Chimeric antigen receptor (CAR) T-cell therapy has proven successful in the treatment of hematological malignancies, notably acute lymphoblastic leukemia and B-cell lymphoma. However, the efficacy of CAR T cells against solid tumors is poor, likely due to tumor-associated immunosuppression. Here, we demonstrated that antagonizing the “inhibitor of apoptosis proteins” with the clinical smac-mimetic, birinapant, significantly enhanced the antitumor activity of CAR T cells in a tumor necrosis factor (TNF)-dependent manner. Enhanced tumor cell death occurred independently of the perforin-mediated granule exocytosis pathway, underscoring the cytotoxic potential of CAR T-cell–derived TNF. Combining CAR T-cell therapy with birinapant significantly reduced established tumor growth in vivo, where either therapy alone was relatively ineffective. Using patient biopsy-derived tumoroids, we demonstrated the synergistic potential of combining CAR T-cell therapy with smac-mimetics. Taken together, we identified CAR T-cell–derived TNF as a potent antitumor effector, which can be further harnessed by smac-mimetics.

Adoptive cellular therapy using chimeric antigen receptor (CAR) T cells is showing great promise for the treatment of cancer (1, 2). However, although CAR T cells have been successful in treating hematological malignancies (2), results in the solid tumor setting have been less positive (3). One reason for the lack of efficacy of CAR T cells in solid tumors is the immunosuppressive microenvironment encountered (1). This includes the expression of ligands on tumor cells that bind to checkpoint receptors on the T cells and inhibit their function. For this reason, combining CAR T cells with additional therapies that target checkpoint receptor–ligand interactions may improve outcomes in solid cancers. PD-1 blockade, for example, has been shown to significantly enhance the efficacy of both CAR T-cell (4) and adoptive cell therapy (5). The efficacy of CAR T-cell therapy may also be enhanced by agents that sensitize tumor cells to T-cell effector mechanisms, including cytokines. Smac-mimetics (SM) are a new class of anticancer agents that can sensitize tumor cells to tumor necrosis factor (TNF)–mediated cell death (6). Because TNF is key to the antitumor effector function of cytotoxic T cells (7), this class of small-molecule drugs may also enhance CAR T-cell efficacy.

During apoptotic cell death, the mitochondrial protein Smac/Diablo binds to and antagonizes the major “inhibitor of apoptosis proteins” (IAPs). Because IAPs are often upregulated in human cancer and facilitate resistance to tumor cell death following therapy, small-molecule mimetics of Smac/Diablo (SMs) were developed (6, 8). Antagonism of IAPs also results in formation of the death-inducing signaling complex upon TNF stimulation, rendering the tumor cell sensitive to TNF-induced cell death via the extrinsic pathway (9). Thus, therapies that boost immunity, particularly in the context of elevated TNF secretion, are likely to synergize with SMs. Indeed, IAP antagonism can promote antitumor immunity in multiple myeloma (10) and glioblastoma (11) when combined with checkpoint inhibition, in part due to TNF secreted by immune cells such as CD8+ T cells.

Based on the positive results from preclinical studies (12, 13), several SMs have completed phase I/II trials in patients with advanced solid and hematological cancers, with the SM birinapant showing the most promise (8). Given the critical role of TNF in antitumor effector functions of cytotoxic T cells (7) and the ability of SMs to sensitize tumor cells to TNF-dependent apoptosis, we investigated the potential of this class of drugs to enhance CAR T-cell efficacy. We demonstrated that IAP antagonism by birinapant sensitized tumor cells to CAR T-cell–derived TNF, significantly enhancing the antitumor activity of CAR T-cell therapy both in vitro and in vivo. The results pave the way for a combination therapy that may improve the efficacy of adoptive cell therapy in solid malignancies.

Mice

All animal studies were performed in accordance with the NHMRC Australian Code for the Care and Use of Animals for Scientific Purposes 8th edition (2013) and with approval from the Peter MacCallum Cancer Centre Animal Experimentation Ethics Committee. C57BL/6 human HER2 transgenic mice (14) and C57BL/6 perforin knockout mice (15) were bred in-house. C57BL/6 and C57BL/6 TNF knockout mice were obtained from the Walter and Eliza Hall Institute (Parkville, Victoria). All mice were housed in the Peter MacCallum Cancer Centre Animal Core Facility under specific pathogen-free conditions.

Human samples

Human PBMCs were isolated from healthy donor buffy coats supplied by Australian Red Cross Blood Service. Biopsy samples from 2 patients with HER2+ metastatic colorectal cancer were included in this study. All human biopsy studies were conducted in accordance with the Declaration of Helsinki under a protocol approved by the Peter MacCallum Cancer Centre Human Ethics Committee, which included informed written consent from all patients.

Antibodies, cytokines, and drugs

Neutralizing antibodies: mouse anti-TNF (BioLegend; clone MP6-XT22), human anti-TNF (BioLegend; clone MAb11). Enbrel (etanercept) was obtained from the Walter and Eliza Hall Institute of Medical Research (Parkville, Victoria). Antibodies for immunofluorescence: anti-TNF (BioLegend; clone MP6-XT22), IFNγ (eBiosciences; clone XMG1.2), Tubulin (Rockland Immunochemicals; clone 600-401-880). Secondary antibodies conjugated to Alexa fluorophores and ProLong antifade with DAPI were purchased from Molecular Probes. For cell stimulation, anti-CD3 (clone 145-2C11) and anti-CD28 (clone 37.51) antibodies were purchased from BD Biosciences and anti-myc tag (clone 2276) from Cell Signaling Technology. IL2 and IL7 were obtained from the NIH and PeproTech, respectively. Anti-human HER2 was purchased from BioLegend (clone 24D2). Birinapant was supplied by Medivir dissolved in 12% Captisol solution in water (vehicle).

Cell lines

The cell lines MC38 and E0771 were retrovirally transduced to express the human HER2 antigen under the control of the mouse stem cell virus LTR promoter as previously described (14). Mouse cell lines were cultured in DMEM medium and the human cell lines HeLa and AU565 in RPMI medium (Gibco), both supplemented with 10% FCS (Thermo Scientific), penicillin/streptomycin (Gibco), and incubated at 37°C in 10% CO2. Tumor lines were verified to be Mycoplasma-negative by the Victorian Infectious Diseases Reference Laboratory by PCR analysis.

