Acquired microRNA-142 Deficit Drives Escape Mechanisms of Anti-Leukemic Surveillance during Blast Crisis Transformation of Chronic Myeloid Leukemia (CML)

CML may evolve from a chronic phase (CP) into blast crisis (BC), but the underlying mechanisms of the transformation remain to be fully elucidated. We reported that microRNA (miR)-142 is downregulated in BC compared with CP patients; and in a murine model of CP CML (i.e., BCR-ABL mouse), miR-142 knock-out (KO) induced a BC-like phenotype, substantiating a mechanistic role of miR-142 deficit in BC transformation. Mechanistically, miR-142 KO induced mitochondrial fusion and increased oxidative phosphorylation (OxPhos) in leukemic stem cells (LSCs), causing a shift of leukemic phenotype from CP to BC ( Nat Commun, in press). Herein, we report that miR-142 deficit also occurs in T lymphocytes of BC patients due to inflammatory cytokines that are aberrantly produced by the proliferating leukemic cells. Using the Mir142 ^−/−BCR-ABL mouse, we observed that miR-142 deficit resulted in loss of T cell number and activity, suppressed the antileukemic immune surveillance, and contributed to BC transformation. In fact, miR-142 KO hampered thymic lymphoid-primed multipotent progenitor (LMPP) differentiation into T cells and rendered mature T cells dysfunctional and exhausted, with increase of PD-1 levels, and decrease of the apoptotic threshold, cell cycling and cytokine production, via blockade of OxPhos/glycolysis switch that regulates the metabolism of otherwise activated T cells. These changes translated into a decrease of T-cell antileukemic surveillance as demonstrated by increased numbers of BC murine Lin ^-Sca-1 ⁺c-Kit ⁺ (LSKs) or human CD34+ blasts cocultured with Mir142 ^−/−T-cells vs those cocultured with Mir142 ^+/+ T-cells. Furthermore, congenic B6 (lethally irradiated to eradicate host T cells) or immunodeficient NSG (no T cells) recipient mice transplanted with Mir142 ^−/−BCR-ABL LSK and Mir142 ^−/− T cells had reduced T cells (both: p<0.0001), increased blasts (B6: p=0.01; NSG: p<0.0001) and a shorter survival (median survival for B6: 58 days vs unreached, p=0.008; for NSG: 28 days vs unreached, p<0.0001) compared with the respective controls transplanted with Mir142 ^−/−BCR-ABL LSK and Mir142 ^+/+ T cells. Transplantation of Mir142 ^−/−BCR-ABL LSKs into Mir142 ^−/− or Mir142 ^flox(f)/fLck-cre+ (i.e., miR-142 KO only in T cells; Mir142T ^Δ/Δ) recipients also resulted in a shorter survival than Mir142 ^+/+ recipients (median survival for Mir142 ^−/− vs Mir142 ^+/+ recipients: 44 vs 53 days, p=0.001; for Mir142T ^Δ/Δ vs Mir142 ^+/+ recipients: 42 vs 53 days, p=0.0002).

To correct miR-142 deficit, we produced a synthetic miR-142 mimic oligonucleotide (CpG-M-miR-142). We treated a cohort of NSG mice transplanted with Mir142 ^−/−BCR-ABL LSKs and Mir142 ^−/− T cells, with CpG-M-miR-142 (30mg/kg/day, IV) or CpG-scramble RNA (SCR) for 3 weeks. We observed increased T cells (p=0.04), decreased blasts (p=0.0029) and longer survival (median survival: unreached vs 37 days, p=0.0005) in CpG-M-miR-142-treated recipients vs SCR-treated controls. Similar results were observed in CpG-M-miR-142-treated Mir142 ^−/−BCR-ABL mice vs SCR-treated controls. To test CpG-M-miR-142 activity in human T cells, a cohort of BC CML patient-derived xenograft (PDX) mice were given patient's autologous T cells and CpG-M-miR-142. Mice receiving T+CpG-M-miR-142 had increased T cell expansion (p=0.04), reduced leukemic cells (p=0.0078), and longer survival (median survival: 53 vs 38 days, p=0.01) than controls receiving T+SCR. In secondary transplants, the recipients of BM from T+CpG-M-miR-142-treated donors survived longer (median survival: 73 vs 53 days, p<0.0001) than the recipients of BM from T+SCR-treated donors, suggesting a decrease in LSC burden.

Finally, we tested the efficacy of tyrosine kinase inhibitors (TKIs), which represent the primary targeting approach for CML, in combination with CpG-M-miR-142 and/or PD-1 inhibitor in Mir142 ^−/−BCR-ABL mice. The mice treated with TKI+CpG-M-miR-142 (median survival: 107 days, p=0.001) or TKI+PD-1 inhibitor (104 days, p=0.002) survived longer than the controls treated with TKI alone (66 days); the triplet combination (TKI+CpG-M-miR-142+PD-1 inhibitor) is expected to improve survival further (ongoing experiment).

In summary, CpG-M-miR-142 alone and in combination with TKI or PD-1 inhibitor rescued T-cell antileukemic activity and prolonged survival in BC CML murine and PDX models, thereby providing a potentially new therapeutic approach for BC CML.