Construct design and RD13-01 generation
RD13-01 contains a CAR comprising a CD7-binding scFv, a 4-1BB costimulatory domain, a CD3ζ signaling domain, and a γc intracellular domain, in addition to a EC1-EC2 extracellular domain of E-cadherin fused to a CD28 costimulatory domain. The CAR and the NKi ligand are linked by a F2A self-cleaving peptide.
The retrovirus was produced in 293GP cells, cryopreserved at -80 °C, and thawed immediately before transduction. CD3+ T cells were isolated and activated using anti-CD3/CD28 Dynabeads (Gibco, Cat no: 40203D) following the manufacturer’s protocols. The cells were cultured in X-VIVO 15 medium (Lonza, Cat no: 04-418Q) supplemented with 5% fetal bovine serum (FBS) and 300 IU/mL IL-2. Two days after T cell stimulation, the Dynabeads were removed, and CD7, TRAC, and RFX5 genes were knocked out using CRISPR/Cas9 before transduction. In brief, the T cells were washed twice with OPTI-MEM (Gibco, Cat no: 11058021) and resuspended in OPTI-MEM at a final concentration of 1~3 × 108 cells/mL. Subsequently, 0.4 mL of these cells were mixed with 120 µg of Cas9 mRNA and electroporated into the cells using a BTX Agile Pulse apparatus. Following electroporation, the cells were immediately cultured at 37 °C and 5% CO2. After 16-18 h, 40 μg of total single guide (sg)RNA targeting TRAC, CD7, and RFX5 at an equal molecular ratio (Genescript) was electroporated as described below. Lastly, the T cells were centrifuged at 2,000g for 2 h at 32℃ twice with the retrovirus on days 4 and 5 in retronectin-coated plates. The CAR-T cells were expanded for an extra 6 days.
The genomic sgRNA target sequences (with the protospacer adjacent motif in bold) were as follows: TRAC: 5'-AGAGTCTCTCAGCTGGTACACGG-3';
CD7: 5'-GTAGACATTGACCTCCGTGATGG-3';
RFX5: 5'-GGGGTTGCGGATCCACCTATAGG-3'.
In vitro cytotoxicity assay
Jurkat-Luci tumor cells were generated and used in a luciferase-based cytotoxic T lymphocyte (CTL) assay. Briefly, a firefly luciferase-encoding lentivirus was transduced into Jurkat tumor cells to generate a parental cell for the target cell line preparation. The resulting target cells were resuspended at 1 × 105 cells/mL in X-VIVO 15 medium and incubated with different ratios of T cells (e.g., 10:1 and 5:1) for 8 h at 37 °C. Subsequently, 100 μL of the mixture was transferred to a 96-well white luminometer plate. Next, 100 μL of substrate (Vazyme, Cat#DD1201-01) was added, and the luminescence was immediately determined by Infinite 200 PRO (TECAN).. The results are reported as the percentage of killing based on the luciferase activity in the wells with tumor cells but no T cells [% killing = 100 – ((relative light units (RLU) from well with effector and target cell coculture)/(RLU from well with target cells) × 100)].17
Cytokine production assay
A total of 1 × 105 CAR7 T cells/well were stimulated with Jurkat cells at a 1:1 ratio in a bottom 96-well tissue culture plate for 8 h. Cytokines produced by CAR7 T cells were measured using an enzyme-linked immunosorbent assay (ELISA) kit and antibodies against human IL-2 (R&D, Cat#DY202), human IFN-γ (R&D, Cat#DY285B), human TNF-α (R&D, Cat#DY210), and human GM-CSF (R&D, Cat#DY215).
A total of 1 × 105 CD7 KO CAR19 T cells/well were stimulated with Nalm6 cells at a 1:1 ratio in a bottom 96-well tissue culture plate for 16-18 h. Cytokines produced by CD7 KO CAR19 T cells were measured using an enzyme-linked immunosorbent assay (ELISA) kit and antibodies against human IL-2 and human IFN-γ.
Flow cytometry
Anti-human CD3 FITC (Cat#555916, DB pharmingenTM), CD7 PE (Cat#343106, Biolegend), HLA-II APC (Cat#361714, Biolegend), PD1 BV421 (Cat#329920, Biolegend), TIM3 FITC (Cat#345022, Biolegend), LAG3 PE (Cat#369306, Biolegend), CD45RO APC(Cat#304210, Biolegend), CD4 FITC (Cat#555346, DB pharmingenTM), and CD8 APC (Cat#555369, DB pharmingenTM) antibodies were used to stain cells in phosphate-buffered saline medium. All antibodies were used according to the manufacturer’s instructions. Flow cytometry was performed on a BD Celesta instrument, and the data were analyzed using the FlowJo 10 software (FlowJo, LLC).
