Breaking symmetry Fakultät für Biologie - Digitale Hochschulschriften der LMU - Teil 05/06

Polarity is a fundamental feature of almost all cells. It generally refers to the asymmetric organization of several cellular components. The plasma membrane, for example, exhibits both a transbilayer and a lateral asymmetry in most eukaryotic cells. Lipids are asymmetrically distributed between the cytoplasmic and the extracellular leaflet of the membrane and segregate laterally together with specific proteins to form dynamic nanoscale assemblies, known as rafts. Polarity can also specifically describe the asymmetric distribution of key molecules within a cell. These molecules, known as polarity determinants, can orient a multitude of specialized cellular functions, such as cell shape, cell division and fate determination.
In the framework of this thesis, we aimed to reconstitute essential features of membrane unmixing and cell polarity with a "bottom-up" synthetic biology approach. We worked with both: pure lipid systems, whose unmixing is driven by the asymmetric distribution of lipids in the two leaflets, and a lipid-protein system, whose polarization is instead due to reaction-diffusion mechanisms. In both cases, we used Giant Unilamellar Vesicles (GUVs) and Sup- ported Lipid Bilayers (SLBs) to model biological membranes and employed modern biophys- ical techniques, such as fluorescence correlation spectroscopy, to quantitatively characterize lipid bilayers and protein-lipid interactions.
In the pure lipid systems, we first reconstituted membrane transbilayer asymmetry, applying a cyclodextrin-mediated lipid exchange method, which enables us to enrich membranes with lipids of choice. The enrichment of the membrane with sphingomyelin and/or cholesterol triggers the segregation of lipids into two coexisting asymmetric phases both in SLBs and GUVs, whereas exchanging different amounts of phosphatidylglycerol with the outer leaflet of the GUV membranes controls vesicle shape. Tuning the lipid content of model membranes revealed that small changes in the composition of one leaflet affect the overall lipid miscibility of the bilayer and that membrane shape transformations are possible also in absence of a protein machinery and as a consequence of the lipid redistribution in the membrane.
In the protein-lipid system, we aimed to reconstitute a minimal polarization system inspired by the C. elegans embryo at one-cell stage, which polarize along the anterior-posterior axis by sorting the PARtitioning defective (PAR) proteins into two distinct cortical domains. In this system polarity is maintained by the mutual inhibition between anterior (aPARs: PAR-3, PAR-6 and PKC-3) and posterior (pPARs: PAR-1, PAR-2 and LGL-1) PARs, which reciprocally antagonize their binding to the cortex, mutually excluding each other. We focused on LGL-1, which acts directly on PAR-6. Submitting LGL-1 to model membranes allowed us to identify a conserved region of the protein that binds negatively-charged membranes and to determine its lipid binding affinity and specificity. Selected LGL-1 mutants were then gen- erated to better understand the electrostatic mechanism involved in the membrane binding. LGL-1 was finally combined with PKC-3 to generate a functional membrane binding switch.