Abstract
Background/Aim: B cell maturation antigen (BCMA) is an ideal target for adoptive T cell therapy of multiple myeloma (MM). In this study, we evaluated self-differentiated monocyte-derived dendritic cells expressing BCMA (SD-DC-BCMA) to activate T cells for killing MM cells. Materials and Methods: Lentivirus-modified SD-DC-BCMA harboring tri-cistronic cDNAs encoding granulocyte-macrophage colony-stimulating factor (GM-CSF), interleukin-4 (IL-4), and BCMA was generated. Cytotoxicity of T cells activated by SD-DC-BCMA against MM cells was evaluated. Results: T cells activated by SD-DC-BCMA exhibited a dose-dependent cytotoxicity against BCMA-expressing MM cells and produced high IFN-γ levels, compared to inactivated T cells or control T cells. A significantly higher killing ability of T cells activated by SD-DC-BCMA was further demonstrated in BCMA-overexpressing cells when compared with BCMA-negative cells. Conclusion: The potency of SD-DC-BCMA to activate T cells for antigen-specific cancer killing provides a framework for therapeutic application of adoptive T cell therapy in MM.
Multiple myeloma (MM) is a malignancy of plasma cells, and it develops in the bone marrow. MM causes excessive production of monoclonal immunoglobulin and defective hematopoiesis, which leads to a broad range of clinical manifestations including hypercalcemia, anemia, renal insufficiency, osteolytic lesions, and immune dysfunction (1). Despite the fact that the combination of high-dose chemotherapy with autologous stem-cell transplantation and novel agents, such as immunomodulatory drugs and proteasome inhibitors, has significantly improved clinical outcomes, the majority of MM patients eventually relapse or become unresponsive to the treatment (2–5). Therefore, the development of new immunotherapeutic strategies to be used as an alternative treatment or as adjuvant therapy to enhance the efficacy of standard treatment is urgently required. Immunotherapy has demonstrated great promise for treatment of MM by stimulating durable antitumor immune responses with limited toxicity (6).
Adoptive T cell therapy represents a potential immunotherapeutic modality for treatments of MM and other cancers (6, 7). Dendritic cells (DCs), the most powerful antigen presenting cells (APCs), provide an effective strategy for enhancing tumor specificity and the cytotoxicity of T cells for adoptive T cell therapy (8, 9). Adoptive transfer of antigen-specific T cells activated by DCs revealed effective anti-tumor activity in acute myeloid leukemia (AML) murine model (10), and complete and safe regression in a metastatic melanoma patient (11). MM-specific T cells were reported to be activated by DCs pulsed with different forms of antigens, including whole antigens, mRNAs, and peptides (12–15). Combination therapy in MM using ex-vivo activated T cell transfer following DC or tumor antigen immunization demonstrated enhanced antigen-specific T cell responses, accelerated immune restoration, and improved survival outcomes in patients (16, 17). These data suggest a concept for development of DCs to activate antigen-specific T cells for killing MM cells.
Self-differentiated monocyte-derived dendritic cells (SD-DCs) have been developed to overcome the difficult processes and low potency associated with large-scale DC production (18–23). Unlike the cytokine-driven DC generation or monocyte-derived conventional dendritic cells (ConvDCs), SD-DCs could be generated by transduction of monocytes with lentiviruses containing cytokine-encoded cDNA sequences important for DC differentiation, including granulocyte-macrophage colony-stimulating factor (GM-CSF), interleukin-4 (IL-4) and tumor-associated antigen (TAA) cDNA. These cells possessed a self-differentiation property, as well as the ability to synthesize and present TAA (18–23). The efficiency of SD-DC technology in promoting DC differentiation and antigen-specific T cell responses has been demonstrated (18–23). In comparison with ConvDCs, SD-DCs provided greater potency to upregulate maturation markers on DCs and proteins involved in immune function, such as toll-like receptors (TLRs) and TNF-related apoptosis-inducing ligand (TRAIL) (20, 21). In terms of T cell activation, SD-DCs could enhance T cell expansion, and importantly, increase cytotoxicity of activated T cells against antigen-positive cancer cells in various cancer models (18–23). Our group previously generated SD-DCs expressing cAMP-dependent protein kinase type I-alpha regulatory subunit (SD-DC-PRKAR1A) and reported the greater killing ability of T cells activated by SD-DC-PRKAR1A against cholangiocarcinoma (CCA) cells compared to T cells activated by other DC generation methods (23). To investigate the potential of this SD-DC technology in the development of adoptive T cell therapy for MM, we created new SD-DCs presenting TAA specific to MM, and we evaluated its efficiency in activating T cells for killing MM cells.
The selection of TAA expressed in MM cells is an essential step for generation of SD-DCs presenting this TAA. B cell maturation antigen (BCMA) is an ideal TAA for generation of SD-DCs in the development of adoptive T cell therapy for MM, due to its selective expression in MM cells (24–27), its correlation with MM progression (28), and its vital roles in growth and survival of MM cells (29, 30). Previous studies in both pre-clinical and clinical trials have shown the ability of DCs loaded with BCMA mRNA and peptides to induce anti-MM specific T cell responses (13, 15, 31). Nevertheless, in this study, we were the first to generate SD-DCs expressing BCMA (SD-DC-BCMA) for activation of T cells specific for BCMA. We created SD-DC-BCMA using lentiviral vector, and examined whether this SD-DC-BCMA could activate T cells to specifically kill BCMA-expressing MM cells.
