High expression of LIN28B is associated with aggressive malignancy and poor survival. Here, probing MYCN-amplified neuroblastoma as a model system, we showed that LIN28B expression was associated with enhanced cell migration in vitro and invasive and metastatic behavior in murine xenografts. Sequence analysis of the polyribosome fraction of LIN28B-expressing neuroblastoma cells revealed let-7–independent enrichment of transcripts encoding components of the translational and ribosomal apparatus and depletion of transcripts of neuronal developmental programs. We further observed that LIN28B utilizes both its cold shock and zinc finger RNA binding domains to preferentially interact with MYCN-induced transcripts of the ribosomal complex, enhancing their translation. These data demonstrated that LIN28B couples the MYCN-driven transcriptional program to enhanced ribosomal translation, thereby implicating LIN28B as a posttranscriptional driver of the metastatic phenotype.
It is unknown which posttranscriptional regulatory mechanisms are required for oncogenic competence. Here, we show that the LIN28 family of RNA-binding proteins (RBPs), which facilitate posttranscriptional RNA metabolism within ribonucleoprotein networks, is essential for the initiation of diverse oncotypes of hepatocellular carcinoma (HCC). In HCC models driven by
NRAS
G12V
/Tp53
, CTNNB1/YAP/Tp53 , or AKT/Tp53 , mice without Lin28a and Lin28b were markedly impaired in cancer initiation. We biochemically defined an oncofetal regulon of 15 factors connected to LIN28 through direct mRNA and protein interactions. Interestingly, all were RBPs and only 1 of 15 was a Let-7 target. Polysome profiling and reporter assays showed that LIN28B directly increased the translation of 8 of these 15 RBPs. As expected, overexpression of LIN28B and IGFBP1-3 was able to genetically rescue cancer initiation. Using this platform to probe components downstream of LIN28, we found that 8 target RBPs were able to restore
NRAS
G12V
/Tp53
cancer formation in Lin28a/Lin28b -deficient mice. Furthermore, these LIN28B targets promote cancer initiation through an increase in protein synthesis. LIN28B, central to an RNP regulon that increases translation of RBPs, is important for tumor initiation in the liver.
MYCN amplification (MNA) and disruption of tumor suppressor microRNA (TSmiR) function are key drivers of poor outcomes in neuroblastoma (NB). While MYCN and TSmiRs regulate glucose metabolism, their role in de novo fatty acid synthesis (FAS) and unsaturated FAS (UFAS) remains poorly understood. Here, we show that FAS and UFAS (U/FAS) genes FASN, ELOVL6, SCD, FADS2, and FADS1 are upregulated in high-risk (HR) NB and that their expression is associated with lower overall survival. RNA-Seq analysis of human NB cell lines revealed parallel U/FAS gene expression patterns. Consistent with this, we found that NB-related TSmiRs were predicted to target these genes extensively. We further observed that both MYC and MYCN upregulated U/FAS pathway genes while suppressing TSmiR host gene expression, suggesting a possible U/FAS regulatory network between MYCN and TSmiRs in NB. NB cells are high in de novo synthesized omega 9 (ω9) unsaturated fatty acids and low in both ω6 and ω3, suggesting a means for NB to limit cell-autonomous immune stimulation and reactive oxygen species (ROS)-driven apoptosis from ω6 and ω3 unsaturated fatty acid derivatives, respectively. We propose a model in which MYCN and TSmiRs regulate U/FAS and play an important role in NB pathology, with implications for other MYC family-driven cancers.
