Sage Journals: Discover world-class research

Abstract

Background:

Immunotherapy using engineered T cells expressing chimeric antigen receptor (CAR) is used in an increasing number of cases. The CAR molecule has an antigen-binding domain usually consisting of an antibody-derived V_L and V_H domains. The arrangement of V_L and V_H domains has been shown to affect CAR characteristics and functions of the CAR-transduced T (CAR-T) cells.

Objectives:

This study aims to clarify whether the relative arrangement of V_L and V_H domains in CAR affects the exhaustion status of CAR-T cells especially bearing herpes virus entry mediator (HVEM)-derived co-stimulatory signal domain (CSSD).

Method:

We generated CARs bearing the HVEM-derived CSSD and the V_L and V_H domains derived from HIV-1 broadly neutralizing antibody VRC01 in different arrangements (V_L-V_H or V_H-V_L) as a model system, and compared effector functions and characteristics of the CAR-T cells.

Results:

In the transduced T cells, the V_L-V_H CAR was expressed at higher levels and led to more potent effector functions than the V_H-V_L CAR. Phenotypic analysis showed that the V_H-V_LCAR-T cells contained two populations: one with higher CAR expression and CAR-mediated tonic signaling associated with more exhausted phenotype, and the other with the opposite characteristics. Interestingly, the V_L-V_H CAR-T cells, although expressing higher levels of CAR, exhibited lower CAR-mediated tonic signaling associated with lower percentage of exhausted population, which led to higher expansion upon cognate antigen simulation. These results suggested that the V_L-V_H CAR-T cells appeared to escape exhaustion induced by CAR-mediated tonic signaling despite the high CAR expression.

Conclusions:

The arrangement of V_L and V_H domains in VRC01-based CAR appears to play important roles in not only functions but also exhaustion status of CAR-T cells with HVEM-derived CSSD. Therefore, preclinical validation of the relative position of V_L and V_H domains combined with various CSSDs in CAR could assist in developing effective CAR-T cell-mediated immunotherapy.

Introduction

Immunotherapy using human T cells transduced with the chimeric antigen receptor (CAR) gene, so-called CAR-transduced T (CAR-T) cell therapy, is expected to become a useful therapeutic intervention against otherwise intractable diseases. For example, the CD19-targeted CAR-T cells demonstrated remarkable efficacy against B cell malignancy,1 and the clinical application has been started. On the other hand, constitutive tonic signaling in the absence of ligands is an increasingly recognized complication for synthetic fusion receptors including CAR and can be a cause of poor antitumor efficacy, impaired survival, and reduced persistence of the CAR-T cells in vivo. 2 Thus, it is necessary to analyze the relationship between CAR characteristics and CAR-T cell activities to further improve CAR-T cell technology.

CAR consists of an extracellular antigen-binding domain combined with an intracellular signal transduction domain. The first generation CAR-T cells that have CD3ζ as a signal transduction domain have been shown to often become anergic and fail to elicit a potent immune response.3 The second and third generation CAR-T cells have been developed by adding one and two co-stimulatory signal domains (CSSDs) in the CAR, respectively.4 These CAR-T cells could successfully mimic the T cell receptor-mediated signal transduction upon stimulation by cognate antigens, leading to proliferation and activation.5 This indicates that a signaling assumed by the CSSD is a key event for CAR-T cells to exhibit potent efficacy.4 So far, the CSSDs derived from various co-stimulatory molecules, such as CD28 and 4-1BB, have been employed for constructing CAR.6,7

Accumulating reports have suggested the roles of herpes virus entry mediator (HVEM) in effector function and memory development of CD8 T cells.8,9 In addition, tumor cells that express anti-HVEM single chain Fv (scFv) induce a potent proliferation and cytokine production of co-cultured T cells.10 These findings have suggested that HVEM serves as a potent co-stimulatory molecule in T cells. Thus, we have tested HVEM as a CSSD in CAR-T cells and demonstrated that the HVEM-derived CSSD could be useful for generating potent CAR-T cells.11

To construct the extracellular antigen-binding domain of CAR, the scFv composed of the light- and heavy-chain variable domains (V_L and V_H, respectively) derived from a monoclonal antibody against target antigens has usually been used.4,12,13 It has previously been shown that the arrangement of V_L and V_H domains could affect the characteristics of the scFv expressed in Escherichia coli, such as expression level, antigen binding, cross-reactivity, immunotoxin activity, and viral neutralization activity.14 –20 It has also been reported that the artificial GD2 and CD3-bispecific antibodies could redirect cytotoxic T cells to tumor cells and the activity depended on the arrangement of V_L and V_H domains.21 In another study, the functions of the CAR-T cells targeted to C-type lectin-like molecule-1 for human acute myeloid leukemia cells have been compared by changing the arrangement of V_L and V_H domains in the context of CAR with 4-1BB-derived CSSDs.22 Thus, the arrangement of V_L and V_H domains in CAR could affect its expression levels and CAR-T cell funcions.21 –24 However, it is still unknown whether the arrangement of V_L and V_H domains in the context of CAR with HVEM-derived CSSD could affect the CAR-T cell functions and exhaustion status in particular.

In order to investigate this issue, we generated the CAR bearing V_L and V_H domains derived from an antihuman immunodeficiency virus type 1 (HIV-1) Env antibody VRC01 in different arrangements, namely, V_L-V_H and V_H-V_L, combined with the HVEM-derived CSSD as a model system and compared the effector functions and exhaustion status of the CAR-T cells expressing the respective CAR molecules. The results showed that the V_L-V_H CAR was expressed at higher levels than the V_H-V_L CAR, leading to enhanced effector functions and killing activities against the HIV-1 Env-expressing cells. Further analysis indicated that the V_L-V_H CAR-T cells contained a smaller proportion of exhausted cells than that of the V_H-V_L CAR-T cells. In addition, the V_H-V_L CAR-T cells contained two distinct populations of the cells: one expressing relatively high levels of the CAR associated with higher CAR-mediated tonic signaling and more exhausted phenotype and the other expressing lower levels of the CAR associated with lower CAR-mediated tonic signaling and less exhausted phenotype. Interestingly, the V_L-V_H CAR-T cells exhibited less exhausted phenotype associated with lower CAR-mediated tonic signaling than the V_H-V_L CAR-T cells, which led to higher proliferation upon cognate antigen simulation. The results suggested that the V_L-V_H CAR-T cells appeared to escape exhaustion induced by CAR-mediated tonic signaling despite the high CAR expression. Taken together, the arrangement of V_L and V_H in the extracellular domain of the CAR appears to play important roles in not only functions but also exhaustion status of CAR-T cells with HVEM-derived CSSDs. For developing potent and effective CAR-T cell-mediated immunotherapy, it may be necessary to properly design the extracellular structure including the arrangement of V_L and V_H domains in CARs with various CSSDs.

Materials and Methods

Ethics approval and consent to participate

This study was conducted according to the principles of the Declaration of Helsinki and with the approval of the Scientific Ethics Committee of the Dokkyo Medical University. Written informed consent was obtained from all subjects after oral explanation of this study.

