Imaging of Genetically Engineered T Cells by PET using Gold Nanoparticle Complexed to Copper-64

Abstract

Adoptive transfer of primary T cells genetically modified to have desired specificity can exert an anti-tumor response in some patients. To improve our understanding of their therapeutic potential we have developed a clinically-appealing approach to reveal their in vivo biodistribution using nanoparticles that serve as a radiotracer for imaging by positron emission tomography (PET). T cells electroporated with DNA plasmids from the Sleeping Beauty transposon/transposase system to co-express chimeric antigen receptor (CAR) specific for CD19 and Firefly luciferase (ffLuc) were propagated on CD19⁺ K562-derived artificial antigen presenting cells. The approach to generating our clinical-grade CAR⁺ T cells was adapted to electro-transfer of gold nanoparticles (GNP) functionalized with ⁶⁴Cu²⁺ using the macrocyclic chelator (1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid, DOTA) and polyethyleneglycol (GNP-⁶⁴Cu/PEG2000). MicroPET/CT was used to visualize CAR⁺EGFPffLucHyTK⁺GNP-⁶⁴Cu/PEG2000⁺ T cells and correlated with bioluminescence imaging and assessment of radioactivity. These data demonstrate that GNPs conjugated with ⁶⁴Cu²⁺ can be prepared as radiotracer for PET and used to image T cells using an approach that has translational implications.

INTRODUCTION

Adoptive cell therapy infusing primary T cells genetically modified to express a chimeric antigen receptor (CAR) specific for a tumor-associated antigen has been shown to be effective against hematologic malignancies and solid tumors.¹ The genetic manipulation of primary T cells can improve potency through the engineering of CAR² to impart a fully-competent activation signal as measured in part by the persistence and homing after infusion. To assess biodistribution of systemically-administered CAR⁺ T cells, investigators typically undertake quantitative PCR and flow cytometry using CAR-specific probes from serially sampled tissues and peripheral blood.³ However, this is invasive and does not provide real-time whole-body spatio-temporal distribution of infused T cells. Longitudinal non-invasive imaging can be undertaken on T cells genetically modified to enforce expression of reporter genes, such as Firefly luciferase (ffLuc) for bioluminescent imaging (BLI)⁴ in animal models and thymidine kinase (TK) and associated muteins from herpes simplex virus-1 for positron emission tomography (PET).⁵ Locally administered human TK⁺ T cells and intravenously-infused macaque T cells have been imaged by PET.⁶ Additional studies are needed regarding improving sensitivity and reducing immunogenicity before systemically administered TK⁺ T cells can be reproducibly imaged by PET in clinical trials. Compounding the difficulties associated with human application of this approach to PET is that ¹⁸F-based probes requires enzymatic trapping of the radiotracer in the cytoplasm by recombinant TK and the infused non-metabolized ¹⁸F creates a background signal from pools within tissues and undermines sensitivity.^{3, 6b, 7} The short radioactive half-life (t_1/2= 109.8 min) of 18F also imposes practical limitations and requires an on-site cyclotron or expedited delivery of up to 500 mCi ¹⁸F-based probes as starting material for single infusion of 10 mCi per patient. ⁶⁴Cu conjugated to lipophilic chelator pyruvaldehyde-bis(N4-methylthiosemicarbazone) for PET tracking has been proposed as an alternative and used to track C6 rat glioma cells in mice for up to 20 hours, although leakage of ⁶⁴Cu from cells was observed.⁸ Gold nanoparticles, however, have been investigated for their uptake and have been found appropriate for intracellular retention.⁹ Therefore, as an alternative to in vivo labeling we developed an ex vivo approach to radiolabel primary T cells with gold nanoparticle (GNPs) conjugated to ⁶⁴Cu (GNP-⁶⁴Cu/PEG2000) using electroporation that renders T cells capable of being imaged by PET.

