Available online at www.sciencedirect.com Protein TAILS: when termini tell tales of proteolysis and function 1,2,3 1,2,3 Philipp F Lange and Christopher M Overall Among the hundreds of posttranslational modifications, limited functions of bioactive proteins will be opaque and hence proteolysis, also known as processing, is special: It is hypotheses based on traditional shotgun analyses, may be irreversible, near ubiquitous, and by trimming peptide chains misleading or even worse, totally wrong. from their ends or cutting proteins into two, proteolysis forms shorter chains displaying new termini. The unique chemistry PTM of proteins constitutes a highly diverse and dynamic and location of a-amino-termini and carboxyl-termini in a regulatory layer affecting all aspects of a protein from protein engender special chemical and physical properties to a protein folding, localization, interaction and bioactivity protein. Hence, modification of protein termini is often to its stability and ultimately degradation. Therefore, each associated with new biological activities of a protein. We distinctly modified version of a protein, also called a protein highlight recent proteomic developments enabling high species, and not just the initial translated version, needs to throughput identification of protein termini. This has be considered as the functional units comprising the pro- revolutionized degradomics and protein characterization by teome [3]. The diversity of reversible and irreversible mapping the specificity of terminal modifications and of modifications as well as the extensive modification machin- proteases, and has been used to directly identify new protease ery [4] and the possibility of combinatorial effects substrates and molecular pathways altered by proteolysis. dramatically increase proteome complexity by several Addresses orders. Organisms as different as worm, fly and man have 1 University of British Columbia, Department of Oral Biological and comparable sized genomes yet show a great discrepancy in Medical Sciences, 4.401 Life Sciences Institute, 2350 Health Sciences phenotypic complexity. While splicing introduces bulk Mall, Vancouver, BC, Canada V6T 1Z3 2 complexity it might well be that the diversity created by University of British Columbia, Department of Biochemistry and Molecular Biology, 4.401 Life Sciences Institute, 2350 Health Sciences pinpoint posttranslational modifications accounts for the Mall, Vancouver, BC, Canada V6T 1Z3 observed phenotypic differences. Hence, advanced pro- 3 University of British Columbia, Centre for Blood Research, 4.401 Life teomics has potential to explain phenotypes where con- Sciences Institute, 2350 Health Sciences Mall, Vancouver, BC, Canada ventional genomics fall short — but it is not easy. V6T 1Z3 Corresponding author: Overall, Christopher M ([email protected]) Every modification adds to the functional diversity of the proteome by reversibly or irreversibly converting one protein species into another that potentially is a function- Current Opinion in Chemical Biology 2013, 17:73–82 ally distinct species. In this regard, limited proteolysis is This review comes from a themed issue on Omics special as it has the unique ability to irreversibly convert Edited by Matthew Bogyo and Pauline M Rudd one into two distinct protein species while at the same For a complete overview see the Issue and the Editorial time generating new protein termini serving as attach- ment sites for even further PTM. Available online 6th January 2013 1367-5931/$ – see front matter, # 2013 Elsevier Ltd. All rights Second only to ubiquitin ligases in number, proteases and reserved. their inhibitors constitute a large enzyme family with 567 http://dx.doi.org/10.1016/j.cbpa.2012.11.025 members in humans. In what has been termed the degradome, the assembly of all elements involved in proteolysis — proteases, inhibitors and the processed Introduction substrates — can now be specifically studied in high Proteomics has made astonishing advances in all areas throughput investigations termed degradomics [5 ]. Pro- including peptide enrichment, fractionation, mass spec- teases modify their substrates by hydrolysis of scissile trometry and data analysis — many of which are reviewed bonds releasing two peptide chains with the two amino in this issue. With mass spectrometry based proteomics it acids adjacent to the cleaved bond now becoming car- is now possible to identify more than 10,000 proteins from boxy-terminal or amino-terminal residues. Unlike most human cells [1,2]. Recent advances in the field of protein PTM attachment sites, the hydrolyzed peptide bond is post-translational modifications (PTMs) have uncovered not amenable for direct assessment. For limited proteol- their widespread occurrence and physiological