Cell culture

Minimum essential medium (MEM), Dulbecco’s modified eagle medium (DMEM), and Roswell Park Memorial Institute (RPMI) medium (Thermo Fisher Scientific, Waltham, MA) were supplemented with 10% FBS (Thermo Fisher Scientific), 2 mM glutamine (Thermo Fisher Scientific), 10 U mL⁻¹ penicillin, and 10 µg mL⁻¹ streptomycin (Thermo Fisher Scientific) and named as M10, D10, and R10, respectively. Chinese hamster ovary (CHO) cells expressing GFP (CHO-GFP) or both HIV-1 Env and GFP (CHO-Env-GFP) were maintained in M10 supplemented with nonessential amino acid (NEAA) containing 5 mg mL⁻¹ blastcidine (BSD).11 To prepare target cells for stimulation of CAR-T cells, subconfluent CHO-GFP cells or CHO-Env-GFP cells on a 150 mm culture dish were cultured in the presence of 10 µg mL⁻¹ mitomycin C (Tokyo Chemical Industry) for 3 h. Cells washed with PBS were harvested with a cell scraper (Corning, Corning, NY). Single cell suspension was obtained through a 70 µm nylon cell strainer (Corning). 293FT cells were cultured in D10, and Jurkat E6.1 cells obtained from European Collection of Authenticated Cell Cultures through DS Pharma were maintained in R10. Human peripheral blood mononuclear cells (PBMCs) from healthy donors were prepared by Ficoll-Paque (GE Healthcare Life Sciences, Pittsburgh, PA) density gradient and cultured in AIM-V (Thermo Fisher Scientific) supplemented with 5% FBS, 10 mM HEPES (complete AIM-V) overnight to remove adherent monocytes. The monocyte-depleted PBMCs were used for transduction experiments. All the cells were grown at 37°C with 5% CO₂.

Vector constructions

The cDNA fragment encoding the antigen-binging domain with the V_H-V_L arrangement was artificially synthesized using Genscript (Piscataway, NJ) from an anti-HIV-1 Env (clone: VRC01).25 A series of PCR reactions were carried out by using KOD-Plus-Ver.2 (TOYOBO) for making the cDNA fragment encoding the VRC01-drived antigen-binding domain with the V_L-V_H arrangement (Supplementary Fig. S1). The GGGGS linker sequence was added to the 5′ end of the V_L region by PCR with primers VL-GS-F1 and VL01-GS-R, yielding the V_L-GS fragment. Then, the XbaI restriction site and the leader sequence were added to the V_L-GS fragment by PCR with primers VL-GS-F2 and VL01-GS-R, yielding the SP-V_L-GS fragment (Supplementary Fig. S1A). The GGGGS linker sequence and the EcoRI restriction site were added to the 5′ and 3′ end, respectively, of the V_H region by PCR with primers GS-VH-F and GS-VH-R, yielding the GS-VH fragment (Supplementary Fig. S1B). The assembly PCR was carried out by using the GS-V_H and the SP-V_L-GS fragments as templates with VL-GS-F2 and GS-VH-R primers (Supplementary Fig. S1C). The purified DNA fragment was incubated with Ampli Taq (Thermo Fisher Scientific, Waltham, MA) at 72°C for 10 min to add A tails. The A-tailed DNA fragment was ligated with the pGEM-T easy vector (Promega, Madison, WI). The sequence was verified by using a BigDye Terminator v3.1 cycle sequencing kit (Thermo Fisher Scientific). To construct the CAR-expressing lentiviral vectors, XbaI/EcoRI digested VRC01 scFv DNA fragments (V_H-V_L or V_L-V_H) and each of EcoRI/BamHI digested DNA fragments encoding signal sequences derived from HVEM and CD3ζ were ligated into XbaI/BamHI digested pTK643-EF1a-IRES-GFP/BSD.11

Preparation of the vector virus and HIV-1

The vector lentivirus was prepared as described previously11 with some modifications. Briefly, 293FT cells were cultured at 80-90% confluency on the 150 mm dish (IWAKI, Shizuoka, Japan) coated with poly-l-lysine (Nacalai Tesque, Kyoto, Japan). The culture medium was replaced with D10 containing 25 µM chloroquine (SIGMA, Darmstadt, Germany). Then, the cells were co-transfected with the lentiviral vector plasmid (39 µg), ΔNRF (26 µg), and pMD.G (13 µg) by using the polyethyleneimine (PEI) “MAX” reagent (Polysciences, Warrington, PA) at the DNA-to-PEI ratio of 2:1. After transfection, the culture medium was replaced with D10 containing 5 mM sodium butylate (WAKO, Osaka, Japan) and 10 µM forskolin (Tokyo Chemical Industry, Tokyo, Japan). The culture supernatants were harvested at 48 h after transfection, cleared by centrifugation, and 0.45 µm filtration (Millipore, Burlington, MA). For preparing the concentrated vector virus, the filtered samples were subjected to high-speed centrifugation at 18,000 rpm for 3 h using Himac CR21N (Hitachi Koki, Tokyo, Japan). Titers of the concentrated vector were measured on HeLa cells based on the percentage of GFP positive cells by flow cytometry. For propagation of HIV-1, 293FT cells were transfected with pNL4-326 obtained from Dr. Malcom Martin through NIH AIDS Reagent Program) by using the PEI “MAX” reagent as described above. The viral titers were measured on U337-MAGI-X4 cells27 (obtained from Dr. Michael Emerman through NIH AIDS Reagent Program) according to the protocol provided by AIDS reagent program. Propagation of HIV-1 was performed in the biosafety level (BSL) 3 facility.

Transduction with lentiviral vectors

Two million Jurkat E6.1 cells were seeded in each well of the six-well plate and inoculated with 1 mL of unconcentrated vector virus preparation in the presence of polybrane (8 µg mL⁻¹). The cells were centrifuged at 5,500 rpm for 3 h at 22°C to enhance viral infection. After removal of the supernatant, the cells were cultured at 37°C in a CO₂ incubator for 48 h. The culture medium was replaced with the media supplemented with 10 µg mL⁻¹ BSD. Thereafter, transduced cells were maintained in the media with 10 µg mL⁻¹ BSD until the following assays were performed.

Human primary CD4⁺ and CD8⁺ T cells were isolated from monocyte-depleted PBMCs by using a FACS Aria II cell sorter (BD Biosciences, Franklin Lakes, NJ) with antihuman CD3-APC, CD4-PE, and CD8-PE/Cy7 antibodies (all from Biolegend, San Diego, CA), and purity of more than 95% was routinely achieved. The purified CD8⁺ T cells were activated with anti-CD3/CD28 beads (Thermo Fisher Scientific) at the bead-to-cell ratio of 3:1 in complete AIM-V supplemented with 40 U/mL recombinant human IL-2 (obtained from Dr. Maurice Gately through NIH AIDS Reagent Program). On the next day, activated CD8⁺ T cells were transduced with concentrated recombinant lentivirus by using the plates coated with Retronectin (TAKARA, Shiga, Japan) according to the manufacturer’s instructions. The cells transduced with the vector expressing only GFP were used as a control. After incubation for 48 h, the supernatant was replaced with complete AIM-V supplemented with 300 U mL⁻¹ of recombinant human IL-2. The anti-CD3/CD28 beads were removed at 48 h after changing the media. Thereafter, the cells were cultured for 16-18 days by replenishing fresh media containing IL-2 (300 U mL⁻¹) every 2 to 3 days.

