ESF ProteomeBinders Workshop: Affinity Proteomics
13-16 April 2007
Alpbach, Austria

Organisers:

Cheryl Smythe, Babraham Insititute, Cambridge, UK
Oda Stoevesandt , Babraham Insititute, Cambridge, UK
Mike Taussig, Babraham Insititute, Cambridge, UK

Report

Summary

The ESF ProteomeBinders workshop, held March 13-15 2007 in Alpbach (Austria)
was attended by 119 academic and industrial participants from Europe and the US.
The meeting reviewed the requirements of tools, binding molecules and technologies
for European affinity proteomics programmes. The meeting was co-funded by the
ESF programme on Frontiers of Functional Genomics and the EC FP6
ProteomeBinders coordination action. The programme consisted of two days of talks,
with time set aside for outdoor activities and continuing discussion during the early
afternoon, a layout which proved very popular with all participants. Conference
sessions reviewed the human proteome in overview, bioinformatics infrastructures for
proteomics, antibodies and alternative binders, affinity tools and methods, intrabodies
as tools for therapeutics and research, applications and quality control of binding
reagents, biomolecular resources in the context of biobanks, and future directions for
large binder resource projects.

Scientific Content

Opening session: Overview of the human proteome
In the opening talk of the meeting, Matthias Wilm (EMBL, Heidelberg, DE) gave an
overview of current developments for time resolved mass spectrometric analysis for
the monitoring of dynamics of protein complex formation in cellular signalling. In the
context of affinity proteomics, he pointed out the impossibility of monitoring whole,
cell lysates by mass spectrometry, due to high backgrounds – acknowledging the need
for panels of specific binding molecules for pre-fractionation. Mathias Uhlén (KTH,
Stockholm, SE) then proceded to describe the human protein atlas project, which
constructs a comprehensive, antibody-based protein atlas for expression and
localisation profiles in 48 normal human tissues and 20 different cancers. So far, more
than 1500 polyclonal, monospecific antibodies towards human proteins have been
raised, rigorously quality controlled and used of immunohistochemical staining on
tissue microarrays. In the human protein atlas programme, binders are raised against
peptide epitopes of ca. 50 amino acids lengths. Since 65 % of human proteins have a
sequence homology of as low as smaller than 15% for peptides of this length, this
approach is capable of generating highly specific binders. Andreas Plückthun
(University of Zürich, CH) then focussed on the generation of specific binders against
integral membrane proteins. Given that in classical immunisation the structural state
of solubilised membrane proteins is uncontrollable, in-vitro selection of binders is
particularly advantageous for this class of proteins.

In the bioinformatics session, Lydie Lane (Swiss Institute for Bioinformatics,
Geneva, CH) presented the UniProtKB/Swiss-Prot protein knowledgebase, with particular emphasis on the future goal of providing complete sets of annotated
proteins for various model organisms, and the ongoing Human Proteome Initiative,
which annotates all know human gene products and their counterparts in other
mammals. Complementing the efforts of SwissProt, Henning Hermjakob (EBI,
Cambridge, UK) reported on the progress of creating a knowledgebase for binding
reagents towards human proteins, based on the existing framework of the molecular
interactions (MI) format built by the HUPO Proteomics Standards Initiative.