Materials and Methods
Cell culture. Human MM cell lines, RPMI8226, NCI H929, and KMS20, were kindly provided by Prof. Seiji Okada of the Center for AIDS Research and Graduate School of Medical Sciences, Division of Hematopoiesis, Kumamoto University, Kumamoto, Japan. KKU-055 cholangiocarcinoma cell lines were obtained from Japanese Collection of Research Bioresources (JCRB) Cell Bank (Osaka, Japan). Raji B lymphocytic cell lines were purchased from American Type Culture Collection (ATCC) (Manassas, VA, USA), and Lenti-X™ human embryonic kidney (HEK) 293T cell lines were purchased from Takara Bio (Mountain View, CA, USA). MM and Raji cells were cultured in Roswell Park Memorial Institute (RPMI) 1640 Medium (Gibco; Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 15% fetal bovine serum (FBS) (Gibco) and 50 units penicillin/50 μg streptomycin (both from Thermo Fisher Scientific). KKU-055 and Lenti-X™ HEK 293T cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM) (Gibco) supplemented with 10% FBS and 50 units penicillin/50 μg streptomycin. All cells were maintained at 37°C with 5% CO2.
Construction of lentiviral vectors and lentivirus production. Replication-defective lentiviral vector backbones (Addgene, Watertown, MA, USA) were kindly provided by Dr. Naravat Poungvarin, Clinical Molecular Pathology Laboratory, Department of Clinical Pathology, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand. The full-length GM-CSF cDNA (Sequence ID: BC113924.1) and IL-4 cDNA (Sequence ID: NM_000589.4) inserted with P2A and F2A elements (GM-P2A-IL4-F2A) were custom-made by Integrated DNA Technologies, Inc. (Coralville, IA, USA), and full-length BCMA cDNA (Sequence ID: NM_001192.2) was reverse transcribed from mRNA extracted from U266 MM cell lines. To generate SD-DC-BCMA lentiviral construct, the full-length GM-P2A-IL4-F2A and BCMA cDNAs were amplified by polymerase chain reaction (PCR) and cloned into pCDH lentiviral transfer vector (Addgene), namely pCDH-GM/IL4/BCMA (the full sequence of SD-DC-BCMA lentiviral construct and the primer sequences available on request). Lentiviral transfer vector expressing GM-CSF, IL-4, and irrelevant protein (a phytochrome-based near-infrared fluorescent protein; iRFP) or SD-DC-iRFP (pCDH-GM/IL4/iRFP), obtained from our previous study (23) was used as a negative control. For generation of mWasabi green fluorescence protein and firefly luciferase-expressing target cancer cells, full-length cDNAs of mWasabi and luciferase enzyme were custom-made (Integrated DNA Technologies, Inc.) and cloned into pCDH lentiviral transfer vector to generate pCDH-mWasabi/luciferase lentiviral vector. Lentiviral transfer construct containing BCMA cDNA (pCDH-BCMA) was generated for production of BCMA-overexpressing target cells. All plasmid constructs were analyzed by Sanger DNA sequencing technique. Calcium phosphate transfection was carried out to test protein expression of the transgene and to produce the lentiviruses. To examine protein expression of the transgene, the Lenti-X™ HEK 293T cells were transfected with 1 μg each of pCDH-GM/IL4/BCMA or pCDH-GM/IL4/iRFP lentiviral transfer construct. After culturing for 48 h, supernatants and transfected cells were collected to examine protein expression by ELISA and immunoblotting, respectively. To produce the lentiviruses, Lenti-X™ HEK 293T cells were transfected with each lentiviral transfer construct together with psPAX lentiviral packaging vector and pMD2G envelope vector (Addgene) at ratio of 30 μg : 21 μg : 6 μg, respectively. Lentiviral vectors were harvested at 24, 48, and 72-h post-transfection and concentrated by ultracentrifugation at 20,000 × g, 4°C, for 90 min. The virus titers in the cultured supernatants of Lenti-X™ HEK 293T cells (expressed as infectious unit/ml or IU/ml) were quantified by using a qPCR Lentivirus Titration (Titer) Kit (ABM, Richmond, BC, Canada), which were calculated by using the online lentivirus titer calculator located at https://old.abmgood.com/viralexp/lentivirus-calculator.php (ABM).
Generation of ConvDCs and SD-DCs. The study protocol using twelve human blood samples from healthy volunteers was approved by the Siriraj Institutional Review Board (SIRB) of the Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand (COA no. Si 370/2020). HLA class I typing was performed in all donor samples using sequence-specific oligonucleotide primed PCR (PCR-SSO) by the Department of Transfusion Medicine, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand. Monocytes were isolated from peripheral blood mononuclear cells (PBMCs) by plastic adhesion (32) and were used to generate DCs. For ConvDC generation, monocytes at 1×106 cells were cultured in AIM-V Medium (Invitrogen, Carlsbad, CA, USA) containing 50 ng/ml GM-CSF and 25 ng/ml IL-4 (both from ImmunoTools, Friesoythe, Germany) for 5 days followed by AIM-V Medium containing 50 ng/ml tumor necrosis factor-alpha (TNF-α) and 50 ng/ml interferon-gamma (IFN-γ) (both from ImmunoTools) for 2 days. For generation of SD-DC-BCMA and SD-DC-iRFP, the transduction at different multiplicities of infection (MOI) of 50, 70, 100, and 150 was preliminarily tested by following the previously described protocols (23, 33). The MOI of 70, which did not affect cell viability and was able to provide mature DC phenotypes, was selected for the SD-DC generation in this study. The volume of viral supernatants required for the transduction at the indicated MOI was determined by the following formula: Total viral supernatants (ml)=Total number of cells × MOI/Virus titer (IU/ml). Monocytes at 1×106 cells were transduced with SD-DC-iRFP or SD-DC-BCMA lentiviral supernatants at an MOI of 70 in AIM-V Medium containing 10 μg/ml protamine sulfate (Sigma-Aldrich, St. Louis, MO, USA). After 16 h of transduction, the transduced medium was removed, and the cells were cultured in AIM-V Medium without cytokine supplementation for 7 days. Immunophenotypes of SD-DCs were characterized as described below. A schematic diagram of SD-DC generation is shown in Figure 1.