Flow cytometry

For flow cytometry, the centrifuged cells were resuspended in the fluorescence-activated cell sorter (FACS) buffer (PBS containing 2% FBS and 0.02% sodium azide) containing the antibodies and incubated on ice for 30 min otherwise indicated. After washing with ice-cold FACS buffer, the cells were fixed with 1% paraformaldehyde/PBS and analyzed by using a flow cytometer (FACS Calibur or LSRFortessa X-20; BD Biosciences). The data were analyzed using the FlowJo software (BD Biosciences). The following antibodies used for flow cytometry were obtained from Biolegend, unless it is indicated otherwise. The CAR expression on the transduced Jurkat E6.1 cells was analyzed with an anti-c-myc tag antibody (Santa Cruz Biotechnology, Dallas, TX) and an antimouse Immunoglobulin(Ig)s-PE antibody (Agilent Technologies, Santa Clara, CA) as primary and secondary antibodies, respectively. The activation of the vector-transduced Jurkat E6.1 cells in the co-culture assay was analyzed with antihuman CD3-APC and CD69-PE antibodies. The CAR expression on the vector-transduced primary CD8⁺ T cells was analyzed with the biotinylated anti-c-myc tag antibody followed by streptavidin-PE (TONBO biosciences, San Diego, CA). For evaluating exhaustion of the CAR-T cells, flow cytometry was performed with antihuman PD-1-APC and antihuman LAG-3-PE/Cy7 antibodies. For detecting CD3z phosphorylation in CAR-T cells, the vector-transduced primary CD8⁺ T cells were stained with the biotinylated anti-c-myc tag antibody followed by streptavidin-PE. Stained cells were fixed at 37°C for 10 min in Fix Buffer I (BD) and permeabilized at 4°C for 30 min in Perm Buffer III (BD). Permeabilized cells were incubated with the anti-CD247(pY142)-PE antibody (BD) at room temperature for 30 min.

Western blot

Cells (5 × 10⁶) were washed once with ice-cold PBS, resuspended in 150 µL of RIPA buffer (10 mM TrisHCl [pH 7.4], 1% NP-40, 0.1% sodium deoxychorate, 0.1% SDS, 0.15 M NaCl, and 1 mM EDTA) supplemented with Complete Mini (Roche, Mannheim, Germany) and 1 mM PMSF and incubated on ice for 30 min. The cell lysate cleared by centrifugation (20 µL) was mixed with the NuPAGE LDS sample buffer (Thermo Fisher Scientific) and 0.1 M dithiothreitol and boiled at 95°C for 5 min. The boiled samples were separated by electrophoresis in a NuPAGE 10% Bis-Tris gel (Thermo Fisher Scientific) and transferred to an Immobilon P membrane (Millipore, Burlington, MA). The membrane was blocked with the blocking buffer (5% nonfat dried milk and 0.1% Tween-20 in Tris-buffered saline [TBS]) for 30 min. The blocked membrane was incubated overnight at 4°C with the anti-Actin antibody (I-19) or anti-CD3ζ antibody (F-3) (Santa Cruz Biotechnology, Dallas, TX) diluted in a blocking buffer. The membrane was then incubated for 1 hour with antigoat IgG-HRP (Millipore) or antimouse IgG-HRP (GE Healthcare UK Ltd., Buckinghamshire, England) diluted in a blocking buffer. The membrane was washed three times with 0.1% Tween-20/TBS after all antibody incubation steps. Immunoreactivity was visualized with Lumi-Light PLUS (Roche) as a substrate and detected with Light-Capture II (ATTO, Tokyo, Japan). The band intensity was quantified using ImageJ (National Institute of Health, MD).

Co-culture assay for effector functions of the CAR-T cells

Activation and IL-2 secretion of the CAR-transduced Jurkat E6.1 cells were determined by co-culturing with target cells (CHO-GFP or CHO-Env-GFP). Target cells (1 × 10⁵) were seeded in each well of a 96 well flat-bottom plate. CAR-transduced Jurkat E6.1 cells (2 × 10⁵) were added and co-cultured for 24 h. On the next day, the culture supernatants including the cells were harvested and centrifuged to separate cells and cell-free supernatants. IL-2 secretion into the supernatant was measured using Human IL-2 ELISA MAX Deluxe (Biolegend) according to the manufacturer’s instructions. The remaining cells were used to determine CD69 expression by flow cytometry.

To determine cytotoxicity of the CAR-T cells, target cells (CHO-GFP or CHO-Env-GFP, 1 x 10⁴) were seeded in each well of 96 well flat-bottom plate. The CAR-T cells were added at different effector-to-target ratios (see also in figure legend) and co-cultured in phenol red-free R10 with NEAA. After 24 h of incubation, the cell-free supernatant was collected by centrifugation. Lactate dehydrogenase release in the supernatant was assayed using CytoTox 96 nonradioactive cytotoxicity assay (Promega). The cytotoxicity (% lysis) was calculated according to the manufacturer’s instructions. In cases of unequal percentage of CAR-transduced cells in the cultures, un-transduced T cells were added to ensure that both the number of CAR⁺ T cells and the total number of T cells remained consistent across the CAR-T cell groups. The collected supernatants were also used to determine IL-2, TNF-α, and IFN-γ secretion using human IL-2, TNF-α, and IFN-γ ELISA MAX Deluxe (all from Biolegend) according to the manufacturer’s instructions.

Co-culture for anti-HIV-1 activity of the CAR-T cells

CD4⁺ T cells isolated from the same donor of CD8⁺ T cells, which have been used for generating CAR-T cells, were activated with anti-CD3/CD28 beads at the bead-to-cell ratio of 3:1 in complete AIM-V supplemented with 40 U mL⁻¹ IL-2. The anti-CD3/CD28 beads were removed at days 3 and cultured for 9 days with replenishing fresh media containing 40 U mL⁻¹ IL-2 every 2–3 days. Activated CD4⁺ T cells (1 × 10⁵) were exposed with HIV-1 (NL4-3 strain) at MOI = 0.1 for 3 h. After removal of unbound viruses, the virus-infected cells were cultured overnight. HIV-1-infected CD4⁺ T cells were co-cultured for 3 days with CAR-T cells at the E:T ratio of 1:1, 0.1:1, and 0.01:1. Cells were harvested and stained with antihuman CD8-PE/Cy7 (Biolegend) as described above. Intracellular HIV-1 p24 expression was detected using Intracellular Fixation and Permeabilization Buffer Set (eBioscience) with anti-HIV-1 p24 antibody KC57-RD1 (Beckman Coulter) according to the manufacturer’s instructions. The levels of HIV-1 p24 secreted into the supernatant of the co-culture were determined by the HIV-1 p24 ELISA Kit (RIMCO CORPORATION, Okinawa, Japan).

