Interferon-gamma constitutively expressed in the stromal microenvironment of human marrow cultures mediates potent hematopoietic inhibition

Clinical and laboratory studies have suggested involvement of interferon-gamma (IFN-gamma) in the pathophysiology of aplastic anemia. T cells from aplastic anemia (AA) patients secrete IFN-gamma in vitro, activated cytotoxic lymphocytes infiltrate aplastic bone marrow (BM), and IFN-gamma mRNA, not detected in normal BM, is present in BM from most AA patients. Many patients respond to immunosuppressive therapy with antithymocyte globulin and cyclosporine. Using long-term BM cultures (LTBMC) as a tissue culture model of hematopoiesis, we show that IFN-gamma is a potent inhibitor in the long-term culture-initiating cell (LTC-IC) assay, the best in vitro surrogate test for human hematopoietic stem cells, as well as of the output of committed progenitor cells (colony-forming unit-granulocyte-macrophage CFU-GM and burst-forming unit-erythroid BFU-E). In LTBMC, continuous addition of relatively high IFN-gamma concentrations (1,000 U/mL weekly or 200 U/mL every 2 days) was required for inhibition of secondary colony formation, a measure of LTC-IC number and clonogenicity. To mimick local production of IFN-gamma, human stromal cells were engineered by retroviral-mediated gene transfer to express a transduced IFN-gamma gene. IFN-gamma secreted by stromal cells was far more potent than exogenous IFN-gamma in its effects in the LTC-IC assay. For purified CD34+ cells culture in the presence of IFN-gamma stroma dramatically reduced secondary colony numbers as well as production of CFU-GM and BFU-E. Supernatants from these cultures contained only about 20 U/mL of IFN-gamma; this quantity of cytokine, when added to LTBMC, had little effect on hematopoiesis. The mechanism of hematopoietic suppression was related to the inhibition of cell cycle progression and induction of apoptosis of CD34+ cells. There was no apparent effect of local low-level IFN-gamma production on stromal cell function, as reflected in cell morphology, cell surface phenotype, or expression of hematopoietic growth factor genes. LTBMC with genetically altered stromal cells offers an in vitro model of immune suppression of hematopoiesis in AA and may be helpful in testing certain therapeutic modalities. We infer from our data that local production of low levels of inhibitory cytokine is sufficient to markedly inhibit hematopoiesis and to destroy stem cells and more mature progenitor cells.