Nucleolar Stress Plays a Key Role in the Pathogenesis of T Cell Acute Lymphoblastic Leukemia

Increased Notch signaling is thought to be central to the pathogenesis of cortical T-cell acute lymphoblastic leukemia (T-ALL), with activating mutations of NOTCH1 in great than 80% of cases. Increased Notch signaling results in markedly increases ribosome biogenesis, inducing the nucleolar stress pathway. CDNK2A, encoding both ARF and INK4A, plays a key role in the nucleolar stress pathway. This is an evolutionarily conserved pathway that includes the translocation of certain "sensor” ribosome proteins to the nuclear cytoplasm where they bind and alter the activity of p14^ARF (ARF) and MDM2, which in turn, alters ribosome biogenesis, activates p53, and induces a growth arrest. To overcome this negative feedback loop, over 80% of T-ALL carry biallelic deletions of CDKN2A (encoding for P16^INK4A and ARF). In addition, ~17% of T-ALL carries mutations of sensor ribosome genes. These observations suggest the hypothesis that increased Notch signaling induces nucleolar stress in T cells and inactivation of the nucleolar stress pathway is essential to T-ALL pathogenesis.

To test this hypothesis, we first generated transduced primary murine CD4⁺ CD8⁺ (DP) thymocytes in culture with retrovirus expressing empty vector, wildtype NOTCH1, or activating NOTCH1 mutants (aNOTCH1) carrying hotspot mutations L1594P or L1601P. Expression of aNOTCH1 induced nucleolar stress, as measured by induction of Arf expression, accumulation of 5’ external transcribed spacer (ETS) containing pre-rRNA, and NPM1 translocation from nucleoli to the nucleoplasm. It also resulted in with p53 activation, with marked increased expression of p53 target genes p21^Cip1/Waf1 and Bax. Of note, expression of aNOTCH in primary DP thymocytes had a negative impact on cellular proliferation, with induction of apoptosis. We next explored the impact of aNOTCH1 mutants in vivo, using a retroviral transplantation model. In brief, wildtype c-Kit+ cells were transduced with wildtype or aNOTCH retrovirus and transplanted into lethally irradiated mice. The contribution of NOTCH1 (GFP+) expressing cells to hematopoiesis was examined over time. The contribution of aNOTCH variants to hematopoietic stem cells (HSCs), common lymphoid progenitors (CLP), and T cells was similar to input percentage and was stable over time. In sharp contrast, there was a near complete loss of aNOTCH expressing myeloid and B cells. We next examined the impact of CDKN2A loss on HSC maintenance. In Cdkn2a^+/- or Cdkn2a^-/- mice the number of phenotypic HSCs was similar to wildtype mice. This is expected since the expression of CDKN2A is very low in HSCs under steady state conditions. In contrast, in competitive transplantation assays, both Cdkn2a^+/- and Cdkn2a^-/- HSCs consistently outperformed wildtype competitors. We hypothesized that replication stress induced by transplantation may active Cdkn2a expression in WT HSCs, suppressing their proliferation. To test this, Cdkn2a^+/- or Cdkn2a^-/- bone marrow chimeras were treated with 5-fluorouracil (5-FU) following stable engraftment. Consistent with this hypothesis, treatment with 5-FU resulted in marked increase in Arf (but not Ink4a) expression and a significant (~ 5-fold) increase in Cdkn2a^+/- or Cdkn2a^-/- hematopoietic output. Finally, we modeled the impact of Cdkn2a loss on aNOTCH-induced leukemogenesis by retrovirally transducing NOTCH variants into Cdkn2a^+/- or Cdkn2a^-/- HSPCs. Compared with wildtype, expression of aNOTCH in Cdkn2a^+/- or Cdkn2a^-/- HSCs resulted in a marked expansion of HSCs (3.70 ± 0.22-fold) and CLPs (6.95 ± 2.28-fold) but did not rescue the aNOTCH-induced block in myeloid or B lymphoid differentiation. A tumor watch confirmed that Cdkn2a loss strongly cooperated with aNOTCH expression to induce T-ALL mice (incidence of T-ALL at 200 days of 17/17 (100%) in NOTCH1 L1601P Cdkn2a^-/- cell recipients vs 3/12 (25%) in NOTCH1 L1601P Cdkn2a^+/+ cell recipients, P < 0.0001).

In summary, these data provide support for a novel model of T-ALL leukemogenesis in which CDKN2A loss is an early event that provides a fitness advantage under conditions of replicative stress that leads to an expanded pool of CDKN2A^+/- (or CDKN2A^-/-) HSCs and CLPs. Loss of CDKN2A provides a permissive cellular environment for activating NOTCH1 mutations by attenuating nucleolar stress. These data suggest that targeting nucleolar stress pathway may have therapeutic activity in T-ALL.

Link:Roche: Other: Roche provided Idasanutlin free of charge. No compensation was provided, and Roche was not involved in design, conduction, or analysis of experiments.