Breaking ribosomes to fight leukemia Free

In this issue of Blood, Damaskou et al¹ identify targeted disruption of ribosome biogenesis by small molecule inhibitors of RNA polymerase I as a therapeutic strategy in nucleophosmin 1 (NPM1)–mutated acute myeloid leukemia (AML).

NPM1 mutation is the most frequent driver genetic alteration in adult AML, occurring in approximately one third of AML cases.² It is a de novo AML-defining mutation, as it is not found in clonal hematopoiesis, myelodysplastic syndrome, or other myeloid malignancies. Because of its characteristic molecular and clinical features, NPM1 mutation alone defines a distinct entity in both the International Consensus Classification and the fifth edition of the World Health Organization classification of myeloid neoplasms.³ NPM1 has been implicated in many cellular functions, including protein chaperoning, ribosome biogenesis, cell growth, cell proliferation and centrosome duplication, among others.⁴ Although normal NPM1 primarily resides in the nucleus and shuttles between the nucleus and the cytoplasm, the NPM1 mutation—a 4 base-pair insertion in exon 12—results in aberrant cytoplasmic localization⁵ (and is thus also referred to as cytoplasmic NPM1 or NPM1c). Even though NPM1 mutations have been recognized as AML driver mutations before the next-generation sequencing era and their prevalence in AML is high, the mechanisms by which they drive AML remain poorly understood.

Here, using a conditional knockin mouse model, allowing them to study the effects of NPM1c as an isolated mutation, Damaskou et al observed that several ribosome biogenesis factors were reduced in the mutant cells. Specifically, proteomics analyses in lineage-negative hematopoietic stem and progenitor cells (HSPCs) from this mouse model revealed that several proteins involved in ribosome biogenesis were reduced without concomitant changes at the messenger RNA level—findings consistent with posttranscriptional regulation. Examination of published proteomics data from bone marrow of patients with AML provided corroborating evidence.⁶ 

The authors thus hypothesized that NPM1c-mutant HSPCs may be more sensitive to drugs that cause nucleolar stress through inhibition of ribosome biogenesis, like actinomycin D (ActD) and 2 other small molecule inhibitors of RNA polymerase I, the polymerase that specifically transcribes ribosomal RNA. They indeed provide evidence for preferential sensitivity of NPM1-mutant cells to these inhibitors using a range of experimental models with NPM1 mutations, including the mouse knockin model, AML cell lines, and a patient-derived xenograft model.

Searching the cancer dependency map database, as well as published data from another genome-wide CRISPR dropout screen in AML cell lines, to discover specific genes related to ribosomal biogenesis as NPM1c-specific dependencies by means of essentiality for the NPM1c-mutant AML cell line OCI-AML3, the authors identified TSR3. TSR3 is a factor involved in the 40S ribosome maturation pathway and its knockout (KO) specifically inhibited NPM1c-mutant cells. Finally, the authors showed that ActD and TSR3 KO synergize with the B-cell leukemia/lymphoma 2 inhibitor venetoclax (VEN) and that pretreatment with ActD resensitizes resistant cells to VEN.

A previous small single-center study of ActD monotherapy provided some evidence of efficacy against NPM1-mutated AML, with reported remissions in patients with NPM1-mutated relapsed/refractory AML, some of which were long-lasting.⁷ Although both the present and past studies showed evidence that ActD may act by exacerbating nucleolar stress induced by the dislocation of mutant NPM1 out of the nucleolus, by further reducing nucleolar NPM1 levels, other mechanisms of the effects of ActD in this context have been postulated. Mutant NPM1 has been shown to impair mitochondrial function, which is further augmented by treatment with ActD.⁸ This metabolic stress could also mediate synergy between ActD and VEN, like Damaskou et al observed. Although the present study, like previous work, found evidence of p53 activation by ActD, and the authors propose p53-mediated apoptosis as the basis for the synergistic action with VEN, other mechanisms may also be at play.

This study did not fully elucidate the mechanism by which the NPM1 mutation reduces ribosome biogenesis factors post-transcriptionally, nor the mechanism underlying the dependency of NPM1-mutated AML on TSR3. Given the promising results presented here and the favorable therapeutic track record of ActD in NPM1-mutated AML,⁷ further studies into its mechanisms of action are warranted and could shed new light into targetable vulnerabilities in this AML subgroup. It might also be of interest to test combinations of ActD with drugs other than, or in addition to VEN, such as FMS-like tyrosine kinase 3 (FLT3) inhibitors—as FLT3 mutations frequently co-occur with NPM1 mutations in AML—and menin inhibitors, a new class of drugs recently US Food and Drug Administration–approved for AML with KMT2A rearrangements, and shown to have preclinical and clinical efficacy in NPM1-mutated AML.^9,10 

Conflict-of-interest disclosure: E.P.P. declares no competing financial interests.