Dynamic Imaging of Chimeric Antigen Receptor T Cells with [¹⁸F]Tetrafluoroborate Positron Emission Tomography/Computed Tomography

Summary

This protocol describes the methodology for non-invasively tracking T cells genetically engineered to express chimeric antigen receptors in vivo with a clinically available platform.

Abstract

T cells genetically engineered to express chimeric antigen receptors (CAR) have shown unprecedented results in pivotal clinical trials for patients with B cell malignancies or multiple myeloma (MM). However, numerous obstacles limit the efficacy and prohibit the widespread use of CAR T cell therapies due to poor trafficking and infiltration into tumor sites as well as lack of persistence in vivo. Moreover, life-threatening toxicities, such as cytokine release syndrome or neurotoxicity, are major concerns. Efficient and sensitive imaging and tracking of CAR T cells enables the evaluation of T cell trafficking, expansion, and in vivo characterization and allows the development of strategies to overcome the current limitations of CAR T cell therapy. This paper describes the methodology for incorporating the sodium iodide symporter (NIS) in CAR T cells and for CAR T cell imaging using [¹⁸F]tetrafluoroborate-positron emission tomography ([¹⁸F]TFB-PET) in preclinical models. The methods described in this protocol can be applied to other CAR constructs and target genes in addition to the ones used for this study.

Introduction

Chimeric antigen receptor T (CAR T) cell therapy is a rapidly emerging and potentially curative approach in hematological malignancies^1,2,3,4,5,6. Extraordinary clinical outcomes were reported after CD19-directed CAR T (CART19) or B cell maturation antigen (BCMA) CAR T cell therapy². This led to the US Food and Drug Administration (FDA) approval of CART19 cells for aggressive B-cell lymphoma (axicabtagene ciloleucel (Axi-Cel)⁴, tisagenlecleucel (Tisa-Cel)³, and lisocabtagene maraleucel)⁷, acute lymphoblastic leukemia (Tisa-Cel)^5,8, mantle cell lymphoma (brexucabtagene autoleuce)⁹, and follicular lymphoma (Axi-Cel)¹⁰. Most recently, the FDA approved BCMA-directed CAR T cell therapy in patients with multiple myeloma (MM) (idecabtagene vicleucel)¹¹. Moreover, CAR T cell therapy for chronic lymphocytic leukemia (CLL) is in late-stage clinical development and is expected to receive FDA approval within the next three years¹.

Despite the unprecedented results of CAR T cell therapy, its widespread use is limited by 1) insufficient in vivo CAR T cell expansion or poor trafficking to tumor sites, which leads to lower rates of durable response^12,13 and 2) the development of life-threatening adverse events, including cytokine release syndrome (CRS)^14,15. The hallmarks of CRS include not only immune activation resulting in elevated levels of inflammatory cytokines/chemokines but also massive T cell proliferation after CAR T cell infusion^15,16. Thus, the development of a validated, clinical-grade strategy to image CAR T cells in vivo would allow 1) CAR T cell tracking in real time in vivo to monitor their trafficking to tumor sites and uncover potential mechanisms of resistance, and 2) monitoring of CAR T cell expansion and potentially predicting their toxicities such as the development of CRS.

Clinical features of mild CRS are high fever, fatigue, headache, rash, diarrhea, arthralgia, myalgia, and malaise. In more severe CRS, patients may develop tachycardia/hypotension, capillary leak, cardiac dysfunction, renal/hepatic failure, and disseminated intravascular coagulation^17,18. In general, the degree of elevation of cytokines, including interferon-gamma, granulocyte-macrophage colony-stimulating factor, interleukin (IL)-10, and IL-6, has been shown to correlate with the severity of clinical symptoms^17,19. However, the extensive application of "real-time" serum cytokine monitoring to predict CRS is difficult due to the high cost and limited availability. To exploit the beneficial characteristics of CAR T cell therapy, non-invasive imaging of adoptive T cells can be potentially utilized to predict the efficacy, toxicities, and relapse after CAR T cell infusion.

Several researchers have developed strategies to use radionuclide-based imaging with positron emission tomography (PET) or single-photon emission computed tomography (SPECT), which provides high resolution and high sensitivity^{20,21,22,23,24,25,26,27,28,29,30} for the in vivo visualization and monitoring of CAR T cell trafficking. Among those radionuclide-based imaging strategies, the sodium iodide symporter (NIS) has been developed as a sensitive modality to image cells and viruses using PET scans^31,32. NIS⁺CAR T cell imaging with [¹⁸F]TFB-PET is a sensitive, efficient, and convenient technology to assess and diagnose CAR T cell expansion, trafficking, and toxicity³⁰. This protocol describes 1) the development of NIS⁺CAR T cells through dual transduction with high efficacy and 2) a methodology for imaging NIS⁺CAR T cells with [¹⁸F]TFB-PET scan. BCMA-CAR T cells for MM are used as a proof-of-concept model to describe NIS as a reporter for CAR T cell imaging. However, these methodologies can be applied to any other CAR T cell therapy.

Protocol

The protocol follows the guidelines of Mayo Clinic's Institutional Review Board, Institutional Biosafety Committee, and Mayo Clinic's Institutional Animal Care and Use Committee.

1. NIS⁺ BCMA-CAR T cell production

NOTE: This protocol follows the guidelines of the Mayo Clinic's Institutional Review Board (IRB 17-008762) and Institutional Biosafety Committee (IBC Bios00000006.04).

