M.Sc Thesis

M.Sc StudentBarkai Oren
SubjectCytoplasmic mRNA Decay Initiates in the Nucleus
DepartmentDepartment of Medicine
Supervisor PROF. Mordechai Choder
Full Thesis textFull thesis text - English Version


mRNA buffering involves factors that mediate both mRNA synthesis and decay, including the 5’ to 3’ exonuclease Xrn1 that functions as the major mRNA decay factor in the yeast cytoplasm.  Previously, we reported that a number of these factors shuttle between the nucleus, where they regulate transcription, and the cytoplasm, where they execute mRNA degradation. We hypothesized that their shuttling regulates the balance between mRNA synthesis and decay. We identified two functional nuclear localization sequences (NLSs) in Xrn1. NLS1 is located in a protruded domain, called “Tail”. I fused Tail to GFP and found, unexpectedly, that it was localized exclusively in the cytoplasm suggesting that NLS1 was not recognized. I therefore randomly mutagenized the tail moiety of Tail-GFP and found a number of point mutants that were localized in the nucleus. Their nuclear localization was dependent on NLS1, indicating that NLS1 is functional. Interestingly, the point mutations mapped to an R3H-like motif. In collaboration with Dr. Shiladitya Chattopadhyay in our lab, we demonstrated that my prediction was correct and the tail binds RNA and mutations in this motif compromise RNA-binding. I present a model, supported by my data and data of my collaborator that the decaying RNA binds the R3H-like motif and the active site while it binds also NLS1. Thus, NLS1 is inaccessible as long as the RNA binds it. Only after RNA degradation, NLS1 is exposed and Xrn1 is imported. This establishes a link between RNA degradation and Xrn1 import and transcription.

In collaboration with Prof. Jose E. Perez-Ortin (U. Valencia), who analyzed the strains that we have created by “Genomic Run On” (GRO) technique, we found that mutations in NLS1/2, which block Xrn1 import (but not its enzymatic activity), do not change steady state mRNA levels, yet they compromise transcription of >3000 genes and, unexpectedly, also decrease the decay rate of their transcripts. Thus, nuclear Xrn1 is required for fast mRNA synthesis and turnover in a manner that maintains mRNA buffering. I have corroborated this high-throughput finding by analyzing a few specific mRNAs by Northern blot hybridization.

While analyzing the Northern results, I serendipitously found that PMA1 mRNA is composed of two isoforms, a long one and a one that is shorter by ~450 b. Interestingly, the ratio between the long and the short isoforms is affected by Xrn1 import capacity. This was corroborated by RT-qPCR. I also found that isoforms of YEF3 mRNAs are similarly affected by disruption of Xrn1 NLS1/2. We propose that, in the nucleus, Xrn1 is involved in the choice of cleavage and polyadenylation sites, i.e., in alternative polyadenylation mechanism.

In summary, Xrn1 shuttling is involved in the linkage between mRNA synthesis and decay. Import of Xrn1 to the nucleus is important not only for transcription, as we have reported previously but, unexpectedly, also for alternative polyadenylation in the nucleus and mRNA decay in the cytoplasm. Xrn1 shuttling seems to integrate several stages of mRNA biogenesis and turnover that occur in the two main compartments in the cell.