Generation of CAR T cells

To generate activated mouse T cells, splenocytes from C57BL/6 mice were filtered through an Acrodisc 70-nm filter with 10 mL red cell lysis buffer (150 mmol/L NH4Cl, 10 mmol/L KHCO3, and 0.1 mmol/L EDTA in 500 mL distilled water pH 7.4) and rinsed with 10 mL RPMI medium (Gibco; Invitrogen). Human PBMCs were isolated from buffy coats using a Ficoll gradient centrifugation. The cell suspension was pelleted at 1400 RPM for 4 minutes, washed twice in 20 mL RPMI, and resuspended in 10 mL enriched T-cell media [RPMI supplemented with 10% FCS (Thermo Scientific), penicillin/streptomycin (Gibco), L-glutamine, nonessential amino acids, 10 mmol/L 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) sodium pyruvate, and (Calbiochem)]. T cells were activated with anti-CD3 (0.5 μg/mL) and anti-CD28 (0.5 μg/mL) in the presence of IL2 (100 IU/mL) and IL7 (2 ng/mL) at a density of 5 × 106 cells/mL.

The generation of murine and human CAR T cells was done as previously described (14, 16). Briefly, retrovirus encoding a CAR composed of an extracellular scFv–anti-human HER2 fused to the transmembrane domains of CD28 and CD3ζ was transduced into activated T cells from either the spleen of mice or from human PBMCs in the presence of IL2 (100 IU/mL) and IL7 (2 ng/mL; PeproTech). The viral packaging GP.E86 cell line containing empty (control) LXSN or LXSN-anti-HER2 CAR retroviral vector was generated as described previously (17). The CAR construct used in this study was a second-generation CAR composed of an extracellular single-chain variable fragment (scFv) specific for human-HER2, a CD8 hinge region, transmembrane and intracellular CD28 and CD3ζ domains, and a c-Myc tag domain allowing for CAR T-cell FACS sorting (18). Primary mouse T cells were collected from splenocytes of C57BL/6 human-HER2 transgenic or wildtype C57BL/6 mice. T cells were retrovirally transduced as described previously (19). After transduction, T cells were maintained in supplemented RPMI media with IL2 (100 U/mL) and IL7 (2 ng/mL). Tnf−/− or perforin−/− CAR T cells were generated as above from C57BL/6 Tnf−/− or perforin−/− mice.

In vitro assays

CAR T-cell activity was measured using a standard chromium release assay as previously described (15). The percentage-specific killing was determined using the formula: (Sample 51Cr release – Spontaneous 51Cr release)/(Total 51Cr release – Spontaneous 51Cr release) × 100, and represented as a Michaelis–Menten kinetic trend. All assays were performed using triplicate wells. The rate of killing in the control is represented as 100% killing, and treatment groups were compared with this value. All killing assays were done over 18 hours unless otherwise specified. To generate relative killing bar graphs, relative killing at the E:T ratio that results in 50% maximal killing of the least cytotoxic condition was compared, using Michaelis–Menten trends, as done previously (15, 20). Cytokines were detected using a mouse inflammation CBA kit (BD Biosciences, 552364) as per manufacturer's instructions and analyzed on a FACSVerse (BD Biosciences). All assays were analyzed using triplicate determinations.

Flow cytometry

Cells assessed via flow cytometry were plated out in V-bottom plates in 100 μL media before being were pelleted by centrifugation at 1400 RPM for 4 minutes and resuspended in 100 μL FACS buffer (PBS plus 2% FCS). Cells were then stained with an anti-human HER2 antibody (clone 24D2; BioLegend) for 30 minutes on ice before being washed twice in FACS buffer, resuspended in 200 μL FACS buffer and transferred to bullet tubes for analysis. Analysis of samples was performed on a Fortessa X20 (BD Biosciences), and data were analyzed with FlowJo (Tree Star).

3′RNA sequencing (RNA-seq)

MC38 cells expressing human HER2 antigen and AU565 cells were left untreated or cocultured with CAR T cells at a 2:1 ratio at 37°C for 6 hours. Cells were washed thoroughly with PBS to remove contaminating CAR T cells, and cells were lysed. Total RNA was isolated as per the kit manufacturer's instructions (Nucleospin RNA extraction kit, Macherey-Nagel).

Single-end 75-bp RNA-seq was performed in-house at the Peter MacCallum Cancer Centre Molecular Genomics Core on the NextSEquation 500 (Illumina). Demultiplexing of the reads was performed using CASAVAv1.8.2, and low-quality reads Q < 30 were removed. Cutadapt (v1.9) was used to trim polyA-derived sequences and biased 3′ reads resulting from random hexamer priming. HISAT2 was used to map the resulting reads to the human reference genome. Read counting was performed using featureCounts; part of the subread package. Voom-LIMMA workflow was used to normalize data for differential gene expression. Gene set enrichment analysis was performed using GSEA2 (v3.0) for identification of enriched signatures obtained from the MSigDB Hallmarks data sets (7). Sequencing data have been deposited into the Gene-Expression Omnibus (www.ncbi.nlm.nih.gov/geo/) under accession number GSE124140.

Fixed and time-lapse microscopy

Microscopy of T-cell–tumor cell conjugates was done as previously described (15). The slides were imaged using a FV1000 confocal microscope (Olympus). CAR T cells selected for quantitation had a single contact site with one tumor cell, indicating a single synapse event. The percentage of T cells that were positive for the indicated cytokine in a synapse or not was quantitated manually by confocal microscopy. A minimum of 100 cells were counted in each condition.