1. Production of BCMA-CAR, NIS, and luciferase-green fluorescent protein (GFP)-encoding lentiviruses.
   NOTE: A second-generation BCMA-CAR construct was synthesized de novo (see the Table of Materials) and cloned into a third-generation lentiviral vector under the control of an elongation factor-1 alpha (EF-1α) promotor. The BCMA-CAR construct (C11D5.3-41BBz) included 4-1BB costimulation and a single-chain variable fragment (scFv) derived from an anti-human BCMA antibody clone C11D5.3^33,34. The NIS is under the control of the EF-1α promotor and binds to the puromycin resistance gene via self-cleaving peptides (P2A). The lentiviral vector encoding luciferase-GFP (see the Table of Materials) is used to transduce tumor cells, which then express GFP and luciferase.
   1. Prepare lentiviral vector plasmids: pLV-EF1α-BCMA-CAR (15 µg), pBMN-CMV-GFP-Luc2-Puro (15 µg), and pLV-EF1α-NIS-P2A-Puro (15 µg).
      NOTE: pBMN-CMV-GFP-Luc2-Puro and pLV-EF1α-NIS-P2A-Puro contain the puromycin resistance gene. Therefore, NIS- or luciferase-GFP-transduced cells can be selected with 1 µg/mL or 2 µg/mL of puromycin dihydrochloride, as described previously^14,35.
   2. Seed 20 × 10⁶ of 293T cells in a T175 flask and incubate for 24 h at 37 °C with 5% CO₂. Confirm that 293T cells are evenly distributed on the flask at 70-90% confluence by direct visualization under the microscope.
   3. Prepare a master mix of 15 µg of the expression vector (e.g., CAR, NIS, or luciferase-GFP linear DNA), 7 µg of the envelope vector (VSV-G), and 18 µg of the packaging vector (gag, pol, rev, and tat). Dilute the DNA master mix in 4.5 mL of the transfection medium, and then add 111 µL of the pre-complexing reagent (Mixture A).
   4. Prepare a new tube, and dilute 129 µL of the liposomal transfection reagent in 4.5 mL of the transfection medium (Mixture B).
   5. Combine Mixtures A and B and flick the tube to mix the contents. Incubate for 30 min at room temperature (RT).
   6. After the incubation, simply aspirate the cell supernatant without detaching the cells and add 16 mL of a growth medium containing 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin-glutamine. Then, add the mixture of Mixtures A and B on 293T cells dropwise. Finally, incubate the transfected cells at 37 °C, 5% CO₂ for 24 h.
   7. On days 1 and 2 post-transfection, harvest the supernatant of 293T, spin down at 900 × g for 10 min, and filter through a 0.45 µm nylon filter. Concentrate the filtrate at 24 and 48 h by ultracentrifugation at 112,700 × g for 2 h, and freeze at -80 °C.
2. Ex-vivo T cell isolation (Figure 1)
   NOTE: Perform all cell culture work in a laminar flow cabinet using aseptic technique and personal protective equipment. Peripheral blood mononuclear cells (PBMCs) are harvested from healthy volunteer donor blood collected during apheresis³⁶. Pathogen screening was performed on human cells used in this study.
   1. Use the standard density gradient technique to isolate PBMCs.
      1. Gently add 15 mL of density gradient medium (density of 1.077 g/mL) (containing alpha-D-glucopyranoside, beta-D-fructofuranosyl homopolymer, and, 3-(acetylamino)-5-(acetylmethylamino)-2,4,6-triodobenzoic acid monosodium salt) to a 50 mL density gradient separation tube without creating air bubbles (see the Table of Materials).
      2. To avoid cell trapping, dilute the blood sample with phosphate-buffered saline (PBS, 0.2 g/L of potassium chloride, 0.2 g/L of potassium phosphate monobasic, 8 g/L of sodium chloride, and 1.15 g/L of sodium phosphate dibasic) containing 2% FBS at a 1:1 volume ratio. Gently transfer the diluted blood on top of the density gradient medium without breaking the interface between the two. Spin down at 1,200 × g for 10 min at RT.
         NOTE: A 50 mL density gradient separation tube can be used for the isolation of 4-17 mL of a blood sample. The 50 mL density gradient separation tube (see the Table of Materials) used in this protocol does not require the "brake off" during centrifugation. However, when standard 50 mL tubes are used, the brake needs to be off and requires 30 min centrifugation.
      3. Transfer the supernatant into a new 50 mL conical tube, wash with PBS + 2% FBS by filling up to 50 mL, and then spin down at 300 × g for 8 min at RT.
      4. Aspirate the supernatant, and resuspend the pelleted cells in 50 mL of PBS + 2% FBS. Count the number of cells, and then spin down at 300 × g for 8 min at RT. Repeat the previous step for a total of 2 washes.
   2. Aspirate the supernatant, and resuspend the pelleted cells to a concentration of 50 × 10⁶ cells/mL with PBS + 2% FBS.
   3. Perform T cell isolation from PBMCs using a negative selection magnetic bead kit.
      NOTE: An ideal negative selection kit includes magnetic beads attached to antibodies against antigens expressed on cells other than T cells. A commonly used kit contains antibodies conjugated to magnetic beads against CD15, CD14, CD34, CD36, CD56, CD123, CD235a, CD19, and CD16 (see Table of Materials).
      1. Transfer PBMCs to a 14 mL polystyrene round-bottom tube. Then, place the PBMCs and the negative selection antibody cocktail in a fully automated cell separator, and perform T cell isolation according to the manufacturer's protocol.
3. T cell stimulation and T cell expansion (Figure 1)
   1. To culture the isolated T cells, prepare T cell expansion medium (TCM) made with serum-free hematopoietic cell medium supplemented with 10% human serum albumin and 1% penicillin-streptomycin-glutamine¹⁴. After T cell isolation, count the cells and culture at a concentration of 2 × 10⁶ cells/mL with TCM.
   2. Wash anti-CD3/CD28 beads three times with TCM before culturing with T cells.
      1. Mix the vial containing the beads by swirling. Then, pipette the required volume of beads (3:1 beads:cell) (e.g., when stimulating 1.0 × 10⁶ cells of T cells, use 3.0 × 10⁶ of anti-CD3/CD28 beads) into a sterile microcentrifuge tube (1.5 mL) and resuspend in 1 mL of TCM.
