You are here:
Research
New enzymes and paradigms in RNA metabolism
Our laboratory follows an integrative approach to identify and characterize, in vitro and in vivo, new players in the intricate world of RNA processing. We are particularly interested in enzymes that ligate and phosphorylate RNA molecules, i.e. RNA ligases and RNA kinases. For this purpose we combine: i) Biochemistry and bioinformatics to reach a deep mechanistic understanding; ii) Cross-linking and immunoprecipitation (CLIP) followed by deep sequencing to identify RNA substrates for these enzymes and assign them to specific RNA metabolic pathways, and iii) Mouse knockout models to monitor their impact in the context of a full organism and to reveal potential connections between RNA metabolism and disease. We recently identified the human tRNA ligase complex, involved in pre-tRNA splicing and probably in the splicing of pre-mRNA molecules during the unfolded protein response. In the future we aim to reveal a general role for the tRNA ligase in mechanisms of RNA repair.
The long sought human tRNA ligase finally identified!
Similar to other RNA molecules, precursor transfer RNA (pre-tRNA) transcripts are subjected to extensive post-transcriptional processing before they are matured to fulfill their biological functions. In particular, intron-containing pre-tRNAs undergo excision of the intervening sequence in two steps: first, the tRNA endonuclease generates 5’- and 3’-exons with 2’, 3’-cyclic phosphate and 5’-hydroxyl ends, respectively. In animals, the second step predominantly entails direct exon ligation in a reaction that preserves the phosphate group at the 3’ end of the 5’ exon (Fig. 1). Yet, who ligates tRNA exons during tRNA splicing?
We purified the tRNA ligase from HeLa cytoplasmic extracts using a sequence of classic chromatographic steps. To monitor ligase activity we assayed inter-strand ligation of a particular siRNA displaying 3’-phosphate and 5’-hydroxyl termini, predicting that the “siRNA-ligase” would in fact be the tRNA ligase. After purification, mass spectrometry and bioinformatics analysis, we arrived at a very promising candidate, HSPC117, a protein of unknown function. RNAi-mediated depletion of HSPC117 inhibited maturation of intron-containing pre-tRNAs both in vitro and in living cells. Importantly, we could ligate exon halves with immunoprecipitates from a stable cell line expressing a myc-tagged wild type, but not a mutant, inactive form of HSPC117 (Fig. 2). We were also able to assess the role of HSCP117 in living cells by monitoring the maturation of an inducible, de novo synthesized reporter pre-tRNA molecule. Therefore, we concluded that HSPC117 is the long sought catalytic component of the human tRNA ligase (Popow et al. 2011, see publication highlights).
The human tRNA ligase is a pentameric complex containing HSPC117 and four extra polypeptides, including the ATP-dependent RNA helicase DDX1 (Fig. 3).
The identification of the human tRNA ligase is the group’s most important recent achievement at IMBA. We value the finding itself and the enormous perspectives that it generates.
What is the plan for the tRNA ligase?
We want to address the function of the tRNA ligase in vivo by generating a conditional-knockout mouse and a mouse encoding an inactive tRNA ligase. We are also performing PAR-CLIP, i.e. Photoactivatable Ribonucleoside Enhanced Crosslinking and Immunoprecipitation, to identify RNA targets of the tRNA ligase in addition to tRNA exon halves. Among others, a potential candidate is the mRNA encoding Xbp1, an essential protein during the mammalian Unfolded Protein Response (UPR). This transcript undergoes non-canonical splicing requiring enzymatic cleavage by the endonuclease Ire1 and ligation by a yet unidentified RNA ligase (Fig. 4). Splicing of Xbp1 leads to a frameshift giving rise to a protein that travels to the nucleus and functions as a transcription factor to orchestrate the expression of genes involved in UPR. It is known that in yeast the tRNA ligase Rlg1 splices both tRNA exon halves and the Xbp1 homolog Hac-1. Therefore it is tempting to speculate that HSPC117 could be the ligase mediating UPR in humans. We are intensively testing this hypothesis with a recently developed in vitro system using human cell extracts.
Taken together, investigating the functions of the tRNA-ligase assures many years of exciting work!

- Figure 4 (click to view legend)
Figure 4: The “RNA side” of the unfolded protein response. The cartoon depicts events that take place upon endoplasmic reticulum stress. The endonuclease Ire1 cleaves Xbp1-mRNA with similar chemistry as the tRNA endonuclease cleaves pre-tRNAs, i.e. generating 2’,3’-cyclic phosphate and 5’ OH. An unknown ligase, which we are trying to identify, joins the two exons leading to a frameshift that results in a longer version of the Xbp1 protein that travels to the nucleus and functions as a transcription factor. Xbp1u, unspliced; Xbp1s, spliced.

- Figure 1 (click to view legend)
Figure 1: The animal tRNA splicing pathway. In humans, pre-tRNA introns are removed by the tetrameric endonuclease TSEN, composed of Sen2, Sen15, Sen34 and Sen54. Exon halves display a 2’,3’-cyclic phosphate at the end of the 5’ exon and a 5’-OH at the 5’ end of the 3’ exon. A tRNA ligase joins exon halves by direct ligation leading to a mature tRNA, with the phosphate group at the phosphodiester bond originating from the 2’, 3’-cyclic phosphate.

- Figure 2 (click to view legend)
Figure 2: HSPC117 is the essential subunit of the human tRNA ligase complex. Affinity-purification of c-myc-HSPC117 from stably transfected HeLa cell lines yields an immunoprecipitate (IP) able to ligate tRNA exon halves. IP of wild type (WT) or C122A mutant c-myc-HSPC117 were incubated with body-labeled tRNA exon halves. Mature tRNA is generated exclusively by IPs of WT c-myc-HSPC117. An IP prepared from a non-expressing clone was used as a negative control. The asterisk denotes an unrelated band.


