Glycan topology-switch of integrin activity – GlycoTopoSwitch

Glycosylation of proteins and lipids is essential for life. Glycans participate in cellular functions as diverse as cell adhesion, signaling, intracellular trafficking, host-pathogen interaction, and immunity. Many of these functions can be ascribed to cell surface glycoconjugates. Their dynamics is modulated in a carbohydrate and sugar-binding protein-dependent manner either by retention in lattices, or by intracellular distribution through endocytic uptake. For the latter, we have previously proposed a novel mechanism (termed GL-Lect) according to which sugar-binding proteins of the galectin family drive the glycolipid-dependent formation of tubular endocytic pits from which so-called clathrin-independent endocytic carriers are formed.

For the GL-Lect mechanism, a conundrum arises: How can a rather static modification such as glycosylation control a cellular process — endocytosis — that is highly dynamic and needs to be acutely regulated? We propose to address this intriguing point at the example of alpha5beta1 integrin. This cell adhesion molecule adopts 2 extreme conformational states: the extracellular matrix-bound fully active conformation, and the non-ligand bound inactive conformation. We have previously discovered that in migratory cells, only the inactive conformation of the integrin is trafficked via the retrograde route from the plasma membrane to the Golgi apparatus, from where it is then secreted in a polarized manner to the leading edge. Inhibition of retrograde transport leads to a loss of its polarized distribution to the leading edge and an inhibition of persistent cell migration.

How would the inactive integrin conformation be specifically recognized? In preliminary studies we have now found that a galectin family member, galectin-3 (Gal3), preferentially interacts on cells and in detergent micelles with the inactive alpha5beta1 integrin conformation. This suggests the exciting possibility that the interaction Gal3-integrin is determined not only by the core glycosylation of the latter, but also by acute glycan rearrangements in space that are induced by conformational changes (glycan topology switch). Only certain glycan topologies on a given alpha5beta1 integrin molecule would enable a proximity configuration of glycans such that Gal3 molecules can oligomerize, an event that is required for operating the GL-Lect mechanism.

We will address this glycan topology switch hypothesis based on a multidisciplinary approach in 4 work packages: i) Advanced live cell imaging using lattice light sheet microscopy with theoretical modeling to monitor the construction of endocytic pits and to determine key elements of cellular machinery that are involved (molecular choreography); ii) reconstitution of alpha5beta1 integrin in model membrane (lipid nanodiscs and giant unilammelar vesicles) to test in a rigorous and quantitative manner whether and how glycolipids can amplify the conformational state-specific recognition of alpha5beta1 integrin by Gal3; iii) high-resolution cryo-EM combined with crosslinking proteomics and glycoproteomics to deduce a prototype building plan underlying the glycan topology switch; iv) site-directed mutagenesis and synthetic biology approaches to quantitatively test predictions from the structural studies on endocytosis and cell migration.

Apart from affording stringent testing of a groundbreaking regulatory paradigm at the forefront of mechanistic glycobiology — the glycan topology switch hypothesis, the current program is also expected to enable a number of other key contributions, such as first complete structures of an integrin and of Gal3, and the first identification of cellular machinery that specifically interacts with the inactive integrin conformation. We thereby ambition to provide a radically new perspective on how acute glycan topology dynamics on cargo proteins controls conformational state-specific endocytosis and endocytic trafficking-dependent cell polarity.