Frameshift mutation of UVRAG: Switching a tumor suppressor to an oncogene in colorectal cancer

Abstract

Colorectal cancer (CRC) ranks as the second leading cause of cancer-related deaths in the Western world. It has a nearly 50% metastasis rate and only a subset of patients respond to current treatment strategy. UVRAG, a key autophagy effector and a guardian of chromosomal stability, is truncated by a frameshift (FS) mutation in CRC with microsatellite instability (MSI). However, the pathological and clinical significance of this UVRAG truncation remains less understood. Our recent study discovered that this FS mutation yields a much shortened form of the UVRAG protein, which counteracts most of the tumor-suppressor functions of wild-type (WT) UVRAG in autophagy, centrosome stability, and DNA repair in a dominant-negative fashion. Whereas this truncated mutation of UVRAG promotes tumorigenesis, epithelial-to-mesenchymal transition, and metastasis, it appears to sensitize CRC tumors to adjuvant chemotherapy, making it a potential molecular marker to individualize therapeutic approach in CRC.

Keywords:

• autophagy
• chemotherapy
• colorectal cancer
• metastasis
• UVRAG

The initial functional description of UVRAG (UV radiation resistance associated gene) focused on its regulation of the autophagy pathway, a quality-control mechanism in the cell. However, more recent investigations have shown that UVRAG governs a plethora of cellular processes to maintain homeostasis and genetic stability, also including ER-Golgi transport, endosomal degradation, DNA repair, and centrosome integrity. Consequently, tumor cells may strive to inactivate UVRAG in some contexts to eliminate an important stumbling block to the development of cancer. Notably, exon 8 of the UVRAG gene contains a microsatellite tract comprised of a cluster of 10 adenine nucleotides (A₁₀), rendering it highly susceptible to mutagenesis from polymerase slippage during DNA replication. Indeed, unlike modifications found in most other tumor suppressors, which are typically deleted or otherwise downregulated in cancers, a heterozygous deletion of one nucleotide (A) in the UVRAG A₁₀ tract has been detected in a third of CRC with MSI. This FS mutation generates a downstream premature stop codon, which is generally considered to be a mere inactivation of WT UVRAG (UVRAG^WT). However, in light of our recent data that this truncated UVRAG (UVRAG^FS) exerts oncogenic activity, the detailed molecular events downstream of this mutation and its implication in CRC pathogenesis are being uncovered.

The UVRAG protein is structurally characterized by a N-terminal proline-rich domain, a lipid-binding C2 domain, a BECN1-binding coiled-coil domain (CCD), and a C-terminal unstructured region responsible for interaction with the centrosome and the DNA repair machinery. The CRC-associated UVRAG FS mutation lies within the CCD, leading to the ablation of entire C-terminal region of UVRAG. Expression of UVRAG^FS can be detected in a series of MSI CRC cell lines with a concomitant reduction in the levels of UVRAG^WT. Notably, the UVRAG^FS proteins seem not to be produced in sufficient amounts likely due to degradation of nonsense UVRAG mRNAs via nonsense-mediated decay. In addition to losing the tumor-suppressing properties of UVRAG^WT, we found that the truncated UVRAG^FS overrides co-existing UVRAG^WT and acts as a dominantly active oncogene, promoting proliferation, anchorage-independent growth, and aggressive cell invasion, all of which are established hallmarks of cancer.

The UVRAG CCD has a central role in the regulation of autophagy owing to its interaction with BECN1, which, however, is truncated in UVRAG^FS, raising the concern as to whether this tumor-derived UVRAG mutant retains partial or complete autophagy function. A previous work compared different CRC cell lines and claimed that the autophagy levels of these cells were not affected by the UVRAG status, leading to a perception that alterations in UVRAG have no effect on autophagy. However, a direct comparison of cellular autophagy across highly variable genetic backgrounds can be misleading under some conditions. To further clarify this, we used a more homogenous experimental model to ectopically express UVRAG^FS in CRC cell lines. The results showed that UVRAG^FS sequesters endogenous UVRAG^WT and BECN1 proteins, resulting in a diminished interaction between them, consequently, blocking activation of BECN1-associated PIK3C3/Vps34 kinase, and therefore abrogating the pro-autophagy activity of UVRAG^WT in a dominant-negative manner. However, UVRAG^FS still promotes cellular transformation in the autophagy-deficient background, as compared to the control group, indicative of an autophagy-independent oncogenic mechanism associated with UVRAG^FS in CRC.

