Decoding ribosomes to keep growing (in AML)

In this issue of Blood, Pauli et al report that SNORD42A, a small nucleolar RNA (snoRNA) required for ribosomal RNA (rRNA) methylation and protein translation, is critical for acute myeloid leukemia (AML) cell proliferation.¹ 

In AML, leukemic cells (also referred to as blasts) are stuck at early stages of differentiation as a result of differentiation block in the myeloid lineage, and accumulate uncontrollably.²  This is the fundamental problem of AML. Although multiple genetic and epigenetic abnormalities that give rise to leukemogenesis have been well described, the molecular mechanisms underlying malignant transformation are still poorly understood.

Emerging evidence show that ribosomes are intimately linked to oncogenesis.³  Protein synthesis underpins cell growth and proliferation. Therefore, there is an increased demand for protein synthesis in rapidly proliferating cells, which can be achieved by modifying ribosome biogenesis and functions. The “onco-ribosomes” on which cancer cells depend for their expansion are an Achilles’ heel of cancer cells and, thus, a potential target for therapeutic intervention.

The eukaryotic ribosome is a ribonucleoprotein complex composed of 80 ribosomal proteins and 4 rRNAs. Upon their synthesis, rRNAs undergo posttranscriptional modifications that can influence rRNA folding, stability, and translation efficiency.⁴  snoRNAs are a class of small noncoding RNAs that guide rRNAs for posttranscriptional modifications.⁵  They associate with diverse partner proteins with catalytic activities and mostly function in the nucleolus within snoRNA ribonucleoprotein complexes. snoRNAs are divided into 2 structurally distinct classes: C/D box snoRNAs and H/ACA box snoRNAs, which direct the site-specific 2′-O-ribose methylation and the conversion of uridines to pseudouridines (ie, pseudouridylation) of rRNAs, respectively.

Aberrant expression or mutations of snoRNAs have been reported in different types of cancer, including AML, but the molecular mechanisms of snoRNAs that result in oncogenesis are not clearly defined.^6,7  Müller-Tidow and colleagues previously made an intriguing observation that the induction of C/D box snoRNAs by leukemogenic fusion proteins is critical for AML cell self-renewal and leukemogenesis. When expressed in hematopoietic cells, the oncogenic fusion protein AML1-ETO augmented the expression of amino-terminal enhancer of split (AES) proteins. Then, AES interacted with the RNA helicase DDX21 and upregulated C/D box snoRNAs, which increase rRNA 2′-O-methylation and protein translation. Such molecular changes ultimately promoted the self-renewal of leukemic cells and leukemia development in vivo. There are >200 snoRNAs in humans, and individual snoRNAs seem to play nonredundant roles. Although this study clearly implicated snoRNAs in AML pathogenesis, the function of individual snoRNAs remained unexplored.

In the present study, Pauli et al, from the same group, expand on their previous findings and provide evidence to support the claim that SNORD42A, which belongs to the C/D box snoRNA family, is an important regulator of AML cell survival and proliferation. In an effort to identify the specific snoRNAs required for leukemic cell survival, they performed a pooled CRISPR-Cas9 screen. SNORD42A was identified as the top snoRNA sustaining AML cell survival and was shown to be most consistently depleted across 4 different AML cell lines. Consistently, loss of SNORD42A significantly impaired proliferation and clonogenic potential of AML cells. To assess the relevance of SNORD42A to human AML, the authors examined its expression in 63 patient samples. SNORD42A was among the most highly expressed snoRNAs in these primary AML cells. Compared with the normal CD34⁺ cells, AML patient blasts displayed higher levels of SNORD42A expression and correlated with blast percentages in the bone marrow. Moreover, patients with an intermediate-risk AML displayed higher levels of SNORD42A than a favorable risk group.

To pursue the molecular consequences of SNORD42A loss, the authors assessed posttranscriptional modification of rRNAs. Consistent with the predicted function of SNORD42A,⁸  SNORD42A deficiency specifically decreased the 2′-O-methylation of uridine 116 of 18S rRNA, with global effects on translation, including reduced translation of ribosomal proteins, suggesting that SNORD42A is an important regulator of ribosome biogenesis. Compatible with these effects, 18S-U116 located in a highly structured region of 40S ribosomal subunit, supporting its role in the maintenance of structural integrity of ribosomes.

Although these findings provide new evidence that increased expression of SNORD42A constitutes an important pathway in leukemogenesis, some questions remain unresolved. For instance, it is not directly addressed whether the maintenance or self-renewal of AML blasts depends on SNORD42A-mediated rRNA modification and translation control or any other pathways. In this context, it is notable that snoRNAs have an expanded role besides ribosome biogenesis and translation control, including the regulation of RNA splicing and chromatin remodeling. Furthermore, recent studies have suggested that they modulate various cancer-related signaling pathways, such as p53 activation.⁹  In addition, although SNORD42A levels are higher in AML blasts than in normal hematopoietic cells and correlate with blast percentage, it is unclear how its expression is upregulated and whether the expression is controlled either uniformly in all AML or specifically in certain subtypes. This is important because depending on the expression profiles, SNORD42A and 2′-O-methylation at its rRNA target site might constitute a promising therapeutic target to limit AML progression. Further studies will be necessary to resolve these issues, ideally using in vivo models of human AML, such as xenograft models. Such studies may facilitate the development of novel therapeutic strategies that impair maintenance of leukemic blasts in AML pathogenesis.