   3. Place the microcentrifuge tube with the beads on a magnet for 1 min, and aspirate the supernatant. Remove the tube from the magnet, and resuspend the washed beads in 1 mL of TCM. Repeat the previous two steps for a total of 3 washes.
   4. Resuspend the beads in 1 mL of TCM and transfer them to the T cells. Then, dilute the T cells to a final concentration of 1.0 × 10⁶ cells/mL with TCM. Transfer the T-cell/bead suspension to a tissue-culture-treated 6-well plate and place it in the incubator (37 °C, 5% CO₂).
4. Titration of lentiviruses (Figure 2)
   1. Prepare T cells for titration assay. Ensure that approximately 1.0 x 10⁶ cells are available to titrate one type of virus.
   2. Stimulate T cells as described in section 1.3.2.
   3. Plate 1.0 × 10⁵ cells of stimulated T cells in a 96-well plate (titer plate) and incubate at 37 °C, 5% CO₂ for 24 h (Figure 2A). Isolate and stimulate the T cells as described in section 1.2.
   4. Prepare a dilution plate (96-well plate) by adding 100 µL of TCM into the wells of the designated columns and the untransduced control wells (Figure 2B).
   5. Thaw one vial of lentiviral particles on ice and gently pipette up and down to mix well. Transfer 50 µL of the virus supernatant into the wells of Column 6 of the 96-well plate (dilution 3) (Figure 2B). Pipette up and down to mix well.
   6. Perform serial dilutions (2-fold serial dilution): transfer 50 µL from well A6 to well B6 and then 50 µL from well B6 to well C6; repeat until G6. Then, add 50 µL of the diluted virus to the titer plate (Figure 2B).
   7. Incubate the titer plate at 37 °C, 5% CO₂ for 48 h, and determine the percentages of CAR-, NIS-, or GFP-positive cells by flow cytometry (Figure 2C).
      1. Wash the wells by spinning down the titer plate two times at 650 × g at 4 °C for 3 min.
      2. Stain the transduced T cells as described in steps 1.5.3 to 1.5.9.
      3. Determine the titers based on the percentages of CAR-, NIS-, or GFP-positive cells by using formula (1):
         Titers = Percentage of BCMA-CAR⁺ or NIS⁺ T cells × T cell count at transduction × the specific dilution / volume (1)
5. Transduction of lentiviruses and NIS⁺BCMA-CAR T cell expansion
   1. Twenty-four to 48 h after T cell stimulation, perform lentiviral transduction on stimulated T cells (T cells should form clusters).
      1. Thaw the frozen lentiviruses encoding CAR or NIS at 4 °C.
      2. Mix the stimulated T cells well to break up the clusters, and simply add freshly thawed virus at a multiplicity of infection (MOI) of 5.0 (when transducing 1.0 × 10⁶ T cells, use 5.0 × 10⁶ of lentivirons). Incubate the transduced cells at 37 °C, 5% CO₂.
      3. On days 3, 4, and 5, count the transduced T cells using a hematocytometer³⁷ or a fully automated cell counter³⁸ and adjust the cell concentration to 1.0 × 10⁶ cells/mL by adding fresh, pre-warmed TCM. For NIS-transduced T cells carrying the puromycin resistance gene, treat the cells with 1 µg/mL of puromycin dihydrochloride on days 3, 4, and 5.
   2. On day 6, remove the anti-CD3/CD28 beads from the transduced T cells (from step 1.3.4) by mixing well to break up the T cell clusters and placing them in a magnet for 1 min. Then, simply place the collected transduced T cells back in culture at a concentration of 1.0 × 10⁶ cells/mL. After removing the beads from the T cells, assess the expression of CAR and NIS by flow cytometry.
      NOTE: As the single-chain variable fragment of the BCMA-CAR is derived from mouse, it can be stained with goat anti-mouse IgG (H+L) conjugated with Alexa Fluor 647. NIS can be detected using anti-human ETNL [synthetic peptide corresponding to aa625-643 (SWTPCVGHDGGRDQQETNL)]. This antibody recognizes the cytosolic C-terminus of NIS. Therefore, T cells must be permeabilized before incubation with an anti-human NIS antibody.
   3. Perform surface staining of BCMA-CAR using goat anti-mouse IgG (H+L).
      1. Take an aliquot of the culture (e.g., 50,000 T cells) and wash with flow buffer (PBS, 1% FBS, and 1% sodium azide). Next, resuspend the cells with 50 µL of flow buffer, and stain the cells with 1 µL of goat anti-mouse antibody for detecting CAR expression and 0.3 µL of live-dead aqua for excluding dead cells.
      2. Incubate for 15 min in the dark at RT, wash the cells by adding 150 µL of flow buffer, and centrifuge the cells at 650 × g for 3 min at 4 °C.
   4. After surface CAR staining, fix and permeabilize the cells by adding 100 µL of fixation medium (PBS with 4.21% formaldehyde) and incubate for 20 min at 4 °C. Wash the cells twice with 100 µL of a buffer that contains a cell-permeabilizing agent such as saponin (650 × g for 3 min at 4 °C).
   5. Resuspend the fixed/permeabilized cells in 50 µL of a permeabilizing buffer. Then, add 0.3 ng of anti-human ETNL NIS antibody in 50 µL of flow buffer and incubate for 1 h at 4 °C.
   6. Add 150 µL of flow buffer, and centrifuge the cells at 650 × g for 3 min at 4 °C. Incubate the cells with 2.5 µL of anti-rabbit secondary antibody in 50 µL of flow buffer for 30 min at 4 °C, wash the cells by adding 150 µL of flow buffer, and centrifuge at 650 × g for 3 min at 4 °C.
   7. Finally, resuspend in 200 µL of flow buffer and perform flow cytometry to determine the transduction efficiency (Figure 3A,B).
   8. On day 8, count and spin down the T cells at 300 × g for 8 min at 4 °C. Resuspend the T cells with freezing medium (90% FBS + 10% dimethylsulfoxide [DMSO]) at a concentration of 10 × 10⁶/mL, then transfer 1 mL each to labeled cryovials.
   9. Place the vials in a -80 °C freezer for 48 h. After 48 h (and by day 10), transfer the T cells to liquid nitrogen.
      NOTE: See Figure 1 for the overview of NIS⁺ CAR T cell production. Figure 3D represents examples of ex vivo T cell expansion from three different donors.