Animal model
5- to 7-week-old NCG mice were engrafted with Jurkat-GFP-Luci cells through intravenously (i.v.) injection, and 1 × 107 of CAR7 T cells were i.v. injected 4 days later. To measure luminescence, we injected the mice with 150 mg/kg of D-luciferin intraperitoneally and monitored tumor burden by recording luminesce in an in vivo imaging system (Caliper Life Sciences) at the indicated time points. Live imaging software was used to visualize and calculate total luminescence. To analyze the tumor cells from the peripheral blood of the mice, 100 μL of blood sample was collected by orbital puncture. After red blood cell lysis, the cells were incubated for subsequent flow cytometric analysis. Survival curves were generated. Mice were euthanized when they developed signs of excessive tumor burden or when the weight loss exceeded 20% of baseline. All animal experiments were conducted in compliance with laboratory animal welfare and ethics committee.
In vivo cytokine release detection
Female, 5- to 7-week-old NCG mice received i.v. 2 × 106 Jurkat-GFP-Luci cells on day 0, while 6 days later, they were i.v. administered 1×106 freshly isolated peripheral blood mononuclear cells and 5 × 106 CAR7 T cells. To track cytokine production of both infused populations in vivo, 150 μL peripheral blood was obtained by tail-vein bleeding at specified time points. Serum samples were measured using ELISA kits for human IFN-γ (R&D, Cat#DY285B), human G-CSF (R&D, Cat#DY214), human M-CSF (R&D, Cat#DY216), human GM-CSF (R&D, Cat#DY215), human TNF-α (R&D, Cat#DY210), human IL-1β (R&D, Cat#DY201-05), human IL-6 (R&D, Cat#DY206-05), mouse IL-1β (R&D, Cat#DY401), mouse IL-6 (R&D, Cat#DY406), mouse CCL2 (R&D, Cat#DY479), mouse G-CSF (R&D, Cat#DY414), and mouse TNF-α (R&D, Cat#DY410).
Recipient CD4 T cell alloreactivity assay
Donor T cells were isolated and activated using anti-CD3/CD28 Dynabeads (Gibco). These cells were further used as stimulators after HLA-II upregulation and subsequent mitomycin-c treatment. To generate primed alloreactive recipient T cells, pretreated donor T cells were mixed with fresh allogeneic donor peripheral blood mononuclear cells at a 1:2 ratio in X-VIVO 15 medium supplemented with 10% FBS and 300 IU/mL IL-2. IL-2 were withdrawn from day 4. Ten days later, the cells were harvested, labeled with CSFE, and cocultured with CAR7 T cells from the same donor with (327KO) or without HLA-II KO (37KO), respectively, for 6 h. Following coculture, the cells were incubated with CD3 and CD4 antibodies and permeabilized for 10 min using BD FACS Permeabilizing Solution 2, followed by incubation with anti-IFN-γ antibody and analysis using flow cytometry.
NK cell alloreactivity measurement
K562-luciferase cells were infected with a lentivirus encoding a chimeric receptor comprising an EC1-EC2 extracellular domain, an E-cadherin transmembrane domain, and the CD28 intracellular domain (K562-Ecad). Three days after infection, the engineered cells were subjected to coculture with NK92 cells. KLRG1 expression was introduced in NK92 cells through KLRG1 mRNA electroporation. Target cells were cocultured with NK92 cells at different effector:target ratios at 37 °C. Luminescence was determined immediately after 4 h of coculture to evaluate the NK cell cytotoxicity. To test the cytokine release from NK cells, 1 × 105 NK92 cells/well were cocultured with K562-Ecad cells at a 1:1 ratio in a bottom 96-well tissue culture plate for 12 h. Cytokines produced by NK92 cells were measured using a human IFN-γ ELISA kit.
Protein phosphorylation assay
A total of 1 × 106 CAR7 T cells/well were stimulated with Jurkat-GFP cells at a 1:1 ratio in a bottom 96-well tissue culture plate. Next, 500 μL of 4% paraformaldehyde in PBS was added to the cells to immediately stop the reaction at the indicated time points and the cells kept at 37 °C for an additional 10 min. CAR7 T cells cocultured with non-targeted K562-GFP cells were used as a baseline control. The cells were cooled on ice for 1 min and centrifuged at 400 × g for 5 min at 4 °C. They were then permeabilized in 1 mL of ice-cold 100% methanol and incubated on ice for 30 min before washing with FACS buffer (1% FBS in PBS) three times. Subsequently, the cells were incubated with CD3 antibody. Phospho-MEK, p-ERK, p-STAT3, and p-AKT levels were evaluated using the respective phospho-specific antibodies. To detect p-CD3ζ, an unconjugated primary antibody (Rb mAb to CD3ζ (Cat#ab68235, Abcam) against its phospho-epitopes was incubated with the cells first, which were then further probed with a fluorescence-conjugated secondary antibody (donkey anti-rabbit IgG FITC (Cat#406403, Biolegend). Next, the cells were washed twice with FACS buffer, and 100,000–300,000 events were acquired on BD FACS celesta (BD Biosciences). Analysis was performed using the FlowJo software.
