Apoptosis of Hematopoietic Stem Cells Contributes to Bone Marrow Suppression Following Chimeric Antigen Receptor T Cell Therapy

Abstract

Chimeric antigen receptor (CAR) T cell (CAR-T) therapy represents a revolutionary treatment for patients with relapsed/refractory hematologic malignancies. However, its use can result in significant toxicities, including cytokine release syndrome (CRS), a potentially life-threatening clinical syndrome resulting from the release of proinflammatory cytokines upon T cell activation. In addition, patients who develop CRS often experience prolonged cytopenias, and those with the most severe CRS also have the longest delays in full marrow recovery. Although an association between CRS and delayed bone marrow recovery has been established, the precise mechanism underlying this phenomenon remains unknown. This study was conducted to test our hypothesis that delayed bone marrow recovery following CAR-T therapy is caused by elevation of proinflammatory cytokines, leading to apoptosis and depletion of hematopoietic stem and progenitor cells (HSPCs). SCID-beige mice bearing intraperitoneal CD19⁺ Raji cell tumors were treated with injection of human CD19.28z CAR T cells. Bone marrow was then harvested for analysis by flow cytometry, and HSPCs were isolated for whole-transcriptome analysis by RNA sequencing. Complete blood counts and serum cytokine levels were measured as well. A second model was developed in which SCID-beige mice were treated with murine IFN-γ (mIFN-γ), murine IL-6 (mIL-6), or both. Bone marrow was harvested, and flow cytometry assays were conducted to evaluate the degree of apoptosis and proliferation on specific HSPC populations. SCID-beige mice bearing intraperitoneal Raji cell tumors that were treated with CAR-T therapy developed CRS, with elevations of several proinflammatory cytokines, including profound elevation of human IFN-γ. Gene set enrichment analysis of RNA sequencing data revealed that genes associated with apoptosis were significantly upregulated in HSPCs from mice that developed CRS. Endothelial protein C receptor (EPCR)-negative HSCs, a subset of HSCs that is poised for terminal differentiation, was found to be specifically decreased in mice that were treated with CAR T cells. Furthermore, HSPCs were found to have increased levels of apoptosis upon treatment with mIFN-γ and mIL-6, whereas short-term HSCs and multipotent progenitors exhibited increases in proliferation with mIFN-γ treatment alone. The results from this study provide evidence that the elevation of proinflammatory cytokines following CAR-T therapy impacts the bone marrow through a combined mechanism: pluripotent HSCs that are exposed to elevated levels of IFN-γ and IL-6 undergo increased cell death, while more committed progenitor cells become more proliferative in response to elevated IFN-γ. These combined effects lead to depleted stores of repopulating HSCs and ultimately cytopenias. © 2023 American Society for Transplantation and Cellular Therapy. Published by Elsevier Inc.

Keywords: Apoptosis; CAR-T therapy; Cytokine release syndrome; Cytopenia; Hematopoietic stem cell; IL-6; INF-γ; Neutropenia.

Figure 1.

A) SCID-Beige mice were treated with intraperitoneal injection of GFP-FFLuc labeled Raji cells or PBS control, followed by an incubation period of 2–3 weeks while monitoring tumor growth by in vivo imaging. After sufficient tumor burden was achieved, the mice were treated with CD19 directed CAR-T cells or PBS control. The mice were then monitored for 3 days at which point the mice were euthanized, blood was collected for CBC and cytokine analysis and bone marrow was isolated for flow cytometry and RNA sequencing analysis. B) Normalized weight trend of mice treated with: PBS control followed by CAR-T cells (n = 13, black), Raji cells followed by PBS control (n = 12, green) or Raji cells followed by CAR-T cells (n = 13, pink). C) Heatmap of Z-scores of human or murine cytokine levels from mice treated with: PBS control followed by CAR-T cells (n = 3), Raji cells followed by PBS control (n = 3) or Raji cells followed by CAR-T cells (n = 3). D-F) Cytokine levels of mice treated with: PBS control followed by CAR-T cells (black), Raji cells followed by PBS control (green) or Raji cells followed by CAR-T cells (pink): D) human interferon gamma E) human GM-CSF or F) murine IL-6. * = p < 0.05, ** = p < 0.01, *** = p < 0.001, **** = p < 0.0001.

Figure 2.

RNA sequencing was performed on HSPCs isolated from bone marrow of mice treated with PBS control followed by CAR-T cells, Raji cells followed by PBS control, or Raji cells followed by CAR-T cells and then enriched for cell surface expression of c-kit by magnetic bead cell isolation. A) Volcano plot comparing expression of genes from mice treated with Raji cells only or Raji cells followed by CAR-T cells; genes with P_adj < 0.05 are highlighted in blue. B) Summary of normalized enrichment scores (NES) from gene set enrichment analysis comparing gene expression from mice treated with Raji cells only or Raji cells followed by CAR-T cells. C) Heat map of Z-scores of expression of apoptosis associated genes from mice treated with Raji cells only, CAR-T cells only or Raji cells followed by CAR-T cells. D-G) Gene set enrichment analysis plots for: D) Hallmark TGF-β signaling gene set (NES = −0.99, p = 0.41), E) Hallmark IL-6/JAK-STAT3 signaling gene set (NES = 1.75, p < 0.001), F) Hallmark IFNγ response gene set (NES = 1.59, p < 0.001), G) Hallmark apoptosis gene set (NES = 1.30, p = 0.04).

Figure 3.

A) Representative flow cytometry gating to quantify HSPCs from mice treated with PBS control followed by CAR-T cells, Raji cells followed by PBS control or Raji cells followed by CAR-T cells. B-F) Percentage of HSPCs from mice treated with: PBS control followed by CAR-T cells (black), Raji cells followed by PBS control (green) or Raji cells followed by CAR-T cells (pink): B) LT-HSC (LSK, CD150+, CD48−, EPCR+), C) ST-HSC (LSK, CD150−, CD48−, EPCR+), D) MPP2 (LSK, CD150+, CD48+), E) MPP3/4 (LSK, CD150−, CD48+) or F) SLAM EPCR− (LSK, CD150+, CD48−, EPCR−). * = p < 0.05.

Figure 4.

A) SCID-Beige mice were treated with retroorbital injection of PBS control (black), mIFNγ (blue), mIL-6 (purple) or both mIFNγ and mIL-6 (orange) on 3 consecutive days. Bone marrow was isolated for flow cytometry analysis 24 hours after the final injection. B) Representative flow cytometry plots showing gating of HSPC for Annexin V or Ki67. C-G) Percentage of HSPCs positive for annexin V staining from mice treated with: PBS control (black), mIFNγ (blue), mIL-6 (purple) or mIFNγ and mIL-6 (orange): C) LT-HSC (LSK, CD150+, CD48, EPCR+), D) ST-HSC (LSK, CD150−, CD48−, EPCR+), E) MPP2 (LSK, CD150+, CD48+), F) MPP3/4 (LSK, CD150−, CD48+), G) SLAM EPCR− (LSK, CD150+, CD48−, EPCR−). H-L) Percentage of HSPCs positive for Ki67 staining from mice treated with: PBS control (black), mIFNγ (blue), mIL-6 (purple) or mIFNγ and mIL-6 (orange): H) LT-HSC (LSK, CD150+, CD48−, EPCR+), I) ST-HSC (LSK, CD150−, CD48−, EPCR+), J) MPP2 (LSK, CD150+, CD48+), K) MPP3/4 (LSK, CD150−, CD48+), L SLAM EPCR− (LSK, CD150+, CD48−, EPCR−).