2. NIS⁺ BCMA-CAR T cell imaging with [ ¹⁸F]TFB-PET scan

NOTE: This protocol follows the guidelines of Mayo Clinic's Institutional Animal Care and Use Committee (IACUC A00001767-16), IRB, and IBC (Bios00000006.04). OPM-2 is a BCMA⁺ MM cell line, which is often used as a target cell line for BCMA-CAR T cells^39,40.

1. Establish luciferase⁺ BCMA⁺ OPM-2 cells.
   1. Seed 500,000 OPM-2 cells in a tissue culture-treated 24-well plate. Thaw lentivirus encoding luciferase-GFP at 4 °C.
   2. Add the freshly thawed virus at an MOI of 3.0 to OPM-2 cells and mix well by pipetting. Place the plate in the incubator (37 °C, 5% CO₂).
   3. Forty-eight hours after the transduction, add 2 µg/mL of puromycin to select the transduced cells. Four days after the transduction, assess the GFP-positive cells (luciferase-positive) by flow cytometry (Figure 3E).
2. Establish BCMA⁺ OPM-2 xenograft mouse models (Figure 4).
   1. Count luciferase⁺ OPM-2 cells and spin them down twice to remove all cell culture medium. Resuspend the OPM-2 cells at a concentration of 10 × 10⁶ cells/mL with PBS.
   2. On day -21, inject 100 µL (1.0 × 10⁶ cells) of luciferase⁺ OPM-2 cells into the tails of 8-to-12-week-old immunocompromised NOD-scid IL2rγ^null (NSG) mice (Figure 4).
   3. On day 20 after the OPM-2 cell injection (day -1 of CAR T cell injection), check the tumor burden via bioluminescence imaging (BLI) (Figure 4).
      NOTE: OPM-2 cells form a slow-growing tumor, which usually takes 2-3 weeks to engraft.
   4. Administer 10 µL/g of D-luciferin to OPM-2 xenograft mice via intraperitoneal (IP) injection. After 10 min, perform BLI on the mice under 2% isoflurane gas (Figure 5A). After confirming tumor engraftment, randomize the mice according to the tumor burden.
   5. On day -1 of the CAR T cell injection, thaw NIS⁺BCMA-CAR T cells, and remove the freezing medium by centrifugation (300 × g, 8 min, 4 °C). Then, the resuspend cells with TCM at 2.0 × 10⁶ cells/mL and incubate overnight (37 °C, 5% CO₂).
   6. On day 0, count and centrifuge the NIS⁺BCMA-CAR T cells (300 × g, 8 min, 4 °C). Resuspend the NIS⁺BCMA-CAR T cells at 50 × 10⁶ cells/mL with PBS.
   7. Administer 100 µL (5.0 × 10⁶ cells) of NIS⁺BCMA-CAR T cells via tail vein injection to the OPM-2 xenograft mice. On days 7 and 15, image the mice using a PET scan.
3. NIS⁺BCMA-CAR T cell in vivo imaging using BCMA⁺OPM-2 xenograft mouse model.
   1. Weigh the mice before the imaging, and remove any metal ear tags to eliminate metal-related artifacts.
   2. Prepare [¹⁸F]TFB as previously described⁴¹.
      NOTE: [¹⁸F]TFB must be produced the day of its use. Radiochemical purity should be >99% and molar activity >5 GBq/mmol.
   3. Inject 9.25 MBq [¹⁸F]TFB intravenously via tail vein injection. Allow an uptake period of ~40 min for the radiotracer to be distributed in the body and clear the blood.
   4. Anesthetize the mouse using isoflurane inhalation (2%).
      NOTE: Isoflurane is the preferred inhaled anesthetic as it has rapid and reliable onset and recovery.
   5. Prior to anesthetizing the mouse, clean all surfaces of the anesthesia machine with disinfectant cleaners.
   6. Place the mouse inside the induction chamber. Turn the vaporizer dial to 2% and wait for the mouse to become recumbent and non-responsive within 1-2 min.
   7. Monitor the mouse to avoid insufficient anesthesia or excessive depression of respiratory functions. In brief, pinch toe to confirm the insufficient anesthesia.
      NOTE: The normal respiratory rate is up to 180/min, and the acceptable drop rate is 50%.
   8. Apply ophthalmic ointment to avoid corneal drying and trauma.
   9. Acquire PET/CT images 45 min post-injection with the anesthetized mouse in a micro PET/CT imaging workstation (see the Table of Materials). Next, acquire static PET images for 15 min followed by CT image acquisition for 5 min with 360° rotation and 180 projections at 500 µA, 80 keV, and 200 ms exposure.
4. Analyzing acquired imaging data
   1. Analyze the images using PET image processing software (Figure 5B and Supplemental Video S1).
   2. Define the volume of interest (VOI).
   3. Calculate the standardized uptake value (SUV) using formula (2).
      SUV in VOI = Concentration of activity in VOI (MBq/mL) × body weight (g) / administered dose (MBq)  (2)
