Microrna 29b Mediates Immune Evasion of Natural Killer Cells in Acute Myeloid Leukemia

Natural killer (NK) cells are innate effector cells that can spontaneously recognize and kill cancer cells.  While important in inducing long-term disease free survival (DFS) in the setting of killer immunoglobulin-like receptor (KIR) mismatch for acute myeloid leukemia (AML), harnessing NK cells to kill autologous or self AML blasts to extend DFS has had no success.  In this study, we uncover one potential mechanism by which AML blasts can evade NK cell cytotoxicity. The aryl hydrocarbon receptor (AHR) is a ligand-activated transcription factor that has now been shown to be expressed in immature human NK cells.  Furthermore, activation of AHR in these immature cells suppresses human NK cell maturation and function.  We and others have shown that human AML blasts secrete ligands that can activate AHR, leading us to hypothesize that AML may evade autologous NK cell cytotoxicity in part by inhibiting NK cell maturation. Expression of microRNA (miR)-29b has previously been shown to suppress the expression of transcription factors Tbx21 (TBET) and Eomesodermin (EOMES), both of which are critical for terminal NK cell differentiation and function.  Here we show that AHR is able to directly regulate the expression of miR-29b and thus may serve as the link between AML and immune evasion of NK cells.  We first identified putative AHR binding sites within the proximal
promoter of miR-29b, suggesting that AHR may directly regulate miR-29b expression.  To test this, we transfected the AHR responsive HepG2 human cell line with a miR-29b promoter driven luciferase reporter and measured luciferase activity after treatment with the known AHR agonist, FICZ, compared to vehicle control.  We discovered that cells treated with 6-formylindolol[3,2-b]carbazole (FICZ) had increased luciferase activity compared to cells treated with vehicle control (P<0.05).  This effect was subsequently shown to be directly mediated by AHR, as mutation of the putative AHR binding site as well as siRNA targeting of AHR mRNA both significantly diminished the induced luciferase activity (P<0.01).  We then established the importance of miR-29b in human NK cell development by transducing immature NK cells, characterized as Lin(-)CD117(+)CD94(-), with a miR-29b knockdown virus and testing for the ability of these cells to become mature [i.e., Lin(-)CD117(-)CD94(+)] NK cells, after two weeks in IL-15 and FICZ.  Indeed, in early experiments, knockdown of miR-29b resulted in increased percentages of mature NK cells compared to cells transduced with control virus (16.8% compared to 3.6%, respectively), despite being cultured with an AHR agonist.  Finally, utilizing a translational in vivo murine AML model developed by our laboratory that recapitulates human AML, we have found that AML blasts harvested from these leukemic mice release an AHR agonist (P < 0.01), similar to previous reports that have described AHR ligands being produced by human AML blasts.  In addition,NK cells isolated from these mice have increased levels of miR-29b when leukemic, compared to NK cells from wild type control littermates, consistent with our hypothesis that AHR regulates miR-29b expression.  Thus, we propose that AHR, when activated by AML-derived ligands, upregulates miR-29b in NK cells to ultimately suppress the primary regulators of NK cell maturation and function, resulting in immune evasion (See Figure).