For time-lapse imaging, CAR T cells were sorted for double-positive CD8 and CAR (anti-myc) cells and labeled with CellTrace Violet (CTV, Molecular Probes) as follows: 1 × 107 T cells were stained with 5 μmol/L CTV in PBS for 20 minutes at 37°C and washed for 5 minutes with media. Cells were then pelleted and resuspended in prewarmed media for 10 minutes. Labeled T cells were then added to adherent targets in media containing propidium iodide (PI) and 1 μmol/L birinapant or vehicle control (Captisol). Chamber slides were mounted on a heated stage within a temperature-controlled chamber maintained at 37°C, and constant CO2 concentration of 5% was infused using a gas incubation system with active gas mixer (“The Brick”; Ibidi). Optical sections were acquired through sequential scans or Brightfield/DIC on a TCS SP5 confocal microscope (Leica Microsystems) using a ×40 (NA 0.85) air objective and Leica LAS AF software. Image analysis was performed using Meta-Morph Imaging Series 7 software (Universal Imaging).

Mouse experiments

C57BL/6 human HER2 transgenic mice were injected subcutaneously with 2.5 × 105 MC38-HER2 cells. At day 7 after tumor injection, mice were preconditioned with total-body irradiation (4 Gy) prior to the administration of 1 × 107 CAR T cells on days 7 and 8. Mice were also treated with 50,000 IU IL2 on days 0 to 4 after T-cell transfer. Mice were treated with either vehicle (Captisol) or birinapant (12 mg/kg per mouse) on days 0, 4, and 8 after T-cell transfer. Tumor growth was monitored approximately every second day using a caliper square to determine the product of 2 perpendicular tumor diameters. The mice were culled when the tumor size reached the ethical limit (150 mm2) and recorded for the survival plot.

Patient-derived tumoroids

Tumoroids were grown from tumor cells derived from colorectal patient biopsies. Tumor cells were plated on μ-Plate 96 Well (ibidi) by adding 20 μL of Matrigel/cell suspension per well, and placed to set in 37°C incubator for 30–60 minutes. Following incubation, 350 μL of OB media supplemented with 500 nmol/L A8301 (Torcis, Bioscience), 2× B27 (Gibco, Thermo Scientific), epidermal growth factor (50 ng/mL; Sigma-Aldrich), gastrin (1 μg/mL; Sigma-Aldrich), 1 mmol/L N-acetyl Cyst (Sigma-Aldrich), 5 μmol/L SB202190 (Sigma-Aldrich), 10 μmol/L SB431542 (Sigma-Aldrich), 10 μmol/L Y27632 (Sigma-Aldrich) was added to each well, and tumoroids were grown in hypoxic conditions (37°C, 5% O2, 5% CO2) for 5 to 7 days. Tumoroids were grown to >50 μm and checked for viability using a light microscope on the day of assay, before proceeding. Tumoroids were deemed suitable for coculture after 10 to 14 days after initial biopsy.

A total volume of 350 μL supplemented OB media was applied to the Matrigel-embedded tumoroids. PI (Sigma-Aldrich) was added to a final concentration of 2 μg/mL. CAR T cells/well (2.5 × 105) were then resuspended in this media. Wells were set up with the following conditions: tumoroids + vehicle control, tumoroids + CAR T cells/TNF (10 ng/mL), tumoroids + birinapant (1 μmol/L), tumoroids + CAR T cells/TNF + birinapant, CAR T cells + birinapant + TNF-neutralizing antibody (80 μg/mL). Tumoroids were incubated at 37°C, 5% CO2 for 24 hours, and then end-point images were acquired. U-Plate 96-well plates were mounted on a heated stage in a temperature-controlled chamber, maintained at 37°C, 5% CO2 (“The Brick”; ibidi). Using the cellSens software (Olympus), imaging of the plate was obtained on an UPLSAPO 4X (NA 0.16) air objective using a Hammamatsu ORCA-Flash 4.0 camera. Z-stack images were acquired by taking 25 sequential images (Z spacing 50 μm) of PI (emission 656 nm, exposure 368.644 ms) through the Matrigel, and Z stack images converted to one image in EFI format. Image quantitation was conducted using Meta-Morph Imaging Series 7 software. An identical region of interest was identified within the Matrigel, and integrated morphometry analysis was used to filter out individual T cells based on size. A threshold was applied to delineate PI-positive pixels, and integrated intensity was measured. Identical conditions were applied to compared imges.

Statistical analyses

Statistical significance was determined using an unpaired Student t test, ANOVA, or log-rank (Mantel–Cox) test for mouse survival data using GraphPad Prism 7 software. Differences were considered significant when P < 0.05.

CAR T cells produce cytokines upon tumor cell recognition

TNF is a key inflammatory cytokine produced by cytotoxic T cells upon target recognition that is critical for their antitumor effector function (7). Here, we generated mouse CAR T cells specific for the human HER2 antigen (Fig. 1A; ref. 21) and found that upon recognition of HER2-expressing MC38 colon carcinoma cells (MC38-HER2), TNF and interferon gamma (IFNγ) were key cytokines produced (Fig. 1B). Human HER2-directed CAR T cells also secreted TNF and IFNγ when exposed to HeLa cells that overexpressed HER2 (Supplementary Fig. S1A). Confocal microscopy of mouse CAR T cells cocultured with MC38-HER2 or parental MC38 cells (Fig. 1C, left) revealed an increase in immunological synapse formation between CAR T cells and MC38-HER2 tumor cells and a significantly increased number of synapses positive for TNF and IFNγ (Fig. 1C, right).

Figure 1.

CAR T cells produce inflammatory cytokines upon tumor cell recognition. A, Schematic representation of the CAR T-cell. B, MC38-HER2 cells were exposed to CAR T cells at the indicated effector-to-target (E:T) ratio. After 4 hours, cytokines in supernatants were measured by cytometric bead array. C, Left, MC38-HER2 or parental cells were seeded in chamber slides, then overlaid with CAR T cells. After 2 hours, cells were fixed and stained as indicated, then visualized by confocal microscopy. Scale bar, 10 μm. Right, the percentage of CAR T cells that formed an immunologic synapse was quantitated by confocal microscopy (>100 cells). The percentage of CAR T cells positive for the indicated cytokines was quantitated by confocal microscopy (>100 cells). MC38-HER2 cells were left untreated or incubated with CAR T cells (in triplicate) at low effector-to-target ratio for 6 hours followed by (D) 3′ RNA-seq of viable cells. Left, heat map showing significantly regulated genes following CAR T-cell treatment (P < 0.05). GSEA enrichment score plots from RNA-seq data. Right, log2 counts/million cells from RNA-seq data in D. Error bars, mean ± SEM of triplicate determinations. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001 by an unpaired Student t test.