Stimulation of the CAR-T cells with the cells expressing the cognate antigen

The CAR-T cells (0.5 × 10⁶) at days 18 were stimulated with mitomycin C-treated CHO-GFP or CHO-Env-GFP cells (1 × 10⁶) in 2 mL of complete AIM-V supplemented with 300 U mL⁻¹ IL-2. The cells were cultured for 7 days by replacing half of the culture media with the fresh media containing 300 U mL⁻¹ IL-2 every 2 to 3 days. On days 7, the cells in half of the culture were stimulated again with mitomycin C-treated CHO-GFP or CHO- Env-GFP cells (1 × 10⁶) in 2 mL of complete AIM-V supplemented with 300 U mL⁻¹ IL-2 and cultured for 7 days. Flow cytometric analyses to determine CAR-T (CD3⁺GFP⁺) cell expansion were performed on days 7 and 14 after the initial stimulation.

Microscopic analysis of CAR cluster formation

The CAR-transduced primary CD8⁺ T cells at days 9 were stained with the biotinylated anti-c-myc tag antibody followed by streptavidin-APC (Biolegend) as described above. The CAR-T cell population with high CAR expression was sorted on FACS Aria II and seeded on a flat bottom 96 well cell culture plate. The sorted cells were observed using a BZ-9000 fluorescent microscopy system (KEYENCE, Osaka, Japan) with a 20× nonoil objective lens. The phase contrast and fluorescent images were taken and analyzed using ImageJ (National Institute of Health, MD).

Statistical analysis

Statistical analysis was described previously.11,28 The two-tailed Mann-Whitney U test and unpaired one-way or two-way analysis of variance (ANOVA) with the Bonferroni multiple comparison test were performed using GraphPad Prism (GraphPad Software, San Diego, CA). The P value of <.05 was considered statistically significant.

Results

Arrangement of V_L and V_H domains affects expression levels of CAR in human T cells

A pair of lentiviral vectors expressing the CAR bearing antigen-binding domains with different arrangements of the V_L and V_H domains derived from the anti-HIV-1 Env monoclonal antibody VRC01 were constructed (Fig. 1A and Supplementary Fig. S1). The viral titers for both CAR-expressing lentiviral vectors were very similar (Supplementary Fig. S2). The human T lymphocyte-derived Jurkat E6.1 cell line was transduced with either one of the V_H-V_L and V_L-V_H vectors and analyzed for the CAR expression. Whereas the transduction efficiencies of the V_H-V_L and V_L-V_H vectors determined by the percentage of GFP⁺ cells were similar to each other (Supplementary Fig. S3), expression levels of the V_L-V_H CAR on the surface of successfully transduced cells were significantly higher than those of the V_H-V_L CAR (Fig. 1B). Western blot analysis also revealed that the amount of the V_L-V_H CAR in the whole cell lysate was larger than that of the V_H-V_L CAR (Fig. 1C). Therefore, the arrangement of the V_L and V_H domains in the CAR appears for the determination of the level of cell surface expression of the CAR possibly through the effects on CAR synthesis or stability in the CAR-T cells.

FIG. 1.

Arrangement of V_L and V_H domains affects expression levels of CARs in human T cells. (A) Schematic representation of the CAR expressing lentiviral vector constructs. The CARs contain different orders of the variable fragments (V_H-V_L or V_L-V_H) derived from the monoclonal antibody VRC01 against HIV-1 Env protein fused with the co-stimulatory signal sequence from HVEM and the human CD3ζ chain (CD3z). Transmembrane (TM) sequences are also derived from HVEM. All lentiviral vector constructs express GFP under the control of the internal ribosomal entry site (IRES). Ig in the figure represents immunoglobulin. (B) Gating strategy for CAR expression analysis in the transduced human T-cell line. GFP⁺-transduced cells were gated (left) to analyze CAR expression (middle) with the anti-c-myc tag antibody by flow cytometry. Typical counter plot and CAR histograms of each transduced cell are shown. Similar transduction efficiencies were achieved by lentiviral transduction (see also Supplementary Fig. S1). The number in each histogram is the mean fluorescence intensity (MFI) of the CAR. The levels of CAR expression were normalized by dividing MFI of CARs by that of GFP. Right graph shows the normalized CAR expression level in each CAR-T cell with different V_L-V_H arrangements in the CAR. Each bar shows the mean with SD (n = 5). ** P < .05 (Mann-Whitney U test). (C) CAR (58 kDa, top), endogeneous CD3ζ (18 kDa, middle), and actin (43 kDa, bottom) expression levels in the CAR-transduced Jurkat E6.1 cells harboring different V_L and V_H arrangements in the CAR determined by western blot with anti-CD3ζ and anti-actin antibodies, respectively.

Arrangement of V_L and V_H domains in the CAR affects antigen-dependent activation of the CAR-T cell line

The Jurkat E6.1 cells were transduced with the V_L-V_H or V_H-V_L lentiviral vector, and the successfully transduced CAR-T cells were co-cultured with the CD3^- target cells expressing HIV-1 Env. Then, CD69 expression and IL-2 secretion of the CAR-T cells were determined by flow cytometry and ELISA, respectively. While no significant upregulation of CD69 in the CAR-T cells was observed upon co-cultivation with the control CHO-GFP cells, the cells co-cultured with CHO-Env-GFP cells showed significant levels of CD69 upregulation, indicating the antigen-dependent activation of the CAR-T cells (Fig. 2A and Supplementary Fig. S4). Consistently, IL-2 secretion by the CAR-T cells was increased upon co-cultivation with the antigen-expressing target cells (Fig. 2B). Antigen-dependent activation determined by CD69 upregulation and IL-2 secretion was higher for the V_L-V_H CAR than that of the V_H-V_L CAR (Fig. 2A and B), indicating that the arrangement of V_L and V_H in the antigen-binding domain of the CAR could affect the level of antigen-dependent activation of the CAR-T cells.

FIG. 2.

Arrangement of V_L and V_H domains in the CAR affects antigen-dependent activation of the CAR-T-cell line. (A) The activated (CD69⁺) cell percentages in the transduced (CD3⁺GFP⁺) or un-transduced (CD3^-GFP⁺) T cells were determined by the co-culture assay with target cells expressing GFP (CHO-GFP) or HIV-1-Env and GFP (CHO-Env-GFP) (see also Supplementary Fig. S3). Each bar shows the mean with SDs from the co-culture assay using the control or CAR-transduced T cells with different V_L and V_H arrangements in the CAR (V_H-V_L or V_L-V_H) (n = 3). **P < .05 (one-way ANOVA; Bonfferoni’s post-hoc). (B) IL-2 secretion in the co-culture assay was measured by ELISA. Each bar shows the mean with SDs from the co-culture assay using the control T cells or CAR-transduced T cells with different V_L and V_H arrangements in the CAR (V_H-V_L or V_L-V_H) (n = 3). **P < .05 (two-way ANOVA; Bonfferoni’s post-hoc).

