Broadly applicable TCR-based therapy for multiple myeloma targeting the immunoglobulin J chain

Abstract

Background: The immunoglobulin J chain (Jchain) is highly expressed in the majority of multiple myeloma (MM), and Jchain-derived peptides presented in HLA molecules may be suitable antigens for T-cell therapy of MM.

Methods: Using immunopeptidomics, we identified Jchain-derived epitopes presented by MM cells, and pHLA tetramer technology was used to isolate Jchain-specific T-cell clones.

Results: We identified T cells specific for Jchain peptides presented in HLA-A1, -A24, -A3, and -A11 that recognized and lysed JCHAIN-positive MM cells. TCRs of the most promising T-cell clones were sequenced, cloned into retroviral vectors, and transferred to CD8 T cells. Jchain TCR T cells recognized target cells when JCHAIN and the appropriate HLA restriction alleles were expressed, while JCHAIN or HLA-negative cells, including healthy subsets, were not recognized. Patient-derived JCHAIN-positive MM samples were also lysed by Jchain TCR T cells. In a preclinical in vivo model for established MM, Jchain-A1, -A24, -A3, and -A11 TCR T cells strongly eradicated MM cells, which resulted in 100-fold lower tumor burden in Jchain TCR versus control-treated mice.

Conclusions: We identified TCRs targeting Jchain-derived peptides presented in four common HLA alleles. All four TCRs demonstrated potent preclinical anti-myeloma activity, encouraging further preclinical testing and ultimately clinical development.

Fig. 1

Jchain target identification and subsequent selection of promising T-cell clones recognizing Jchain peptides in HLA-A1, A24, A3 and A11. A JCHAIN microarray data (probe ILMN_2105441) from an Illumina HT12.0 microarray dataset (GSE76340) displaying gene expression levels in mean fluorescence intensities (MFI) in healthy tissue of non-hematopoietic origin (in gray), healthy tissue of hematopoietic origin (in blue), healthy subsets containing B cells (yellow) and B- and plasma cell malignancies (in red). Per healthy tissue gene expression was measured in 2–7 samples (mean; 3.4). Abbreviations: ALL, acute lymphoblastic leukemia; CLL, chronic lymphocytic leukemia; MM, multiple myeloma; EBV-LCL, Epstein–Barr virus-transformed lymphoblastoid cell lines; PBMC, peripheral blood mononuclear cells; BMMC, bone marrow mononuclear cells; HPC, hematopoietic precursor cells; DC, dendritic cells; MΦ1, type 1 macrophages; MΦ2, type 2 macrophages; IFN-γ, interferon-γ; HUVEC, human umbilical vein endothelial cells; PTEC, proximal tubular epithelial cells. B Violin plot of JCHAIN expression (top graph) in 133 MM samples and 5 healthy plasma cell (PC) samples. Samples were enriched for CD138; therefore, CD138 (SCD1) expression is displayed as positive control (bottom graph). Data were extracted from publicly available dataset GSE13591. Dashed lines indicate background level for negative expression values. Dotted line indicates an arbitrary cutoff for samples expressing JCHAIN (expression value > 1000). C Example of verification of peptide identification by mass spectrometry. Tandem mass spectra of ISDPTSPLRTR peptide eluted from HLA-A3 positive U266 MM cells as well as HLA-A11 positive UM9 MM cells (top graph). Corresponding tandem mass spectra of synthetic ISDPTSPLRTR peptide (bottom graph). D, E IFN-γ production by Jchain-specific T cells clones after overnight co-culture measured by ELISA. Graphs are separated based on peptide-HLA specificities. Averages of technical duplicates are depicted. D Jchain T-cell clones co-cultured with K562 target cells transduced with target HLA alleles (-A1, -A24, -A3, or –A11) without or with additional transduction of the JCHAIN gene (+ Jchain). E T-cell clones overnight stimulated with antigen negative K562 cells transduced with target HLA molecules HLA-A1, -A24, -A3 or -A11 loaded with decreasing concentrations of Jchain peptides

Fig. 2

Investigation of cross-reactivity by Jchain-specific T-cell clones. IFN-γ production by T-cell clones measured by ELISA after overnight co-culture, technical duplicates are depicted. Graphs are separated based on peptide-HLA specificities. A T-cell clones were stimulated with a panel of cell lines of non-B cell origins transduced with target HLA (+ A1, A24, A3 or A11) or naturally expressing target HLA molecules (A1^pos, A24^pos, A3^pos or A11^pos). HLA-transduced K562 cells were included as negative control (neg ctrl), target gene and HLA-transduced K562 cells were included as positive control for T-cell function (pos ctrl). THP-1 A24 was not recognized by allo-HLA-A24 T-cell clone; therefore, this cell line was excluded from data as stimulatory capacity is lacking. B T-cell clones were stimulated with an EBV-LCL panel containing EBV-LCLs expressing HLA-I alleles with an allele frequency over 1% that do not express target HLA alleles. Controls as in A

Fig. 3

Killing of MM cell line U266 by Jchain-specific T-cell clones. A Killing of JCHAIN expressing U266 cells transduced with HLA-A1 or A24 by Jchain A1 and A24 specific T-cell clones in a 6-h 51Cr release assay (top graphs). Killing of WT (HLA-A3 positive) or HLA-A11-transduced U266 cells by Jchain HLA-A3 or A11-specific T-cell clones (bottom graphs). Killing assays were performed using E/T ratios 10:1 and 1:1. Allo-HLA T-cell clones, recognizing peptides derived from housekeeping proteins in specific HLA alleles, were included as positive controls. Average values and standard deviations of technical triplicates are shown. B IFN-γ production measured by ELISA after overnight co-culture using the same target cells as in A in an E/T of 1:6. Values and means of technical duplicates are shown