Figure 1.

CAR T cells produce inflammatory cytokines upon tumor cell recognition. A, Schematic representation of the CAR T-cell. B, MC38-HER2 cells were exposed to CAR T cells at the indicated effector-to-target (E:T) ratio. After 4 hours, cytokines in supernatants were measured by cytometric bead array. C, Left, MC38-HER2 or parental cells were seeded in chamber slides, then overlaid with CAR T cells. After 2 hours, cells were fixed and stained as indicated, then visualized by confocal microscopy. Scale bar, 10 μm. Right, the percentage of CAR T cells that formed an immunologic synapse was quantitated by confocal microscopy (>100 cells). The percentage of CAR T cells positive for the indicated cytokines was quantitated by confocal microscopy (>100 cells). MC38-HER2 cells were left untreated or incubated with CAR T cells (in triplicate) at low effector-to-target ratio for 6 hours followed by (D) 3′ RNA-seq of viable cells. Left, heat map showing significantly regulated genes following CAR T-cell treatment (P < 0.05). GSEA enrichment score plots from RNA-seq data. Right, log2 counts/million cells from RNA-seq data in D. Error bars, mean ± SEM of triplicate determinations. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001 by an unpaired Student t test.

Close modal

We also detected significant transcriptional changes in the MC38-HER2 tumor cells in response to CAR T-cell treatment (Fig. 1D, left), with GSEA analysis revealing significant gene enrichment in pathways relating to TNF signaling through NF-κB and the IFNγ response (Fig. 1D, center). Further analysis of the data confirmed upregulation of specific genes known to be induced by TNF and IFNγ signaling, including CXCL10, NFKB1A, IRF1, and TAP1 (Fig. 1D, right). Transcriptional analysis of the human breast cancer line AU565 (which endogenously expresses HER2) cocultured with human CAR T cells also showed gene enrichment pathways relating to TNF signaling (Supplementary Fig. S1B). These data demonstrated that CAR T cells can produce the inflammatory cytokines, TNF and IFNγ, which elicits a transcriptional cytokine response in tumor cells.

Birinapant enhances tumor cell death in the presence of CAR T cells

Because the SM, birinapant, has previously been shown to enhance the death of tumor cells expressing a model antigen in the presence of transgenic OT-I T cells (15), we wanted to determine if birinapant could enhance the antitumor activity of CAR T cells to tumor cells expressing the tumor antigen HER2. Mouse CAR T cells were cocultured with MC38 or breast cancer E0771 tumor cells (both engineered to express human HER2), and human CAR T cells with HeLa or AU565 tumor cells (both endogenously express HER2), and cell death was measured using a chromium release killing assay. In all cell lines, a significant increase in cell death in the presence of birinapant compared with vehicle was observed (Fig. 2A). Birinapant did not induce cell death of the CAR T cells (Supplementary Fig. S1C), as has been previously observed for other T cells (15, 22).

Figure 2.

IAP antagonism enhances tumor cell death in the presence of CAR T cells. A, Chromium release assay (18 hours) using mouse CAR T cells with E0771-HER2 and MC38-HER2 targets and human CAR T cells with HeLa-HER2 and AU565 targets (at the indicated E:T ratios) in the presence of vehicle control (Captisol) or birinapant (1 μmol/L). B, Left, MC38-HER2 cells were seeded in chamber slides, then overlaid with mouse CAR T cells (2:1 ratio) labeled with CTV in PI-containing media. Representative still images at the indicated time points are depicted (hr:min). Scale bar, 50 μm. Right, percentage PI-positive cells without direct CAR T-cell contact was manually quantitated by live imaging. The number of tumor cells that died throughout the duration of the movie (as measured by PI uptake) without contact with a T-cell was quantitated and presented in the graph. Quantitation data are pooled from 3 separate camera positions. C, Schematic of chromium release assay to quantify bystander killing. D, Relative killing of MC38 parental cells by bystander killing in the presence vehicle control (Captisol) or birinapant (1 μmol/L), as depicted in C. All relative killing data are pooled from 3 independent experiments. Error bars, mean ± SEM; *, P < 0.05; **, P < 0.01 by an unpaired Student t test.

Figure 2.

IAP antagonism enhances tumor cell death in the presence of CAR T cells. A, Chromium release assay (18 hours) using mouse CAR T cells with E0771-HER2 and MC38-HER2 targets and human CAR T cells with HeLa-HER2 and AU565 targets (at the indicated E:T ratios) in the presence of vehicle control (Captisol) or birinapant (1 μmol/L). B, Left, MC38-HER2 cells were seeded in chamber slides, then overlaid with mouse CAR T cells (2:1 ratio) labeled with CTV in PI-containing media. Representative still images at the indicated time points are depicted (hr:min). Scale bar, 50 μm. Right, percentage PI-positive cells without direct CAR T-cell contact was manually quantitated by live imaging. The number of tumor cells that died throughout the duration of the movie (as measured by PI uptake) without contact with a T-cell was quantitated and presented in the graph. Quantitation data are pooled from 3 separate camera positions. C, Schematic of chromium release assay to quantify bystander killing. D, Relative killing of MC38 parental cells by bystander killing in the presence vehicle control (Captisol) or birinapant (1 μmol/L), as depicted in C. All relative killing data are pooled from 3 independent experiments. Error bars, mean ± SEM; *, P < 0.05; **, P < 0.01 by an unpaired Student t test.

