Post-translational modifications (PTMs) come in many flavors and broadly influence protein behavior. Mapping the positions of a specific PTM from the single-protein to the genome-wide level represents an important problem in modern biochemistry. One powerful method for such mapping is mass spectrometry (MS).

A cartoon depiction of a designer protease cleaving at a PTM. Credit: Marina Corral Spence/Springer Nature

In a typical MS workflow, proteins are digested by the protease trypsin, which selectively cleaves the C-terminal peptide bonds of Arg and Lys. The mass-to-charge ratio for the resulting peptides is then determined and offset from the expected ratio based on the primary sequence can often be specifically attributed to one or more PTMs. The PTM must occur on a peptide of appropriate size, which is determined by the positions of Arg and Lys within the protein. When Arg and Lys spacing do not yield the correct peptide sizes, coverage gaps can conceal PTMs and hinder mapping. Surprisingly, very few proteases exist as backups for trypsin.

To address this challenge, Brian Paegel and Duc Tran, a postdoc fellow in Paegel's laboratory at Scripps Research Institute in Florida, sought to develop better protease tools for studying modified proteins. According to Paegel, the work was born from conversations he had in 2008 with colleague and coauthor Valerie Cavett about the limitations in mass analysis of proteins, which led him to “start sketching schemes to evolve proteolytic function.”

Their idea was to develop proteases that specifically cleave proteins at the site of target PTMs, which would facilitate precise determination of the PTM positions by MS. The notion of such 'designer proteases' was not new, but no such mutants of trypsin exist—possibly because trypsin has many residues involved in binding, making it a challenging target for directed evolution.

The group developed a trypsin mutant that cleaves at citrulline, a PTM implicated in several important epigenetic and immunological functions. To develop this mutant, they had to develop a method for identifying protease mutants with the desired activity and high specificity. “We tried and failed at many strategies to identify functional mutants before arriving at the activity-based screening strategy,” recalls Paegel. The activity-based screen involved using a home-made bisamide rhodamine probe, which fluoresces only upon cleavage at citrulline and allowed them to identify mutants that were highly specific.

Paegel says they got “a lucky break” when they found a mutation in the protease that improves the expression both in vitro and in cells; and they now include this mutation in all their designer proteases. He calls this a stark reminder of Leslie Orgel's second rule, “evolution is cleverer than you are.”

After successfully generating the citrulline-specific protease, the team demonstrated that this protease allows them to precisely determine 12 sites of citrullination in the enzyme protein arginine deiminase 4 and 25 sites of citrullination in fibrinogen, two of which were previously unknown. They are currently testing the protease on a proteome-wide scale.

Future directions involve generalizing the directed evolution approach to proteases that target other PTMs. Paegel and his team are interested in developing proteases that recognize PTMs that are labile in MS and are thus difficult to observe by changes in mass-to-charge ratio. Examples include phosphorylations of serine, threonine or sulfated tyrosine. These upcoming tools, as well as the citrulline-specific protease, signal a promising future for precision PTM mapping.