You'd Prefer An Argonaute

Translational Pausing Ensures Membrane Targeting and Cytoplasmic Splicing of XBP1u mRNA

Kota Yanagitani, Yukio Kimata, Hiroshi Kadokura, Kenji Kohno

Science Vol. 331 no. 6017 pp. 586-589, 4 February 2011.
DOI: 10.1126/science.1197142

This week’s articulate summary and analysis by Anna Drinnenberg:

This paper from Yanagitani et al. further characterizes a mechanism involving an unconventional splicing event of the XBP1 mRNA that controls a cellular response to the accumulation of unfolded proteins in the endoplasmic reticulum (ER). For this splicing event to occur, it is thought that the nascent XBP1u (u – unspliced) protein, while still part of the mRNA-ribosome-nascent chain (R-RNC) complex, recruits the whole complex to the ER membrane, where a protein localized within the membrane processes the XBP1u mRNA into its spliced XBP1s (s – spliced) form. The HR2 region of the XBP1u protein that was suggested to be important for this recruitment, however, is located at the very C-terminus of the protein. Therefore HR2 is exposed from the ribosomal tunnel for only a brief period before translation is finished, which leads to the question of how the R-RNC complex can still persist while being recruited to the ER by HR2.  Therefore, the authors hypothesize that a translational pause must occur to ensure sufficient time for the ER-recruitment of the R-RNC complex and splicing of the XBP1u mRNA.

Using in vitro studies, they convincingly showed a pause during translation of the XBP1u protein by detection of translational intermediates composed of a tRNA covalently attached to the nascent polypeptides, whereas translation of XBP1s protein that had a different C-terminal region lacking HR2 showed no delay. Furthermore, the authors narrowed down the region responsible for the translational pause to the evolutionary conserved C-terminal part of the XBP1u protein, namely the last 26-amino acids. Exchanging many of these amino acids for alanine decreased or abolished translational pausing, whereas mutating a serine residue at position 255 (S255A mutant) increased pausing, interestingly. The authors try to explain the effect of the S255A mutant by hypothesizing that this residue might ensure an appropriate efficiency of translational pausing to recruit the R-RNC complex, while preventing undesired translational arrest (which would not relieve the spliced mRNA). For all subsequent analysis they included two mutant constructs (in addition to the S255A construct) that nearly completely abolish translational pausing.

After showing that translational pausing also happens in vivo, they demonstrated that in vitro it also appeared to be required for efficient membrane recruitment through the HR2 region. An in vivo demonstration of the R-RNC recruitment would still be worthwhile since the complexity of the intracelluar environment through which such an R-RNC complex would have to traverse is certainly much greater than their in vitro system.

Returning to the molecular effects of membrane recruitment of the R-RNC complex, the authors showed that translation pausing is important, but not absolutely necessary, to ensure efficient splicing of the XBP1u mRNA. While there was certainly a decrease in splicing efficiency without pausing, the effect seemed to be relatively small. However, it is still possible that this subtle decrease in splicing efficiency has greater physiological consequences during a response to ER stress. Mutating a combination of the amino acids that contributed to translational pausing, instead of one at a time, might have also yielded bigger effects on splicing.

Overall, the authors performed a very thorough study showing translation pausing of the XBP1u mRNA and demonstrating its importance for splicing of the XBP1u mRNA. The authors speculate that a physical interaction between the nascent peptide and the ribosomal tunnel might explain the translational pause as it has been observed for the bacterial SecM and TnaC proteins. An important follow-up question is: Is translational pausing a more widespread phenomenon than can be predicted based on the amino acid composition of a protein? As was suggested in an accompanying perspective by David Ron and Koreaki Ito in Science, recent data mapping the progression of ribosomes across mRNAs at single nucleotide resolution (Ingolia et al., Science 2009) will be crucial in answering this question.