Neuroblastoma (NB) is a highly metastatic pediatric cancer arising from the neural crest lineage. Genetic amplification of the MYCN proto-oncogene is a defining feature of NB, present in about 20% of all cases. The let-7 tumor suppressor microRNA targets the 3’ UTR of MYCN mRNA. We previously demonstrated that the 3’ UTR of MYCN mRNA acquires the ability to sequester let-7 in MYCN -amplified ( MA ) disease, thus inhibiting its function. This work established that a noncoding element within an oncogenic mRNA can contribute independently to disease pathology and genetic patterning. To further investigate the roles of noncoding RNA elements within the MYCN mRNA, we engineered cells expressing either MYCN-ORF (MYCN open reading frame only), MYCN-GL (full-length MYCN mRNA from the intact genetic locus), and Null-GL (full-length MYCN mRNA variant where EGFP replaces MYCN protein). We observe that all constructs enhance growth compared to controls in vitro and in vivo. MYCN-GL -expressing cells displayed the most robust growth in vitro despite containing multiple regulatory RNA elements. Remarkably, the Null-GL construct induces cells to grow as fast or faster than MYCN-ORF -expressing cells. Animal studies further confirmed these observations, where the Null-GL -driven tumors had the highest incidence and lowest latency, followed by MYCN-GL and then MYCN-ORF . Further, through NGS analysis, let-7, miR-101 , and miR-34a targets are enriched in both MYCN-GL and Null-GL expressing cells. Thus, the 3’UTR of MYCN , which is also targeted by these microRNAs, may interact with them in MYCN-GL and Null-GL cells to deliver a protective effect for the mRNA targets of these microRNAs. We also observed more dynamic differential gene expression in the -GL constructs than in ORF and GFP-expressing cells. In addition, MYCN-GL and Null-GL expressing cells are similarly enriched in gene ontology pathways for cancer, RNA metabolism, and microRNA processing pathways as compared to ORF and GFP . Whole genome sequencing also revealed more similarities in copy number variation in MYCN-GL and Null-GL than in ORF and GFP , suggesting that these constructs may provide selective pressure to favor specific CNV patterns. These observations show that full-length MYCN mRNA containing noncoding regulatory elements are more robust drivers of cell growth and oncogenicity than MYCN protein alone and provide insights into the mechanisms of oncogenic contribution. These results open an exciting door for our understanding of NB pathology and genetic patterning and have broad implications for other oncogene-driven cancers.
Background/Objectives: Neuroblastoma is a genetically diverse, highly metastatic pediatric cancer accounting for 15% of childhood cancer deaths despite only having ~8% of childhood cancer incidence. The current standard of care for high-risk diseases is highly genotoxic. This, combined with less than 50% survival in high-risk diseases and an abysmal 5% survival in relapsed cases, makes discovering novel, effective, and less toxic treatments essential. Methods: A prophylactic syngeneic mouse model was used to test high-dose lipid-mediator highly unsaturated fatty acids on tumorigenesis. Wildtype mice were gavaged with 12.3–14.6 g/d (adult human equivalent) omega-3 EPA, DHA, or oxidation-resistant bis allylic deuterated DHA (D-DHA) and 4.6–6.0 g/d arachidonic acid (ARA). At seven days, MYCN-expressing murine neuro-2a cells syngeneic to the gavaged mice were injected subcutaneously. Oral gavage continued for 10–20 d post-injection when tumors and tissues were harvested. Results: Fifty percent of control (not gavaged) animals form tumors (4/8) at about 10 d. High-dose DHA, D-DHA, and EPA block tumor formation completely in n = 8 or 10 animals. In contrast, ω6 arachidonic acid (4.6–6.0 g/d) enhances tumor formation (6/10 tumors) and reduces latency (5.5 to 10 days) compared to the control. The co-delivery of ARA and EPA results in a reduced tumor burden analogous to the control group, suggesting that EPA directly opposes the mechanism of ARA-mediated tumor formation. DHA acts through a non-oxidative mechanism. Conclusions: Sustained high-dose ω3 (weeks/months) is safe and well-tolerated in humans. These results suggest that ω3 DHA and EPA delivery at ultra-high doses may represent a viable low-toxicity therapy for neuroblastoma.
Functional gene disruption is a central tenet of cancer research, where novel drug targets are often identified and validated through cell-growth based knockdown studies or screens. Short hairpin RNA (shRNA)-mediated mRNA knockdown is widely used in both academic and pharmaceutical settings. However, off-target effects of shRNAs as well as interference with endogenous small RNA processing have been reported. We show here that lentiviral delivery of both gene-specific and non-targeting control shRNAs impair in vitro cell growth in a sequence independent manner. In addition, exogenous shRNAs induce a depressed cell-cycle-gene expression signature that is also shRNA-sequence independent and present across several studies. Further, we observe an shRNA mediated general repression of microRNAs belonging to polycistronic genetic clusters, including microRNAs from established oncogenic microRNA clusters. The collective impact of these observations is particularly relevant for cancer research, given the widespread historical use of shRNAs and the common goal of interrogating genes that regulate proliferation. We therefore recommend that when employing shRNA for target validation, care be taken to titrate shRNA dose, use hairpin-expressing controls, perform gene-of-interest rescue experiments and/or corroborate shRNA-derived results by small interfering RNA (siRNA) knockdown or CRISPR/Cas9-mediated genetic knockout. Minimizing these deleterious sequence independent effects will improve research fidelity and help address reported challenges in experimental reproducibility.