The following afternoon session was devoted to the presentation of different types of
affinity reagents, both classical antibodies and alternatives. Ronald Frank
(Helmholtz Centre for Infection Research, Braunschweig, DE) reported on the
identification and application of low molecular weight compounds (peptides and
organics) as binding reagents to probe cellular functions by interference with normal
protein function; an approach termed “Chemical Biology”. In the context of
generation of chemical binders for wide coverage of the human proteome, he particularly underlined the impact of SPOT synthesis for both peptide and other chemistry and of scaffolded peptides mimicking multiple noncontinous binding sites. Larry Gold (Somalogic, Boulder, CO, USA) reported on the latest developments of nucleic acid aptamers as binding reagents, and their application for the detection of disease markers. While unspecific electrostatic interactions between polyanionic aptamers and basic patches of proteins have a half-life on the order of 10 s, half lives of cognate aptamer – protein pairs are on the order of 100 min. Moreover, use of photoaptamers with photoreactive nucleobase analoga within the binding site allows photoactivated covalent linking of the aptamer to its specific target protein. Arne Skerra (Technical University Munich, DE) then highlighted the use of scaffold binders, in particular anticalins. Protein scaffolds possess enough intrinsic folding stability to allow for the multiple substitutions or even insertions at the primary structural level. Their lack of immune effector functions (in contrast to classical antibodies) may be a significant advantage in theraputical applications. Anticalins are based on the the lipocalin beta-barrel, with 4 loops amenable to randomisation through PCR with degenerate primers. Being single chain, anticalins can be produced in high amounts in E.coli. Anticalins exhibit high thermal stability and target specificitiy and up to subnanomolar affinities. Stefan Dübel (Technical University of Braunschweig, DE) introduced the Antibody Factory, part of the German National Genome Research Network (NGFN). The overall aim of this platform of four closely cooperating projects is to develop a system for the generation of wide panels of binding reagents, which is affordable, automated and independent of antigen-delivery by end users. To this end, improved phage display vectors and large human antibody gene libraries (5 x 10^9 independent clones) have been constructed and evaluated. In a pilot study, > 300 binders to > 60 target proteins have been obtained. Andrew Bradbury (Los Alamos National Laboratory, NM, USA) described the successful selection of a sequence-independent antisulfotyrosine scFv from a phage display library of 8000 clones. The scFv was converted into a full length IgG and into an scFv-Alkaline Phosphatase fusion, both increasing the stability. Since it has proved to be extremely difficult to generate
antibodies able to recognise post-translational modifications independently of
sequence context by immunisation, the use of phage display provides proof of
principle for the use of this technology to develop similar reagents against other posttranslational modifications. Carl Borrebaeck (Lund University, SE) reported on
recent technological development of recombinant antibody microarrays and their
clinical application in the search for protein signatures that are associated with a particular malignancy. Such biomarkers can subsequently be used for, (i) early
diagnosis; (ii) predictive diagnosis, i.e. assessing the risk for tumor relapse and
disease progression; and (iii) predicting treatment resistance.

The morning session on the second meeting day reviewed novel affinity tools and
methods for proteomics applications. Thomas Laurell (University of Lund, SE)
presented two microarray formats enabling readout by both fluorescence and mass
spectrometry. In one format, a nanoporous silicon surface serves as a carrier for non-
contact printed antibody microarrays of ˜ 14000 spots/cm2. The 3D porous silicon
protein chip surface has excellent properties for manufacturing and assay
reproducibility. From the arrays, both fluorescence detection and, after high-speed
digestion, mass spectrometric analysis of bound species are possible. As a second
assay format, a MALDI target with both on-target solid phase extraction and in-target
fluid handling has been developed, enabling the enrichment of low-abundance
proteins from an integrated microcolumn of nanobeads. The target format is currently
designed in a 96-array format to meet industrial standards for analyte handling by
robotic pipetting. Christof Niemeyer (University of Dortmund, DE) presented a
variety of semisynthetic DNA-Protein conjugates for diagnostic applications. As a
common principle, DNA oligomers are used as building blocks for the sequence-
directed self-assembly of nanostructured architecture. For instance, conjugates of
DNA and proteins are being utilised for the functionalisation of microarray surfaces
via nucleic acid hybridisation. Ulf Landegren (University of Uppsala, SE) gave an
update on proximity ligation techniques for the sensitive detection of very low
abundant proteins in diagnostics, and of in-situ patterns of protein-protein interactions
and covalent posttranslational modifications. The technique employs antibodies or
other binding reagents coupled to oligonucleotides for proximity-dependent ligation
reactions depending on dual or triple recognition of target molecules. Current
detection limits are single infectious agents in solution, and even single protein
molecules or interacting pairs of proteins in situ. By a suitable choice of binding
reagents, non-protein molecules can also be detected, e.g. to monitor binding of
transcription factors to specific DNA sequences.