Schematic diagram of SD-DC generation and T cell activation for the cytotoxicity assay. Monocytes and lymphocytes were isolated from PBMCs of healthy donors. Monocytes isolated by plastic adherence were differentiated to SD-DCs by lentiviral transduction followed by in vitro culture without cytokine supplementation for 7 days. Cryopreserved lymphocytes were thawed and activated by SD-DCs for 3 days followed by expansion with IL-2, IL-7, and IL-15 cytokines for 6 days before being used to evaluate cytotoxic activities.
Generation of activated effector T cells. Non-adherent lymphocytes containing effector T cells were activated by co-culturing with SD-DC-iRFP or SD-DC-BCMA at an effector cell to DC ratio of 10:1 in AIM-V Medium containing 5% human AB serum for 3 days. After activation by SD-DCs, effector T cells were separated and further expanded in AIM-V Medium containing 5% human AB serum, 20 ng/ml of IL-2, 10 ng/ml of IL-7, and 20 ng/ml of IL-15 (all from ImmunoTools) for 6 days. Medium and cytokines were refreshed every other day. Inactivated T cells, later used as negative control, were cultured in AIM-V Medium containing 5% human AB serum without SD-DCs for 3 days followed by expansion in the medium containing cytokines as previously described. A schematic diagram of T cell activation for the cytotoxicity assay is shown in Figure 1.
Cytotoxicity assay. Inactivated T cells and T cells activated by SD-DC-iRFP or SD-DC-BCMA were co-cultured with mWasabi-luciferase-expressing target cells, Raji, RPMI8226, or NCI H929, at effector cell-to-target cell (E:T) ratios of 2.5:1, 5:1, and 10:1. After 24 h of co-culturing, the remaining target cells were lysed using Pierce® Luciferase Cell Lysis Buffer, and cytotoxicity was determined using Pierce® Firefly Luciferase Glow Assay Kit (both from Thermo Fisher Scientific) and a Lumat LB 9507 Ultra-sensitive Tube Luminometer (Berthold Technologies GmbH & Co. KG., Bad Wildbad, Germany). The BCMA-negative KMS20 and KKU-055 or BCMA-overexpressing KMS20 and KKU-055 target cells were co-cultured with each effector T cell condition at the same E:T ratios for 8 h. The cytotoxicity against KMS20 cells was determined by the percentage of dead cells using flow cytometric analysis of annexin V/PI-stained cells whereas the cytotoxicity against KKU-055 cells was assayed by crystal violet staining of the remaining viable attached cells.
Immunoblot analysis. Cells were harvested and lysed with radioimmunoprecipitation assay (RIPA) buffer. NCI H929 cells were treated with 17.8 μM cycloheximide (CHX) and 5 μM proteasome inhibitor MG132 (both from Sigma-Aldrich) for inhibition of protein synthesis and proteasome-dependent degradation, respectively, before harvesting. Proteins were separated using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellulose membrane. The membranes were then probed with 1:500 mouse anti-human BCMA monoclonal antibody (mAb) (Clone D-6; Santa Cruz Biotechnology, Dallas, TX, USA) or 1:5,000 mouse anti-human GADPH mAb (Clone 0411; Santa Cruz Biotechnology) as an internal control. The immunereactive proteins were visualized by SuperSignal® West Pico Chemiluminescent Substrate (Thermo Fisher Scientific) and detected using X-ray film exposure.
Flow cytometry analysis. Cell surface expression of the BCMA protein was determined by staining cells with APC-conjugated human mAb against BCMA (Clone REA315; Miltenyi Biotec, Bergisch Gladbach, Germany). Immunophenotypes of monocytes, ConvDCs, and SD-DCs were examined using fluorescent-conjugated mouse mAbs, including CD14-APC (Clone 18D11), CD14-FITC (Clone 18D11), CD11c-APC (Clone BU15), CD40-FITC (Clone HI40a), HLA-DR-FITC (Clone LT-DR) (all from ImmunoTools), CD80-PE (Clone 2D10.4), and CD86-PE (Clone IT2.2) (both from eBioscience, San Diego, CA, USA). Lymphocyte and T cell subsets were examined using fluorescent-conjugated mouse mAbs including CD3-FITC (Clone UCHT-1), CD3-PerCP (Clone UCHT-1), CD4-APC (Clone MEM-241), CD8-APC (Clone UCHT-4), CD16-APC (Clone 3G8), CD19-APC (Clone LT19), CD45RO-FITC (Clone UCHL1), CD62L-APC (Clone LT-TD180) (all from ImmunoTools), and CD56-PE (Clone 5.1H11; BioLegend, San Diego, CA, USA). The percentage of KMS20 cell death in cytotoxicity assay was determined by using annexin V/PI staining method (ImmunoTools). All cells were analyzed using a BD Accuri C6 Plus Flow Cytometer (BD Biosciences, San Jose, CA, USA) and FlowJo software V10 (Becton, Dickenson and Company, Franklin Lakes, NJ, USA).
Quantitative real-time polymerase chain reaction (qPCR). Genomic DNA was extracted from monocytes and SD-DCs using GeneJET Genomic DNA Purification Kit (Thermo Fisher Scientific). The qPCR reactions containing DNA templates, specific primers targeting to albumin (ABL) gene for a single-copy control gene and puromycin-resistance (Puro) gene for a specific gene in the lentiviral vector (primer sequences available on request), and LightCycler® 480 SYBR Green I Master (Roche Diagnostics GmbH, Mannheim, Germany) were prepared and analyzed by a LightCycler® 480 Real-time PCR System (Roche Applied Science, Penzberg, Germany). The vector copy number (VCN) per cell was determined in three independent donors using cycle threshold (Ct) value of puromycin-resistance gene compared to albumin gene.
