The TGF-β/SMAD Signaling Pathway As a Mediator of NK Cell Dysfunction and Immune Evasion in Myelodysplastic Syndrome

Myelodysplastic syndrome (MDS) is a spectrum of clonal hematopoietic disorders affecting the myeloid lineage, characterized by ineffective hematopoiesis and dysplastic features. In about a third of patients, MDS transforms to secondary AML, and for this subgroup the overall survival is dismal. Immune responses against MDS have the potential to prevent disease progression; however, immune evasion mechanisms pose a formidable barrier. In fact, NK cells from MDS patients show significant impairment in their effector function but the mechanism underlying this immune dysfunction is not well understood. Although a multiplicity of mechanisms contribute to NK cell immune evasion, two of the most important mechanisms are down-regulation of activating receptors and the production of inhibitory cytokines such as transforming growth factor β1 (TGF-β), a broadly immunosuppressive cytokine that renders NK cells unable to kill tumor cells. We hypothesized that the MDS microenvironment evades NK cell immune surveillance through a mechanism involving TGF-β, and that this immune dysfunction can be reversed by agents that interfere with this inhibitory pathway.

Thus, we isolated NK cells from the peripheral blood of MDS patients (n=16) and compared them to the NK cells from healthy donors (n=10). We observed that NK cells from high-risk MDS patients are dysfunctional, produce significantly less cytokines IFNg (p<0.0001) and CD107a (p=0.0029) and display significantly lower cytotoxicity (p=0.04) towards K562 as demonstrated by 51-Cr release assay compared to their healthy counterparts. This was associated with an abnormal phenotype with downregulation of many activating receptors and upregulation of exhaustion/inhibitory molecules such as PD-1 and the TGF-β induced molecules, CD9 and CD103. To show direct involvement of the TGF-β pathway in mediating MDS-induced NK dysfunction , w e examined activation of the TGF-β pathway by measuring phosphorylation of SMAD 2/3and showed that phospho-SMAD 2/3 is constitutively expressed in NK cells from MDS patients but not in healthy donor NK cells (p=0.03). We noted high expression of TGF-β LAP on the surface of myeloid blasts cells from MDS patients (80-90%) but not healthy myeloid cells (p <0.0001). Interestingly, when myeloid blasts were cultured together with NK cells, we detected significantly larger amounts of TGF-β in the supernatant, suggesting that NK-blast cross-talk may modulate TGF-β release by myeloid blasts. To further explore the mechanism of TGF-β-induced NK dysfunction and the contribution of NK-blast cross-talk to this phenomenon, we co-cultured healthy donor NK cells with myeloid blasts either in direct contact or separated using a transwell membrane and measured their cytotoxicity and cytokine release after 72h of co-culture. NK cells cultured in direct contact with myeloid blasts had significantly impaired effector function as evidenced by lower cytokine production (TNFa p<0.0001, IFNg p=0.0003) and CD107a degranulation (p=0.0002), whereas NK cells separated from blasts by a transwell membrane did not show any evidence of dysfunction. These data point to an important role for cell-cell contact in mediating blast-induced NK dysfunction.

We next investigated whether targeted inhibition of TGF-β can protect NK cells from TGF-β-mediated dysfunction, using two different approaches. First, we tested the role of a TGFβR1 kinase inhibitor, Galunisertib, currently in clinical trials for MDS. The addition of Galunisertib to cocultures of healthy NK cells and myeloid blasts prevented NK cell dysfunction and blast-induced SMAD2/3 phosphorylation. For the second approach, we tested if TGF-βR2 knockout by CRISPR-CAS9 gene editing can render them resistant to the suppressive effect of TGF-β. When co-cultured with 50ng/ml of recombinant TGF-β or myeloid blasts for 48 hrs, TGF-βR2 KO NK cells retained their ability to mediate a strong effector function. In vivo experiments to assess the in vivo activity of TGF-βR2 KO NK cells in a xenograft model of myeloid leukemia are underway.

Together, these data support an important role for the TGF-β/SMAD signaling pathway as a mediator of NK cell dysfunction and hence immune evasion in MDS, and support strategies to target the TGF-β/SMAD signaling axis for next-generation NK cell therapies in MDS.