Hematopoiesis by iPSC-derived hematopoietic stem cells of aplastic anemia that escape cytotoxic T-cell attack

Abstract

Hematopoietic stem cells (HSCs) that lack HLA-class I alleles as a result of copy-number neutral loss of heterozygosity of the short arm of chromosome 6 (6pLOH) or HLA allelic mutations often constitute hematopoiesis in patients with acquired aplastic anemia (AA), but the precise mechanisms underlying clonal hematopoiesis induced by these HLA-lacking (HLA^-) HSCs remain unknown. To address this issue, we generated induced pluripotent stem cells (iPSCs) from an AA patient who possessed HLA-B4002-lacking (B4002^-) leukocytes. Three different iPSC clones (wild-type [WT], 6pLOH⁺, and B*40:02-mutant) were established from the patient's monocytes. Three-week cultures of the iPSCs in the presence of various growth factors produced hematopoietic cells that make up 50% to 70% of the CD34⁺ cells of each phenotype. When 10⁶ iPSC-derived CD34⁺ (iCD34⁺) cells with the 3 different genotypes were injected into the femoral bone of C57BL/6.Rag2 mice, 2.1% to 7.3% human multilineage CD45⁺ cells of each HLA phenotype were detected in the bone marrow, spleen, and peripheral blood of the mice at 9 to 12 weeks after the injection, with no significant difference in the human:mouse chimerism ratio among the 3 groups. Stimulation of the patient's CD8⁺ T cells with the WT iCD34⁺ cells generated a cytotoxic T lymphocyte (CTL) line capable of killing WT iCD34⁺ cells but not B4002^- iCD34⁺ cells. These data suggest that B4002^- iCD34⁺ cells show a repopulating ability similar to that of WT iCD34⁺ cells when autologous T cells are absent and CTL precursors capable of selectively killing WT HSCs are present in the patient's peripheral blood.

Graphical abstract

Figure 1.

Establishment of iPSCs with different HLA genotypes from the monocytes of a patient with acquired AA. (A) HLA-A*24:02–lacking cells in the different lineages of cells. Histograms of HLA-A*24:02-expressing cells in different lineages of the patient (green lines) and a healthy individual (blue lines) are shown. (B) HLA-B4002 expression in monocyte fractions of the patient. Three different populations including A24⁺B4002⁺ (WT), A24⁺B4002⁻ (B*40:02^mut), and A24⁻B4002⁻ (6pLOH⁺) monocytes are shown. (C) Results of TaqMan PCR on 14 iPSC clones showing 10 6pLOH⁺ iPSC clones (red dots) and 4 6pLOH⁻ iPSC clones (green dots) that include 1 B4002⁻ clone; the blue dots represent control DNA. The x-axis and y-axis represent the amplified product amounts of B*40:02 and B*35:01, respectively. (D) Next-generation sequencing findings for B4002⁻ iPSC clone G2 showing a start loss mutation in B*40:02. The results of the HLA-B allelic sequencing of WT iPSC clone E3 is shown as a control. The green bar represents the start loss mutation locus of the B*40:02^mut clone. (E) The percentages of WT, B*40:02^mut, and 6pLOH⁺ iPSC clones are shown.

Figure 2.

Induction of HSC differentiation from iPSCs using different culture systems. (A) Representative FCM results showing the expression of various surface markers on iPSCs obtained by coculturing WT iPSC clone G1 with OP9 cells for 21 days. (B) The percentages of SSEA-4^–, CD34⁺, CD108⁺, CD184⁺, and CD45⁺ expressed by iPSCs during the differentiation process of iPSCs cultured with OP9 cells. The percentages represent the mean ± standard deviation (SD) of the data from 5 experiments using 3 different WT iPSC clones analyzed at 7, 14, and 21 days of culture. (C) Changes in the SSEA-4 expression of WT clone G1 with time after culture. (D) HSC- and iPSC-related gene expression patterns of iCD34⁺ cells. Total RNA extracted from original iPSC clones (WT and 6pLOH clones), iCD34⁺ cells induced from the same iPSCs, CD34⁺ cord blood cells, and a stem cell-like acute monoblastic leukemia (AML) cell line (AML-MT) was subjected to an RT-PCR analysis for 4 genes: HoxA9, CD45, CD41, and OCT3. (E) Representative images of hematopoietic cells induced from iPSCs in various hematopoietic differentiation media (StemPro-34 serum-free medium [SFM] and OP9, WEHI, and OP9 and WEHI conditioned media). Images were captured by using a light microscope at day 14. (F) The expression of cell surface markers on iPSCs cultured under the 4 different methods. Cells were harvested and analyzed by FCM at the indicated time points (0, 7, 14, 21, and 28 days). The data represent results from 2 independent experiments and are expressed as the mean ± standard error of the mean (SEM) of the percentage of cells expressing the given marker. FSC-W, forward scatter width; SFM, serum-free medium; SSC-H, side scatter height; UCB, umbilical cord blood.

Figure 3.