Fig. 4

Functionality of Jchain TCRs in CD8 and CD4 T cells. A–C Jchain A1, A24, A3 and A11 restricted TCRs were sequenced and introduced with murine constant beta domains (mTCR) into CD4 and CD8 T cells. After mTCR enrichment functionality was assessed. A TCR T cells were stained with the respective Jchain pHLA tetramers and analyzed by FACS. TCR T cells were gated on mTCR +. Parental T-cell clones were included as positive controls and CMV TCR T cells were included as negative controls. B Endogenous recognition of JCHAIN and HLA (A1, A24, A3, or A11) transduced K562 cells by Jchain TCR CD4 and CD8 T cells. HLA only transduced K562 cells were included as negative control. Values and means of technical duplicates are shown. C TCR-transduced CD8 T cells were used for 6-h chromium release assays to study target cell lysis in E/T ratios 10:1 and 1:1. T cells were co-cultured with U266 MM cells, UM9 MM cells or antigen negative K562 cells transduced with (+ HLA) or naturally expressing (HLA^pos) target HLA molecules. CMV TCR T cells were used as a negative control. Values and means of technical triplicates are shown

Fig. 5

Recognition of healthy hematopoietic and non-hematopoietic subsets by Jchain TCR-transduced CD8 T cells. A IFN-γ production after overnight co-culture of Jchain TCR Td CD8 T cells with CD40L activated B cells, immature dendritic cells (immDCs), mature dendritic cells (mDCs), PHA-activated T cells (PHA T cells), and keratinocytes or fibroblasts pre-treated for 48 h with 100 IU/ml IFN-γ. Symbols represent the average value (from technical duplicates) of target cells isolated from different donors. Target cells not expressing the relevant HLA restriction allele are depicted in gray, cells expressing the HLA restriction alleles are depicted in color. Per panel T cells with one of the Jchain TCRs are shown as indicated in the graph titles. K562 + HLA and peptide loaded K562 + HLA are included as negative and positive controls. B, C FACS-based killing experiment of CD40L activated peripheral blood B cells from healthy donors with Jchain A1 and Jchain A24 TCR T cells in an E/T ratio of 3:1, samples were measured using fixed acquisition times and fixed flow rates. B Example of B cell survival and killing after overnight co-culture of HLA-A1^neg/HLA-A24^pos B cells with CMV TCR CD8 T cells (left), Jchain A1 TCR T cells (middle) and Jchain A24 TCR T cells (right). Gated on SYTOX blue-, single cells, CD3-, CD19 + . C Quantification of percentage surviving B cells of data in B and additional donors, HLA-A1 and -A24 typing is indicated in graph titles. Percentage surviving cells was calculated relative to cells in the negative control CMV TCR T-cell culture. Technical triplicates are shown

Fig. 6

Killing of MM cells in patient-derived bone marrow samples. Killing of patient-derived bone marrow MM samples was assessed by FACS-based cytotoxicity experiments in which TCR Td T cells were co-cultured with patient BM samples in an E/T ratio 3:1. MM cell survival was analyzed after overnight co-culture. A Example of survival of an HLA-A1^pos/A24^pos patient sample after co-culture with CD8 T cells transduced with a CMV (negative control), Jchain HLA-A1 or Jchain HLA-A24 restricted TCR. In red MM cells are displayed, highlighted by black boxes for clarity, MM cells were gated on: live cells → single cells → CD3-negative cells to exclude co-cultured T cells → CD45 negative-intermediate, CD19 negative → CD56 positive, CD38 positive. MM cells (in red) were backgated on total CD3-negative cells (in blue). B MM.J1-MMJ.10 codes in graph titles represent different MM patients, additionally expression of target HLA molecules and JCHAIN expression as a fold increase relative to housekeeping genes is displayed. JCHAIN expression was measured by qPCR on sorted MM cells. Numbers of surviving MM cells acquired per 2500 Flow-Count Fluorospheres are displayed. Jchain HLA-A1 and HLA-A24 TCR Td T CD8 T cells co-cultured with MM patient samples from different individuals expressing HLA-A1, A24 or both (top row). Jchain HLA-A3 and HLA-A11 TCR Td T CD8 T cells co-cultured with MM patient samples from different individuals expressing HLA-A3 or A11 (bottom row). Technical triplicates are shown. Data representative of two independent experiments

Fig. 7

In vivo anti-tumor efficacy of Jchain HLA-A1, A24, A3 and A11 restricted TCR-transduced CD8 T cells. NSG mice engrafted with 2 × 10⁶ U266 multiple myeloma cells transduced with Luc2 luciferase and HLA-A1, -A11, -A24 or wildtype (WT) were i.v. injected with 3–6 × 10⁶ TCR-transduced CD8 T cells after 21 days. CD8 T cells were separately transduced with Jchain HLA-A1 (5D12), -A3 (5C8), -A11 (16C7), -A24 (A24) or control CMV (pp65-NLV-HLA-A2) TCR and enriched for mTCR expression by MACS. T cells were infused 10 days after re-stimulation. Tumor outgrowth was frequently tracked by bioluminescence imaging. Data representative of two independent experiments. A Raw bioluminescent images of Jchain HLA-A1 TCR-treated mice and control CMV TCR-treated mice. B Mean and standard deviations of tumor signal (average radiance) over time on the ventral side in Jchain HLA-A1 TCR-treated (upper left), Jchain HLA-A3 TCR-treated (lower left), Jchain HLA-A24 TCR-treated (upper right) and Jchain HLA-A11 TCR-treated (lower right) mice compared to CMV TCR control mice. Statistics depict two-way ANOVA comparing groups per timepoint with Sidak’s multiple comparisons post hoc test