The session on Intrabodies was opened by Terence Rabbitts (MRC Laboratory of
Molecular Biology, Cambridge and Leeds University, UK). He described the
screening for and development of an optimal single domain antibody fragment
scaffold, for optimised interfering with protein-protein interactions in cancer cells.
The preclinical proof of principle for this therapeutic approach is provided by an anti-
RAS single VH domain that prevents tumour formation in a mouse model by
interfering with the activated RAS-PI3K interaction. Silvère van der Maarel (Leiden
University Medical School, NL) presented the application of single domain VHH
intrabodies in the context of a cellular model for oculopharyngeal muscular dystrophy
(OPMD), a paradigm for protein aggregation disorders. Intracellular expression of
specific VHH could prevent or reduce nuclear protein aggregation. Current focus is
on the (in vivo) use of VHH in the early diagnosis and therapy of disorders of
misfolded protein aggregation. Heinrich Leonhardt (Ludwig Maximilians
Universität Munich, DE) presented the application of intrabodies for the targeting and
tracing of antigens in living cells, by coupling of the epitope-recognition fragment of
Heavy-chain antibodies from Camelidae with a fluorescent protein domain. The
resulting “chromobodies” can be expressed in mammalian cells and recognise
antigens in different subcellular compartments. Even antigens embedded into inaccessible structures such as chromatin could be traced throughout S phase and
mitosis, demonstrating the suitability of chromobodies for live cell studies.

In the afternoon session on Applications and quality control, Erich Chevet
(Universite Bordeaux, FR) introduced the Alphascreen technology for the detection of
molecular interactions. Donor and acceptor beads are functionalised with binding
reagents. If the targets of the reagents interact, proximity of the beads can be detected by a chemoluminescence reaction. The platform provides a versatile, sensitive, time-resolved, homogeneous and miniaturisable means for binder assay development in high throughput. André Bernard (Institute for Micro- and Nanotechnologie, Buchs, CH) presented affinity contact printing and compact disc readers for protein patterning and affinity sensing. Affinity contact printing uses a structured elastomeric stamp derivatised with binders against the target molecules. After the target molecules have been captured, they are printed in a monomolecular layer from the elastomer onto a variety of surfaces. Feature-sizes in the sub-micrometer range can be achieved. For the detection of binding events, inexpensive classical silver or gold staining and standard compact disc (CD) reader technology (e.g. from a laptop) can be combined into a novel binding-assay system, substituting the established, but expensive, signal tags and microarray scanners used for classical microarray readout. Markus Templin (NMI at the University of Tübingen, DE) discussed binder characterisation and quality control for large binder resources in terms of specificity, affinity and cross-reactivity. In particular, anticipated applications for binders and the methods available for characterisation define the quality requirements. Screening systems and key parameters describing the properties and characteristics of binders were presented. The following session on Biomolecular Resources was opened by Leena Peltonen (University of Helsinki, FI) with an overview of possibilities in large scale population studies in Europe. The lack of integration of competence and resources to maximally utilise these distributed national resources was identified as the major
bottleneck in the European biobanking landscape. Additionally, she reviewed new
funding opportunities in the biobanking field provided by the 7th framework
programme of the EC, FP7. Gert-Jan van Ommen was replaced as a speaker by Kurt
Zatloukal (Medical University of Graz, AT) who presented the ESFRI roadmap
process and current plans for the joint European Research Infrastructure for
Biobanking and Biomolecular Resources. Whereas biobanks contain biological and
clinical samples themselves, biomolecular resources (including protein binding
reagents, ORF clone collections, siRNA libraries) are a crucial prerequisite to analyse
samples and extract the full value from biobanks. Martin Yuille (University of
Manchester, UK) reported on experience from the UK in the construction of research
infrastructure networks for biology in the age of “-omics”.

