I have always been fascinated by the fact that close to one half of the single domain proteins in the Protein Data Bank have their N- and C-terminal elements in close proximity. Some years ago, Krishna and Englander pointed out that this number is rather high. In fact, it is much higher than what one would expect on a random probability basis. The exact reasons for this peculiar observation are still being debated:
Now, if we go 30 years back, we would find a classic study by Creighton, which showed a clean cyclization of the BPTI protein (its structure is shown below). Remarkably, the cyclization was triggered by a “middle of the road” carbodiimide reagent, so there is nothing fancy in this chemistry. In the graphic below you can clearly see that the ends of BPTI are fairly close to each other and can be forced to cyclize without much trouble.
I wonder why we do not see more naturally occurring cyclic enzymes (although David Craik has been talking about some really cool ones of late)? The artificially cyclized versions can be significantly more stable than the corresponding non-cyclized ones, which can be seen time and again, for instance in the following paper by Howarth, although here the authors used a fairly “fancy” cyclization brew (you just can’t beat Creighton’s carbodiimide…):