In Vivo Nanopatterning of Membrane Proteins

Specific activation of T cells is initiated in the contact zone with an antigen presenting cell, the so-called immunological synapse. The T cell receptor recognizes ist cognate antigen presented by MHCII molecules on the antigen-presenting cell, which initiates a plethora of highly regulated protein interactions. Early signaling is accompanied by spatial reorganization of proteins within the synapse, leading to a large-scale segregation into microclusters and supramolecular activation clusters. Currently, interactions are inferred from the colocalization of proteins within the developing synapse via two-color fluorescence microscopy. That approach, however, does not measure the interaction per se, and thus is highly prone to false positive results. We propose here to measure protein-protein interactions during early T cell signaling directly by slightly retarding one of the interaction partners ("bait"), and recording the co-retardation of a second fluorescently tagged protein ("prey"). The idea is based on a method previously introduced by us to study protein-protein interactions in the live cell plasma membrane using protein micropatterning (Schwarzenbacher et al., Nat. Methods 5:1053 (2008)). For this, a bait-specific ligand is immobilized on a glass surface in a characteristic micropattern. When cells are grown on such substrates, the bait located in the plasma membrane follows these patterns. Interaction with a fluorescent prey leads to the rearrangement of the prey to form the same patterns, which can be easily detected via total internal reflection fluorescence microscopy. Up to now, the method was predominantly used to obtain a static and averaged picture of a particular interaction, i.e. no spatial or temporal variations were addressed. However, as our approach allows for live cell analysis, spatiotemporal changes of protein interactions can be directly recorded, what shall be the subject of this proposal. We will study interactions between the T cell receptor and a selected choice of potential interaction partners during the formation of the immunological synapse between the T cell and a functionalized surface. Such mimicries of an antigen-presenting cell are frequently used in the community, and have been successfully established also in our lab (Huppa et. Al, Nature, in press). Our idea is to provide the cell with a micro- or nanostructured substrate, which allows on the one hand for the formation of spatiotemporal heterogeneities, and on the other hand for sufficient retardation of the bait molecules so that interaction with fluorescent prey can be analyzed by ist co-patterning. To achieve this, it will be necessary to attach the bait to the surface via a reversible linkage with an adjustable off-rate. The NTA-Ni2+-oligohistidine complex will be used for reversible ligation, and ist affinity will be down-modulated to the value required for proper functioning by applying free histidine. Analysis of spatial heterogeneities in the protein-protein interaction requires the interrogation of contrast values at a spatial frequency which is higher than the desired resolution of the map. Our method was established using a feature size of 3m, which is insufficient to resolve details within spatial features of the immunological synapse. To improve the resolution down to length-scales characteristic of the synapse (~1m), feature sizes of 250nm or smaller have to be employed. We propose here to use nanocontact printing for this purpose, which shall be set up as the basis for our patterning technology. The miniaturization strategies will be implemented in two steps: i) Features will be generated, which can be imaged by classical diffraction-limited optics (~250nm). Ii) Features below ~250nm, referred to as nanopatterns, cannot be imaged by classical optics devices. Nanopatterns will be read out by photoactivation localization microscopy (PALM), a new technology developed to obtain subdiffraction resolution in light microscopy by photoactivation and localization of single fluorophores. The resolution is then only limited by the localization precision of single molecules, which is ~50nm.

The specific activation of T cells begins in the area of contact with the antigen-presenting cell (APC), the so-called immunological synapse. The central process is the binding of the T cell receptor to antigen, which is presented by MHCII molecules on the surface of APCs. This binding stimulates the recruitment of a large number of signaling molecules to the synapse and their activation. The aim of this project was to investigate the timing of this recruitment process and the spatial arrangement of the proteins involved. On the one hand we used the method of protein micropatterning recently developed by us, on the other hand high-resolution single-molecule microscopy. The micropatterning method allows proteins to be captured in the living cell membrane and immobilized at specific sites by growing the cells on microstructured glass slides. With this method we wanted to find out how far cellular proteins and lipids influence each other. In fact, there have been suspicions that very small, dynamic lipid rafts exist in the cell membrane for around 20 years; however, their detection was not experimentally possible. Using the micropatterning technique, we were able to show for the first time that such rafts do not exist in the local environment of specific lipid-anchored proteins, which were considered to be markers of rafts. Next, we applied the method to T cells. Here we succeeded in precisely determining the time course of the recruitment of signaling molecules. Finally, we attempted to further miniaturize protein microstructures on glass slides. New stamping materials were developed for this purpose, enabling us to produce structures of less than 100 nm by microcontact printing. This makes it possible to measure molecular recruitment processes even in cellular structures whose size is below the resolution of light microscopy. To carry out such measurements, it was also necessary to establish novel microscopy methods whose resolution is below the diffraction limit. For this, we set up single-molecule localization microscopy in our lab. However, our analysis has shown that this internationally established method is susceptible to imaging artifacts resulting from the multiple counting of single molecular signals. Thus, the claim of seemingly aggregated proteins can be the result of overcounting problems. As a solid basis for the interpretation of our measurement data, we therefore first developed a new analysis method for such high-resolution images. Furthermore, we applied this method to the T cell receptor: contrary to previous studies, we could find no evidence for an non-uniform distribution of the molecule in the cell membrane, which sheds a completely new light on the T cell signaling processes.