The concluding session of the meeting reviewed large binder projects and possible
future directions. Adam Clark (National Cancer Institute, MD, USA) had kindly
agreed to have his talk presented by Mike Taussig (Babraham Institute, Cambridge,
UK). The NCI’s Clinical Proteomic Technologies Initiative for Cancer is designed to
accelerate the translation of proteomics from a research tool into a reliable and robust
clinical application by improving protein measurement capabilities, evaluating
promising technologies for applicability in both analytical and clinical validation
studies, and providing a reagent resource composed of well characterised peptides,
proteins, and antibodies to the scientific community. Michael Lisurek (FMP, Berlin,
DE) introduced the German ChemBioNet, a network of chemists and biologist for chemical biology. The mission of ChemBioNet is to ensure that every chemist should
be able to know about the biological activity profile of their compounds and every
biologist should be able to get a small molecule tool to manipulate biological systems
in a dosage, time and spatially dependent manner. Marius Ueffing (GSF,
Neuherberg, DE) concluded the meeting by highlighting recurrent aspects of the
discussion. The availability of reliable binding reagents will enable quantitative
analytical approaches and will therefore potentiate systems analysis. “The binder of
the 21st century” has yet to be identified; however the type of binder chosen will
depend on the cost effectiveness of efforts to generate a binder and on fidelity
(sensitivity/selectivity) of the binder. As methods and technology drive biology, the
building of binding reagent infrastructures will be of crucial importance to realise the
potential of affinity proteomics.

Assessment of the results & impact of the event

Essential to the ambition of characterising fully the human proteome are systematic
and comprehensive collections of specific affinity reagents directed against all human
proteins, including splice variants and modifications. Although a large number of
affinity reagents are available commercially, their quality is often questionable and
only a fraction of the proteome is covered. In order for more targets to be examined,
there is a need for broad availability of panels of affinity reagents, including binders
against proteins of unknown functions. The most familiar affinity reagents are
antibodies and their fragments, but engineered forms of protein scaffolds and nucleic
acid aptamers with similar diversity and binding properties are becoming viable
alternatives. Recent initiatives in Europe and the USA have been established to
improve both the availability and quality of reagents for affinity proteomics, with the
ultimate aim of creating standardised collections of well validated binding molecules
for proteome analysis. These initiatives were presented side by side during the ESF
ProteomeBinders workshop on Affinity Proteomics. As well as coordinating affinity
reagent production through existing resources and technology providers, these
projects aim to benchmark key molecular entities, tools and applications, and
establish the bioinformatics framework and databases needed. The benefits of such
reagent resources will be seen in basic research, medicine and the biotechnology and
pharmaceutical industries.

The scientific and economic importance of affinity binder-based initiatives are
substantial, both for the near and the more distant future. Although the goals are
ambitious, the meeting identified no fundamental technological reasons why they
cannot be achieved. However, future efforts must be thoroughly coordinated and
organised. Given the numbers of binders required, an internationally coordinated
action would seem to be the only way to tackle this tremendous but feasible task, the
successor and similar in scale to the human genome sequencing project and
indispensable to deriving the full benefit from the vast of genomic information now
available. The ESF ProteomeBinders workshop on Affinity Proteomics brought
together a critical mass of leading European and US academic and industrial players
in affinity proteomics, contributing to global networking in the field.

Programme
Tuesday 13.3.2007 – Arrival and Registration

14:00 onwards: Registration open at the Conference Centre Alpbach.
18:30 onwards: Get together at Jakober
19:30 Workshop Opening Dinner at Jakober

Wednesday 14. 3. 2007 – First day of Workshop

Session 1 Overview of the human proteome
Chair: Mike Taussig

8:30 Mike Taussig: Welcome
The Babraham Institute, Cambridge, UK
8:35 Matthias Wilm: Proteomics is chasing the time line
EMBL, Heidelberg, Germany
9:00 Mathias Uhlén: A human protein atlas for normal and cancer tissues
Department of Biotechnology, AlbaNova University Center, Royal Institute of Technology, Stockholm, Sweden
9:25 Andreas Plückthun: Specific binders against integral membrane proteins
Dept Biochemistry, University of Zurich, Switzerland

Session 2 Bioinformatics infrastructures for proteomics
Chair: Toby Gibson

10:20 Lydie Lane: The UniProtKB-Swiss-Prot Human Proteomic Initiative
(HPI) as a support for the characterisation of the human proteome in health and disease
Swiss-Prot group, Swiss Institute of Bioinformatics, Geneva
10:45 Henning Hermjakob: Designing a binder database
Proteomics Services, European Bioinformatics Institute (EBI), Hinxton, UK
11:05 Discussion