Close modal

To investigate the mechanism of this enhanced tumor cell death in the presence of birinapant, we monitored the interaction of mouse CAR T cells with MC38-HER2 cells by time-lapse imaging. We observed a significant increase in the death of MC38-HER2 cells, indicated by red fluorescence, in the birinapant-treated cocultures (Fig. 2B; Supplementary Movies S1–S2). Further analyses of the imaging revealed a significant increase in the percentage of cell death that occurred independently of direct CAR T-cell interaction, described as “bystander killing” (refs. 7, 15; Fig. 2B, right). To confirm this, we conducted a killing assay with a 50:50 mix of unlabeled MC38-HER2 and chromium-labeled MC38 parental cells. CAR T cells do not recognize non-HER2–expressing tumor cells (14), and therefore, any chromium release occurred independently of direct CAR T-cell interaction (Fig. 2C). These data demonstrated a significant increase in bystander killing in the presence of CAR T cells and birinapant (Fig. 2D).

Birinapant enhances CAR T cell efficacy in a TNF-dependent manner

We next wanted to determine the mechanism by which birinapant enhances tumor cell death in the presence of CAR T cells. Given the ability of birinapant to sensitize cells to TNF-mediated cell death, we first added an anti-TNF–neutralizing antibody to the cocultures of both mouse and human CAR T cells and their cognate target tumor cells (Fig. 3A). In all cases, the addition of anti-TNF significantly reduced the ability of birinapant to enhance tumor cell death in the presence of CAR T cells.

Figure 3.

Birinapant enhances tumor cell death in the presence of CAR T cells in a TNF-dependent, perforin-independent manner. A, Chromium release assay (18 hours) using mouse CAR T cells with E0771-HER2 and MC38-HER2 targets and human CAR T cells with HeLa-HER2 and AU565 targets (at the indicated E:T ratio) in the presence of vehicle control (Captisol) or birinapant (1 μmol/L) and presence or absence of a TNF-neutralizing antibody (10 μg/mL). B, Chromium release assays using WT or Tnf−/− CAR T cells and MC38-HER2 cells as targets (at the indicated E:T ratio) in the presence or absence of vehicle control (Captisol) or birinapant (1 μmol/L). C, WT or Tnf−/− CAR T cells were exposed to MC38-HER2 cells at the indicated E:T ratio. After 4 hours, cytokines in supernatants were measured by CBA. D, Left, chromium release assay (4 hours) using WT or perforin−/− CAR T cells and MC38-HER2 cells as targets (at the indicated E:T ratio) in the presence of vehicle control (Captisol) or birinapant (1 μmol/L). Center, chromium release assay after 18 hours. Right, chromium release assay using MC38 HER2 cells and WT or perforin−/− CAR T cells (at the indicated E:T ratio) untreated (vehicle) in the presence or absence of anti-TNF–neutralizing antibody (10 μg/mL). Error bars, mean ± SEM of triplicate determinations from a representative of 2 independent experiments. *, P < 0.05 by an unpaired Student t test.

Figure 3.

Birinapant enhances tumor cell death in the presence of CAR T cells in a TNF-dependent, perforin-independent manner. A, Chromium release assay (18 hours) using mouse CAR T cells with E0771-HER2 and MC38-HER2 targets and human CAR T cells with HeLa-HER2 and AU565 targets (at the indicated E:T ratio) in the presence of vehicle control (Captisol) or birinapant (1 μmol/L) and presence or absence of a TNF-neutralizing antibody (10 μg/mL). B, Chromium release assays using WT or Tnf−/− CAR T cells and MC38-HER2 cells as targets (at the indicated E:T ratio) in the presence or absence of vehicle control (Captisol) or birinapant (1 μmol/L). C, WT or Tnf−/− CAR T cells were exposed to MC38-HER2 cells at the indicated E:T ratio. After 4 hours, cytokines in supernatants were measured by CBA. D, Left, chromium release assay (4 hours) using WT or perforin−/− CAR T cells and MC38-HER2 cells as targets (at the indicated E:T ratio) in the presence of vehicle control (Captisol) or birinapant (1 μmol/L). Center, chromium release assay after 18 hours. Right, chromium release assay using MC38 HER2 cells and WT or perforin−/− CAR T cells (at the indicated E:T ratio) untreated (vehicle) in the presence or absence of anti-TNF–neutralizing antibody (10 μg/mL). Error bars, mean ± SEM of triplicate determinations from a representative of 2 independent experiments. *, P < 0.05 by an unpaired Student t test.

Close modal

To confirm this observation, we also utilized TNF-deficient (Tnf−/−) CAR T cells, which had significantly decreased effector function compared with wild-type (WT) CAR T cells, an effect that was not ameliorated by the administration of birinapant (Fig. 3B). We detected no TNF production from the Tnf−/− CAR T cells, as expected, but equivalent amounts of IFNγ were secreted between Tnf−/− and WT CAR T cells (Fig. 3C). To confirm that birinapant did not enhance tumor cell death via the perforin-mediated granule exocytosis pathway, we repeated the assay with perforin−/− CAR T cells. Perforin−/− CAR T cells had decreased effector function compared with WT CAR T cells at 4 hours, and tumor cell death was unaffected by the addition of birinapant (Fig. 3D, left). However, over an 18-hour assay, tumor cell death in the presence of perforin−/− CAR T cells was increased, and this was further augmented in the presence of birinapant (Fig. 3D, center). This increase in cell death was significantly reduced by the addition of a TNF-neutralizing antibody (Fig. 3D, right), demonstrating that the increase in tumor cell death in the presence of CAR T cells and birinapant is perforin-independent and TNF-dependent.

Combination CAR T-cell therapy and birinapant enhances tumor control

Using a human-HER2 transgenic mouse model (14), we next investigated the combination therapy of CAR T cells and birinapant in vivo. MC38-HER2 tumor–bearing C57BL/6 transgenic human-HER2 mice were treated with 1 × 107 CAR T cells on days 0 and 1, and birinapant (12 mg/kg) on days 0, 4, and 8. Control mice established robust tumors, which were not cleared by CAR T-cell or birinapant therapy alone. However, the combination of CAR T-cell therapy and birinapant significantly reduced tumor growth, with complete tumor clearance observed in 50% mice. Long-term overall survival was significantly increased following the combination therapy (Fig. 4A), demonstrating that the addition of birinapant enhanced CAR T-cell therapy in vivo.

