Peptidomics: Bridging the gap between proteome and metabolome

744 Proteomics 2006, 6, 744–747

REPORTPeptidomics: Bridging the gap between

proteome and metabolome

Mikhail Soloviev and Paul Finch

School of Biological Sciences, Royal Holloway University of London, London, UK

Studies of naturally occurring peptides and protein profiling by ‘classical’ proteomics are linkedby common analytical objectives and methodologies. The first international workshop “Pepti-domics: methods and applications” held on 6–7th September 2005 at Royal Holloway Universityof London confirmed that the science of peptidomics is a rapidly developing activity of highinterest to both academia and industry. This meeting featured talks by over 20 leading interna-tional scientists detailing methods and typical applications, including newly-developed cap-abilities for protein and peptide analyses. It provided a definition of the scope of the subject interms of current and future technologies together with applications ranging from studies ofdefined biological extracts to complete ecosystems. The proceedings of this meeting, speakers’contact details and other relevant information can be accessed at: www.rhul.ac.uk/biosci/meet-ings.

Received: December 2, 2005Accepted: December 4, 2005

Keywords:

Biomarker / Peptide mixtures / Peptidomics / Protein degradation

Introduction

Peptides per se play a central role in many biological pro-cesses and some classes of such biologically active peptides,e.g. hormones, cytokines and growth factors, have beenknown and studied for years. The term peptidomics is rela-tively new and was first mentioned in the literature less thanfive years ago. Logically, and to the uninitiated, peptidomicsis the specification of the complement of peptides of a cell(organelle, tissue or organism). The importance of smallpolypeptides in biological processes and the technical chal-lenges posed by their analysis are sufficient to justify therecognition of this activity, but, as established by MikhailSoloviev (Royal Holloway University of London) in his open-ing address and subsequent talk, there are other dimensionsto the concept. Conventional approaches to quantitative pro-teomics involving affinity-based assays at the protein level oranalysis of tryptic digests at the peptide level suffer from anumber of serious limitations; these can potentially be over-

come by working on complete or selectively depleted peptidemixtures (“affinity peptidomics” or “combinatorial pepti-domics” respectively). The meeting thus embraced a range ofobjectives and methodologies linked by the analysis andproperties of peptides.

Peptidomics – the missing link betweenproteome and metabolome

The first two speakers overviewed the recent history andoutlined the challenges of peptide analysis. Andrey Karelin(Shemyakin-Ovchinnikov Institute, Moscow) provided adetailed analysis of tissue specific peptide pools as well aspeptide pools generated by cell cultures. The peptides (over300 were examined) showed biological activity towards cul-tured cells in respect of proliferation rate, toxicity and differ-entiation. These effects, in some cases the same as those ofthe complete protein, were concentration–dependent andnot simply additive when mixtures of peptides were tested.

Michael Juergens (Biovision, Hannover) described thecurrent state of the art in peptide analysis and the company’sproprietary Peptidomics technologies, which are in factknown protein analysis techniques (e.g. multidimensionalpeptide separation, MS/MS analysis or Edman sequencing)applied to the study of native peptides.

Correspondence: Dr. Mikhail Soloviev, School of BiologicalSciences, Royal Holloway University of London, Egham, Surrey,TW20 0EX, UKE-mail: [email protected]: +44-1784-434326

Abbreviation: ETD, electron transfer dissociation

2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.proteomics-journal.com

DOI 10.1002/pmic.200500878

Proteomics 2006, 6, 744–747 Report 745

Dr. Pedro R. Cutillas (Ludwig Institute for Cancer Re-search, London) described the advantages of expanding tra-ditional proteomics technologies to the analysis of peptidepools in studying renal Fanconi Syndrome (a conditionresulting from a variety of genetic or environmental causes).Dr. Cutillas used a combination of MALDI-TOF-MS, LC-MS/MS (with and without labelling with ICAT reagents), and 2–DE to compare isolated protein and peptide fractions fromnormal and Dent’s Disease urines.

Harald Mischak (Mosaiques diagnostics, Hannover)described the use of capillary electrophoresis coupled tomass spectrometry and proprietary software solutions for ahigh throughput analysis and evaluation of peptides pres-ent in urine samples. Typically the method allows theanalysis of over 1000 polypeptides per sample with highreproducibility and yields sets of 10–100 peptide bio-markers for each distinct disease. To this end Dr. Mischakand co-workers have generated a human urinary polypep-tide map–a database which contains information on thefrequency and abundance of polypeptides present in urine,as well as disease-specific changes of individual polypep-tides.

A convincing case presented by Klaus Rumpel (Pfizer)put peptidomics along with other established “omics” tech-nologies, namely transcriptomics, proteomics and metabo-lomics and highlighted its growing role in the pharmaceu-tical industry. The human peptidome (.10 000) and meta-bolome (,10 000) are relatively small compared to thetranscriptome (,100 000) or the proteome (.1 000 000) andtherefore can be used to great advantage (a major one beingthe speed of analysis) in combination with other “omics”technologies in the discovery of biomarkers, of which “pat-terns” are becoming increasingly common.