relevance. ysis, termed processing, the site of modification is there- However, for comprehensive analysis of PTMs specific fore determined by identification of the ‘neo’ termini of peptide enrichment approaches and dedicated analyses the products. Consequently the degradome and termi- are required, without which PTMs are usually under- nome are mutually dependent, with the identification of sampled and overlooked, respectively. In the absence of termini adding considerable functional annotation to the functional annotation of proteins from PTMs many key proteome. www.sciencedirect.com Current Opinion in Chemical Biology 2013, 17:73–82 74 Omics Enriching for post-translational modifications based on the observation of one single peptide only. PTMs often occur at low stoichiometry and thus efficient For proteins having two or more such modifications, enrichment techniques are key for their successful and protein identification can often be made by two or more comprehensive identification. In general different chemi- different and unique peptides. However, for single cal affinities between the modified and unmodified peptides bearing a PTM, such as phosphopeptides, species are utilized for differential binding to a resin or unambiguous protein identification is problematic. chromatographic media yielding positive or negative For the identification of protein termini we and others selectivity and enrichment. All approaches share the introduced high confidence protein identification from common hurdle of unspecific carryover and loss following single peptide identifications based on multiple peptide binding to surfaces. A great advance for the enrichment of variants [7,8]. In the past ten years since its introduc- peptides bearing PTMs is the replacement of resins by tion [5] degradomics and its subfield, terminomics, have soluble hyper branched polyglycerol polymers leading to developed from a small field covered by only a few massively decreased nonspecific binding while increasing publications a year to a vibrant community publishing binding capacity [6 ]. over 40 papers in 2011 (Figure 1). For in depth com- parison of available mass spectrometry based methods Upon successful peptide enrichment mass spectrom- for the proteome-wide analysis of limited proteolysis etry is used for peptide and PTM identification. Unlike and their subsequent modification we refer to a recent identification of the entire protein by multiple peptides review by Huesgen and Overall [9 ] and by the in one shotgun experiment, identification of a specific accompanying paper in this issue from the Gaevert modification and often the protein bearing the PTM, is laboratory [10]. Figure 1 (a) •cell culture •lysate / secretome •cell culture •in vivo tissue + overexpressed protease + recombinant protease •in vivo cell + recombinant protease •in vivo fluid •+/+ vs. -/- 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 •Degradomics coined •ICAT protein shedding •positive selection • N-acetylation •hyperbranched •N-cyclization •cleavage kinetics •manual substrate curation polymers •negative selection for •automated •negative selection •positive selection terminal enrichment topography mapping •unbiased substrate •unbiased substrate identification identification (b)(c) Functional 50 Competence Assignment Substrate activated Identification 45 N pro Limiting peptide Complexity 40 10 N 35 P P Me Ac Cleavage Site 5 Specificity 30 10 C N 25 Terminal DAEF... Modifications ...EVKM Publications 20 Protein Termini Ac LTIS... Me MAEF... 15 Cleavage Kinetics 10 Data Integration 5 Protease Web 0 TopFIND 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Year Current Opinion in Chemical Biology Proteome-wide identification of protein termini over the decade: (a) timeline of key steps in degradomics and terminomics since its introduction. The top stream covers the type and complexity of the analytes, the bottom shows key advances. (b) Count of publications covering degradomics and terminomics over the last 10 years. (c) Summary of areas driven by advances in protein terminus identification covered in this review. Current Opinion in Chemical Biology 2013, 17:73–82 www.sciencedirect.com Protein termini and proteolysis Lange and Overall 75 Characterization of the proteolytic machinery maturation (removal of the initiator methionine, signal Since the function of a protease is inherently linked to the peptide and pro-peptide) [31 ]. More recently, in skin effect of proteolysis on its substrates, and since more than 50% of the >2000 proteins identified had evidence of half of all proteases have no annotated substrates in stable cleavage products in vivo [29 ]. MEROPS, the protease database (http://merops.sanger.- ac.uk), since 2000 a major focus has been in the identi-