5. Confirmation of NIS⁺BCMA-CAR T cell trafficking to the tumor sites with the flow cytometry
   1. After [¹⁸F]TFB-PET imaging, place the mouse back into the cage. Following cessation of anesthesia, monitor the animals until they are capable of purposeful movement and ensure that they have access to food and water.
   2. Monitor the mice until the decay of the injected [¹⁸F]TFB. Once the radioisotope is not detectable, euthanize the mice with CO₂.
   3. To euthanize, place the mice into the cage (no more than 5 mice per cage).
   4. Expose the mice to CO₂ until complete cessation of breathing in approximately 5-10 min.
      NOTE: This euthanasia method must be consistent with the AVMA Guidelines for the Euthanasia of Animals (2020 Edition).
   5. To ensure the death of the mice, perform cervical dislocation by grasping the back of the head and the base of the tail of the animal, then pulling hands apart/away from each other.
   6. Harvest the bone marrow to confirm that the NIS⁺BCMA-CAR T cells efficiently traffic to the tumor site.
   7. Transfer the harvested femurs and tibia to a 6-well plate containing 5 mL of cell culture medium. Remove the muscles and tendons from the femurs and tibia, and simply cut both ends (above the joints) of the femurs and tibia.
   8. Fill an insulin syringe with the cell culture medium, and flush bone marrow onto the 6-well plate. For femurs, use 22 G needles and 5 mL syringes because the femur diameter is larger than that of the tibia.
   9. Use the flat end of the plunger to grind the bone marrow. Place a 70 µm cell strainer on a sterile 50 mL conical tube, and filter the ground bone marrow. Then, fill up the tube with the flow buffer, and centrifuge the tube at 300 × g for 8 min at 4 °C.
   10. Aspirate the supernatant, and resuspend the bone marrow with 5 mL of flow buffer. Transfer 200 µL of bone marrow to a 96-well plate. Centrifuge the plate at 300 × g for 3 min at 4 °C. Decant the supernatant, and resuspend the cells with 50 µL of flow buffer.
   11. Stain the bone marrow with flow antibodies against 0.25 µg of mouse CD45, 0.03 µg of human CD45, 0.06 µg of human CD3, 0.24 µg of human BCMA, and 0.3 µL of live/dead aqua. Incubate the plate for 15 min at RT in the dark.
   12. Centrifuge the plate at 300 × g for 3 min at 4 °C. Then, add 200 µL of flow buffer, and run on the flow cytometer (Figure 5C).

Representative Results

Figure 1 represents the steps of generating NIS⁺BCMA-CAR T cells. On day 0, isolate PBMCs and then isolate T cells by negative selection. Then, stimulate T cells with anti-CD3/CD28 beads. On day 1, transduce T cells with both NIS and BCMA-CAR lentiviruses. On days 3, 4, and 5, count T cells and feed with media to adjust the concentration to be 1.0 × 10⁶/mL. For NIS-transduced T cells, add 1 μg/mL of puromycin to select NIS⁺ cells. On day 6, remove the beads by placing the cells in the magnet for a minute. Then, put the de-beaded cells from the tube into a new flask. Take an aliquot of cells (e.g., 50,000 cells) and stain with antibodies to check the expression of NIS and CAR on the surface of T cells using flow cytometry. On day 8, count the T cells and cryopreserve with freezing media at the concentration of 10 × 10⁶/mL.

Figure 2 represents the outline of titrating the lentiviruses. On day 0, resuspend T cells at the concentration of 1.0 × 10⁶/mL in TCM. Then, stimulate T cells with anti-CD3/CD28 beads at a 1:3 cell:beads ratio. Add 100 μL (100,000 cells) of T cells to the colored wells as indicated in Figure 2A. This plate is called a "titer plate." Incubate the titer plate at 37 °C, 5% CO₂ for 24 h. On day 1, prepare the dilution plate. Add 100 μL of TCM to the colored wells as indicated in Figure 2B. Then, add 50 μL of freshly thawed lentiviruses into the first row (e.g., A6, A7, or A8 as depicted in Figure 2B). Perform serial dilution by transferring 50 µL from A6 to B6, and then 50 µL from B6 to C6, repeating until G6. Perform serial dilution for A7 and A8 as well. Then transfer 50 µL of the diluted virus to the titer plate. Twenty-four hours after transferring the virus from the dilution plate to the titer plate, feed the cells with 100 μL of TCM. On day 3, stain cells with antibodies and analyze the expression of NIS and BCMA on the T cells via flow cytometry.