Enzyme-linked immunosorbent assay (ELISA). GM-CSF, IL-4, and IFN-γ cytokines in culture supernatants were measured using a Quantikine® ELISA Human GM-CSF Immunoassay (DGM00; R&D Systems, Inc., Minneapolis, MN, USA), a Quantikine® ELISA Human IL-4 Immunoassay (D4050; R&D Systems), and a Quantikine® ELISA Human IFN-γ Immunoassay (DIF50; R&D systems), respectively. The cytokine concentrations were determined according to the manufacturer’s instructions.
Statistical analysis. Experimental data were collected and pooled from at least three independent experiments, and the results are presented as mean±standard deviation (SD). Statistical analyses were performed using one-way analysis of variance (ANOVA) and least significant difference (LSD) post hoc test with GraphPad Prism software version 7 (GraphPad Software, San Diego, CA, USA) and SPSS Statistics version 18.0 (SPSS, Inc., Chicago, IL, USA). The difference of data sets with a p-value <0.05 was considered to be statistically significant.
Results
BCMA expression in human MM cells. Expression of BCMA was examined by immunoblot analysis and flow cytometry. Immunoblot analysis showed BCMA expression in the two MM cell lines, RPMI8226 (low expression level) and NCI H929 (higher expression level); however, there was no BCMA expression in the non-MM cell line (Raji; Burkkitt’s lymphoma B cell line) (Figure 2A, B). Flow cytometric analysis indicated that cell surface BCMA expression levels were consistent with total BCMA expression levels detected by immunoblot analysis (Figure 2C, D). BCMA was expressed approximately 2-fold greater in NCI H929 (relative mean fluorescence intensity; rMFI: 12.37±0.27, % positive cells: 79.67±8.67%) than in RPMI8226 (rMFI: 5.26±1.44, % positive cells: 50.20±9.21%), but was not expressed in Raji (rMFI: 0.74±0.04, % positive cells: 0.10±0.06%) (Figure 2C, D).
Expression of BCMA protein in human B cell line and MM cell lines. BCMA expression level was determined by immunoblot analysis and flow cytometry in B cell line (Raji) and MM cell lines (RPMI8226 and NCI H929). (A) Immunoblot analysis showing bands of BCMA and GAPDH (loading control). (B) Densitometric analysis of BCMA bands relative to those of GAPDH bands obtained. (C) Flow cytometry histogram showing staining of surface BCMA (dark gray) in comparison to a matched isotype control (light gray). Numbers denote relative mean fluorescence intensity (rMFI). (D) Percentage of BCMA-positive cells obtained. All data with error bars represents mean±SD from three independent experiments. Differences between the data sets were analyzed by Student’s t-test. Asterisks represent p-values: *p<0.05, ** p<0.01, ***p<0.001.
BCMA was processed via proteasome-dependent degradation. Intracellular proteins are typically processed in antigen presenting cells (APCs) via the proteasome pathway and presented as antigenic peptides on MHC class I molecules to activate CD8+ T cell responses (34). To determine whether BCMA is processed via this mechanism, proteasome degradation of BCMA was investigated in NCI H929 cell line, which contained the highest level of BCMA expression (Figure 2). Cells were treated with 17.8 μM cycloheximide (CHX) to inhibit new protein synthesis and to allow degradation of BCMA. Relative BCMA quantities after CHX treatment were determined at 0, 1, 2, 4, and 6 h by immunoblot analysis. BCMA half-life and turn-over rate were approximately 1 h and 6 h, respectively (Figure 3A, B). Then, proteasome-dependent degradation of BCMA was determined by adding a proteasome inhibitor – MG132. Treatment of cells with 17.8 μM CHX in combination with 5 μM MG132 for 6 h markedly stabilized BCMA, compared to CHX treatment alone (Figure 3C, D). These results demonstrated that BCMA had a short half-life, and that its degradation was processed via the proteasome pathway.
Proteasome degradation of BCMA protein in NCI H929 cell lines. BCMA degradation was investigated in NCI H929 cells by treatment with protein synthesis inhibitor – cycloheximide (CHX), and proteasome inhibitor – MG132. (A) Immunoblot analysis showing bands of BCMA and GAPDH after CHX treatment at 0, 1, 2, 4, and 6 h. (B) Densitometric analysis of BCMA bands relative to those of GAPDH bands after CHX treatment plotted versus time. (C) Immunoblot analysis of BCMA and GAPDH after CHX and/or MG132 treatment at 6 h. (D) Densitometric analysis of BCMA bands relative to those of GAPDH bands after CHX and/or MG132 treatment at 6 h. All data with error bars represents mean±SD from three independent experiments. Differences between the data sets were analyzed by Student’s t-test. Asterisks represent p-values: *p<0.05, **p<0.01.
Lentiviral construct for generation of SD-DC-BCMA. Lentiviral construct containing cDNA sequences encoding GM-CSF, IL-4, and BCMA (pCDH-GM/IL4/BCMA) was generated (Figure 4A) and used to produce lentiviruses by transfection into Lenti-X™ HEK293T cells for the generation of SD-DC-BCMA. Lentiviral vector expressing GM-CSF, IL-4, and irrelevant protein iRFP (pCDH-GM/IL4/iRFP) was also created and used to produce lentiviruses for generation of SD-DC-iRFP as a negative control (Figure 4A). After 48 h of transfection, GM-CSF and IL-4 cytokines in culture supernatants, and BCMA protein in cell lysates were determined by ELISA and immunoblot analysis, respectively. GM-CSF and IL-4 were produced at similar levels from Lenti-X™ HEK293T cells transfected with lentiviral vectors for generation of SD-DC-iRFP or SD-DC-BCMA. The concentrations of GM-CSF secreted from SD-DC-iRFP and SD-DC-BCMA transfections were 38,799±8,765 pg/ml and 38,243±14,125 pg/ml, respectively, whereas the concentrations of IL-4 secreted from SD-DC-iRFP and SD-DC-BCMA transfections were 47,385±20,529 pg/ml and 41,976±10,773 pg/ml, respectively. There was no GM-CSF and IL-4 detected in untransfected Lenti-X™ HEK293T cells. Specifically, expression of BCMA was detected only in Lenti-X™ HEK293T cells transfected with the lentiviral vectors for generation of SD-DC-BCMA. These results confirmed the potencies of these lentiviral vectors for generation of SD-DC-iRFP and SD-DC-BCMA to mediate transgene expression.
