such as tracking time, tracking intervals, sensitivity in the detection, probe, and reporter of choice.We recommend you reading two recent publications discussing each of these aspects and subsequent repercussions on successful in vitro tracking of immune cells [5,53].The same reviews discuss the pros and cons of both direct and indirect labeling approaches and will help you decide what is best for your experimental needs.How a certain radiotracer distributes in a living animal, its metabolism and clearance route are essential pre-requisites for designing your experiment and the correct interpretation of data.For example, 89 Zr-based radiopharmaceuticals distribute primarily in the liver.Upon cell death, or as a result of demetallation, free 89 Zr will be largely accumulating in bones, joints, and marrow because of its high affinity for electronegative donor atoms (i.e., oxygen and phosphorus) in the hydroxyapatite of the bone matrix [60].While free 89 Zr homing to the bone and 89 Zr-labeled cells homing to the bone marrow can be reliably differentiated in humans (due to the large size of the bone marrow cavity), it is not possible to do so in a preclinical setting.However, when tracking immunotherapeutics in living animals using 89 Zr, one could exploit a chelating agent also used in clinical practice, deferoxamine, to capture the free 89 Zr and excrete it through the renal route [30].As the above shows, current literature provides numerous examples of immune cell tracking and troubleshooting tips, some of which were also incorporated in this protocol.However, sometimes experiments will require further optimization for both direct and indirect approaches as many factors can influence labeling efficiency, including the radiotracer of choice, cell type, cell size, buffer, temperature, potential risks of radiotoxicity and associated DNA damage.
Mechanisms by which radioactive probes can enter a cell can also differ.When directly labeling immune cell therapies, the probe can be internalized (i) through small molecules transporters, or (ii) can passively diffuse through the lipophilic cell membrane; whereas, when a reporter gene is employed, this can be (i) a transporter, in which case the probe is internalized, (ii) a receptor, thus irreversibly binding to the probe, or (iii) an enzyme, which then modifies the probe structure, resulting in the entrapment and intracellular accumulation of the probe.
In vivo cell tracking is a constantly growing and multidisciplinary field, fueled by the recent advancements in imaging technology.The development of a total-body PET, with its unprecedented sensitivity and resolution [61], permits to extend the in vivo tracking time of directly labeled cell-based immunotherapies.Multiplexing capabilities can now be achieved not only via multi-modal imaging (e.g., PET-CT or PET-MRI) [62][63][64], but also by multiplexed PET [9], with the imaging and quantification of two PET tracers within the same PET acquisition, whereby allowing the simultaneous and independent visualization and targeting of cellular immunotherapies to the tumor using either direct (Fig. 3

Fig. 1
Fig. 1 Schematic representation of the direct labelling experimental workflow

Fig. 3 3 .
Fig. 3 Multiplexed PET-CT imaging of tumor and PSMA-specific CAR-T cell therapy in a preclinical model of prostate cancer.| 3 x 10 6 PSMA-positive and PSMA-negative tumors were implanted in NOD SCID gamma males (6-8 weeks old).(Right) 5 x 10 6 89 Zr-oxine labelled PSMA-targeting CAR-T cells were systemically infused 25 days post tumor establishment.Alongside 89 Zr-oxine to image CAR-T cells, an [ 86 Y]Y-DOTA-PSMA tracer, retro-orbitally injected, was used to simultaneously image the PSMA-positive tumor using a newly described PET image reconstruction method.[ 86 Y]Y-DOTA-PSMA uptake was also seen in the ocular cavity (OC) and bladder (B), the latter as a result of tracer renal excretion. 89Zr-oxine signal was also observed in liver (Li) and bones (Bo).(Left) 25 days post tumor establishment, a single dose of 5 x 10 6 PSMA-targeting CAR-T cells engineered to express hNIS reporter gene were systemically Wash cells twice with ice-cold PBS 2.3.2.3.Incubate cells with desired amount of radioactivity (suggested is 1.35-1.50μCi per sample) in 1 mL PBS for 30 min-1 h at 37 °C and 5% CO 2 (Note 2) 2.3.2.4.Blocking or competitive studies as "specific- [65]t) or indirect (right) labeling approaches.This is rendered possible by separating and reconstructing double coincidences from recovered triple coincidences via an additional reconstruction algorithm[65].
3.1.Benchtop centrifuge 3.2.1.5 mL Eppendorf tubes 3.3.Ice bucket for cooling cells prior to administration 3.4.Anesthetic (e.g., isofluorane) for inhalation 3.5.Syringes 0.3 mL U-100 insulin (29 gouge preferred, 0.33 x 12.7 mm; sterile, single-use 3.6.Rodent anesthesia induction chamber equipped with three-way valves, tube mount connector for inlet, PVC tubing for gas inlet, and 22 mm scavenging tube 3.7.Rodent anesthesia system, including animal facemask suitably sized for either mouse or rat and isofluorane vaporizer 3.8.Veterinary scavenging unit -240V with automatic weighing mechanism, variable speed control and automatic temperature compensation 3.9.Preclinical multi-modal SPECT-and PET-CT scanner (microSPECT and PET-CT) 4. Nuclear imaging by microSPECT-CT and microPET-CT and data analysis: 4.1.For in vivo imaging, please refer to material in 2. 4.2.For in vivo data reconstruction, an image analysis software based on full 3D iterative algorithm is required 4.3.For in vivo data analysis, please refer to Note 28 for a full list of available analysis software packages (free or commercially available).Commercially available licenses can often be purchased with the scanner.All are able to co-register SPECT or PET with CT anatomical reference