Modelling complexity and redundancy in endocytic actin polymerisation

Actin is one of the most ubiquitous proteins of life and can form filaments which play crucial roles in a wide range of processes from cell division to intracellular trafficking. Formation of these filament networks is tightly controlled using a wide array of protein types, chief among them being nucleators. Nucleators facilitate the unfavourable first steps of filament formation and thus their regulation dictates when and where filamentous networks are produced. The central proline-rich region of Las17 (yeast homologue of human WASp) is thought to nucleate actin “mother filaments” at the endocytic sites. Arp2/3 – a potent nucleator activated by Las17 – can branch these mother filaments. Proline also constitutes the core binding region of SH3 domains which leaves the nucleating region of Las17 open to competitive regulation. Eleven Las17-binding SH3 domains are recruited to yeast endocytic sites. Five of these bind via a tandem of domains (three SH3s in Sla1 and two SH3s in Bzz1). We hypothesise that this “cloud” of SH3 domains can regulate the access of actin to the proline-rich region of Las17. However, the high number of proteins and interactions involved renders a purely experimental approach challenging.

Throughout this thesis, two agent-based models are built (one being a progression of the other) to test the veracity of our regulatory cloud hypothesis. Binding affinities were experimentally obtained to build the model, demonstrate the power of avidity conferred through tandem SH3 binding, and refine our Las17 nucleating mechanism. We identify that the weak interactions of the SH3 cloud can combine in effect – particularly complemented by the tandem binding of Sla1 and Bzz1 – to define a window of Las17 nucleating activity. This work suggests how endocytic SH3 domains can regulate endocytic progression whilst also furthering our understanding of the relatively unexplored nucleating mechanisms employed by Las17