Arrangement of V_L and V_H domains in the CAR also affects car expression levels and effector functions of the primary human CAR-T cells

Next, human primary CD8⁺ T cells were transduced with the V_L-V_H or V_H-V_L lentiviral vector for preparing CAR-T cells. Transduction efficiencies determined by the percentage of GFP⁺ cells were similar to each other (Supplementary Fig. S5). “Similar to the results of the Jurkat E6.1 T cell line (Fig. 1 and Fig. 2), the amount of the V_L-V_H CAR in the whole cell lysate of the transduced primary T cells was larger than that of the V_H-V_L CAR (Fig. 3A), Whereas both the V_L-V_H and V_H-V_L CAR-expressing T cells exhibited cytotoxicity against the HIV-1 Env-expressing target cells, and the cytotoxic activities of the V_L-V_H CAR-T cells were higher than those of the V_H-V_L CAR-T cells (Fig. 3B). Similarly, the levels of secreted cytokines, such as IL-2, interferon (IFN)-γ, and tumor necrosis factor (TNF)-α, were higher for the V_L-V_H CAR-T cells than for the V_H-V_L CAR-T cells (Fig. 3C). Similar results were also obtained when the HIV-1-infected CD4⁺ T cells were used as a target. Namely, the V_L-V_H CAR-T cells eliminated HIV-1-infected CD4⁺ T cells more efficiently than the V_H-V_L CAR-T cells (Supplementary Fig. S6 and Fig. 3D). Consistently, the levels of HIV-1 p24 in the co-culture supernatants were lower for the V_L-V_H CAR-T cells than for the V_H-V_L CAR-T cells (Fig. 3E). These data suggested that the arrangement of V_L and V_H in the antigen-binding domain could affect expression levels of the CAR and the effector functions in human primary CD8⁺ T cells as well.

FIG. 3.

Arrangement of V_L and V_H domains in the CAR also affects CAR expression levels and effector functions of the primary human CAR-T cells. (A) CAR (58 kDa, top), endogeneous CD3ζ (18 kDa, middle), and actin (43 kDa, bottom) expression levels in CAR-T cells with different V_L and V_H arrangements in the CAR (V_H-V_L or V_L-V_H) were determined by western blot with anti-CD3ζ and anti-actin antibodies, respectively. (B) The CHO-Env-GFP cells were co-cultured with the control T cells (open circle) or CAR-T cells with different V_L and V_H arrangements in the CAR (V_H-V_L; gray filled circle or V_L-V_H; black filled circle) at effector-to-target (E:T) ratios of 10:1, 5:1, and 1:1. The lactate dehydrogenase release into the culture supernatant was measured and the cytotoxicity (% lysis) was calculated. The results show the representative of two separate experiments with primary CD8⁺ T cells from two different healthy human donors (n = 3). *P < .05 (two-way ANOVA; Bonfferoni’s post-hoc). (C) Bar graph shows IL-2 (left), IFN-γ (middle), and TNF-α (right) secretion measured by ELISA at an E:T ratio of 10:1 in the co-culture assay. Each bar shows the mean with SDs (n = 3). *P < .05 (two-way ANOVA; Bonfferoni’s post-hoc). (D) HIV-1-infected CD4 T cells were co-cultured for 3 days with the control T cells (open circle) or CAR-T cells with V_L and V_H arrangements in the CAR (V_H-V_L; gray filled circle or V_L-V_H; black filled circle) at E:T ratios of 1:1, 0.1:1, and 0.01:1. Graph shows the mean with SDs of the percentage of HIV-1 p24⁺ cells in CD8^- population. The results show the mean of two separate experiments with primary CD8⁺ T cells from two different healthy human donors (n = 6). **P < .05 (two-way ANOVA; Bonfferoni’s post-hoc). (E) HIV-1 p24 released into the supernatant of the co-culture was measured by ELISA. Bar graph shows the mean with SDs of two separate experiments with primary CD8⁺ T cells from two different healthy human donors (n = 6). **P < .05 (two-way ANOVA; Bonfferoni’s post-hoc).

Arrangement of V_L and V_H domains in the CAR affects exhaustion status of the CAR-T cells

To examine the effects of V_L-V_H arrangements in the antigen-binding domain of the CAR on exhaustion status of the CAR-T cells, we have compared the V_L-V_H and V_H-V_L CAR-T cells for expressions of the inhibitory receptors, such as lymphocyte activation gene 3 (LAG-3) and programed cell death 1 (PD-1). Flow cytometric analysis revealed that the LAG-3⁺/PD-1⁺ cells, which represent terminally exhausted population, were induced in the V_H-V_L CAR-T cells but not in the V_L-V_H CAR-T cells at days 9 (Fig. 4A) and 16 (Fig. 4B) of the culture period. A recent study has suggested that clustering of CAR molecules on the cell membrane might induce strong tonic signaling, which in turn induces CAR-T cell exhaustion.7 To examine whether tonic signaling has been associated with CAR-T cell exhaustion, the levels of phosphorylation in tyrosine 142 in CD3ζ of the CAR were analyzed by flow cytometry. As CD3ζ phosphorylation was not elevated in nontransduced population of the control T cells and the CAR-T cells with different V_L-V_H arrangements (Supplementary Fig. S7), CD3ζ phosphorylation was specifically occurred in the CAR-T cells (Fig. 4C). The level of CD3ζ phosphorylation was higher in the V_H-V_L CAR-T cells than in the V_L-V_H CAR-T cells at days 9 and 16 (Fig. 4C). These data suggested that the strength of tonic signaling was associated with the level of CAR-T cell exhaustion. Since poor proliferation upon stimulation with the cognate antigen is one of the hallmarks of the T cell exhaustion,29,30 we next examined CAR-T cell proliferation upon cognate antigen stimulation. When the CAR-T cells were stimulated with the mitomycin C-treated antigen-expressing cells, the V_L-V_H CAR-T cells proliferated more vigorously than the V_H-V_L CAR-T cells (Fig. 4D). Thus, these data showed that the V_H-V_L CAR-T cells appeared to be exhausted at higher level than the V_L-V_H CAR-T cells, suggesting that arrangement of V_L and V_H in the antigen-binding domain of the CAR affects exhaustion status of the CAR-T cells.

FIG. 4.

