We have three main research interests:
  1. to understand and manipulate peptide selection by major histocompatibility complex (MHC) class I molecules;
  2. to understand what governs the trafficking of transmembrane proteins (especially MHC class I) between intracellular compartments and the cell surface; and
  3. to develop novel methods for the analysis and manipulation of cells with microparticles or engineered surfaces for basic research and practical application.

These three areas often overlap, and we are especially interested in any synergies that develop from this interaction. We use methods from cell biology, biochemistry, biophysics, biotechnology, and computational biology.

Short summary:

Our favorite model system are major histocompatibility complex (MHC) class I molecules, which are present on all nucleated cells. They play a central role in the mammalian immune defence against viruses, intracellular bacteria, and cancer, since they carry fragments of many intracellular proteins to the surface of the cell, where they can be surveyed by cytotoxic T lymphocytes (CTL). If the CTL find in this way that a cell makes unusual proteins (for example, originating from a virus or tumor), then they induce it to undergo apoptosis. This way, the production sites of viruses are eliminated.

Class I molecules don’t just bind any peptide but only those that have the right length and sequence. Binding of such high-affinity peptides stabilizes the structure of class I molecules. We would like to know how the structure and dynamics of class I changes upon peptide binding. We have also designed small molecules that help class I molecules to fold and to bind and even exchange bound peptide.

In live cells, peptide binding depends on a set of chaperones called the peptide loading complex, and especially on one protein called tapasin. We are very interested in how this chaperone works.

The trafficking of class I molecules to and from the cell surface is very interesting since it is dependent on the presence of the bound peptide ligand. We are asking, for example, what changes are brought about in the structure of the class I molecule by the peptide, how these changes are read out by the cell, and how the decision is made to localize the class I molecule either inside the cell, or on its surface. We also investigate why different class I molecules are transported to the cell surface at different speeds.

Viruses like to inhibit class I peptide loading and transport. We investigate the protein gp40/m152 from the murine cytomegalovirus in order to find out how it keeps class I molecules inside the cell.

For our biotechnological delivery and sensing approaches, we mostly use micrometer-sized polyelectrolyte capsules. They are, for example, modified with antibodies to detect small molecules, or they are introduced into cells and opened inside. We also use especially patterned surfaces to manipulate membrane proteins of cells.

Some news articles that feature our work:
Our papers

are found on the Publications page, with some video abstracts on the Videos page.

Detailed description with links to our publications:

MHC class I molecules are ideal model systems for protein folding and trafficking since many well-studied allotypes exist and many reagents to study them are available. They are at the core of the antiviral and antitumor response and in therefore the focus of modern tumor immunotherapy.

Class I molecules select their peptides from a large intracellular pool. The binding of peptide changes the structure and the movements of a class I molecule. No crystal structures of peptide-free class I exist, but our molecular dynamics (MD) simulations have shown that without peptide, some parts of the peptide binding site are intrinsically mobile (natively unfolded)9,34. Occupation of one part of the binding site, the F pocket, by a partial ligand is sufficient to quench this flexibility29,41. Different class I allotypes (there are >10 000 in humans) have different degrees of flexibility, and we can predict this flexibility from the sequence by MD simulation23.

The understanding of peptide binding opens up the door to its manipulation. Using biochemistry with recombinant proteins and cells, we found that small molecules targeted to the F pocket can block the reassociation of partially dissociated peptides from class I and accelerate their dissociation by a thousandfold29,37. Such small molecules can be used for tumor epitope discovery, diagnosis, and therapy; they are being further developed with partner companies.

Peptide binding in the lumen of the endoplasmic reticulum (ER) is aided by the proteins of the peptide loading complex (see a great video from the group of David Williams), among them the chaperone tapasin, which imposes thermodynamic control on the peptide binding process and thus leads to stable peptide binding15,18. For peptide exchange by tapasin, we developed an exciting in vitro system18. We showed that binding and function of tapasin respond to the native unfolding of the class I peptide binding site; thus, some class I molecules require tapasin to a greater extent than others for peptide loading23,38.

Class I molecules bind peptides in the endoplasmic reticulum (ER) and bring them to the cell surface, traveling along the secretory pathway like all other membrane proteins. We have shown that trafficking of different class I allotypes is regulated by the ER chaperone matrix in a proteostatic fashion36. Along this way, quality control mechanisms exist that make sure that only folded and properly assembled proteins are transported. We investigate how suboptimally loaded class I molecules, i.e. those without peptides or with ill-fitting peptides, are recognized for intracellular retention or retrieval to the ER12. We found that empty class I molecules are recognized by calreticulin in the cis-Golgi17, and that this quality control system surveys the flexibility of the peptide binding site34.

Class I trafficking is also influenced by viruses. We have found that the gp40/m152 protein of the murine cytomegalovirus (mCMV) holds class I molecules in the ER-Golgi intermediate compartment by interacting with it43. We continue to investigate the molecular mechanism of this interaction, and the retention of gp40 itself.

At the end of their lifespan, class I molecules are endocytosed from the cell surface and destroyed in lysosomes. We have shown by microscopy and flow cytometry that loss of the light chain prevents recycling from early endosomes and leads to the rapid destruction of class I40, and we are finding out how the ‘free’  heavy chains are recognized inside the cell for routing to endosomes. This ongoing work is significant because class I endocytosis is known to affect the surface levels of other receptors.

Our biotechnology-related projects are mostly done in collaboration with Mathias Winterhalter at Jacobs University. They aim to develop novel techniques for cell research and application. We have demonstrated the introduction of hydrophilic solutes into mammalian cells by microencapsulation and controlled release16,20. Currently, we investigate novel assays that employ polyelectrolyte microcapsules for sensing and delivery, and the manipulation of cells through growth on patterned surfaces.

Some important collaborators past and present:
  • Adnane Achour, Karolinska Institute
  • Tim Elliott and Denise Boulanger, Southampton University Medical School
  • Peter van Endert, Institut Necker Enfants Malades, Paris
  • Hartmut Hengel, Freiburg University Medical School
  • Randy Schekman, University of California, Berkeley
  • Stefan Stevanović and Hans-Georg Rammensee, Tübingen University
  • Mathias Winterhalter, Jacobs University
  • Martin Zacharias, Munich Technical University

Thank you very much for cooperating with us!