Engineering of immunotherapies is a burgeoning field with active contribution from physical technologies¹⁰ and can be used to address and important challenge and an unmet clinical need for the in vivo tracking of tumor-targeting T cells.¹¹ Assessments of T-cell trafficking kinetics to tumor locations have been made.¹² A recent study by Koya et al.^12b using signal correlation from PET and BLI reporter genes showed that T cells home within 2 to 5 days leading to reduction in tumor sizes. We demonstrated that cultured CAR⁺ T cells pre-labeled with ⁶⁴Cu before infusion could be tracked in vivo using μPET/CT. Although room for improvement still remain, (e.g. towards the impact of electroporation process on T-cell death and contribution of free GNP-⁶⁴Cu/PEG2000 released from necrotic cells) our work has translational implications as we use an approach that can be undertaken in compliance with current good manufacturing practice for Phase I/II trials.

RESULTS & DISCUSSION

We have developed a strategy to genetically modify primary peripheral blood mononuclear cells (PBMC) and propagate CD19-specific CAR⁺ T cells that have application in clinical trials for patients with B-cell malignancies (INDs# 14193, 14577, and 14739). Our approach uses the Sleeping Beauty (SB) transposon/transposase system,¹³ a non-viral gene delivery method based on electro-transfer of DNA plasmids, to introduce the CAR and designer artificial antigen presenting cells (aAPC) to retrieve and numerically expand genetically modified primary T cells.¹⁴ We adapted the electroporation process to co-express a 2^nd generation CAR (that signals through CD28 and CD3-ζ) and ffLuc. This was achieved using a process we dubbed “double transposition” (Figure 1A).^13–14 Thawed PBMC isolated from the peripheral blood of healthy volunteer donors were electroporated with three SB DNA supercoiled plasmids expressing three codon optimized¹⁵ (CoOp) genes: (i) CAR transposon (Figure 1B), (ii) fusion of enhanced Green Fluorescent Protein (EGFP)_, ffLuc, hygromycin phosphotransferase (Hy), thymidine kinase (TK) to create the EGFPffLucHyTK transposon (Figure 1C), and (iii) SB11 transposase (Figure 1D). The genetically modified T cells (designated CAR⁺EGFPffLucHyTK⁺) were selectively propagated on γirradiated designer aAPC in the presence of recombinant human cytokines IL-2 and IL-21 in the presence of cytocidal concentration of hygromycin B. The aAPC (clone #4)¹⁶ are derived from K562 and genetically modified to express truncated CD19 along with desired co-stimulatory molecules (CD64, CD86, CD137L, and membrane bound IL-15 (mIL-15) co-expressed with EGFP). The electroporated and propagated T cells homogeneously expressed CD3 with 12% co-expressing CD4 and 84% co-expressing CD8. Flow cytometry also revealed 68% expression of CAR and 78% expression of EGFP (Figure S1). To track the T cells in vivo we modified the electroporation procedure to load the genetically modified T cells with GNPs modified to function as a reporter for PET. ⁶⁴Cu²⁺ was chosen as a radioisotope for PET based on longer half-life of ⁶⁴Cu (t_1/2= 12.7 hr) compared to 18F, and absence of high-energy γ-emission that could otherwise lead to DNA damage in the cells.^{17 64}Cu²⁺ was conjugated to 7 nm GNPs using the macrocyclic chelator, 1,4,7,10-tetraazacyclododecane-1,4,7,10- tetraacetic acid (DOTA)¹⁸ and coated with long chain polyethylene glycol to make it biocompatible (GNP-⁶⁴Cu/PEG2000) for the final size of 35 nm (Figure 1E).¹⁹ Although, endocytosis has been investigated as a usual mode for internalization of GNP, it requires extended periods of time,²⁰ and is a practical limitation with radioactive material which have short half life. Therefore, we developed electro-transfer as a method to label T cells with GNP-⁶⁴Cu/PEG2000 due to its ability to instantly label the cells.

Figure 1.

(A) Schematic of the processes for the generation of CAR⁺EGFPffLucHyTK⁺GNP-⁶⁴Cu/PEG2000⁺ T cells. Components used include (B) CAR transposon (CD19RCD28mZ(CoOp)/pSBSO). (C) SB11 transposon (pKan-CMV-SB11) (D) ffLuc transposon (EGFPffLucHyTK) (E) GNP-⁶⁴Cu/PEG2000; Synthesis of DOTA-TA). 7 nm GNP (shown) was compared with 15 nm GNP. All other parameters stayed the same during this comparison. TEA, triethylamine; DCM, dichloromethane.