Generation of SD-DCs from monocytes using lentiviral transduction. Human monocytes were converted to ConvDCs using recombinant cytokines, and to SD-DC-iRFP and SD-DC-BCMA using lentiviral transduction. Monocytes at the day of isolation (Day 0), and ConvDCs, SD-DC-iRFP, and SD-DC-BCMA at day 7 post-transduction were analyzed. (A) Schematic representation of lentiviral constructs for generation of SD-DC-iRFP and SD-DC-BCMA. (B) Lentiviral vector copy number (VCN) per cell relative to albumin (single-copy control) quantified by qPCR. (C) Levels of secreted GM-CSF in culture supernatants measured by ELISA. (D) Levels of secreted IL-4 in culture supernatants measured by ELISA. (E) Representative cell morphologies. Scale bars represent 50 μm. The results were summarized from three independent experiments. All data with error bars represents mean±SD. Differences between the data sets were analyzed by Student’s t-test. Asterisks represent p-values: *p<0.05, **p<0.01, ***p<0.001.
SD-DC-BCMA was successfully generated from human monocytes. The schematic representation of lentiviral constructs for generation of SD-DC-iRFP and SD-DC-BCMA are illustrated in Figure 4A. Human monocytes were isolated from PBMCs with a purity of 70-80% and were self-differentiated to SD-DC-iRFP and SD-DC-BCMA by lentiviral transduction. ConvDCs were also generated from human monocytes by treatment with recombinant cytokines. To examine lentiviral gene integration, at day 7 after transduction, DNA samples extracted from SD-DC-iRFP and SD-DC-BCMA were analyzed by qPCR. At a MOI of 70, SD-DC-iRFP and SD-DC-BCMA contained approximately 1 vector copy number (VCN) per cell (SD-DC-iRFP: 0.97±0.02 VCN and SD-DC-BCMA: 1.05±0.09 VCN) (Figure 4B). SD-DC differentiation was confirmed by GM-CSF and IL-4 production using ELISA, morphological changes, and immunophenotypic changes by flow cytometry. Similar levels of GM-CSF and IL-4 were produced by SD-DC-iRFP and SD-DC-BCMA at day 7 after transduction, while untransduced monocytes showed no GM-CSF or LI-4 production (Figure 4C, D). The GM-CSF level produced by SD-DC-iRFP and SD-DC-BCMA was 3,220±459 pg/ml and 3,784±467 pg/ml, respectively (Figure 4C). The IL-4 level produced by SD-DC-iRFP and SD-DC-BCMA was 806±198 pg/ml and 1,232±661 pg/ml, respectively (Figure 4D). In comparison with monocytes, all types of DCs, including ConvDCs, SD-DC-iRFP, and SD-DC-BCMA, demonstrated the formation of dendrites (Figure 4E). In accordance with the morphologies, immunophenotypic changes were also observed by flow cytometry. Expression level of monocyte marker, CD14, was specifically found only in monocytes (positive cells: 74.66±10.85%), whereas the expression level was reduced in ConvDCs (positive cells: 1.57±1.55%), SD-DC-iRFP (positive cells: 2.83±2.02%), and SD-DC-BCMA (positive cells: 3.58±2.32%) (Figure 5A, B). In contrast, expression levels of DC marker and co-stimulatory molecules, including CD11c, CD40, CD80, CD86, and HLA-DR, were highly increased in ConvDCs (positive cells of CD11c: 88.20±5.66%, CD40: 91.96±7.43%, CD80: 16.01±21.74%, CD86: 98.93±0.80%, and HLA-DR: 51.70±2.20%), SD-DC-iRFP (positive cells of CD11c: 59.93±21.92%, CD40: 90.96±5.66%, CD80: 14.42±9.76%, CD86: 94.63±3.84%, and HLA-DR: 48.10±7.18%), and SD-DC-BCMA (positive cells of CD11c: 80.86±11.49%, CD40: 92.53±1.35%, CD80: 9.61±10.52%, CD86: 96.6±1.80%, and HLA-DR: 37.50±9.28%) compared to monocytes (positive cells of CD11c: 55.90±16.03%, CD40: 22.22±23.13%, CD80: 1.21±0.84%, CD86: 48.73±27.56%, and HLA-DR: 4.80±5.98%) (Figure 5A, B). The morphologies and immunophenotypes of SD-DC-iRFP and SD-DC-BCMA were similar to those of ConvDCs. These results supported the efficiency of the SD-DC strategy for DC generation.
Immunophenotypic analysis of monocytes, ConvDCs, and SD-DCs. Monocytes at day 0, and ConvDCs, SD-DC-iRFP, and SD-DC-BCMA at day 7 post-transduction were collected and stained to determine immunophenotypic markers, including CD14, CD11c, CD40, CD80, CD86, and HLA-DR by flow cytometry. (A) Flow cytometric histogram showing representative mean fluorescence intensity of each immunophenotypic marker in monocytes and DCs generated by different conditions in comparison to matched isotype control conditions. (B-G) Percentages of positive cells for each immunophenotypic marker obtained from three independent experiments. All data with error bars represents mean±SD.