Besides sequestration of UVRAG^WT in the cytoplasm where it cannot form a fully functional autophagy complex, we show that UVRAG^FS also displaces UVRAG^WT from the centrosome by disrupting the UVRAG^WT-CEP63 interaction, which triggers centrosome abnormalities and increases levels of aneuploidy that potentially favor tumor formation. In accord, MSI cancers harboring the FS mutation exhibit substantially enhanced genomic instability and are associated with advanced TNM (Tumor, Node, Metastasis) stage. Furthermore, granting the consent that centrosome amplification per se confers an advantage to cell invasion, we show that UVRAG^FS promotes metastatic tumor growth and cell invasion via the mechanism of RAC1 activation and cytoskeleton remodeling. Although autophagy is considered to promote metastatic colonization by providing anoikis resistance, inhibiting autophagy has minimal effect on UVRAG^FS-mediated metastatic growth in our work. Additionally, and importantly, we observed that aberrant RAC1 activation as a result of UVRAG^FS expression leads to the loss of cell-cell adhesion and induction of epithelial-mesenchymal transition during CRC metastasis. Thus, UVRAG^FS is not only an autophagy inhibitor, but also functions as a metastasis-promoting molecule; the latter property suggests that, at least in some contexts, this protein may be eminently exploitable in cancer treatment.

DNA damage-inducing 5-fluorouracil-based chemotherapy is the most common regimen used for patients with advanced CRC. Although improved survival rate is achieved, this treatment strategy is still based on a one-size-fits-all approach that can only benefit specific populations. Given the oncogenic behavior of UVRAG^FS, one might intuitively expect that tumors harboring the FS mutation would be inherently more resistant to chemotherapy-induced cell death; surprisingly, this does not seem to be the case in our study. Using a xenograft CRC model, we observed an enhanced cytotoxic effect of chemotherapeutic agents on UVRAG^FS-tumors, as compared to UVRAG^WT-tumors. Subsequent work shows that UVRAG^FS-tumors accumulate higher levels of DNA double-strand breaks following treatment. Although UVRAG^FS lacks the domain required for interaction with PRKDC/DNA-PK (protein kinase, DNA-activated, catalytic polypeptide) complex, it attenuates UVRAG^WT-mediated assembly of a stable PRKDC complex and therefore the activation of PRKDC for double-strand break repair. In this scenario, UVRAG^FS serves as a “sensitizer” that might force cancer cells with mild DNA damage into a death program, thereby enhancing the efficacy of DNA-damaging genotoxic agents. In this respect, UVRAG^FS can serve as a potential predictive biomarker to identify patients who will maximally benefit from this treatment strategy.

Collectively, our study demonstrates that UVRAG represents a new MSI target gene in CRC and, likely, other MSI-related tumors, and that the CRC-associated truncating mutant virtually acts as a dominant-negative inhibitor of residual UVRAG^WT toward aggressive tumor growth, which is exceptionally linked to a better DNA-based treatment response. Of note, it remains to be tested whether this mutant might also possess activities of its own, not present in its normal counterpart, which can actively contribute to various aspects of tumor progression. It is also of interest how different oncogenic mechanisms of UVRAG^FS are primed to engage in distinct stages of cancer development. Although UVRAG^FS subverts the normal repair process induced by DNA-damaging drugs, whether such effects are anticancer drug-dependent awaits further investigation. Regardless of these uncertainties, in deciphering the oncogenic mechanisms of the truncated UVRAG, our work reveals a CRC-related oncogene and a potential predicative biomarker to facilitate the personalization of treatment for advanced CRC.