Figure 3A,B show the representative flow plots of BCMA-CAR T or NIS⁺BCMA-CAR T cells. T cells are gated on FSC/SSC, followed by singlet and live-cell discrimination. Over 90% of cells are NIS⁺ cells (Figure 3B). Figure 3C shows the representative flow plot for the composition of NIS⁺BCMA-CAR T cells. Similar to Figure 3A,B, T cells are gated on FSC/SSC, followed by singlet and live-cell discrimination. Figure 3D shows the T cell expansion curve from days 0 to 8. There are no fold expansion differences between UTD, BCMA-CAR T, or NIS⁺BCMA-CAR T cells. Figure 3E depicts the GFP expression on OPM-2 cells after the transduction of lentivirus that encodes GFP and luciferase followed by puromycin selection.

Figure 4 is the outline of imaging NIS⁺BCMA-CAR T cells in vivo with [¹⁸F]TFB-PET. Inoculate six to eight-week-old mice with 1.0 × 10⁶ cells of luciferase⁺ OPM-2 cells via tail vein injection on day -21. Assess the tumor burden by BLI on day -1. On day 0, randomize the mice according to the tumor burden to be treated with NIS⁺BCMA-CAR T cells through tail vein injection or monitored without any treatment (untreated xenograft). Image the mice with BLI on day 6. Perform [¹⁸F]TFB-PET on day 7 to image NIS⁺BCMA-CAR T cells.

Figure 5A shows the representative BLI 20 days after the inoculation of luciferase⁺OPM-2 cells into the NSG mice. Figure 5B shows the representative PET imaging a week after administration of NIS⁺BCMA-CAR T cells. [¹⁸F]TFB uptake is observed in the sternum, spines, pelvis, and femurs. In addition, physiological uptake of [¹⁸F]TFB is seen in the thyroid and stomach. Figure 5C shows the representative flow plots of femur-derived bone marrow harvested from the untreated xenograft or NIS⁺BCMA-CAR T cell-treated mice. Bone marrow samples are stained with mouse CD45, human CD45, human CD3, and human BCMA. Cells are gated on FSC/SSC, followed by singlet, live, and human-cell discrimination. Bone marrow samples derived from untreated xenograft show BCMA⁺ cells whereas NIS⁺BCMA-CAR T cell treated mouse shows CD3⁺ cells, which support the [¹⁸F]TFB-PET finding.

Figure 1: NIS⁺BCMA-CAR T cell production schema. Normal donor CD3 T cells (isolated from peripheral blood mononuclear cells) using negative bead selection. T cells are plated at 1.0 × 10⁶/mL and expanded in TCM using anti-CD3/CD28 beads added on day 0 of culture and removed on day 6. T cells are dually transduced with lentiviruses encoding NIS or BCMA-CAR on day 1 (MOI=5.0). NIS⁺BCMA-CAR T cells are treated with 1 µg/mL of puromycin on days 3, 4, and 5. T cells are expanded in culture for 8 days. T cells are cryopreserved in FBS with 10% DMSO for future experiments. T cells are thawed and rested overnight at 37 °C before all experiments. Abbreviations: CD = cluster of differentiation; TCM = T cell expansion medium; BCMA = B cell maturation antigen; CAR = chimeric antigen receptor; NIS = sodium iodide symporter; GFP = green fluorescent protein; PBMCs = peripheral blood mononuclear cells; MOI = multiplicity of infection; FBS = fetal bovine serum; DMSO = dimethylsulfoxide. Please click here to view a larger version of this figure.

Figure 2: Lentivirus titration. (A) On day 0, stimulate T cells with anti-CD3/CD28 beads at a 3:1 beads:cell ratio. Add 100 µL of 1.0 × 10⁶/mL of stimulated T cells to the colored wells as indicated in the cartoon. Then, incubate the titer plate at 37 °C, 5% CO₂ for 24 h. (B) Prepare a dilution plate by adding 100 µL of TCM to the colored wells as indicated in the cartoon. Add 50 µL of freshly thawed virus to A6, A7, or A8 (e.g., BCMA-CAR to A6, NIS to A7, and luciferase-GFP to A8). Then, serially dilute the virus by transferring 50 µL from A6 to B6, and then 50 µL from B6 to C6, repeating until G6. Perform serial dilution for A7 and A8 as well. Transfer 50 µL of the diluted virus to the titer plate. (C) On day 3, stain the cells with corresponding antibodies and analyze the titer plate by flow cytometry. Abbreviations: CD = cluster of differentiation; TCM = T cell expansion medium; BCMA = B cell maturation antigen; CAR = chimeric antigen receptor; NIS = sodium iodide symporter; GFP = green fluorescent protein. Please click here to view a larger version of this figure.