Figure 4.

Birinapant enhances CAR T-cell therapy in vivo and in patient-derived tumoroids. C57BL/6 human HER2 transgenic mice were injected subcutaneously with MC38-HER2 cells (2.5 × 105 cells/mouse) and tumors were allowed to establish for 7 days. Mice received total body irradiation (4 Gy) before adoptive transfer of anti-HER2 CAR T cells intravenously (1 × 107 cells/mouse), and a total of 5 doses of intraperitoneal IL2 (50,000 IU/injection) over 5 consecutive days. Birinapant was administered at 12 mg/kg by intraperitoneal injection on days 0, 4, and 8 after CAR T-cell injection. A, Growth of MC38 tumors at indicated time points and tumor size at 12 days after treatment. Data are mean ± SEM of 14–15 mice/group, pooled from 2 independent experiments. *, P < 0.05 by ANOVA. Right, survival of mice to day 50 after treatment. Data are mean ± SEM of 14–15 mice/group, pooled from 2 independent experiments. **, P < 0.01; ****, P < 0.0001 by log-rank (Mantel–Cox) test. B, Fluorescent microscopy of patient-derived colorectal tumoroids embedded in Matrigel and administered vehicle control (Captisol) or soluble TNF (10 ng/mL) in the presence or absence of birinapant (1 μmol/L) in PI-containing media. C, Top, schematic representation of CAR T-cell treatment of Matrigel-embedded tumoroids. Dashed lines, production of TNF by CAR T cells and bystander killing of tumoroids by TNF-mediated apoptosis. Bottom, tumoroids were dissociated into single-cell suspensions and analyzed for HER2 expression by flow cytometry compared with isotype control. D, Tumoroids were treated with vehicle control (Captisol) or CAR T cells in the presence or absence of birinapant (1 μmol/L), or in the presence of birinapant and a neutralizing TNF antibody, in PI-containing media. Total fluorescent intensity was quantified using Meta-Morph software. Data are mean ± SEM from tumoroids derived from 2 individual patients in 3 independent experiments. *, P < 0.05; **, P < 0.01 by ANOVA.

Figure 4.

Birinapant enhances CAR T-cell therapy in vivo and in patient-derived tumoroids. C57BL/6 human HER2 transgenic mice were injected subcutaneously with MC38-HER2 cells (2.5 × 105 cells/mouse) and tumors were allowed to establish for 7 days. Mice received total body irradiation (4 Gy) before adoptive transfer of anti-HER2 CAR T cells intravenously (1 × 107 cells/mouse), and a total of 5 doses of intraperitoneal IL2 (50,000 IU/injection) over 5 consecutive days. Birinapant was administered at 12 mg/kg by intraperitoneal injection on days 0, 4, and 8 after CAR T-cell injection. A, Growth of MC38 tumors at indicated time points and tumor size at 12 days after treatment. Data are mean ± SEM of 14–15 mice/group, pooled from 2 independent experiments. *, P < 0.05 by ANOVA. Right, survival of mice to day 50 after treatment. Data are mean ± SEM of 14–15 mice/group, pooled from 2 independent experiments. **, P < 0.01; ****, P < 0.0001 by log-rank (Mantel–Cox) test. B, Fluorescent microscopy of patient-derived colorectal tumoroids embedded in Matrigel and administered vehicle control (Captisol) or soluble TNF (10 ng/mL) in the presence or absence of birinapant (1 μmol/L) in PI-containing media. C, Top, schematic representation of CAR T-cell treatment of Matrigel-embedded tumoroids. Dashed lines, production of TNF by CAR T cells and bystander killing of tumoroids by TNF-mediated apoptosis. Bottom, tumoroids were dissociated into single-cell suspensions and analyzed for HER2 expression by flow cytometry compared with isotype control. D, Tumoroids were treated with vehicle control (Captisol) or CAR T cells in the presence or absence of birinapant (1 μmol/L), or in the presence of birinapant and a neutralizing TNF antibody, in PI-containing media. Total fluorescent intensity was quantified using Meta-Morph software. Data are mean ± SEM from tumoroids derived from 2 individual patients in 3 independent experiments. *, P < 0.05; **, P < 0.01 by ANOVA.

Close modal

Birinapant enhances CAR T-cell efficacy using patient tumoroids

To determine if the combination of CAR T cells and birinapant would enhance tumor cell death in a human cancer, we utilized patient biopsy-derived colorectal tumoroids that expressed HER2 (Supplementary Fig. S2A). The tumoroids were not sensitive to TNF or birinapant alone, but the combination significantly enhanced tumor cell death, as indicated by red fluorescence (Fig. 4B; Supplementary Fig. S2B). To test the effect of birinapant on CAR T-cell efficacy using this model, human CAR T cells were applied to Matrigel-embedded tumoroids (Fig. 4C, top), which endogenously expressed HER2 (Fig. 4C, bottom). The addition of CAR T cells resulted in minimal tumoroid death, whereas the combination of CAR T cells and birinapant resulted in a significant increase in tumoroid death compared with CAR T-cell treatment alone (Fig. 4D). The addition of an anti-TNF–neutralizing antibody to the combination-treated tumoroids significantly reduced the amount of tumoroid death, suggesting that TNF-mediated bystander killing was the effector pathway of combination CAR T-cell and birinapant therapy, not direct T-cell contact (Supplementary Fig. S2C). Together, these data demonstrate that the addition of birinapant to CAR T-cell treatment can significantly enhance tumor cell death and is likely to be a combination for more effective adoptive cellular therapy.

Although the development of CAR T-cell therapy has been revolutionary within the field of immunotherapy, the clinical success has been limited to hematological cancers (2) and a handful of solid cancers (23, 24). A limited number of studies have reported combination therapies that enhance CAR T-cell therapy in solid cancers. Here, we demonstrated that the SM, birinapant, significantly enhanced the efficacy of CAR T cells in a solid tumor setting, in a TNF-dependent manner.