Arrangement of V_L and V_H domains in the CAR affects exhaustion status of the CAR-T cells. (A) and (B) Representative counter plots shows cell surface expressions of LAG-3 (y-axis) and PD-1 (x-axis) on the GFP⁺-transduced control T cells or CAR-T cells with different V_L and V_H arrangements in the CAR (V_H-V_L or V_L-V_H) at days 9 (A) and 16 (B). Control shows CD8 T cells transduced only with GFP. The numbers in each counter plot show percentages of each population in the CAR-T cells with different V_L and V_H arrangements in the CAR (V_H-V_L or V_L-V_H). Right bar graphs show the percentages of the CAR-T cell populations that express one or two inhibitory receptors (LAG-3 and PD-1) at days 9 (A) and 16 (B). The mean with SDs of two separate experiments with primary CD8 T cells from two different healthy human donors is shown (n = 6). **P < .05 (two-way ANOVA; Bonfferoni’s post-hoc). (C) Left panels show the typical phosphorylated CD3ζ histogram of the successfully transduced GFP⁺ populations of the control T cells or the CAR-T cells with different V_L-V_H arrangements at days 9 and 16 analyzed by flow cytometry. Control shows CD8 T cells transduced only with GFP. The number in each histogram shows MFI of phosphorylated CD3ζ in the CAR. Right bar graphs show the MFI of phosphorylated CD3ζ in the GFP⁺ populations of control T cells or CAR-T cells with different V_L-V_H arrangements at days 9 (left) and 16 (right). The mean with SDs of two separate experiments with primary CD8 T cells from two different healthy human donors is shown (n = 6). *P < .05 (one-way ANOVA; Bonfferoni’s post-hoc). (D) The control T cells (open circle) or the CAR-T cells with different V_L and V_H arrangements in the CAR (V_H-V_L; gray filled circle or V_L-V_H; black filled circle) were stimulated with mitomycin C-treated CHO-GFP (left) or CHO-Env-GFP cells (right). The CAR-T cell (CD3⁺GFP⁺) expansions were monitored by flow cytometry. Control shows CD8 T cells transduced only with GFP. Line graphs shows fold increase of the percentage of CD3⁺GFP⁺ cells at each time point to pre-stimulation. The mean with SDs of two separate experiments with primary human CD8⁺ T cells from two different healthy human donors is shown (n = 6). *P < .05 (two-way ANOVA; Bonfferoni’s post-hoc).

The V_L-V_H CAR-T cells resist exhaustion despite the high CAR expression

Next, we examined the characteristics of the CAR-T cell with different V_L-V_H arrangements in more detail during the culture period. It was noted that while the V_L-V_H CAR was stably expressed at a high level during the culture period, the V_H-V_L CAR expression at day 9 became bimodal and that the proportion of the cells with the lower V_H-V_L CAR expression level enlarged at day 16 (Fig. 5A and B, upper panels). Flow cytometric analysis revealed that the population of the CAR-T cells with the higher levels of V_H-V_L CAR expression contained much larger percentage of the LAG-3⁺/PD-1⁺ cells compared to the other population with the lower levels of V_H-V_L CAR expression (Fig. 5A and B, lower panels). Immunostaining analysis revealed that the T cells expressing the V_H-V_L CAR at higher levels indeed had clustering of the CAR molecules on the cell membrane (Fig. 5E, middle panels). Interestingly, the V_L-V_H CAR expressed at high levels without causing significant T cell exhaustion was also shown to form CAR clusters on the cell membrane (Fig. 5E, lower panels). There was no significant difference in the number of CAR-T cells with the CAR cluster between the V_H-V_L CAR-T cell and the V_L-V_H CAR-T cell (Fig. 5E, right graph). We next examined the level of CD3ζ phosphorylation in each CAR-T cell population (Fig. 5F). The V_H-V_L CAR-T cell with lower CAR expression did not increase the level of CD3ζ phosphorylation in the CAR, whereas the one with higher CAR expression significantly increased the level of CD3ζ phosphorylation in the CAR during the culture period. Interestingly, the level of CD3ζ phosphorylation in the CAR was slightly increased in the V_L-V_H CAR-T cell which expressed the CAR at high level but was significantly lower than that in the V_H-V_L CAR-T cell which expressed the CAR at high level. Thus, the V_L-V_H arrangements in the context of the CAR construct used in this study may allow the T cells to evade tonic signaling-induced CAR-T cell exhaustion despite the high CAR expression.

FIG. 5.

The V_L-V_H CAR-T cells resist exhaustion despite the high CAR expression. (A) and (B) Representative histograms show cell surface CAR expression in GFP⁺-transduced primary human CD8 T cells with different V_L and V_H arrangements in the CAR at days 9 (A, upper panels) and 16 (B, upper panels). Inhibitory receptor expressions on the CAR-T cell with low CAR expression (Low CAR) and with high CAR expression (High CAR) were analyzed by flow cytometry. Representative counter dot plots show cell surface expressions of LAG-3 (y-axis) and PD-1 (x-axis) on the GFP⁺-transduced control T cells of CAR-T cells with different V_L and V_H arrangements in the CAR at days 9 (A, lower panels) and 16 (B, lower panels). Control shows CD8 T cells transduced only with GFP. The numbers in each plot show percentages of each population. (C) and (D) Bar graph shows the percentages of cell populations that express one or two inhibitory receptors (LAG-3 and PD-1) of the CAR-T cell with low or high CAR expression at days 9 (C) and 16 (D). The mean with SDs of two separate experiments with primary human CD8⁺ T cells from two different healthy human donors is shown (n = 6). **P < .05 (two-way ANOVA; Bonfferoni’s post-hoc). (E) Imaging analysis of cell surface CAR distributions on sorted high CAR population of the CAR-T cells with different V_L and V_H arrangements in the CAR. Control is the sorted T cells, which only expresses GFP. The representative images of GFP (left) and CAR (right) are shown. Scale bars = 25 µm. Right bar graph shows the number of GFP⁺-transduced cells with CAR puncta. Data of 100 randomly chosen GFP⁺-transduced cells in each image are shown as mean with SDs (n = 3). (F) Upper panels show a typical phosphorylated CD3ζ histogram of each CAR-T cell populations at days 9 and 16 analyzed by flow cytometry. The lower graph shows the MFI of phosphorylated CD3ζ in the V_H-V_L CAR-T cell with low CAR expression (open circle), the V_H-V_L CAR-T cell with high CAR expression (filled black circle), and the V_L-V_H CAR-T cell with high CAR expression (filled gray circle) at days 9 and 16. The mean with SDs of two separate experiments with primary CD8 T cells from two different healthy human donors is shown (n = 6). **P < .05 (two-way ANOVA; Bonfferoni’s post-hoc).

Discussion

The CAR-T cell-mediated immunotherapy is expected to become useful therapeutic intervention against diseases with impaired cellular immunity such as malignant diseases. In order to improve the technology, it is important to design an optimal CAR molecule with maximal efficacy. For this purpose, newer generations of CARs are being developed by modifying its components such as CSSD. For example, a recent study from our group demonstrated that the CSSD derived from HVEM could be more useful than previously used CD28- or 4-1BB-derived CSSD.11 Although there is a previous study which compared different arrangements of V_L and V_H in the context of CAR with 4-1BB-derived CSSD,21 –24 it has been unknown how the arrangement of V_L and V_H affects CAR-T cell functions and exhaustion status in particular in the context of the CAR bearing the HVEM-derived CSSD.

In order to investigate this issue, we constructed the lentiviral vectors for expressing the CAR bearing V_L-V_H or V_H-V_L derived from an anti-HIV-1 Env antibody and the HVEM-derived CSSD as a model system. Analysis of the human T cell line and primary T cells transduced with these vectors demonstrated that the levels of cell surface expression were higher for the V_L-V_H CAR than the V_H-V_L CAR apparently reflecting the different abundance in the cells. As to scFv molecules, it was previously reported that the arrangement of V_H and V_L could affect its expression levels or stability.14 –18 It is possible that the level of synthesis or stability of the V_L-V_H CAR is higher than the V_H-V_L CAR in our system, although the detailed mechanism is unknown yet. Another issue is that the CAR with different V_L and V_H arrangement might have different binding affinities against the antigen. To clarify this, it may be necessary to perform the biophysical analysis for both V_L-V_H and V_H-V_L scFvs in the future. This might be able to reveal precise mechanisms of the differences in the potencies of the CAR-T cells with different V_L and V_H arrangements.