In vitro testing of electroporated and propagated CAR⁺EGFPffLucHyTK⁺ primary T cells revealed ffLuc activity at 2.17 ± 0.08 CPM/cell upon administration of D-Luciferin, compared with 0.011 ± 0.001 CPM/cell when no D-Luciferin was administered (data not shown). The ffLuc activity is a measure of T-cell viability as this enzyme requires ATP as a co-factor which is present only in live cells. Electroporation parameters were adjusted to improve uptake of GNPs. Initially, acute T-cell leukemia cell line (Jurkat cells) were used to assess the efficiency of electro-transfer by EGFP expression using control DNA plasmid (designated as pmaxGFP). The highest transfection efficiency, as measured by transient GFP expression 24 hours after electroporation was observed at the field strength of 1 kV/cm (200 V across 2 mm electrode spacing) applied for 5 msec using square-wave pulse generator (Figure 2A). (This finding was supported by repeating the experiment (Figure S2). Although, conditions for electroporation may differ for cell types, 1 kV/cm has been previously reported in literature as an appropriate value for transfection in mammalian cells and was therefore adopted for future experiments.²¹ GNP/PEG2000 (synthesized in a similar fashion as GNP-⁶⁴Cu/PEG2000 but without ⁶⁴Cu²⁺ addition) was used to determine the concentration of GNP-⁶⁴Cu/PEG2000 to be used in the electroporation reaction. Using the conditions determined for electro-transfer of DNA plasmid into T-cell line (Jurkat cells), we first investigated the effect of GNP/PEG2000 concentration and its electro-transfer into 5x10⁶ GFP⁺ffLuc⁺ T-cell lines (Jurkat cells) (Figure 2B). Since primary cell response differ from donor to donor, we repeated this investigation on 20x10⁶ CAR⁺EGFPffLucHyTK⁺ T cells using the cells from the donor that was used in vivo (Figure 2C). GNP/PEG2000 concentration of 1.5×10¹² particles in 100 μL of electroporation reaction was considered as appropriate for electro-transfer into 20x10⁶ CAR⁺EGFPffLucHyTK⁺ primary T cells in context of cell recovery, ffLuc expression and cell viability determined by trypan blue exclusion method. To assess the impact of electro-transferred GNPs on BLI, we electroporated CAR⁺EGFPffLucHyTK⁺ T cells with 1.5×10¹² GNP/PEG2000 using 1 kV/cm for 5 msec and intravenously injected 20x10⁶ T cells in a mouse. There was no significant difference in average radiance from mouse infused with T cells that did (1.33e4 p/s/cm2/sr) and did not (1.12e4 p/s/cm2/sr) carry the GNPs (Figure 2D) which supports the premise that although the process of electroporation may have the potential to affect cell viability, the presence of GNP/PEG2000 in the cells did not detract from their ability to affect ffLuc activity.

Figure 2.

(A) Effect of electric field intensity and pulse duration on efficiency of electro-transfer of DNA plasmid (pmaxGFP) into T-cell lines (Jurkat cells). (B,C) Cell recovery, cell viability and ffLuc expression after electroporation of (B) GFP⁺ffLuc⁺ T-cell lines (Jurkat cells) and (C) CAR⁺EGFPffLucHyTK⁺ T cells with GNP/PEG2000 (normalized against non-electroporated sample) as a function of GNP/PEG2000 concentration. (D) Effect of intracellular GNP/PEG2000 on in vivo BLI signal from ffLuc. (E) Transverse, coronal, and sagittal planes of mice infused with T cells labeled with 15 nm and 7 nm GNP-⁶⁴Cu/PEG2000, respectively, imaged after 10 min and 18 hours by μPET/CT.

To assess the impact of size of GNPs to cross pores in the T-cell membrane introduced by electroporation, we electro-transferred 15 nm and 7 nm GNPs conjugated with ⁶⁴Cu²⁺ and PEG2000 into 20x10⁶ CAR⁺EGFPffLucHyTK⁺ primary T cells which resulted in 1 μCi and 5.8 μCi of radioactivity for the two sets of cells. T cells were injected into two mice and μPET imaging of transverse, coronal, and sagittal planes was undertaken beginning at 10 minutes post infusion (Figure 2E). Although initial PET signal was observed from both mice, longer term in vivo PET signal was observed only from mice that received T cells bearing 7 nm GNP-⁶⁴Cu/PEG2000. We hypothesize that 15 nm GNP-⁶⁴Cu/PEG2000 were less efficiently electro-transferred into T cells due to the increased size of the GNP.