Characterization of HSPs derived from iPSCs with different HLA genotypes. (A) Expression of A*02:01 and A*24:02 by WT and 6pLOH⁺ iCD34⁺ cells. WT and 6pLOH⁺ iPSCs cultured with OP9 cells for 21 days were examined for HLA-A24 and A2 expression on the gated CD34⁺ cells. (B) HSC induction from WT and 6pLOH⁺ iPSCs and HLA expression by the iCD34⁺ cells. Cultured WT (blue column) and 6pLOH⁺ (red column) with OP9 cells were analyzed for the percentage of CD34⁺ cells and HLA expression by the gated CD34⁺ cells. The data show the mean ± SEM from 3 independent experiments. (C-D) B4002 expression by CD34⁺ cells derived from 3 different iPSCs with different HLA genotypes. iPSCs cultured with a feeder system (OP9 cells) for 21 days were examined for the expression of A2402 and B4002. HSCs derived from a B4002⁻ iPSC clone lacked B4002 but retained A2402 as expected. A representative set of (C) scattergrams and (D) the percentages of CD34⁺ cells and HLA-A allele⁺ cells are shown. The columns represent the mean ± SEM of the values determined by FMC. (E) Hematopoietic colony-forming capacities of iCD34⁺ cells. The pie chart shows CFU, granulocyte, erythrocyte, megakaryocyte, macrophage (CFU-GEMM), CFU-granulocyte (CFU-G), CFU, granulocyte-macrophage (CFU-GM), CFU, macrophage (CFU-M), and burst-forming unit erythroid (BFU-E)–derived colonies generated from the WT-iCD34⁺ cells. (F) The plating efficiencies of iCD34⁺ cells derived from (left panel) 3 WT iPSC clones are compared among (middle panel) 3 different feeder-free systems with StemPro or CM from OP9 or WEHI cells. (Right panel) Summary of the plating efficiencies of iCD34⁺ cells with 3 different HLA genotypes. The data indicate the mean ± SEM of the CFU percentages obtained from 3 independent experiments. The plating efficiency was defined as the frequency of colonies generated from 5000 iCD34⁺ seeded cells (total number of colonies per 5 × 10³ cells seeded). *P < .05; **P < .01; ***P < .001. NS, not significant; SSC-W, side scatter width.

Figure 4.

Engraftment of AA iPSC WT, B*40:02^mut, and 6pLOH⁺HSCs (CD34⁺iPSC HSPCs) in immunodeficient mice. (A) Human CD34⁺ and CD45⁺ cells in different organs of young BRGS mice at 10 to 12 weeks after transplantation of WT, B*40:02^mut and 6pLOH⁺ CD34⁺ iPSCs. (Left panel) The percentages of human hematopoietic cells represent the mean ± SEM of 3 WT, 2 B*40:02^mut, and 3 6pLOH⁺ cell recipients. (Right panel) A representative set of FCM results is shown. The numbers below the gated area indicate the mean ± SEM of the positive cell percentages. (B) Human myeloid cells in the BM of a young mouse transplanted with 6pLOH⁺ iCD34⁺ cells. Representative results of (upper panel) hematoxylin and eosin staining and (lower panel) human myeloperoxidase staining are shown. (C) Engraftment of myeloid and lymphoid lineage cells in the different organs. The percentages of each cell lineage in (top left panel) the BM, (top right panel) spleen, and (bottom panel) PB are shown. The data represent the mean ± SEM of the percentages of each marker-positive cell obtained from 3 WT, 2 B*40:02^mut, and 3 6pLOH⁺ cell recipient mice. (D) HLA expression profiles of human B cells isolated from the spleen of the transplanted mice with the indicated iCD34⁺, (upper panels) HLA-A2402, and (lower panels) HLA-B4002 expression by CD19⁺ lymphoid cells in the spleens. (E) Morphology of CFU-GEMM–derived colonies from the BM iCD34⁺ cells. (Left) CFU-GEMM–derived colonies and (right) the constituent cells prepared by the cytospin are shown. FSC-A, forward scatter area; FSC-H, forward scatter height.

Figure 5.

Escape of B4002⁻iCD34⁺cells from the attack by autologous CTLs specific to B4002⁺iCD34⁺cells. (A) CTLs were cocultured with iCD34⁺ cells possessing either the 6pLOH⁺ or the WT genotypes in 96-well round-bottom plates for 5 days and visualized using a light microscope. Representative images of experiments are shown; black arrows indicate iCD34⁺ cells; white arrows indicate CTLs. Top panels, original magnification ×60; bottom panels, original magnification ×20. (B) CTLs were cocultured with iCD34⁺ cells possessing either the 6pLOH⁺ or the WT genotypes in 96-well round-bottom plates at a 10:1 effector:target (E:T) ratio for 4 hours and then analyzed by FCM. The analysis was performed by gating cells as 7-aminoactinomycin D–negative (live cells), CD3⁺, CD8⁺, or CD107a⁺ cells. (C) CTLs were cocultured in 96-well round-bottom plates with iCD34⁺ cells with 6pLOH⁺, WT, or B*40:02^mut genotypes at the indicated E:T ratios. The levels of IFN-γ in the supernatants from overnight cocultures were determined by enzyme-linked immunosorbent assay. Summarized data (mean ± SEM) from 3 independent experiments are shown. (D) iCD34⁺ cells possessing the 6pLOH⁺, B*40:02^mut, or WT genotypes were labeled with ⁵¹Cr and then cocultured with autologous CTLs in 96-well round-bottom plates for 4 hours. Summarized data (mean ± SEM) from 3 independent experiments are shown. (E) Clonotype analysis (Vβ cDNA sequencing) of the CTL line and BM T cells of the patient before receiving immunosuppressive therapy using single T-cell RT-PCR indicate that the third-most recurrent TCRVβ CDR3 sequence of the CTL line was identical to that of the most predominant T-cell clone in the BM. (F) CDR3α sequence of T cells with the same CDR3β sequence derived from the CTL line and BM T cells. The T cell expressed 2 TCRα chain mRNAs; 1 was functional and the other was unproductive. *P < .05.