RNA sequencing analysis
Briefly, 5 × 106 CAR7 T cells/well were stimulated with Jurkat-GFP cells at a 4:1 ratio. The T cells were collected after 16 h, and total RNA was isolated from each sample. Libraries were prepared using SMART-Seq v4 Ultra Low Input RNA Kit (Takara). PCR products were then indexed using Nextera XT DNA Library Prep Kit (Illumina) and sequenced across 75 base pairs (bp) using a paired-end strategy with a 150-cycle high-output flow cell on NextSeq 550 (Illumina). Fastq files from replicate sequencing runs were concatenated and aligned to hg38 using spliced transcript alignment to a reference (STAR) software version 2.5.2a. A mapped read depth of 60 million reads per sample was used. The reads were aligned to hg38 using the STAR alignment software.
Detection of RD13-01 chromosomal translocation
qPCR assays were used to detect the potential occurrence of a chromosomal translocation for RFX5, CD7, and TRAC. These six translocations were labeled as TRAC:CD7, TRAC:RFX5, CD7:TRAC, CD7:RFX5, RFX5:CD7, and RFX5:TRAC. The template contained tandem sequences for the four forward and four reverse primer-flanking sequences corresponding to the four probes. The concentration of the standard plasmid was determined using its stock solution, and its copy number was calculated accordingly. Genomic DNA was extracted using Fast Pure Cell/Tissue DNA Isolation Mini Kit (Vazyme Biotech). qPCR was performed, and the gene copy numbers of the genomic DNA were calculated according to the standard curve and Cq value.
Trial design and oversight
We conducted a first-in-human, single-arm, dose-escalation phase I clinical trial of a CD7 targeting “off-the-shelf” CAR-T cell (RD13-01) therapy for patients with r//r CD7+ hematological malignancies at the Bone Marrow Transplantation Center, the First Affiliated Hospital, Zhejiang University School of Medicine. The primary objective of the study was to assess the safety of RD13-01, while the secondary objective was to assess its efficacy.
The study protocol was approved by the First Affiliated Hospital, School of Medicine, Zhejiang University Institutional Review Board and registered in ClinicalTrials.gov (NCT04538599). The 13 participants or their guardians provided written informed consent in accordance with the Declaration of Helsinki. Patient 11 (diagnosed with AML) also consented for compassionate use. One patient was excluded because of CNS involvement. A total of 12 patients were enrolled in this study. The enrolled patients received one infusion of RD13-01 between September 2020 and January 2021, except for the first enrolled patient who received split-dosing infusion for safety concerns.
Patients were given a lymphodepleting condition regimen consisting of fludarabine, cyclophosphamide, and etoposide before the RD13-01 infusion. Patient number two received a second dose on day 22 due to suboptimal expansion and unsatisfactory clinical response.
Assessment of toxic effects and response
CRS and ICANS were graded according to the ASTCT consensus47. Other toxicities during and after therapy were assessed according to the National Institutes of Health Common Terminology Criteria for Adverse Events Version 5.0 (http://ctep.cancer.gov/). Therapy responses were assessed using flow cytometry as well as morphological and imaging analyses according to National Comprehensive Cancer Network Guidelines (http://www.nccn.org).
RD13-01 manufacturing and expansion assessment
Following RD13-01 infusion, serial peripheral blood samples were collected in K2EDTA BD vacutainer tubes (BD Biosciences). The persistence of RD13-01 in fresh peripheral blood was determined using flow cytometry and qPCR. Circulating RD13-01 expansion was measured based on the proportion of CAR+ cells among CD45+ CD7- T lymphocytes. CAR DNA copies were also evaluated as another method of determining RD13-01 expansion and persistence. Genomic DNA was extracted using AxyPrep Blood Genomic DNA Miniprep Kit (Axygen) from cryopreserved peripheral blood and bone marrow samples. qPCR was performed in triplicate using AceQ qPCR Probe Master Mix (Vazyme Biotech) on a CFX ConnectTM real-time PCR system (Bio-Rad). Copy numbers per microgram of genomic DNA were calculated from a standard curve of 10-fold serial dilutions of purified CAR plasmid containing 2 × 101-2 × 106 copies/μL.
Statistical analysis
The data cutoff date was March 31, 2021. Variables were tabulated and summarized with descriptive statistics. For the time-to-event analyses, the Kaplan-Meier method was used to describe the LFS. Serum concentrations of cytokines in clinical samples were compared using the Mann-Whitney method whereas other comparisons were performed using the Student’s t-test. All P-values presented are two-tailed. P-values lower than 0.05 were considered statistically significant. Statistical analysis was performed using GraphPad Prism 7 software.