Role of Long Noncoding RNAs in Aggressive Brain Tumors

Author:
Reon, Brian, Experimental Pathology - Graduate School of Arts and Sciences, University of Virginia
Advisor:
Dutta, Anindya, Department of Biochemistry and Molecular Genetics, University of Virginia
Abstract:

Long noncoding RNAs (lncRNAs) are emerging as key regulators of cellular processes and are commonly found to be misregulated in many human pathologies, including cancer. Glioblastomas (GBM) and lower-grade gliomas (Grade II and III) are aggressive, difficult to treat brain tumors that have high mortality rates. Although much of the research into glioma biology has focused on identifying the molecular drivers of oncogenesis, these efforts have primarily focused on alterations in protein coding genes and the role of lncRNAs has not been sufficiently characterized. This dissertation focuses on expanding our knowledge of lncRNAs in gliomas by analyzing their expression globally and more targeted studies examining the role of two lncRNAs.
In our initial study, we analyzed the expression data from over 750 RNA-seq datasets from GBMs, Grade III and II gliomas and normal brain tissue. We found that hundreds of lncRNAs are differentially expressed in gliomas compared to normal brain tissue. Furthermore, many lncRNAs were found to be preferentially expressed within certain GBM and LGG subtypes. Using these subtype specific lncRNAs we identified similarities between the highly aggressive IDH1/2 wt LGG subtype and mesenchymal GBMs. We also used Cox regression to create a survival algorithm that is capable of separating LGG patients into two distinct prognostic groups. Lastly, we identified all lncRNAs that are associated with GBM patient survival, to aid in identifying which lncRNAs might play critical roles in brain tumors.
Using our brain tumor lncRNA expression and survival association data, we identified two oncogenic lncRNAs, GS1-124K5.4 and LINC00152, for further study. Both LINC00152 and GS1-124K5.4 are upregulated in GBMs and aggressive gliomas, and high expression of either lncRNAs is associated with negative patient outcomes in GBMs. High levels of LINC00152 has no affect on cell growth but lead to an increase invasion in U87 cells. In contrast, expression of GS1-124K5.4 does not affect invasion but high expression of GS1-124K5.4 leads to increased cell growth. Secondary structure analysis of LINC00152 suggests that a protein-bound stem-loop in the 3’ end of LINC00152 is partially involved in LINC00152’s proinvasive function. Furthermore, LINC00152 is upregulated in 10 other tumors types and high expression of is associated with a poor prognosis in 7 other tumors. These results suggest that LINC00152 potentially functions as an oncogene in many cancers.

Degree:
PHD (Doctor of Philosophy)
Language:
English
Issued Date:
2017/04/27