In our system, the V_L-V_H CAR-T cells were shown to be more potent than the V_H-V_L CAR-T cells in terms of their functions, such as antigen-dependent activation and cytotoxicity against the target cells. In our previous study comparing different CSSDs, the levels of CAR expression and CAR-T cell functions were correlated.11 Therefore, it is possible that the difference in the CAR-T cell functions is simply caused by the quantitative difference in the expression levels of CAR molecules. However, it is unlikely because the exhaustion status of the CAR-T cells was also affected by the arrangement of V_L and V_H in the CAR. It is thought that tonic signaling mediated by the CAR molecule is responsible for induction of CAR-T cell exhaustion.2 A previous study has shown that the CAR-T cells expressing high levels of CAR exhibited exhausted phenotype and that this phenotype had been rescued by experimentally reducing CAR expression.31 It has also been suggested that clustering of the CAR molecules on the cell membrane could induce tonic signaling, which leads to exhaustion of CAR-T cells.7,31 In consistent with the previous reports, our data indicated that the levels of V_H-V_L CAR expression were correlated with the strength of the CAR-mediated tonic signaling and the proportion of exhausted cells. Interestingly, the V_L-V_H CAR expressed at higher levels than the V_H-V_L CAR did not significantly induce CAR-mediated tonic signaling and lead to CAR-T cell exhaustion although it formed clusters on the cell membrane. In addition, the high level of V_L-V_H CAR expression was not associated with the increase in the exhausted cells (Fig. 5A-D). Thus, the abundance of the V_L-V_H CAR molecules by itself may not be responsible for induction of the exhausted phenotype. It has been reported that framework regions within certain CAR scFv are sufficient to induce tonic signaling.7 Thus, it is possible that the V_L-V_H arrangement in the context of the CAR construct used in this study might be hard to convey the CAR-mediated tonic signaling which induces CAR-T cell exhaustion. Homology modeling of the V_L-V_H and V_H-V_L CARs suggested the distinct structural differences in the antigen binding domain of the CAR molecules (Supplementary Fig. S8). This difference might also determine the level of tonic signaling.

Although the V_L-V_H arrangement appeared to be more useful than the opposite arrangement in our system, the limitation is that the results of this study could not directly be applied for other CAR constructs.21 –24,32 However, the evidence showed in this study might provide useful information for designing potent CARs and for developing CAR-T cells, which could evade exhaustion. It would be desirable to compare different CAR constructs with both arrangements of V_L and V_H domains combined to various CSSDs at preclinical stage for generating effective CAR-T cells against various targets.

Conclusion

In summary, this study demonstrated that the arrangement of V_L and V_H in the antigen-binding domain could affect expression levels of the VRC01-based CAR with HVEM-derived CSSD as well as the functions and exhaustion status of the CAR-T cells by modulating CAR-mediated tonic signaling. This could provide useful information for designing optimal CARs and for developing effective CAR-T cell-mediated immunotherapy.

Authors’ Contributions: J.N. conceptualized and designed the study. J.N. and E.T. assisted with the acquisition of data. J.N., E.T., and M.M. carried out analysis and interpretation of data. J.N., E.T., M.M., and L.S. assisted in drafting and revising the manuscript. All the authors have approved the final article.

Author Disclosure Statement: No competing financial interests exist.

Funding Information: Installation of FACS Aria (BD Biosciences) and BSL3 facility to Dokkyo Medical University was initially supported by Promotion and Mutual Aid Corporation for Private Schools of Japan. This work was supported in part by Dokkyo Medical University, Investigator-Initiated Research Grant (Nos. 2014-03 and 2017-03), JSPS KAKENHI Grant-in-Aid for Young Scientists (B) (No. JP16K19617), and Grant-in-Aid for Scientific Research (C) (No. JP18K07272) to J.N.

Footnotes

Acknowledgments

We thank Yuki Shinozaki, Satomi Yoshida, and Kayoko Kobayashi for technical and secretarial assistance; Yasuko Nonaka for cell sorting; and Takashi Namatame for DNA sequencing. We also thank members of Department of Microbiology, Dokkyo Medical University, for critical discussion of the paper and their input and assistance.

References

Maude

, Teachey

, Porter

, et al. CD19-targeted chimeric antigen receptor T-cell therapy for acute lymphoblastic leukemia. Blood. 2015; 125:4017–4023. DOI: 10.1182/blood-2014-12-580068.

Ajina

, Maher

Strategies to address chimeric antigen receptor tonic signaling. Mol Cancer Ther. 2018; 17:1795–1815. DOI: 10.1158/1535-7163.MCT-17-1097.

Kershaw

, Westwood

, Parker

, et al. A phase I study on adoptive immunotherapy using gene-modified T cells for ovarian cancer. Clin Cancer Res. 2006; 12:6106–6115. DOI: 10.1158/1078-0432.CCR-06-1183.

Dotti

, Gottschalk

, Savoldo

, et al. Design and development of therapies using chimeric antigen receptor-expressing T cells. Immunol Rev. 2014; 257:107–126. DOI: 10.1111/imr.12131.

Maus

, Grupp

, Porter

, et al. Antibody-modified T cells: CARs take the front seat for hematologic malignancies. Blood. 2014; 123:2625–2635. DOI: 10.1182/blood-2013-11-492231.

Kawalekar

, O'Connor

, Fraietta

, et al. Distinct signaling of coreceptors regulates specific metabolism pathways and impacts memory development in CAR T cells. Immunity. 2016; 44:380–390. DOI: 10.1016/j.immuni.2016.01.021.

Long

, Haso

, Shern

, et al. 4-1BB costimulation ameliorates T cell exhaustion induced by tonic signaling of chimeric antigen receptors. Nat Med. 2015; 21:581–590. DOI: 10.1038/nm.3838.

Flynn

, Hutchinson

, Murphy

, et al. CD8 T cell memory to a viral pathogen requires trans cosignaling between HVEM and BTLA. PLoS One. 2013; 8:e77991. DOI: 10.1371/journal.pone.0077991.

Steinberg

, Huang

, Wang-Zhu

, et al. BTLA interaction with HVEM expressed on CD8(+) T cells promotes survival and memory generation in response to a bacterial infection. PLoS One. 2013; 8:e77992. DOI: 10.1371/journal.pone.0077992.

10.

Park

, Anand

, Zhao

, et al. Expression of anti-HVEM single-chain antibody on tumor cells induces tumor-specific immunity with long-term memory. Cancer Immunol Immunother. 2012; 61:203–214. DOI: 10.1007/s00262-011-1101-8.

11.

Nunoya

, Masuda

, Ye

, et al. Chimeric antigen receptor T cell bearing herpes virus entry mediator co-stimulatory signal domain exhibits high functional potency. Mol Ther Oncolytics. 2019; 14:27–37. DOI: 10.1016/j.omto.2019.03.002

12.

