Sophie Jackson (University of Cambridge): Knots in Proteins

Most proteins have to fold into a highly specific three-dimensional structure in order to function. It has been known for decades that all the information necessary for a protein to fold correctly is embedded in its primary sequence, the sequence of amino acids within the protein chain. From the 1970s to the 1990s, scientists characterised the folding pathway of a number of proteins with very different structures and functions. During this time, it was considered impossible for a protein chain to fold into a knotted structure, however, in 2000 a number of deeply knotted protein structures were discovered overturning this long-held belief. Now, we know of many proteins that adopt a variety of knotted structures including some which form a trefoil knot (3-1), a figure-of-eight knot (4-1), a Gordian knot (5-2), a Stevedore knot (6-1) as well as a 7-1 knot. Thus, Nature has been able to find suitably efficient pathways by which these structures can form. The talk will discuss what we currently know from experimental and computational studies, on the stability (thermodynamic, kinetic, mechanical and against degradation by cellular recycling machines). In addition to the folding pathways of these unsual topologically complex proteins, as well as their functions. In particular, the evidence on whether there are any evolutionary advantages for a protein to be knotted will be highlighted. Finally, the talk will cover our current ability to either predict novel knotted protein structures using machine-learning techniques (AlphaFold) and our efforts at designing knotted proteins do novo.