Birinapant is currently undergoing clinical trials as a single agent for a number of hematological and solid tumor cancers (25, 26). Many tumors upregulate IAPs (27) to facilitate resistance to apoptosis following standard chemotherapies (28). The degradation of IAPs by SMs results in the secretion of TNF, which can then induce apoptosis in an autocrine manner via the extrinsic pathway (6). Tumor cells that do not secrete TNF following IAP degradation can also be killed by the exogenous addition of TNF (29). We have previously shown the potential of cytotoxic lymphocytes as a source of TNF that can be utilized by SMs in vitro (15). Using a human-HER2 self-antigen mouse model, we demonstrated here that HER2+ MC38 tumor–bearing mice treated with CAR T cells and birinapant had a significant survival advantage compared with either monotherapy, with complete remission achieved in half the mice. Given the lack of robust responses to CAR T-cell therapy observed in this model (14), these data may have significant implications for previously difficult-to-treat solid malignancies.

In order to demonstrate the potency of this combination therapy on patient tissue, we utilized HER2+ patient-derived colorectal tumoroids embedded in Matrigel as a model of a solid tumor. We have previously uncovered a critical role of the TNF-mediated apoptotic pathway as a mechanism of inducing tumor cell death in the absence of direct T-cell contact (7, 15). We demonstrated that, even with low CAR T-cell penetration, in the presence of birinapant, CAR T-cell–derived TNF can penetrate through the Matrigel and trigger apoptosis of buried tumor cells through this “bystander” effect (7, 15). This may therefore represent a therapy that overcomes the limitation of small numbers of CAR T cells reaching and penetrating the tumor efficiently, while allowing for potent killing of antigen-negative tumor cells by TNF. Capitalizing on the enhanced TNF produced by T cells following checkpoint blockade (15) may also provide a therapy by combining CAR T-cell administration along with birinapant and anti–PD-1 treatment. Given all 3 therapies are either in the clinic or currently being assessed in clinical trials, we expect there can be rapid translation of this therapy into patients.

R.W. Johnstone reports receiving commercial research grants from Roche and AstraZeneca and is a consultant/advisory board member for MecRx. No potential conflicts of interest were disclosed by the authors.

Conception and design: J. Michie, C.J. Kearney, J. Oliaro

Development of methodology: J. Michie, P.A. Beavis, V. Narasimhan, R.G. Ramsay, J. Oliaro

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): J. Michie, P.A. Beavis, A.J. Freeman, K.M. Ramsbottom, V. Narasimhan, E.J. Lelliott, R.G. Ramsay, J. Silke, J. Oliaro

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): J. Michie, S.J. Vervoort, K.M. Ramsbottom, V. Narasimhan, N. Lalaoui, R.G. Ramsay, P.K. Darcy, I. Voskoboinik, J. Oliaro

Writing, review, and/or revision of the manuscript: J. Michie, P.A. Beavis, V. Narasimhan, R.G. Ramsay, R.W. Johnstone, P.K. Darcy, C.J. Kearney, J. Oliaro

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): V. Narasimhan, J. Silke

Study supervision: S.J. Vervoort, R.G. Ramsay, I. Voskoboinik, C.J. Kearney, J. Oliaro

C.J. Kearney is funded by an NHMRC Early Career Fellowship, S.J. Vervoort by a Rubicon Fellowship (Netherlands Organization for Scientific Research), R.G. Ramsay by an NHMRC project grant, P.K. Darcy by an NHMRC program grant, I. Voskoboinik by an NHMRC fellowship and project grant, and J. Oliaro by an NHMRC and an NBCF project grant.

We thank Medivir (Sweden) for supply of birinapant, A/Professor Michael Kershaw for the HER2-expressing tumor lines, Professor Joseph Trapani for perforin-deficient mice, and Elhadi Lich for assistance with tumoroid analyses.