The 7 nm GNP-⁶⁴Cu/PEG2000 was then used to label genetically modified CAR⁺EGFPffLucHyTK⁺ T cells using the established electroporation conditions (20x10⁶ T cells, 1.5×10¹² GNP-⁶⁴Cu/PEG2000, 100 μL, 1 kV/cm, 5 msec) to deliver 3 μCi. 20x10⁶ radiolabeled primary T cells were intravenously injected and imaged using μPET/CT and BLI. The PET signals from transverse, coronal, and sagittal planes of the mouse co-localized with the BLI signal (Figure 3A). As the BLI signal indicates the presence of live cells, the co-localized PET/CT and BLI data supports the hypothesis that viable (metabolically active) genetically modified T cells can be tracked in vivo using positron emitter, ⁶⁴Cu, and imaged by μPET/CT scanner.

Figure 3.

(A) PET images (transverse, coronals and sagittal planes) from CAR⁺EGFPffLucHyTK⁺GNP-⁶⁴Cu/PEG2000⁺ T cells infused in a mouse correlated with BLI signal. (B) Comparison of post-mortem biodistribution of the GNP-⁶⁴Cu/PEG2000 when enveloped inside the T cells compared to when infused directly, 14 hours after intravenous infusion. (C) Lung-to-liver and lung-to-spleen ratios of GNP-⁶⁴Cu/PEG2000 when enveloped inside the T cells compared to direct injection, 14 hours after intravenous injection.

GNPs loaded within primary T cells are expected to bypass clearance by the liver and spleen which are part of the reticuloendothelial system (RES). This was assessed by comparing the biodistribution of 20x10⁶ T cells radiolabeled with GNP-⁶⁴Cu/PEG2000 compared with free GNP-⁶⁴Cu/PEG2000 (90 μCi) also administered via tail vein. The post-mortem biodistribution of the radioactivity, 14 hours after infusion, revealed that the T cells preferentially carried the GNP-⁶⁴Cu/PEG2000 to the lungs. Whereas, when free GNP-⁶⁴Cu/PEG2000 were administered, comparatively larger fraction was found outside the pulmonary system (Figure 3B) and into major RES organs (liver and spleen)²². Kennedy et al.²³ also reported four-fold more efficient delivery of GNP to the tumor site when transported within T cells to the tumor locations, compared to accumulation via enhanced permeability and retention effect. The loading of GNP-⁶⁴Cu/PEG2000 in T cells resulted in less uptake by RES and desired delivery to the lungs. Cell-free GNP-⁶⁴Cu/PEG2000 was taken up 17.5-fold more in liver and 25-fold more in spleen compared to when packaged in T cells (Figure 3C). This supports the premise that T cells can act as delivery vehicles for nanostructures without being sequestered by the RES.^23–24

In this report, as another example of contribution of engineering approaches to immunotherapy,¹⁰ we show that ⁶⁴Cu conjugated to GNP and coated with PEG could be used to image primary T cells in vivo. A barrier to successful implementation of cell-based therapies is determination of optimal dosage and schedule for administering the biologic and ascertaining that these cells reach their desired target locations.²⁵ As a result, non-invasive real-time monitoring of infused cells has been an area of active research as it can inform on the dosage needed to achieve therapeutically effective numbers of infused cells within desired tissues. Various methods under consideration for tracking include ex vivo and in vivo labeling of cells for imaging and each has pros and cons.¹¹ An advantage of our approach using GNPs is that the conjugation to ⁶⁴Cu as well as the electro-transfer into T cells can be undertaken using methods in compliance with current good manufacturing practice for Phase I/II trials.

