T-cell receptor-driven lymphomagenesis in mice derived from a reprogrammed T cell

Abstract

The conversion of mature somatic cells into pluripotent stem cells, both by nuclear transfer and transduction with specific "reprogramming" genes, represents a major advance in regenerative medicine. Pluripotent stem cell lines can now be generated from an individual's own cells, facilitating the generation of immunologically acceptable stem cell-based therapeutics. Many cell types can undergo nuclear reprogramming, leading to the question of whether the identity of the reprogrammed cell of origin has a biological consequence. Peripheral blood, containing a mixture of T, B, NK, and myeloid cell types, represents one potential source of reprogrammable cells. In this study, we describe the unique case of mice derived from a reprogrammed T cell. These mice have prerearranged T-cell receptor (TCR) genes in all cells. Surprisingly, ≈50% of mice with prerearranged TCR genes develop spontaneous T cell lymphomas, which originate in the thymus. The lymphomas arise from developing T cells, and contain activated Notch1, similar to most human and mouse T-cell acute lymphoblastic lymphomas. Furthermore, lymphomagenesis requires the expression of both prerearranged TCRα and TCRβ genes, indicating a critical role for TCR signaling. Furthermore, inhibitors of multiple branches of TCR signaling suppress lymphoma growth, implicating TCR signaling as an essential component in lymphoma proliferation. The lymphomagenesis in mice derived from a reprogrammed T cell demonstrates the deleterious consequences of misregulation of the TCR rearrangement and signaling pathways and illustrates one case of cellular reprogramming where the identity of the cell of origin has profound consequences.

Fig. 1.

LN3 lymphomas are located in lymphoid organs and have an immature thymocyte phenotype. (A) Surface phenotype of nonleukemic (Left) and three separate leukemic thymuses stained for molecules normally associated with T-cell development, indicating that the individual lymphomas have heterogeneous phenotypes that partially overlap with several stages of immature thymocyte development. (B) Offspring from a cross of wt × LN3 mice were analyzed over a 7-mo period for onset of lymphoma. Mice that showed signs of illness were euthanized, and lymphomas were verified by dissection as greatly enlarged thymuses, spleens and/or lymph nodes. Lymphomas were verified by flow cytometry. LN3 mice contain both prerearranged TCRα and TCRβ; LN3α contain only the prerearranged TCRα; LN3β contain only the prerearranged TCRβ; wt indicates littermates that lacked prerearranged TCR loci; LN3 β2m^−/− mice do not express surface MHCI molecules; LN3 I-A^b−/− mice lack MHCII molecules; and LN3 RAG^−/− mice are RAG2-deficient. In all crosses studied, only mice that expressed both the prerearranged TCRα and TCRβ chains developed lymphomas. Statistical P values were calculated by using Fisher's exact test.

Fig. 2.

LN3 lymphomas have activated Notch and NFAT signaling pathways. (A) Western blots were performed on cell lysates from 17 individual LN3 lymphomas, using an antibody that recognizes the activated (cleaved) form of Notch. Fifteen of 17 LN3 lymphomas stained positive for the cleaved form of Notch1 (indicated by arrow), showing that this pathway is activated in the majority of lymphomas derived from LN3 mice, regardless of RAG expression (RAG^−/−) or β2M and IA^b expression (MHC^−/−). (B) NFAT-eGFP is partially nuclear localized in untreated lymphoma cells. Two independent LN3αβ lymphoma-derived cell lines (LN3A and LN3B) were transduced with a retrovirus encoding an NFAT-eGFP fusion protein. Representative data are shown from one of these lines (LN3B). Cells were either left untreated, or treated in culture for 30 min with PMA (1 μM) + ionomycin (250 ng/mL) or cyclosporin A (10 μM), and then plated onto glass slides with coverslips and immediately analyzed for on a fluorescence microscope. Untreated cells have a partially nuclear localized NFAT-eGFP, as compared to cells that have almost entirely nuclear NFAT-eGFP (PMA/ionomycin treated) and cells that have almost entirely cytoplasmic NFAT-eGFP (cyclosporin A treated).

Fig. 3.

Inhibitors of Notch1 activation, NFAT signaling, LCK, and PKCθ all block LN3 lymphoma proliferation. Four independent LN3 lymphoma cell lines: RAGLN3 #64 (white bars), RAGLN3 #31 (light gray bars), LN3 A (dark gray bars), LN3 B (black bars), and one B cell hybridoma (dotted bars, as a control for nonspecific toxicity) were plated at 5,000 cells per well and were treated with a γ-secretase inhibitor (A), cyclosporin A (B), an Lck inhibitor (C), or a PKCθ inhibitor (D). Three days later, live cells were counted by flow cytometry, and percent growth was calculated as: 100 × (number of cells in treated well) / (number of cells in untreated wells). Results are representative of at least four independent experiments for each inhibitor and each cell line. A titration of inhibitors were used in each assay, and concentrations that did not affect B cell hybridoma growth were chosen where possible to ensure that the inhibitors were not acting through nonspecific toxicity (this control for toxicity was not possible for cyclosporin A, which is known to also inhibit B-cell lymphoma growth).