Phenotype of effector T cells after activation by SD-DCs. Mature dendritic cells possess the ability to induce naïve T cell activation and effector differentiation (35). Once activated, effector T cells undergo proliferation and exhibit an effector-memory phenotype (35). Non-adherent lymphocytes containing 60-80% T cells (the remaining 20-40% were B cells, NK cells, and others) were used as effector T cells. To investigate the changes in population and phenotype of effector T cells after activation by SD-DCs, effector T cells were examined for lymphocyte and T cell subsets by flow cytometry. Inactivated T cells that were not co-cultured with SD-DCs were used as negative control. The results showed no significant differences in lymphocyte subsets, including CD4+ population containing cells with helper activity (CD3+CD4+), CD8+ population containing cells with cytotoxic activity (CD3+CD8+), B cells (CD3-CD19+), and NK cells (CD3-CD16+CD56dim and CD3-CD16-CD56bright), between conditions. On the other hand, slight differences in T cell subsets including naïve T cells (Tn) (CD3+CD45RO-CD62L+), central memory T cells (Tcm) (CD3+CD45RO+CD62L+), effector memory T cells (Tem) (CD3+CD45RO+CD62L-), and T effector cells (Te) (CD3+CD45RO-CD62L-) were observed between inactivated T cells (Tn: 60.20±9.15%, Tcm: 10.15±5.48%, Tem: 10.01±5.53%, and Te: 9.87±3.07%), effector T cells activated by SD-DC-iRFP (Tn: 53.70±8.74%, Tcm: 14.18±5.40%, Tem: 21.33±14.38%, and Te: 21.50±15.33%), and effector T cells activated by SD-DC-BCMA (Tn: 41.33±12.58%, Tcm: 13.84±7.59%, Tem: 24.10±12.83%, and Te: 23.66±14.04%). These results demonstrated slight decrease of the Tn cells and slight increase of the Te cells after the activation by SD-DCs; however, these differences were not statistically significant.
Cytotoxicity of effector T cells against BCMA-expressing MM cells was enhanced by SD-DC-BCMA. To investigate whether effector T cells activated by SD-DC-BCMA could exhibit and enhance specific killing against BCMA-expressing MM cells, inactivated T cells and effector T cells activated by SD-DC-iRFP or SD-DC-BCMA were co-cultured with target cells that expressed different levels of BCMA, including Raji, RPMI8226, and NCI H929 cells, at effector-to-target (E:T) ratios of 2.5:1, 5:1, and 10:1. After 24 h of co-culturing, the cancer killing activity was determined by luciferase assay. The results showed low killing activity of effector T cells activated by SD-DC-BCMA against BCMA-negative Raji cells (Figure 6A). In contrast, specific killing against low and high levels of BCMA-expressing cell lines, RPMI8226 and NCI H929, respectively, in an E:T ratio-dependent manner was demonstrated compared to inactivated T cells and those activated by SD-DC-iRFP (Figure 6B, C). At the highest E:T ratio of 10:1, effector T cells activated by SD-DC-BCMA significantly caused the highest specific lysis in RPMI8226 (69.60±6.34%) and NCI H929 (60.20±16.06%) cells. Culture supernatants at 24-h post-co-culturing of effector T cells and target cells at an E:T ratio of 10:1 were collected for analysis of IFN-γ cytokine production by ELISA. The results showed that IFN-γ produced by effector T cells activated by SD-DC-BCMA was increased in response to the exposure with BCMA-expressing MM cells (RPMI8226: 5,021±2,539 pg/ml, and NCI H929: 7,195±4,143 pg/ml) compared to the exposure with BCMA-negative cell line (Raji: 1,334±1,062 pg/ml) (Figure 6D). Noticeably, in consistence with specific killing observed with RPMI8226 and NCI H929 target cells, significantly higher levels of IFN-γ were detected in the condition using effector T cells activated by SD-DC-BCMA than in the conditions using effector T cells activated by SD-DC-iRFP and in the conditions using inactivated T cells (Figure 6D). These results clearly demonstrated that SD-DC-BCMA could potentially activate effector T cells to elicit specific cytotoxicity against BCMA-expressing MM cells.
Cytotoxicity of effector T cells against BCMA-expressing MM cells after activation by SD-DC-BCMA. Inactivated T cells and effector T cells activated by SD-DC-iRFP or SD-DC-BCMA were co-cultured with target cells that contained different BCMA expression levels at E:T ratios of 2.5:1, 5:1, and 10:1 for 24 h. The cytotoxic activities were determined using luciferase assay. (A) BCMA-negative Raji cells. (B) Low BCMA-expressing RPMI8226 cells. (C) High BCMA-expressing NCI H929 cells. (D) IFN-γ levels of inactivated T cells and effector T cells activated by SD-DC-iRFP or SD-DC-BCMA in response to each target cell measured by ELISA. Culture supernatants of each effector T cell condition after co-culturing with different target cells at an E:T ratio of 10:1 for 24 h were collected and determined for IFN-γ production. The results were summarized from four independent experiments. All data with error bars represents mean±SD. Differences between the data sets were analyzed by Student’s t-test. Asterisks represent p-values: *p<0.05, **p<0.01.
Cytotoxicity of effector T cells was specific to BCMA expression. To further investigate whether the killing ability of effector T cells activated by SD-DC-BCMA was specific to BCMA expression of the target cells or not, BCMA-overexpressing KMS20 and KKU-055 cells were generated from the BCMA-negative KMS20 and KKU-055 cells for the cytotoxicity assays. The effector T cells activated by SD-DC-BCMA were co-cultured with BCMA-negative KMS20 and KKU-055 or BCMA-overexpressing KMS20 and KKU-055 target cells at E:T ratios of 2.5:1, 5:1, and 10:1 for 8 h. Then, the cytotoxicities of BCMA-negative/BCMA-overexpressing KMS20 and KKU-055 cells were determined by annexin V/PI staining and crystal violet staining, respectively. The results showed that the effector T cells activated by SD-DC-BCMA, but not the inactivated T cells or the effector T cells activated by SD-DC-iRFP, specifically killed BCMA-overexpressing KMS20 and KKU-055 target cells, compared to the BCMA-negative KMS20 and KKU-055 target cells, at both E:T ratios of 5:1 and 10:1 (Figure 7A–D). The percentages of cell death induced by the effector T cells activated by SD-DC-BCMA at the highest E:T ratio of 10:1 were 51.16±7.93% and 51.05±11.72% in the BCMA-overexpressing KMS20 and KKU-055 target cells, respectively, compared to 24.43±3.97% and 17.27±11.36% in the BCMA-negative KMS20 and KKU-055 target cells, respectively (Figure 7A–D). These results indicated that the cytotoxic responses of the effector T cells activated by SD-DC-BCMA were specific to BCMA expression of the target cells.