MATERIALS AND METHODS

DNA plasmids

Three DNA plasmids expressing kanamycin resistance gene were constructed coding for three codon optimized (CoOp) transgenes flanked by IR/DR sequences and used to generate CAR⁺EGFPffLucHyTK⁺ T cells. (1) CD19-specific CAR expressed under human elongation factor 1-α (hEF-1α) promoter (CD19RCD28mZ(CoOp)/pSBSO) (Figure 1B),²⁶ (2) multi-function protein fusing ffLuc downstream of EGFP and upstream of hygromycin phosphotransferase (Hy) and thymidine kinase (TK, HyTK) expressed under hEF-1α promoter (EGFPffLucHyTK) (Figure 1C), (3) SB11 transposase expressed under cytomegalovirus (CMV) promoter (pKan-CMV-SB11) (Figure 1D).²⁶ Supercoiled DNA plasmid pmaxGFP, coding for EGFP, was used to assess the efficiency of electro-transfer by measuring fluorescence.

GNP-⁶⁴Cu/PEG2000

GNPs with average diameter of 7 nm were purchased from Nanopartz, Inc, CO. ⁶⁴CuCl₂ was produced on a CS-15 biomedical cyclotron at Washington University School of Medicine. DOTA-TA was synthesized by linking p-NH₂- Bn-1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (Macrocyclics Inc., TX, USA) and thioctic acid (TA) (Acros, NJ, USA) with a di-amine linker.²⁷ GNPs were functionalized by forming gold-thiol bond with DOTA-TA (GNP-DOTA).²⁷ GNP-DOTA were incubated with ⁶⁴Cu²⁺ at pH 6.5 at 50°C followed by PEGylation using thiolated PEG (Item# MPEG-SH-2000; Laysan Bio, Inc, AL, USA) (GNP-⁶⁴Cu/PEG2000)²⁸ (Figure 1E). Radiolabeled GNP-⁶⁴Cu/PEG2000 was measured 35 nm in diameter by dynamic light scattering and the labeling efficiency was > 95% determined by radio thin layer chromatography.

Artificial antigen presenting cells (aAPC)

K562-derived aAPCs (referred to as clone #4) weretransduced with lentivirus to co-express CD19, CD64, CD86, CD137L (CD137L), and membrane-bound (m) interleukin (IL)-15 (mIL-15, co-expressed with EGFP, Figure 1A) and used to support the ex vivo numeric expansion of CD19-specific CAR⁺EGFPffLucHyTK⁺ T cells.¹⁶

Peripheral Blood Mononuclear Cells (PBMC)

PBMC were isolated from blood by Ficoll- Paque density gradient centrifugation (Cat # 17-1440-02, GE Healthcare Bio-sciences AB, Sweden), re-suspended in freezing medium (40% RPMI-1640, 50% fetal bovine serum (FBS), 10% DMSO) at the density of 20x10⁶ cells/mL/vial, and stored in liquid nitrogen.^14a

Generation of EGFP⁺ acute T-cell leukemia cell line (Jurkat cells)

On the day of electroporation (Day 0), 10⁶ Jurkat cells were re-suspended in 100 μL of phosphate buffered saline (PBS) (Cat# D8537, Sigma-Aldrich, Inc., MO, USA) along with pmaxGFP (2 μg supercoiled DNA, Lonza Cologne GmbH, Germany), transferred to a single cuvette, and electroporated with specific pulse duration and amplitude using ECM 830 BTX electroporation apparatus (BTX Instrument Division, Harvard Apparatus, Inc., MA, USA). The cells were cultured for 24 hours at 37°C in phenol-free RPMI-1640 containing 10% FBS and 2 mM L-alanyl-L-glutamine (GlutaMAX^™, Cat # 35050, Invitrogen Corp., CA, USA) before analyzing for EGFP expression.

Generation of GFP⁺ffLuc⁺ acute T-cell leukemia cell line (Jurkat cells)

Jurkat cells were cultured in RPMI-1640 containing 10% FBS and 2 mM L-alanyl-L-glutamine (GlutaMAX^™, Cat # 35050, Invitrogen Corp., CA, USA) and transduced with lentiviral particles carrying ffLuc-F2A-GFP reporter gene expressed under blasticidin resistance gene (Cat# LVP323, GenTarget Inc., San Diego, CA, USA) using manufacturer’s protocol. Briefly, lentiviral particles were thawed at room temperature and 50 μL of virus was added to 0.5 mL cells (diluted to 10⁶ cells/mL). Blasticidin S (Cat # ant-bl-5, InvivoGen, CA, USA) was added to the culture beginning 4 days after transduction and maintained at a concentration of 0.1 mg/mL until cryopreservation or use.