Technology development for highthroughput protein and peptide analysis

Unlike proteins, peptide pools appear to be more dynamicin their nature. Peptides are often produced by degradationof precursor proteins, both specialised proteins and proteinswith other functions (such as receptors, enzymes, structuraland transport proteins, etc.) in addition to the direct trans-lation from mRNAs. The dynamic nature of peptidomesoffers an opportunity to get an instant insight into the bio-chemical state of the cell or organism, but also presentssignificant difficulties of analysis. This meeting reported ona number of developments in peptidomics technologies.Mikhail Soloviev (Royal Holloway University of London)reviewed affinity peptidomics and combinatorial pepti-domics technologies and presented some examples of theirapplications. Affinity and combinatorial peptidomicsapproaches require only the availability of small peptidefragment(s), e.g. from proteolytic digestion, and are there-fore capable of a reliable analysis of denatured, partially-degraded proteins and protein fragments. These methods

are especially suitable for larger scale and routine (diagnos-tic) applications where inconsistency in storage and sampletreatment exists, or where samples are degraded to differentdegrees, as in forensic applications.

The development of protein microarray technologiesbased on 3-D microporous nitrocellulose was described byJens Beator (Whatman/Schleicher & Schuell), who alsoreviewed a few key applications, including FAST Quant

cytokine arrays, Mercator phosphoarray and serum bio-marker chip for multiplex protein and peptide quantification(including for breast cancer studies). Other technologicaldevelopments from Whatman/Schleicher & Schuell includeantibody screening arrays, surface antigen arrays for diag-nostic applications (e.g. microbial surface antigens or humanIg isotyping), allergy/antigen arrays, autoimmune diagnosticbiochip (which is currently in clinical validation stage),kinase target identification and substrate profiling arrays aswell as whole yeast proteome arrays.

A new approach for performing true quantitative func-tional validation of gene knockdown by coupling RNAi tech-nology to PROTEIN-AQUA and 18O peptide labelling tech-nologies was reported by Graham Scott (Sigma-Aldrich). Theapproach favoured by researchers from Sigma-Aldrich Bio-technology involves application of 18O labelling to geneknockout experiments to identify proteins involved in theknockdown cascade, followed by AQUA measurements ofthe absolute concentrations of the differentially expressedproteins and peptides.

An entirely opposite (to quantitative analysis) conceptwas presented by Ben Reed from Ciphergen. SELDI deriva-tized ‘affinity’ surfaces used in Ciphergen Protein Chips (theprinciple not dissimilar to affinity peptidomics) appears towork better following the normalisation of protein level inthe samples (e.g. sera) using ProteinEqualizer beads. Thisapproach might help in the identification of proteins, butwill obviously disallow quantification, which might severelylimit the range of the applications.

Laura C. Main (Bruker Daltonics, Coventry) describedrecent technological developments, namely monolithic col-umns integrated with nano LC-MS/MS mass spectrometerscapable of electron transfer dissociation (ETD). These newtools allow an ultra high-speed protein and peptide identifi-cation, de novo sequencing and the discovery and localiza-tion of posttranslational modifications (a collaborative effortwith Moredun Institute, Edinburgh and LC Packings, Dio-nex, UK). The unique advantage of the ETD is its capabilityto fragment and sequence peptides with ‘difficult’ sequencesor posttranslational modifications.

The development of the LOPIT (Localization of Organ-elle Proteins by Isotope Tagging) approach was described byKathryn Lilley (Cambridge Centre for Proteomics). LOPITuses the iTRAQ technique to localise organelle proteins byisotope tagging. LOPIT has enabled simultaneous assign-ment of proteins to multiple organelles; it has the potentialof wider applications in a range of eukaryote systems and canbe performed using standard LC-MS/MS equipment.


746 M. Soloviev and P. Finch Proteomics 2006, 6, 744–747

A truly integrated solution to protein and peptide studieswas presented by Brian Keenan (Protein Function Division,Bio-Rad Laboratories). PDQuest 8.0 is the latest release of theintegrated software package for 2-D analysis with gel multi-plexing and DIGE compatibility, spot cutter control and MSdata import capabilities. PDQuest 8.0 uses refined andimproved algorithms, which include advanced spot detectionand editing tools, Gaussian Quantitation (currently the mostaccurate method) and auto removal of unmatched spots inGaussian Image, auto gel warping and new statistical tools.

The significance of integrated software solutions in sys-tems biology applications was the focus of a talk by StephenDavid from Nonlinear Dynamics. The “modas” suite devel-oped by Nonlinear Dynamics is a scalable system that allowsconsistent processing of multiple 2-D experiments through acontrolled workflow.