Cytotoxicity of effector T cells against BCMA-overexpressing target cells. Inactivated T cells, effector T cells activated by SD-DC-iRFP, or effector T cells activated by SD-DC-BCMA were co-cultured with BCMA-negative KMS20 or BCMA-overexpressing KMS20 target cells, and also co-cultured with BCMA-negative KKU-055 or BCMA-overexpressing KKU-055 target cells at E:T ratios of 2.5:1, 5:1, and 10:1 for 8 h. The cytotoxic activities were determined by annexin V/PI staining for BCMA-negative KMS20 and BCMA-overexpressing KMS20 target cells, and by crystal violet staining for BCMA-negative KKU-055 and BCMA-overexpressing KKU-055 target cells. (A) BCMA-negative KMS20 target cells. (B) BCMA-overexpressing KMS20 target cells. (C) BCMA-negative KKU-055 target cells. (D) BCMA-overexpressing KKU-055 target cells. The results were obtained from four independent experiments. The histograms with error bars represent mean±SD. Differences between the data sets were analyzed by Student’s t-test. Asterisks represent p-values: *p<0.05, **p<0.01, ***p<0.001.
Discussion
An important challenge of MM is that it is an incurable disease with high relapse and refractory rates. Although adoptive T cell therapy exhibited promising results in MM (6), further research and development are still required to achieve well-optimized protocols for clinical applications. The efficiency of SD-DCs to activate effector T cells for antigen-specific cancer killing has been reported by our group and others (18–23). The present study, thus, aimed to develop a SD-DC approach for adoptive T cell therapy targeting MM. Since selection of an appropriate antigen for each cancer is crucial for achieving success in adoptive T cell therapy, BCMA was selected as a potential target for MM due to its persistent expression throughout the disease course, and it has only limited expression in normal tissues (24–27) except normal plasma cells. Moreover, targeting BCMA by various immunotherapeutic strategies, including monoclonal antibodies, vaccine, and chimeric antigen receptor (CAR) T cells, demonstrated efficient therapeutic effects in MM patients - even in the patients with relapsed/refractory MM (RRMM) (13, 36–39). Low expression of BCMA in normal plasma cells might raise significant concern for safety of BCMA-targeting therapies. However, BCMA is not expressed in other early B cell developmental stages, and there was no abnormal B cell development and humoral immune defect in BCMA knockout mice (30, 40, 41). Even though the absence of normal plasma cells and low immunoglobulin level after anti-BCMA CAR T cell infusion were reported in some patients, supplemental immunoglobulin can be applied (42). BCMA-targeting therapy therefore represents an effective approach with controllable safety.
In this study, we generated SD-DC expressing BCMA antigen (SD-DC-BCMA) to potentiate the cytotoxicity of effector T cells against MM. Our study demonstrated that BCMA is highly expressed in MM cell lines, RPMI8226 and NCI H929, but it was not expressed in Raji – B lymphocytic cell line (Figure 2). Apart from expression, we also showed that BCMA was degraded via the proteasome pathway (Figure 3C, D), which supported the use of BCMA in SD-DC-BCMA generation, and it ensured the possibility of antigen processing for MHC class I presentation in CD8+ T cell activation (34). Since SD-DC-BCMA expressed the full-length of BCMA protein, its distinct T-cell epitope peptides were able to be presented on multiple MHC molecules. Our preliminary study identified HLA-A*11:01-restricted BCMA epitopes, which were able to stimulate specific T cell responses (data to be published).
The morphologies and immunophenotypes of the generated SD-DC-iRFP or SD-DC-BCMA were similar to ConvDCs that were generated by using standard recombinant cytokines (Figure 4E and Figure 5A, B). Similar lentiviral VCN and levels of GM-CSF and IL-4 in SD-DC-iRFP and SD-DC-BCMA indicated that these two SD-DCs were comparable (Figure 4B–D). Although the amount of GM-CSF and IL-4 produced by both SD-DCs was approximately 13-20 times lower than that of the cytokines used in ConvDC generation, these amounts were adequate for DC differentiation (18–23). In consistent with previous SD-DC reports (21–23), more than 70-80% of monocytes could be differentiated to be SD-DCs as evidenced by the reduction of CD14-positive cells (Figure 5A, B).
Although a change in lymphocyte subsets was not observed between conditions, increased trends of memory and effector T cell subsets were observed after effector T cells were activated by SD-DCs. The differentiation of naïve T cells into memory and effector T cells indicated the features of activated T cells (35). Our previous study reported the ability of SD-DC-PRKAR1A to produce IL-12, which is a signature cytokine produced from mature DCs for T cell activation (23, 43). Full activation signals provided by SD-DCs, including antigen presentation via MHC molecules, co-stimulatory molecules, and secretary cytokines, leads to the development of memory and effector populations of T cells (44).