Figure 3: The generation of NIS⁺BCMA-CAR T cells and luciferase-GFP positive OPM-2 cells. (A and B) Cells are gated on FSC/SSC, followed by singlet and live-cell discrimination. Representative flow plots of (A) untransduced T and BCMA-CAR T cells and (B) UTD and NIS⁺BCMA-CAR T are shown. NIS⁺BCMA-CAR T cells are generated by co-transduction of two viruses on day 1 of T cell expansion, as described in Figure 2. On day 6, cells are stained for CARs and NIS. (C) The phenotypic analysis of NIS⁺BCMA-CAR T. The representative flow plot of NIS⁺BCMA-CAR T cells is shown. (D) Summary of the UTD, BCMA-CAR T, or NIS⁺BCMA-CAR T cell growth kinetics. Incorporation of BCMA-CAR and/or NIS does not impact T cell expansion (two-way ANOVA, n=3 biological replicates, mean ± SD). (E) Flow cytometric analysis of luciferase-GFP-transduced OPM-2. OPM-2 cells are transduced with lentivirus encoding luciferase-GFP with puromycin resistance. Forty-eight hours after transduction, OPM-2 cells are treated with 2 µg/mL of puromycin. Cells are expanded for two more days, and the expression of GFP is analyzed via flow cytometry. Abbreviations: CD = cluster of differentiation; BCMA = B cell maturation antigen; CAR = chimeric antigen receptor; NIS = sodium iodide symporter; GFP = green fluorescent protein; UTD = untransduced; FSC/SSC = forward scattering/side scattering; ANOVA = analysis of variance; n.s.= not significant; SD = standard deviation; FL-1-A = area of fluorophore 1; FITC-A = area of fluorescein isothiocyanate. Please click here to view a larger version of this figure.

Figure 4: Scheme for in vivo trafficking assay in a systemic OPM2 xenograft model. Inject six to eight-week-old NSG mice with 1.0 × 10⁶ of luciferase-positive OPM-2 cells via the tail vein on day -21. On day -1, perform bioluminescent imaging on the mice to confirm the engraftment of OPM-2 cells. On day 0, inject mice with 5.0 × 10⁶ of NIS⁺BCMA-CAR T cells. Image mice with [¹⁸F]TFB-PET/CT on day 7 to assess the trafficking of NIS⁺BCMA-CAR T cells. Abbreviations: BCMA = B cell maturation antigen; CAR = chimeric antigen receptor; NIS = sodium iodide symporter; BLI = bioluminescent imaging; NSG = immunocompromised NOD-scid IL2rγ^null; luc = luciferase; [¹⁸F]TFB-PET/CT = [¹⁸F]tetrafluoroborate positron emission tomography/computed tomography. Please click here to view a larger version of this figure.

Figure 5: In vivo trafficking assay in a systemic OPM2 xenograft model. (A and B) BCMA⁺luciferase⁺OPM2 cells are intravenously injected into NSG mice. Mice receive NIS⁺BCMA CAR T cells three weeks after the inoculation of OPM-2 cells. (A) BLI confirms the engraftment of OPM-2 cells. (B) [¹⁸F]TFB-PET reveals NIS⁺BCMA-CAR T cell trafficking to the bone marrow. (C) To confirm that OPM-2 cells engraft in the bone marrow and NIS⁺BCMA-CAR T cells traffic to the tumor site, mice are euthanized after the imaging, and the bone marrow is harvested. Flow cytometric analysis revealed that OPM-2 cells are engrafted in the bone marrow (left), and NIS⁺BCMA-CAR T cells are present in the bone marrow (right). Abbreviations: BCMA = B cell maturation antigen; CAR = chimeric antigen receptor; NIS = sodium iodide symporter; BLI = bioluminescent imaging; NSG = immunocompromised NOD-scid IL2rγ^null; [¹⁸F]TFB-PET/CT = [¹⁸F]tetrafluoroborate positron emission tomography/computed tomography; IVIS = in vivo imaging system; CD = cluster of differentiation; SUV = standardized uptake value. Please click here to view a larger version of this figure.

Supplemental Video S1: Three-Dimensional (3D) rendering of PET/CT data showing the in vivo distribution of [¹⁸F]TFB in the thyroid, stomach, and bone marrow. Abbreviations: BCMA = B cell maturation antigen; CAR = chimeric antigen receptor; NIS = sodium iodide symporter; PET/CT = positron emission tomography/computed tomography. Please click here to download this Video.

Discussion

This paper describes a methodology for incorporating NIS into CAR T cells and imaging infused CAR T cells in vivo through [¹⁸F]TFB-PET. As proof of concept, NIS⁺BCMA-CAR T cells were generated via dual transduction. We have recently reported that incorporating NIS into CAR T cells does not impair CAR T cell functions and efficacy in vivo and allows CAR T cell trafficking and expansion³⁰. As CAR T cell therapies continue to expand beyond the current B cell malignancies to applications in CLL, there will be a greater need for tools that allow non-invasive in vivo imaging and monitoring of infused adoptive T cells. Dynamic imaging of T cells will enable the validation of adoptive T cell trafficking and potentially allow earlier detection of efficacy and toxicity.

NIS has been investigated and validated as a sensitive modality to image cells and viruses in clinical trials^32,42. Physiological accumulation of tracers for NIS is mainly seen in the thyroid/salivary glands, stomach, and bladder, which are not common organs affected by liquid tumors⁴⁴. Especially in MM, malignant plasma cells are often distributed in the bone marrow or bones, and an extramedullary plasmacytoma in lesions, where the physiological accumulation of tracers for NIS occurs, is a rare phenomenon^44,45. Furthermore, NIS is non-immunogenic and therefore suitable for longitudinal imaging studies⁴⁶. NIS is an intrinsic membrane protein that transports iodide into the cytosol and contains 13 putative transmembrane segments with an extracellular amino terminus site and cytosolic carboxy terminus⁴⁷.