The importance of the contemporary bioinformatics wasreiterated by Simon Hubbard (University of Manchester)who delivered an exciting presentation on chicken genomicsand proteomics/peptidomics studies, where these wereclosely interlinked through a massive informatics effort(covering 330 000 ESTs’ project, EORF programme advancedMS data analysis). His conclusion was that peptidomicsstudies can provide a significant feedback towards the geno-mics and that the bioinformatics and especially machinelearning approaches (such as information theory and Baye-sian classifiers) are promising tools at revealing peptidefragmentation pathways.

A shorter session was devoted to definitions and investi-gations of complex systems to which the methods of prote-omics are making vital contributions. John Doonan (JohnInnes Centre, Norwich) described a range of experimentalapproaches, including those of proteomics, to the dissectionof interactions and functions of proteins, especially cyclin-dependent protein kinases, in plant cell structure and divi-sion.

Functional peptides and proteins

Research efforts at the Shemyakin & Ovchinnikov Instituteof Bioorganic Chemistry and Vavilov Institute of GeneralGenetics (Moscow) and presented by Tsezi Egorov, resultedin the identification and purification of 49 antimicrobialpeptides (including 24 novel peptides) from seeds of Triticumkiharae. These form an important and ancient mechanism ofinnate resistance providing rapid and metabolically inex-pensive first line of defence against pathogens.

Of special interest to biomolecular scientists were severalexamples of integrating peptidomics and transcriptomicsreported by Chris Shaw from the School of Pharmacy,Queen’s University in Belfast. The unique and robust strate-gy involves transdermal electrical stimulation of frogs toharvest skin secretions from granular glands located in thetegument, followed by “shotgun” cDNA library cloning and

peptide purification and identification by using combina-tions of HPLC, LC-MS and MALDI-TOF-MS. A majority ofspecies reported are thought to secrete a biomolecular cock-tail containing up to a few hundred different peptide com-ponents, some of which may lead to the development ofnovel drugs.

Paul Wilmes (University of East Anglia) introduced theaudience to the concept of metaproteomics and its applica-tion to the activated sludge system, which led to the identifi-cation of a number of proteins linked to the metabolic trans-formations important for phosphorus removal. The applica-tion of peptidomics to environmental microbiology researchhas greatly expanded the study of ecosystems in whichmicroorganisms play a fundamental role (e.g. soil, oceans,freshwater, wastewater) and for which lab models are unat-tainable or unrealistic.

Another presentation on the theme of functional pro-tein expression was delivered by MingYue He (The Babra-ham Institute, Cambridge), who described ribosome dis-play technique and cell-free protein arrays, capable of sin-gle-step generation of protein arrays directly fromindividual PCR fragments or from immobilised DNA arraytemplates.

Targeting disease

The successful application of population based proteomics toinvestigate the plasma proteome in cancer patient popula-tions and in healthy subjects receiving nutritional supple-ments (e.g. antioxidants, folate, alpha-tocopherol) was report-ed by Helen Griffiths from Aston University. The reportedapproach relies on sample pooling and other preparativestrategies for improving signal to noise ratio, a combinationof 2-DE and MALDI-TOF analysis for detection and ELISAand western blotting for the validation of the identified bio-markers.

The approach taken by researchers form the Institute ofNeurology (London) and presented by Parmjit Jat was a highresolution differential proteomic analysis of conditionallyimmortalized rat embryo fibroblasts that have previouslybeen shown to undergo synchronous senescence versus pri-mary rat embryo fibroblasts which undergo replicativesenescence upon serial passaging. Parmjit Jat and co-work-ers moved closer to the unravelling of the underlying mech-anism that regulates the finite mitotic life span of somaticcells and the mechanism of changes occurring in cancer cellsby identifying a number of proteins, none of which havepreviously been recognized to be involved with cellularsenescence.

Per Andrén (Uppsala University, Sweden) described thediscovery of a number of markers potentially important inthe development and treatment of Parkinson’s Disease usingthe combination of proteomics (DIGE based), peptidomics(Ettan nanoLC MS/MS based) approaches and MALDI im-


Proteomics 2006, 6, 744–747 Report 747

aging. The remarkable developments and achievements ofthe Uppsala University team in the field of neuropepti-domics include the identification of novel neuropeptidesfrom hypothalamic tissue of rat and mouse, characterisationof their PTM status and their quantification in MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine) treated (dopa-mine depleted) brains.

Summary

The first international meeting devoted to peptidomics con-firmed that the science of peptidomics is a rapidly developingactivity of high interest to both academia and industry. Thismeeting provided a satisfying and intuitive synthesis of thediverse aspects of the methods and benefits of the study ofpeptides and peptide pools.


Documents

Peptidomics: Bridging the gap between proteome and metabolome