Generation and analysis of CAR⁺EGFPffLucHyTK⁺ T cells

CD19-specific CAR⁺EGFPffLucHyTK⁺ T cells were propagated from PBMC after SB transposition as previously described^14a and depicted in Figure 1A. On the day of electroporation (Day 0), 20x10⁶ PBMC were thawed at 37°C in a water bath and re-suspended in 100 μL of Amaxa Nucleofector solution (Human T cell Kit, Cat # VPA-1002) along with CAR transposon (CD19RCD28mZ(CoOp)/pSBSO, 7.5 μg supercoiled DNA, Figure 1B), ffLuc transposon (EGFPffLucHyTK, 7.5 μg supercoiled DNA, Figure 1C), and SB transposase (pKan-CMV-SB11, 5 μg supercoiled DNA, Figure 1D) and transferred to a single cuvette and electroporated (Program U-14) using a Nucleofector II (Lonza Cologne GmbH, Germany). The cells were rested for 2 to 3 hours at 37°C in incomplete phenol-free RPMI-1640 and subsequently cultured overnight in phenol-free RPMI-1640 containing 10% FBS and 2 mM L-alanyl-L-glutamine (GlutaMAX^™, Cat # 35050, Invitrogen Corp., CA, USA). The following day γ-irradiated (100 Gy) K562-aAPC (clone #4) were added at a 1:2 T cell/aAPC ratio. Hygromycin B (HygroGold^™, Cat # ant-hg-1, InvivoGen, CA, USA) was added to the co-culture beginning 5 days after electroporation and maintained at a concentration of 0.2 mg/mL, as suggested by the manufacturer, until cryopreservation of T cells. Additional γ-irradiated aAPC were added at 1:2 T cell/aAPC on days 7, 14, 21, 28, and 35 after electroporation. Soluble recombinant human IL-21 (Cat # 34-8219-85, eBioscience, Inc., CA, USA) was added at a concentration of 30 ng/mL beginning the day after electroporation, and soluble recombinant human IL-2 (Chiron Corp (Novartis V&D), CA, USA) was added to the cultures at 50 U/mL beginning 7 days after electroporation. These exogenous cytokines (IL-2, IL-21) continued to be supplemented on a Monday-Wednesday-Friday schedule. The cultures were monitored by flow cytometry for the unwanted presence of a CD3^neg CD56⁺ cell population and if the percentage exceeded 10% of the total population (which usually occurred between day 10 and 14) a depletion for CD3^negCD56⁺ NK cells was accomplished using CD56 beads (Cat # 130-050-401, Miltenyi Biotech Inc., CA, USA) on a LS column (Cat # 130-042-401, Miltenyi Biotech Inc., CA, USA) according to the manufacturer’s instructions. T cells were enumerated every 7 days and viable cells counted based on trypan blue exclusion using Cellometer automated cell counter (Auto T4 Cell Counter, Nexcelom Bioscience, MA, USA). 35 days after electro-transfer of SB plasmids the CAR⁺EGFPffLucHyTK⁺ cells were re-suspended in freezing medium at 20x10⁶ cells/mL/ vial and stored in liquid nitrogen.

Generation of CAR⁺EGFPffLucHyTK⁺ T cells loaded with GNP-⁶⁴Cu/PEG2000

20x10⁶ CD19-specific CAR⁺EGFPffLucHyTK⁺ T cells were thawed and stimulated with γ-irradiated aAPC (clone #4) in presence of IL-2/21 and hygromycin B as described above.^14a 37 days later the cells were centrifuged and 20x10⁶ T cells were re-suspended in 100 μL in presence of 1X10¹² GNP-⁶⁴Cu/PEG2000 (Figure 1E) in PBS (Cat# D8537, Sigma-Aldrich, Inc., MO, USA) and transferred to a single cuvette and electroporated at 200 V with a 5 msec pulse using ECM 830 BTX electroporation apparatus (BTX Instrument Division, Harvard Apparatus, Inc., MA, USA).