Enhanced cytotoxicity of effector T cells after SD-DC-BCMA activation was then demonstrated. Killing of BCMA-expressing MM cells by effector T cells activated by SD-DC-BCMA was demonstrated in both low BCMA-expressing RPMI8226 and high BCMA-expressing NCI H929 cells in an E:T ratio-dependent manner (Figure 6B, C). At the highest E:T ratio of 10:1, there was a two-fold increase in killing ability of effector T cells activated by SD-DC-BCMA compared to those of inactivated T cells and effector T cells activated by SD-DC-iRFP. These increased killing abilities are consistent with the previous results of effector T cells activated by SD-DC-PRKAR1A in killing CCA cells (23). In addition, the most robust IFN-γ cytokine production was specifically observed in the condition using effector T cells activated by SD-DC-BCMA upon exposure to BCMA-expressing target cells, indicating antigen-specific T cell responses (Figure 6D). Interestingly, effector T cells activated by SD-DC-BCMA showed slightly greater killing capacity in low BCMA-expressing RPMI8226 than in high BCMA-expressing NCI H929 cells. Since RPMI8226 cells were obtained from a patient at the early stage of diagnosis of MM, whereas NCI H929 cells were obtained from a patient who had relapsed, other different phenotypic profiles of each cell line apart from antigen expression may affect killing efficiency (45, 46). Moreover, upregulation of MM-related immune inhibitory molecules, including programmed cell death ligand-1 (PD-L1), interleukin-10 (IL-10), and transforming growth factor-β (TGF-β), was found to be induced by BCMA overexpression and may lead to T cell suppression (47). Expression of PD-L1, IL-10, and TGF-β was previously reported in both RPMI8226 and NCI H929 cells (47–51); however, further experiments to compare expression levels of these molecules in the same batch, and to investigate their effects on killing efficiency are required to validate this hypothesis.
Non-specific killing of effector T cells activated by SD-DC-iRFP to each target MM cells, and effector T cells activated by SD-DC-BCMA to BCMA-negative Raji cells (Figure 6A–C) may be explained by human leukocyte antigen (HLA) mismatch between effector T cells and target MM cells. Since the three target cells have different HLA profiles, the mismatch of HLA molecules between effector T cells derived from healthy donors and target cancer cells was inevitable, so non-specific killing could occur. We have performed HLA class I typing in all donors to match the HLA between the donors and the target cancer cells. However, polymorphism of HLA made it difficult to find fully HLA-matched unrelated donors; only partially HLA class I-matched donors were selected to conduct the experiments in this study. To further investigate BCMA-specific killing of the effector T cells activated by SD-DC-BCMA to the target cancer cells, we conducted the killing assays of the effector T cells activated by SD-DC-BCMA on the BCMA-overexpressing KMS20 and KKU-055 cells, compared to the BCMA-negative KMS20 and KKU-055 cells. Our results demonstrated that the effector T cells activated by SD-DC-BCMA specifically killed the BCMA-overexpressing KMS20 and KKU-055 target cells but did not effectively kill the BCMA-negative KMS20 and KKU-055 target cells (Figure 7A–D). These results strongly supported the specific cytotoxicity of the effector T cells activated by SD-DC-BCMA to the BCMA-expressing target cancer cells.
The advantages of ex vivo activation of effector T cells by SD-DC-BCMA are that multiple T cell clones can be generated leading to broad antitumor T cell responses and decreased tumor immune evasion. Moreover, adoptive transfer of these non-engineered T cells would be effective and safe for clinical use in the MM patients. Nevertheless, the efficacy of T cells activated by SD-DC-BCMA may be limited by other tumor escape mechanisms including antigen loss, MHC downregulation, and upregulation of T cell inhibitory molecules. The combination of this approach with current standard therapies or other immune-targeted therapies would enhance the efficacy and durability for MM treatment.
In conclusion, we present the development of SD-DC-BCMA for T cell activation, which is capable of enhancing antigen-specific cancer killing against MM cells. Our findings provide a framework for therapeutic application of adoptive T cell therapy in MM.
Acknowledgements
This work was financially supported by the TRF-International Research Network (TRF-IRN) (grant number IRN58W0001), the Center of Excellence on Medical Biotechnology (CEMB), S&T Postgraduate Education and Research Development Office (PERDO), Office of Higher Education Commission (OHEC), Thailand (grant number CB-61-006-01), the Siriraj Research Fund of the Faculty of Medicine Siriraj Hospital, Mahidol University (grant number R016034008), and Mahidol University (Basic Research Fund: Fiscal Year 2022, grant number BRF1-029/2565). WC was supported by a TRF-IRN Scholarship (scholarship number IRN5801PHDW05) and a Siriraj Graduate Scholarship. PL was supported by the National Research Council of Thailand (NRCT) (grant number N41A640177). AP was supported by Office of the Permanent Secretary, Ministry of Higher Education, Science, Research and Innovation (Grant no. RGNS 63-068). NJ was supported by a Siriraj Graduate Scholarship. MJ was supported by a TRF Grant for New Researcher (grant number TRG5780173) and a Siriraj Chalermprakiat Grant. PY was supported by a Siriraj Chalermprakiat Grant. The authors gratefully acknowledge Prof. Peter Hokland, Department of Clinical Medicine, Aarhus University, Aarhus, Denmark, for manuscript review and suggestions; Asst. Prof. Kevin Jones for professional language editing; Dr. Jatuporn Sujjitjoon, Dr. Chutamas Thepmalee, Ms. Nunghathai Sawasdee, and Ms. Yupanun Wutti-in for technical advice and laboratory assistance.
Footnotes
Authors’ Contributions
WC, PL, AP, TC, MJ, and PY conceptualized and designed the study. WC optimized and conducted experiments, collected and analyzed data, and drafted the manuscript. PL and NJ partly assisted in performing experiments and data analysis. WC, PL, AP, PP, TC, MJ, and PY interpreted data and revised the manuscript. All Authors read and approved the final version of manuscript for publication.
Conflicts of Interest
The Authors declare no conflicts of interest.
- Received November 4, 2021.
- Revision received January 15, 2022.
- Accepted February 10, 2022.
- Copyright © 2022 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
