Upregulation of Indoleamine 2,3-Dioxygenase Enzymes in Leukemic Mesenchymal Stromal Cells (MSCs) Can Influence MSC/Acute Myeloid Leukemia Cell Cross Talk

Mesenchymal stromal cells (MSCs), an essential element of both normal and leukemic hematopoietic microenvironment, are multipotent cells with a unique immune-modulating ability. Thus, MSCs play a crucial role for both the proliferation and differentiation of hematopoietic stem cells (HSCs) and induce an immune-tolerant milieu. Indoleamine 2, 3-dioxygenase (IDO1 and IDO2) enzymes catabolize tryptophan to kynurenines and play a key role in the induction of immune tolerance in different settings, including acute myeloid leukemia (AML). Furthermore, IDO1/IDO2 pathway is a well described mechanism by which MSCs exert their mmunomodulatory properties. We hypothesized that: 1) MSC-dependent mechanisms are involved in leukemia initiation, maintenance and progression; 2) the expression of IDO1 and IDO2 by MSCs is part of a MSC-dependent mechanism able to create a tumor-supportive milieu. To this aim, we isolated MSCs from the bone marrow of AML patients (AML-MSCs) at diagnosis. We first analyzed their phenotypic and functional properties compared to that of healthy donor-derived MSCs (HD-MSCs). We found that AML-MSCs showed a reduced proliferative capacity but normal immunophenotype, differentiative and immunomodulatory capacity as compared to HD-MSCs. Furthermore, AML-MSCs did not show the chromosomal abnormalities identified in the primary blast counterpart (FISH analysis). We next investigated IDO1/2 expression and functions in MSCs. We demonstrated that IDO enzymes are expressed in AML-MSCs as well as in HD-MSCs. IDO1 is efficiently upregulated by different inflammatory stimuli, and IDO1 protein expression parallels mRNA in both HD-MSCs and AML-MSCs. Interestingly, IDO2 mRNA is expressed at low basal level in all analyzed conditions in HD-MSCs, while it is upregulated, in particular after IFN-gamma stimulation, in AML-MSCs, although the level of induction varies between different patients. When T-cell proliferation was tested in MSC co-cultures, w/ or w/out IDO1/2 inhibitor, 1-methyltryptophan, we found that MSC immunomodulatory potential is IDO-dependent both in HD-MSCs and AML-MSCs. Finally, we found that in co-culture assay with primary AML blasts, MSCs stimulated blast proliferation and this effect is, at least in part, IDO-mediated. These data suggested that IDO enzymes, in particular IDO2, may be differentially expressed in AML-MSCs as compared to HD-MSCs and IDO inhibition has an impact on MSC/AML cell cross talk. These findings may help to discover novel niche-target prognostic/therapeutic factors and to provide novel applications for drugs already under active clinical investigation (i.e. IDO-inhibitors).