Function and Regulation of tRNA fragments

The advent of small RNA sequencing has revealed a previously hidden layer of cellular small RNA diversity. In addition to well characterized microRNAs, we and others have characterized small RNAs derived from tRNAs, known as tRNA fragments or tRFs, which make up a substantial portion of the cellular small RNA milieu. How tRFs and other non-microRNA small RNAs are regulated is not well understood. In this dissertation (1) we identify key mechanisms of tRF regulation, (2) develop a target prediction tool for tRFs, (3) establish a tool to identify novel small RNA regulators, and (4) implicate tRFs in aging.
First, we sought to elucidate the mechanism by which tRF entry into the microRNA effector pathway is regulated. Although several tRFs can enter the microRNA effector pathway, tRFs derived from trailers do not enter the microRNA effector pathway. This observation suggests the presence of a microRNA surveillance mechanism that ensures proper functioning of the microRNA effector pathway. We identify 5'-3' exoribonuclease 2 (XRN2) as a key microRNA effector pathway surveillance enzyme. XRN2 renders tRFs derived from precursor tRNAs highly unstable, thereby preventing microRNA effector pathway entry. Intriguingly, stabilized precursor tRFs correlate with reduced expression of mRNAs with predicted target sites within introns and 3' UTRs.
Next, we developed a tRF target prediction tool based on the finding that many tRFs enter the microRNA effector pathway. Our tool, tRForest, used Ago bound tRF-mRNA pairs as ground-truth for training a random forest algorithm. tRForest outperformed current tRF target prediction tools. We also proposed biological processes that tRFs might be involved in based on Gene Ontology analyses of targets.
Realizing that there are likely undiscovered small RNA regulators, we leveraged publicly available data to identify novel regulators. We found that by combining short hairpin RNA (shRNA) knockdown followed by RNA sequencing with enhanced crosslinking and immunoprecipitation (eCLIP) sequencing data coupled with machine learning, we could predict known and novel microRNA effectors. We validated the role of BMI1 and STK33 in microRNA regulation. A similar approach can be used to identify novel tRF effectors in the future.
Finally, we found that tRFs, particularly tRNA halves, are increased during aging in Saccharomyces cerevisiae, suggesting a role in tRNA half accumulation in aging. Since tRNA halves have been shown to reduce translation, and translation is reduced during aging, we hypothesize that tRNA halves reduce translation during aging. Together, this dissertation identifies novel tRF and microRNA regulatory mechanisms and suggests novel biological functions for tRFs.