Session 3 Antibodies and alternative binders
Chair: Marius Ueffing

16:15 Ronald Frank: ProteomeBinders link to Chemical Biology
Department of Chemical Biology, Helmholtz Centre for Infection Research, Braunschweig, Germany
16:40 Larry Gold: SomaLogic's progress with aptamer-based proteomic arrays
SomaLogic, Boulder, CO, USA
17:05 Arne Skerra: Anticalins and other alternative scaffolds for molecular recognition of proteome targets
Lehrstuhl für Biologische Chemie, Technische Universität München, Germany
17:35 Stefan Dübel: The German NGFN Antibody Factory
Department of Biotechnology Institute for Biochemistry and Biotechnology Technical University of Braunschweig, Germany
17:55 Andrew Bradbury: Selecting affinity reagents which recognise specific post-translational modifications independently of sequence context: the sulfotyrosine example
Los Alamos National Laboratory, New Mexico, USA
18:20 Carl Borrebaeck: Detecting cancer with recombinant antibody microarrays
Dept. of Immunotechnology, Lund University, Sweden
18:45 Discussion
19:30 Dinner at Reblaus

Thursday, 15.3.2007 – Second day of Workshop

Session 1: Affinity tools and methods
Chair: Dolores Cahill

8:30 Thomas Laurell: Microchip based affinity probing - Strategies for on-chip fluorescence and MALDI MS readout
Dept. Electrical Measurements, Div. Nanobiotechnology, Lund University
8:55 Christof Niemeyer: Semisynthetic DNA-Protein conjugates for diagnostic applications
Universität Dortmund, Biologisch-Chemische Mikrostrukturtechnik, Dortmund, Germany 9:20 Ulf Landegren: Assay formats for enhanced protein analyses
Department of Genetics and Pathology, the Rudbeck Laboratory, Uppsala University, Uppsala, Sweden.

Session 2: Intrabodies
Chair: Silvère van der Maarel

10:15 Terence Rabbitts: Interfering with protein-protein interactions in cancer cells with single domain antibody fragments
MRC Laboratory of Molecular Biology, Hills Road, Cambridge, UK and Leeds Institute of Molecular Medicine, Section of Experimental Therapeutics, St. James’s University Hospital, Leeds, UK
10:40 Silvère van der Maarel: Single domain antibody fragments for biomedical applications
Leiden University Medical Center, Department of Human Genetics
10:55 Heinrich Leonhardt: Targeting and tracing of antigens in living cells
with fluorescent nanobodies
Ludwig Maximilians University Munich, Department of Biology II, Planegg-Martinsried, Germany
11:10 Discussion

Session 3: Applications and quality control
Chair: Jörg Hoheisel

16:15 Eric Chevet: Alphascreen™: principles and applications
Team Avenir, INSERM U889, Université Bordeaux 2, Bordeaux, France
16:30 André Bernard: Affinity contact printing and compact disc readers for protein patterning and affinity sensing
Institute for Micro- and Nanotechnologie, Interstate University of Applied Sciences Buchs, Buchs, Switzerland
16:45 Markus Templin: Binder characterisation and quality control: specificity, affinity, cross-reactivity
Naturwissenschaftliches und Medizinisches Institut an der Universität Tübingen, NMI, Germany

Session 4: Biomolecular resources
Chair: Gert-Jan van Ommen

17:15 Leena Peltonen: European possibilities and bottlenecks in large scale population studies: Any new opportunities provided by FP7?
University of Helsinki and National Public Health Institute, Finland & The Broad Institute, MIT and Harvard, Cambridge, MA,USA
17:35 Gert-Jan van Ommen: The ESFRI process and research
infrastructures 'the biology way'
Center for Human and Clinical Genetics, Leiden University Medical Center, Center For Medical Systems Biology, Leiden NL
17:50 Martin Yuille: Building research infrastructure for big biology: some experiences from the UK
University of Manchester, UK DNA Banking Network

Session 5: Large binder projects – Future directions
Chair: Mathias Uhlén

18:00 Adam Clark: The National Cancer Institute’s Clinical Proteomic
Technologies Initiative for Cancer
NCI, National Institutes of Health (NIH), Bethesda, MD, USA
18:30 Michael Lisurek: ChemBioNet: Chemical biology supported by a network of chemists and biologists
FMP, Berlin-Buch, Germany
18:45 Marius Ueffing: Overview and …
GSF - National Research Center for Environment and Health, Neuherberg, Germany

19:00 … final discussion