Huston

, Levinson

, Mudgett-Hunter

, et al. Protein engineering of antibody binding sites: Recovery of specific activity in an anti-digoxin single-chain Fv analogue produced in Escherichia coli . Proc Natl Acad Sci U S A. 1988; 85:5879–5883. DOI: 10.1073/pnas.85.16.5879.

13.

Bird

, Hardman

, Jacobson

, et al. Single-chain antigen-binding proteins. Science. 1988; 242:423–426. DOI: 10.1126/science.3140379.

14.

Luo

, Mah

, Krantz

, et al. V_L-linker-V_H orientation-dependent expression of single chain Fv-containing an engineered disulfide-stabilized bond in the framework regions. J Biochem. 1995; 118:825–831. DOI: 10.1093/oxfordjournals.jbchem.a124986.

15.

Kim

, Neelamegam

, Heo

, et al. Improving the productivity of single-chain Fv antibody against c-Met by rearranging the order of its variable domains. J Microbiol Biotechnol. 2008; 18:1186–1190. [PMC][18600066]

16.

Weatherill

, Cain

, Heywood

, et al. Towards a universal disulphide stabilised single chain Fv format: Importance of interchain disulphide bond location and vL-vH orientation. Protein Eng Des Sel. 2012; 25:321–329. DOI: 10.1093/protein/gzs021.

17.

Zhang

, Geddie

, Kohli

, et al. Comprehensive optimization of a single-chain variable domain antibody fragment as a targeting ligand for a cytotoxic nanoparticle. MAbs. 2015; 7:42–52. DOI: 10.4161/19420862.2014.985933.

18.

Liu

, Ye

, Chen

, et al. Expression of V(H)-linker-V(L) orientation-dependent single-chain Fv antibody fragment derived from hybridoma 2E6 against aflatoxin B1 in Escherichia coli . J Ind Microbiol Biotechnol. 2015; 42:255–262. DOI: 10.1007/s10295-014-1570-9.

19.

Bühler

, Wetterauer

, Gierschner

, et al. Influence of structural variations on biological activity of anti-PSMA scFv and immunotoxins targeting prostate cancer. Anticancer Res. 2010; 30:3373–3379. [PMC][20944111]

20.

Cheng

, Li

, Xi

, et al. A V_L-linker-V_H orientation dependent single chain variable antibody fragment against rabies virus G protein with enhanced neutralizing potency in vivo. Protein Pept Lett. 2016; 23:24–32. DOI: 10.2174/0929866522666151026122752.

21.

Cheng

, Ahmed

, Xu

, et al. Structural design of disialoganglioside GD2 and CD3-bispecific antibodies to redirect T cells for tumor therapy. Int J Cancer. 2015; 136:476–486. DOI: 10.1002/ijc.29007.

22.

Laborda

, Mazagova

, Shao

, et al. Development of a chimeric antigen receptor targeting C-type lectin-like molecule-1 for human acute myeloid leukemia. Int J Mol Sci. 2017; 18:2259.

23.

Burns

, Zhao

, Frankel

, et al. A high molecular weight melanoma-associated antigen-specific chimeric antigen receptor redirects lymphocytes to target human melanomas. Cancer Res. 2010; 70:3027–3033. DOI: 10.1158/0008-5472.CAN-09-2824.

24.

Richman

, Nunez-Cruz

, Moghimi

, et al. High-affinity GD2-specific CAR T cells induce fatal encephalitis in a preclinical neuroblastoma model. Cancer Immunol Res. 2018; 6:36–46. DOI: 10.1158/2326-6066.CIR-17-0211.

25.

, Yang

, Li

, et al. Rational design of envelope identifies broadly neutralizing human monoclonal antibodies to HIV-1. Science. 2010; 329:856–861. DOI: 10.1126/science.1187659

26.

Adachi

, Gendelman

, Koenig

, et al. Production of acquired immunodeficiency syndrome-associated retrovirus in human and nonhuman cells transfected with an infectious molecular clone. J Virol. 1986; 59:284–291. DOI: 10.1128/jvi.59.2.284-291.1986

27.

Vodicka

, Goh

, Wu

, et al. Indicator cell lines for detection of primary strains of human and simian immunodeficiency viruses. Virology. 1997; 233:193–198. DOI: 10.1006/viro.1997.8606.

28.

Nunoya

, Washburn

, Kovalev

, et al. Regulatory T cells prevent liver fibrosis during HIV type 1 infection in a humanized mouse model. J Infect Dis. 2014; 209:1039–1044. DOI: 10.1093/infdis/jit548.

29.

Virgin

, Wherry

, Ahmed

Redefining chronic viral infection. Cell. 2009; 138:30–50. DOI: 10.1016/j.cell.2009.06.036.

30.

Wherry

EJ.

T cell exhaustion. Nat Immunol. 2011; 12:492–499. DOI: 10.1038/ni.2035.

31.

Gomes-Silva

, Mukherjee

, Srinivasan

, et al. Tonic 4-1BB costimulation in chimeric antigen receptors impedes T cell survival and is vector-dependent. Cell Rep. 2017; 21:17–26. DOI: 10.1016/j.celrep.2017.09.015.

32.

Smith

, Harrington

, Staehr

, et al. GPRC5D is a target for the immunotherapy of multiple myeloma with rationally designed CAR T cells. Sci Transl Med. 2019; 11:eaau7746.

Arrangement of V L and V H Domains in VRC01-Based Chimeric Antigen Receptor (CAR) Affects Functions and Exhaustion Status of CAR-T Cells

Abstract

Abstract

Background:

Objectives:

Method:

Results:

Conclusions:

Introduction

Materials and Methods

Ethics approval and consent to participate

Cell culture

Vector constructions

Preparation of the vector virus and HIV-1

Transduction with lentiviral vectors

Flow cytometry

Western blot

Co-culture assay for effector functions of the CAR-T cells

Co-culture for anti-HIV-1 activity of the CAR-T cells

Stimulation of the CAR-T cells with the cells expressing the cognate antigen

Microscopic analysis of CAR cluster formation

Statistical analysis

Results

Arrangement of VL and VH domains affects expression levels of CAR in human T cells

Arrangement of VL and VH domains in the CAR affects antigen-dependent activation of the CAR-T cell line

Arrangement of VL and VH domains in the CAR also affects car expression levels and effector functions of the primary human CAR-T cells

Arrangement of VL and VH domains in the CAR affects exhaustion status of the CAR-T cells

The VL-VH CAR-T cells resist exhaustion despite the high CAR expression

Discussion

Conclusion

Footnotes

Acknowledgments

References

Arrangement of V_L and V_H domains affects expression levels of CAR in human T cells

Arrangement of V_L and V_H domains in the CAR affects antigen-dependent activation of the CAR-T cell line

Arrangement of V_L and V_H domains in the CAR also affects car expression levels and effector functions of the primary human CAR-T cells

Arrangement of V_L and V_H domains in the CAR affects exhaustion status of the CAR-T cells

The V_L-V_H CAR-T cells resist exhaustion despite the high CAR expression