In vivo measurement of PET signal from CAR⁺EGFPffLucHyTK⁺GNP-⁶⁴Cu/PEG2000⁺ T cells

Micro-PET/CT imaging was acquired by Inveon Preclinical Multimodel SPECT/PET/CT System (Siemens AG, USA). Immunocompromised (nu/nu) mice were ordered from Charles River Laboratories International Inc., MA. CAR⁺EGFPffLucHyTK⁺GNP-⁶⁴Cu/PEG2000⁺ T cells were intravenously injected (20x10⁶ cells/mouse). All mice imaging was performed at The Methodist Hospital in accordance with guidelines from Animal Care and Use Committee. Mice were anesthetized, placed on an imaging bed attached with heating pad, laser-aligned to the center of the scanner field-of-view and serially imaged by μPET/CT scan. Images were acquired at 10 min after injection of T cells. CT image reconstruction was achieved using a Common Cone-Beam Reconstruction (COBRA) method (Siemens) and PET images were reconstructed by 2D Ordered Subset Expectation Maximization (OSEM2D) algorithm. The PET and CT images were co-registered and viewed using Inveon Research Workplace software (Siemens).

In vivo measurement of BLI signal from CAR⁺EGFPffLucHyTK⁺GNP-⁶⁴Cu/PEG2000⁺ T cells

After PET imaging, mice were injected with 150 mg/kg D-Luciferin (XenoLight^™, Cat # 122796, Caliper Life Sciences, MA, USA) and BLI was obtained using an IVIS-200 optical imaging system (Caliper Life Sciences, MA, USA). BLI were acquired using 1 minute exposure time and displayed in the same scale of intensity.

Biodistribution studies

Two mice were injected intravenously with CAR⁺EGFPffLucHyTK⁺GNP-⁶⁴Cu/PEG2000⁺ T cells and GNP-⁶⁴Cu/PEG2000. Mice were euthanized at 14 hours post injection. Blood, heart, liver, spleen, kidney, lung, stomach, intestine, muscle were harvested and weighed. Radioactivity of different organs was measured with the gamma counter (Perkin Elmer, MA). Uptake of radiolabeled T cells in tissues was calculated as the percentage of the injected dose per gram of tissue (%ID/g).

Supplementary Material

ESI

Figure S1. Flow cytometry analyses of CD19-specific CAR⁺EGFPffLucHyTK⁺ T cells before cryopreservation 35 days after electro-transfer of SB plasmids. 10⁶ T cells in 100 μL volume were stained with fluorochrome-conjugated to antibody [phycoerythrin (PE), peridinin chlorophyll protein conjugated to cyanine dye (PerCPCy5.5), allophycocyanin (APC)]: anti–CD3 PE (Cat. # 347347, 2.5 μL, BD Biosciences), anti–CD4 APC (Cat. # 555349, 2.5 μL, BD Biosciences), anti–CD8 PerCPCy5.5 (Cat. # 341051, 4 μL, BD Biosciences), and PE-conjugated F(ab′)₂ fragment of goat anti-human Fcγ (Cat. # H10104, 2.5 μL, Invitrogen), used to detect cell surface expression of the CAR. EGFP expression was analyzed and used as marker for estimating ffLuc expression. Blocking of nonspecific antibody binding was achieved using wash buffer (2% FBS and 0.1% Sodium azide in PBS). Data acquisition was on a FACSCalibur (BD Biosciences) using CellQuest version 3.3 (BD Biosciences).

Figure S2. Repeat of Figure 2A: Effect of electric field intensity and pulse duration on efficiency of electro-transfer of DNA plasmid (pmaxGFP) into T-cell lines (Jurkat cells).

INSIGHT, INNOVATION, INTEGRATION.

Clinical-grade primary T cells can be used to prevent and treat malignancies. Ex vivo manipulation of T cells improves in vivo effector functions. The link between what can be achieved in the manufacturing suites and what is achieved after infusion is compromised by an inability to assess T-cell spatio-temporal distribution. A clinically-appealing approach is needed to assess the biodistribution of T cells. Therefore, we developed gold nanoparticles (GNPs) for positron emission tomography (PET) by coupling to ⁶⁴Cu. Since we have adapted electroporation to genetically modify T cells for human application, we electro-transferred GNPs into T cells. These T cells could report their distribution in vivo by PET and represents a step towards developing GNPs as radiolabels for cell-based therapies.