1.
Beavis
PA
,
Milenkovski
N
,
Henderson
MA
,
John
LB
,
Allard
B
,
Loi
S
, et al
Adenosine receptor 2A blockade increases the efficacy of anti-PD-1 through enhanced antitumor T-cell responses
.
Cancer Immunol Res
2015
;
3
:
506
17
.
2.
Maude
SL
,
Teachey
DT
,
Porter
DL
,
Grupp
SA
. 
CD19-targeted chimeric antigen receptor T-cell therapy for acute lymphoblastic leukemia
.
Blood
2015
;
125
:
4017
23
.
3.
Kershaw
MH
,
Westwood
JA
,
Darcy
PK
. 
Gene-engineered T cells for cancer therapy
.
Nat Rev Cancer
2013
;
13
:
525
41
.
4.
John
LB
,
Kershaw
MH
,
Darcy
PK
. 
Blockade of PD-1 immunosuppression boosts CAR T-cell therapy
.
Oncoimmunology
2013
;
2
:
e26286
.
5.
Blake
SJ
,
Ching
AL
,
Kenna
TJ
,
Galea
R
,
Large
J
,
Yagita
H
, et al
Blockade of PD-1/PD-L1 promotes adoptive T-cell immunotherapy in a tolerogenic environment
.
PLoS One
2015
;
10
:
e0119483
.
6.
Vince
JE
,
Wong
WW
,
Khan
N
,
Feltham
R
,
Chau
D
,
Ahmed
AU
, et al
IAP antagonists target cIAP1 to induce TNFalpha-dependent apoptosis
.
Cell
2007
;
131
:
682
93
.
7.
Kearney
CJ
,
Vervoort
SJ
,
Hogg
SJ
,
Ramsbottom
KM
,
Freeman
AJ
,
Lalaoui
N
, et al
Tumor immune evasion arises through loss of TNF sensitivity
.
Sci Immunol
2018
;
3
.
8.
Condon
SM
,
Mitsuuchi
Y
,
Deng
Y
,
LaPorte
MG
,
Rippin
SR
,
Haimowitz
T
, et al
Birinapant, a smac-mimetic with improved tolerability for the treatment of solid tumors and hematologicalal malignancies
.
J Med Chem
2014
;
57
:
3666
77
.
9.
Ashkenazi
A
,
Dixit
VM
. 
Death receptors: signaling and modulation
.
Science
1998
;
281
:
1305
8
.
10.
Chesi
M
,
Mirza
NN
,
Garbitt
VM
,
Sharik
ME
,
Dueck
AC
,
Asmann
YW
, et al
IAP antagonists induce anti-tumor immunity in multiple myeloma
.
Nat Med
2016
;
22
:
1411
20
.
11.
Beug
ST
,
Beauregard
CE
,
Healy
C
,
Sanda
T
,
St-Jean
M
,
Chabot
J
, et al
Smac mimetics synergize with immune checkpoint inhibitors to promote tumour immunity against glioblastoma
.
Nat Commun
2017
;
8
.
12.
Min
DJ
,
He
S
,
Green
JE
. 
Birinapant (TL32711) improves responses to GEM/AZD7762 combination therapy in triple-negative breast cancer cell lines
.
Anticancer Res
2016
;
36
:
2649
57
.
13.
Krepler
C
,
Chunduru
SK
,
Halloran
MB
,
He
X
,
Xiao
M
,
Vultur
A
, et al
The novel SMAC mimetic birinapant exhibits potent activity against human melanoma cells
.
Clin Cancer Res
2013
;
19
:
1784
94
.
14.
Mardiana
S
,
John
LB
,
Henderson
MA
,
Slaney
CY
,
von Scheidt
B
,
Giuffrida
L
, et al
A multifunctional role for adjuvant anti-4–1BB therapy in augmenting antitumor response by chimeric antigen receptor T cells
.
Cancer Res
2017
;
77
:
1296
309
.
15.
Kearney
CJ
,
Lalaoui
N
,
Freeman
AJ
,
Ramsbottom
KM
,
Silke
J
,
Oliaro
J
. 
PD-L1 and IAPs co-operate to protect tumors from cytotoxic lymphocyte-derived TNF
.
Cell Death Differ
2017
;
24
:
1705
16
.
16.
Beavis
PA
,
Henderson
MA
,
Giuffrida
L
,
Mills
JK
,
Sek
K
,
Cross
RS
, et al
Targeting the adenosine 2A receptor enhances chimeric antigen receptor T cell efficacy
.
J Clin Invest
2017
;
127
:
929
41
.
17.
Darcy
PK
,
Haynes
NM
,
Snook
MB
,
Trapani
JA
,
Cerruti
L
,
Jane
SM
, et al
Redirected perforin-dependent lysis of colon carcinoma by ex vivo genetically engineerd CTL.
J Immunol
2000
;
164
:
3705
12
.
18.
Haynes
NM
,
Trapani
JA
,
Teng
MW
,
Jackson
JT
,
Cerruti
L
,
Jane
SM
, et al
Single chain antigen recognition receptors that costimulate potent rejection of established experimental tumours
.
Blood
2002
;
100
:
3155
63
.
19.
Wang
LX
,
Westwood
JA
,
Moeller
M
,
Duong
CP
,
Wei
WZ
,
Malaterre
J
, et al
Tumor ablation by gene modified T cells in the absence of autoimmunity
.
Cancer Res
2010
;
70
:
9591
8
.
20.
House
IG
,
Thia
K
,
Brennan
AJ
,
Tothill
R
,
Dobrovic
A
,
Yeh
WZ
, et al
Heterozygosity for the common perforin mutation, p.A91V, impairs the cytotoxicity of primary natural killer cells from healthy individuals
.
Immunol Cell Biol
2015
;
93
:
575
80
.
21.
Kershaw
MH
,
Jackson
JT
,
Haynes
NM
,
Teng
MW
,
Moeller
M
,
Hayakawa
Y
, et al
Gene-engineered T cells as a superior adjuvant therapy for metastatic cancer
.
J Immunol
2004
;
173
:
2143
50
.
22.
Dougan
M
,
Dougan
S
,
Slisz
J
,
Firestone
B
,
Vanneman
M
,
Draganov
D
, et al
IAP inhibitors enhance co-stimulation to promote tumor immunity
.
J Exp Med
2010
;
207
:
2195
206
.
23.
Ahmed
N
,
Brawley
VS
,
Hegde
M
,
Robertson
C
,
Ghazi
A
,
Gerken
C
, et al
Human epidermal growth factor receptor 2 (HER2)-specific chimeric antigen receptor-modified T cells for the immunotherapy of HER2-positive sarcoma
.
J Clin Oncol
2015
;
33
:
1688
96
.
24.
Louis
CU
,
Savoldo
B
,
Dotti
G
,
Pule
M
,
Yvon
E
,
Myers
GD
, et al
Antitumor activity and long-term fate of chimeric antigen receptor-positive T cells in patients with neuroblastoma
.
Blood
2011
;
118
:
6050
6
.
25.
Noonan
AM
,
Bunch
KP
,
Chen
JQ
,
Herrmann
MA
,
Lee
JM
,
Kohn
EC
, et al
Pharmacodynamic markers and clinical results from the phase 2 study of the SMAC mimetic birinapant in women with relapsed platinum-resistant or -refractory epithelial ovarian cancer
.
Cancer
2016
;
122
:
588
97
.
26.
Amaravadi
RK
,
Schilder
RJ
,
Martin
LP
,
Levin
M
,
Graham
MA
,
Weng
DE
, et al
A Phase I study of the SMAC-mimetic birinapant in adults with refractory solid tumors or lymphoma
.
Mol Cancer Ther
2015
;
14
:
2569
75
.
27.
Deveraux
QL
,
Reed
JC
. 
IAP family proteins—suppressors of apoptosis
.
Genes Dev
1999
;
13
:
239
52
.
28.
Hong
X
,
Lei
L
,
Glas
R
. 
Tumors acquire inhibitor of apoptosis protein (IAP)-mediated apoptosis resistance through altered specificity of cytosolic proteolysis
.
J Exp Med
2003
;
197
:
1731
43
.
29.
Petersen
SL
,
Wang
L
,
Yalcin-Chin
A
,
Li
L
,
Peyton
M
,
Minna
J
, et al
Autocrine TNFalpha signaling renders human cancer cells susceptible to Smac-mimetic-induced apoptosis
.
Cancer Cell
2007
;
12
:
445
56
.