A new method introduces ubiquitin or ubiquitin-like proteins at specific sites in any protein without the requirement of the cellular ubiquitylation machinery. This will help decipher the code by which these modifications control cellular processes.
The enzymatic attachment of small proteins related to ubiquitin is one of the most complex post-translational modifications. Ubiquitin and ubiquitin-like (Ubl) modifiers target proteins to begin degradation, control signaling pathways, recruit DNA repair machinery, regulate RNA splicing, and tune a multitude of other cellular processes1,2,3,4. Curiously, the same modifiers can be involved in different processes. However, the underlying code that specifies the different fates of the target proteins is unclear. Several reasons may contribute to this. First, the modifications themselves are baroquely complex. Second, the machinery that makes ubiquitin and Ubl modifications is promiscuous. Third, and finally, it can be hard to characterize ubiquitin modifications analytically. Fottner et al.5 describe a method to introduce defined modifications at specific places in proteins in vitro and in cells, making it possible to dissect the biochemical mechanism by which they act.
Ubiquitin and Ubl modifiers are attached to their targets through isopeptide bonds between the C terminus of the Ubl protein and ε-amino groups of Lys residues. In some proteins, a single lysine is modified, whereas in others, multiple lysines are modified. Ubiquitin contains seven lysine residues, and these provide attachment sites for further ubiquitin molecules, allowing formation of polyubiquitin chains. These chains branch when several Lys residues in the same ubiquitin moiety are modified. Ubiquitin and Ubl modifications are reversed by a large class of deubiquitylating enzymes (DUBs), allowing cells to fine-tune these modifications dynamically6. As a result, at least in the case of ubiquitin, one must think in terms of broad classes of modifications triggering related fates rather than of single, precisely defined signals corresponding to unique and specific processes. Ubiquitin and Ubl modifications are attached by a cascade of three enzymes, referred to as E1, E2, and E3, acting sequentially2 (Fig. 1a). The E1 activates the modifier, and the E2, with the help of an E3, then transfers it to the target protein. Mammalian cells encode a handful of E1 enzymes, tens of E2 enzymes and hundreds of E3 enzymes. One ubiquitylating enzyme can synthesize different modifications depending on substrate and context, and different enzymes can act on the same target protein.
Deconvoluting these signals has been complicated by the fact that it is difficult to synthesize proteins with specific ubiquitin or Ubl modifications. This is where Fottner et al.5 have brought us an important step forward by providing a versatile method to attach Ubl proteins to specific Lys residues in native target proteins through the natural isopeptide bond. The authors achieve this by elegantly combining enzymatic protein ligation by sortase7 with genetic-code expansion (amber suppression) to introduce unnatural amino acids8 (Fig. 1b). Extensive optimization and refinement of the individual steps made it possible to implement the method robustly both in vitro and in cells. Sortase is a bacterial transpeptidase that can be used to ligate protein sequences to each other. The enzyme recognizes a five-amino-acid sequence, cleaves it, and transfers the resulting N-terminal fragment to the α-amino group of Gly-Gly peptides as the receptor. Fottner et al.5 noticed that the C-terminal ends of Ubl proteins resemble the sortase recognition sequence and can be detected by sortase efficiently through a couple of conservative mutations. They then synthesized a Lys derivative in which a Gly-Gly peptide is attached to the ε-amino group of the Lys residue. The primary amino group of the first Gly is protected by an azido group that can be removed by cell-permeable phosphine. The Lys derivative is introduced into the target protein by amber suppression using a pyrrolysyl tRNA synthetase evolved for this purpose8. An optimized sortase enzyme can then fuse the Ubl protein with the recognition motif to the Gly-Gly-receptor in the target protein efficiently and specifically.
The technology is versatile and robust. In vitro, the authors were able to synthesize di-ubiquitin conjugates linked through Lys6, Lys11, Lys33, Lys48 and Lys63. These conjugates are resistant to cleavage by deubiquitylating enzymes but seem to otherwise behave like natural diubiquitin molecules chemically and biologically. The overall reaction proceeds surprisingly efficiently, so that sufficiently large amounts of modified proteins can be synthesized to characterize their interaction with components of the cellular machinery quantitatively. It should be possible to use the same method to attach defined ubiquitin chains to target proteins by combining this approach with existing technology9,10.
The method also works in Escherichia coli and in mammalian culture cells when the modified sortase, the target protein with the amber codon at the desired location, and the modified Ubl protein are co-expressed, at least when GFP or proliferating cell nuclear antigen (PCNA) is the target and ubiquitin or small ubiquitin-like modifier (SUMO) is the modifier. Doing this in E. coli is much easier than purifying the individual components and may thus be a convenient way to produce modified proteins for biochemical or structural analysis. Implementation in mammalian cell culture (HEK293T cells here) is still reasonably efficient, leading to more than 10% of the target protein being modified. The fact that the target protein only reacts after the modified Lys residues is deprotected makes it possible to control the timing of the modification to some extent. Thus, the system may also turn out to be a useful tool to decipher the mechanism of the biological processes controlled by Ubl modification in cells.
van der Veen, A. G. & Ploegh, H. L. Annu. Rev. Biochem. 81, 323–357 (2012).
Komander, D. & Rape, M. Annu. Rev. Biochem. 81, 203–229 (2012).
Herhaus, L. & Dikic, I. EMBO Rep. 16, 1071–1083 (2015).
Yu, H. & Matouschek, A. Annu. Rev. Biophys. 46, 149–173 (2017).
Fottner, M. et al. Nat. Chem. Biol. https://doi.org/10.1038/s41589-019-0227-4 (2019)
Mevissen, T. E. T. & Komander, D. Annu. Rev. Biochem. 86, 159–192 (2017).
Mazmanian, S. K., Liu, G., Ton-That, H. & Schneewind, O. Science 285, 760–763 (1999).
Liu, C. C. & Schultz, P. G. Annu. Rev. Biochem. 79, 413–444 (2010).
Pickart, C. M. & Raasi, S. Methods Enzymol. 399, 21–36 (2005).
Martinez-Fonts, K. & Matouschek, A. Biochemistry 55, 1898–1908 (2016).
A.M. is a paid consultant for Kymera Therapeutics.
About this article
Cite this article
Singh Gautam, A.K., Matouschek, A. Decoding without the cipher. Nat Chem Biol 15, 210–212 (2019). https://doi.org/10.1038/s41589-019-0230-9