Technologies
	Team
	Home
	Publications
	Funding
	Links

Projects

Antibodies, or more commonly binder molecules, are inevitably important tools for the functional characterisation of human gene products. Next to immunisation of various laboratory animals, a number of in vitro selection techniques exist to obtain highly specific binders, of which phage display is the most common. Within the last few years, we have worked on a conveyer-belt type production pipeline for the generation of human recombinant antibody molecules applying the phage display technology.

Selection of phage display-derived antibodies

Within the NGFN 2 funding period we successfully participated in the SMP “Antibody Factory”, which set out to evaluate all necessary processes for establishing phage display of human antibody fragments as a routine method to generate binders against human gene products on large scale. Our subproject was concerned with the semi-automation of panning and evaluation procedures. We considered all necessary aspects ranging from the expression of in vivo biotinylated antigens in E. coli or mammalian cells and more recently Leishmania tarentolae as hosts. Further, we developed a selection process based on robotic manipulation of streptavidin-coated magnetic bead, as well as protein microarray based screening methods for fast determination of target-specificity by applying the Multiple Spotting Technique. Additionally, we developed a laboratory information management system (LIMS) to store all generated clones and relevant experimental data. Finally, we succeeded with the assembly of streamlined selection pipeline which now could go into production if funding were available.

Selection of binder molecules with inhibitory properties

This project is dedicated to the selection of binder molecules, which allow specific blocking of protein-protein interactions and was developed in close collaboration with the group of Dr. Sylvia Krobitsch in our department (now Otto Warburg Laboratory). We successfully established a combined in vitro / in vivo selection scheme to obtain inhibitory antibodies, called intrabodies. First, phage display libraries are enriched in vitro on immobilised target proteins and once the diversity of the libraries are reduced to a complexity amenable to reverse yeast-2-hybrid screens, the selection is continued in vivo, where the selection is carried out with interaction pairs of the target proteins. Resulting binders not only bind their respective target proteins, but also inhibit defined protein-protein interaction of the targets in vivo. The method is applied using protein-protein interaction pairs involved in Spinocerebellar Ataxia type 2 (Ataxia UK).

Exploring autoimmune antibody libraries

On the quest to obtain the best antibody scaffold for phage display, we have generated several new semi-synthetic antibody libraries in different formats and are currently assembling a large naïve antibody library exclusively from donors with autoimmune disorders. This group of individuals possess large number of antibodies directed against self-proteins and should therefore be a great resource for developing binders against human gene products in future. Furthermore, we are interested to find out if there is a correlation between individual V-gene usage and autoimmune disorders. For this purpose, we intend to select specific binders towards known autoantigens in different autoimmune disorders and combine the outcome of these selections with the knowledge of general antibody repertoire of autoimmune patients we obtain applying next generation sequencing technologies.

Discovery of antibody–antigen interaction pairs in human disease

Many human diseases are characterised by the presence of antibodies directed towards self-proteins. In most cases, the existence of self-reactive antibodies is not fully understood, as the pathogenic role of such antibodies is – with some exceptions – not known. This is also clearly demonstrated by the fact, that every individual has autoantibodies even without being affected by disease and that autoantigenicity patterns overlap (Figure 1). Furthermore, our knowledge about the role of certain autoantibodies in disease progression, whether being of significance or simply a bystander effect, is still vague.

In this respect, we focused our work in recent years in close collaboration with the Charité on the characterisation of autoantigenicity patterns in healthy and diseased individuals. We concentrated primarily on autoimmune disorders and the diseases under investigation ranged from systemic to organ-specific autoimmune diseases (BMBF-Nutrigenomics project), including systemic lupus erythematosus, rheumatoid arthritis, celiac disease and thyroiditis (Graves’ and Hashimoto). Further, screening with sera of dilated cardyomyopathy patients were performed (SFB-TR19 project) and currently autoantigenicity profiling for multiple sclerosis and Alzheimer’s disease are on the way (BMWi-ZIM project). Additionally, we expanded our efforts to identify biomarker sets for therapy response prediction in systemic autoimmune disorders on the example of second line treatment of rheumatoid arthritis with tumour necrosis factor alpha (TNFα) blocking agents.

Next to protein arrays, screening for autoantigens is additionally performed using cDNA expression libraries cloned in M13 or T7 display vectors presenting the recombinant proteins on the bacteriophage surface. Selection is carried out in an iterative process – essentially based on affinity enrichment – using patient-derived immunoglobulin fractions as selection targets and finally, mass sequencing the cDNA inserts applying next generation sequencing technology of individual bacteriophage molecules identifies the putative autoantigens (Figure 2).

Generating autoantigenicity patterns applying protein arrays

For basic screening we apply high-density protein macroarrays carrying >38.000 clones expressing human recombinant proteins, which are comprised of >10.000 different genes and splice variant thereof. Using this technology, which was developed in our department in the late 90’s, lead to the identification of ~1.600 clones expressing potential autoantigens against which healthy and diseased individuals reacted. Notably, the number of autoantigens detected in different disorders decrease with organ specificity. For instance, in rheumatoid arthritis and celiac disease, more than 500 different antigens were scored positive, while in thyroiditis, only ~ 200 clones reacted.

To elucidate the autoreactivity pattern of celiac disease patients in more detail, we have generated protein microarrays with a selection of 160 purified recombinant human proteins and controls, such as commonly used diagnostic markers for the disease (tissue transglutaminase and wheat gliadin) and compared the screening results of 142 patients with that of 50 healthy individuals. Finally, we could identify a number of autoantigens which might serve for diagnostic purposes in patients with IgA-deficiency in future. This is particularly useful, since coeliac disease patients with IgA-deficiency remain frequently undetected in routine diagnosis.

Identification of diagnostic markers for therapy response prediction

In systemic autoimmune disorders, treatment comprises the administration of corticosteroids or immunosuppressive medication. In rheumatoid arthritis, first line treatment is carried out with disease-modifying antirheumatic drugs (DMARDs), such as Methotrexate (MTX). Second line treatment of MTX-resistant patients is carried out with biologicals, mainly tumour necrosis factor alpha (TNFα) blocking agents, such as monoclonal antibodies (infliximab, adalimumab) or soluble TNFα-receptor (etanercept). The major drawback of treating rheumatoid arthritis with TNFα inhibitory biologicals is its high cost as well as the fact that 30-40% of treated individuals do not respond or only poorly respond to treatment. In a pilot study, we have identified a set of biomarkers according which we can potentially discriminate between therapy responders and non-responders and are currently evaluated (BMBF-KMU Innovativ). Furthermore, a clinical study is on its way to recruit 120 patients treated with etanercept (sponsored by Wyeth BioPharma GmbH). Screening autoantigenicity patterns with these specific patient sera shall allow the identification of therapy prediction biomarkers specific for etanercept.

last updated: 31 December, 2009