NIS can be visualized with gamma- or positron-emitting radioisotopes such as technetium-99m (^99mTc) pertechnetate, iodide-123 (¹²³I), ¹³¹I, ¹²⁴I, and [¹⁸F]TFB⁴³. Recently, [¹⁸F]TFB has emerged as a promising iodide analog for NIS-based PET imaging, as it has similar biochemical properties and is radiosynthesized⁴⁸. One advantage of TFB is that it does not undergo organification in thyroid cells and therefore has a comparatively mild uptake in normal thyroid tissue⁴⁸. Another advantage of TFB is its short half-life of 109.8 min, while the half-lives of other tracers range from 12 h to 8 days, which could present safety issues for clinical applications⁴⁹. The main limitation of NIS-based CAR T cell imaging is that tracers, including TFB, do not penetrate the blood-brain barrier (BBB), making it difficult to assess neurotoxicity after CAR T cell treatment^50,51,52,53.

Neurotoxicity is associated with the infiltration of T cells and the activation of myeloid cells in the central nervous system. However, in most cases of neurotoxicity after CAR T cell therapy, the integrity of the BBB is disrupted^50,54. Therefore, it is unclear whether the tracer is unable to cross the BBB in this compromised setting. Further studies need to be carried out to validate whether neurotoxicity after CAR T cell therapy can be imaged with [¹⁸F]TFB-PET. Although the short half-life of [¹⁸F]TFB is safe for patients and staff, it makes its procurement difficult for the hospital. Therefore, institutes must be equipped with cyclotrons or have access to a regional facility. The methodology described in the protocol here can potentially be applied to a variety of other CAR T cells via dual transduction to visualize and assess CAR T cells in vivo using [¹⁸F]TFB-PET scan.

Materials

Name	Company	Catalog Number	Comments
22 Gauge needle	Covidien	8881250206
28 gauge insulin syringe	BD	329461
96 well plate	Corning	3595
Anti-human (ETNL) NIS	Imanis	REA009	ETNL antibody binds the cytosolic C-terminus of NIS
Anti-human BCMA, clone 19F2, PE-Cy7	BioLegend	357507	Flow antibody
Anti-human CD45, clone HI30, BV421	BioLegend	304032	Flow antibody
Anti-mouse CD45, clone 30-F11, APC-Cy7	BioLegend	103116	Flow antibody
Anti-rabbit IgG	R&D	F0110	Secondary antibody for NIS staining
BCMA-CAR construct, second generation	IDT, Coralville, IA
BD Cytofix/Cytoperm Fixation/Permeabilization Solution Kit	BD	554714
CD3 Monoclonal Antibody (OKT3), PE, eBioscience	Invitrogen	12-0037-42
CTS (Cell Therapy Systems) Dynabeads CD3/CD28	Gibco	40203D
CytoFLEX System  B5-R3-V5	Beckman Coulter	C04652	flow cytometer
Dimethyl sulfoxide	Millipore Sigma	D2650-100ML
Disposable Syringes with Luer-Lok Tips	BD	309646
D-Luciferin, Potassium Salt	Gold Biotechnology	LUCK-1G
D-PBS (Dulbecco's phosphate-buffered saline)	Gibco	14190-144
Dulbecco's Phosphate-Buffered Saline	Gibco	14190-144
Dynabeads MPC-S (Magnetic Particle Concentrator)	Applied Biosystems	A13346
Easy 50 EasySep Magnet	STEMCELL Technologies	18002
EasySep Human T Cell Isolation Kit	STEMCELL Technologies	17951	negative selection magnetic beads; 17951RF includes tips and buffer
Fetal bovine serum	Millipore Sigma	F8067
Goat anti-Mouse IgG (H+L) Cross-Adsorbed Secondary Antibody, Alexa Fluor 647	Invitrogen	A-21235
Inveon Multiple Modality PET/CT scanner	Siemens Medical Solutions USA, Inc.	10506989 VFT 000 03
Isoflurane liquid	Piramal Critical Care	66794-017-10
IVIS Lumina S5 Imaging System	PerkinElmer	CLS148588
IVIS® Spectrum In Vivo Imaging System	PerkinElmer	124262
Lipofectamine 3000 Transfection Reagent	Invitrogen	L3000075
LIVE/DEAD Fixable Aqua Dead Cell Stain Kit, for 405 nm excitation	Invitrogen	L34966
Lymphoprep	STEMCELL Technologies	07851
Nalgene Rapid-Flow 500 mL Vacuum Filter, 0.22 uM, sterile	Thermo Scientific	450-0020
Nalgene Rapid-Flow 500 mL Vacuum Filter, 0.45 uM, sterile	Thermo Scientific	450-0045
NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ	Jackson laboratory	05557
OPM-2	DSMZ	CRL-3273	multiple myeloma cell line
pBMN(CMV-copGFP-Luc2-Puro)	Addgene	80389	lentiviral vector encoding luciferase-GFP
Penicillin-Streptomycin-Glutamine (100x), Liquid	Gibco	10378-016
PMOD software	PMOD	PBAS and P3D
Pooled Human AB Serum Plasma Derived	Innovative Research	IPLA-SERAB-H-100ML
Puromycin Dihydrochloride	MP Biomedicals, Inc.	0210055210
RoboSep-S	STEMCELL Technologies	21000	Fully Automated Cell Separator
RPMI (Roswell Park Memorial Institute (RPMI) 1640 Medium)	Gibco	21870-076
SepMate-50 (IVD)	STEMCELL Technologies	85450	density gradient separation tubes
Sodium Azide, 5% (w/v)	Ricca Chemical	7144.8-16
T175 flask	Corning	353112
Terrell (isoflurane, USP)	Piramal Critical Care Inc	66794-019-10
Webcol Alcohol Prep	Covidien	6818
X-VIVO 15 Serum-free Hematopoietic Cell Medium	Lonza	04-418Q