Biochemistry, physiology and disease of the tRNA splicing pathway in mammalian cells
In mice and humans some tRNAs are exclusively encoded as intron-containing genes, thus highlighting the importance of tRNA splicing as an essential process. Introns are excised in the nucleus by the tRNA splicing endonuclease (TSEN) complex. The resulting exon halves are subsequently ligated to generate mature tRNAs. We have combined chromatography, mass spectrometry, RNAi and phylogenetic analysis to identify key components of the human tRNA splicing pathway, and discovered the RNA kinase CLP1 as part of the TSEN complex (S. Weitzer and J. Martinez, 2007). We also discovered the tRNA ligase complex which has been elusive for 30 years, established RTCB/HSPC117 as the catalytic subunit (J. Popow et al., 2011 and 2012), and archease as an essential co-factor of the tRNA ligase complex (J. Popow et al., 2014).
Recently we showed that the tRNA ligase and archease play critical roles in the unfolded protein response and that CLP1, when mutated, causes neurological diseases both in humans and mice. In this annual report we describe the unexpected sensitivity of the tRNA ligase complex to oxidative stress and our attempts to identify a novel RNA processing activity associated with the mammalian RNA 3’ phosphate cyclase RTCD1.
The tRNA ligase complex is inhibited by oxidative stress
In collaboration with Josef Penninger’s group, we recently generated and analyzed a mouse encoding a catalytically dead version of the RNA-kinase CLP1, the product of the single point mutation K127A (Hanada et al., 2013). This mutation affects the interaction between CLP1 and the subunits of the tRNA splicing endonuclease. Consequently, the removal of tRNA introns is severely impaired in vitro. Hence it was surprising to note the accumulation of tRNA fragments, largely derived from tyrosine tRNAs and composed of the 5’ leader sequence and the 5’ exon. Interestingly, the overexpression of such tRNA fragments results in enhanced p53 activation in response to oxidative stress, a possible cause of p53-dependent cell death of motor neurons in Clp1 kinase-dead mutant mice. In principle, 5’ tRNA fragments should not accumulate in the presence of defective tRNA splicing endonuclease, unless the tRNA ligase is unable to join them to 3’ exon sequences. Importantly, the same type of fragments accumulate massively when cells are exposed to agents known to cause oxidative stress, such as hydrogen peroxide (H2O2) or menadione. Therefore we hypothesized that the tRNA ligase might be inhibited by oxidative stress, leading to the accumulation of tRNA fragments (Figure 1).
We tested tRNA ligation in vitro by exposing cells to H2O2. As shown in Figure 2A, the ligase activity was severely inhibited. We obtained a similar result by depleting RTCB by RNAi and monitoring the accumulation of tRNA fragments (Fig. 2B). These experiments indicate that 5’-leader-exon tRNA fragments accumulate due to the lack of ligation activity. We are currently investigating the chemical basis of such inhibition by means of mass spectrometry. In terms of biology, it is interesting to note that Nature has equipped the tRNA ligase complex with the ability to sense or be the target of oxidative stress.
A putative cyclic phosphodiesterase activity is associated with the RNA cyclase RTCD1
Last year we reported initial studies on RTCD1, a mammalian enzyme that resides primarily in the cytoplasm and acts on 3’ phosphate-ended RNAs to generate 2’, 3’-cyclic phosphates. While its in vivo function remains elusive – a knockout mouse is being analyzed in our laboratory – we have detected a novel enzymatic activity associated with RTCD1 that converts terminal 2’, 3’-cyclic phosphates into nucleotides displaying a 3’ OH group; the chemistry at the 2’ position remains to be elucidated. To identify such activity, we purified a FLAG-tagged RTCD1 complex on a cation exchange column and managed to separate two enzymatic activities: FLAG-RTCD1 eluted at ~250mM NaCl (Figure 3, upper panel) and the novel activity - a putative cyclic phosphodiesterase - eluted at ~500mM NaCl (Figure 3, lower panel). Candidates obtained from an ongoing mass spectrometry analysis will be depleted by means of RNA interference and cellular extracts will be inspected for the absence of such enzymatic activity. Identifying the RTCD1-associated putative cyclic phosphodiesterase should reveal a new component of the mammalian RNA repair system. One possible function of the dual activity complex is to target 3’ phosphate-ended RNAs to exonucleolytic degradation by successive conversion into 2’, 3’-cyclic phosphates, and further into nucleotides containing a 3’ OH group.