BtgZI is a unique Type IIS restriction enzyme. It cuts far away from it’s recognition sequence. So far, in fact, that if you put a BsaI site next to a BtgZI site, the BtgZI cut location will be completely separate from the BsaI cut location!
The BtgZI sites can therefore act as a standard connection. If all BtgZI sites are defined, then any part is “switchable”. If your part, using GoldenGate definitions, is defined as a “fragment 1”, and you want it to be “fragment 3”, you can simply switch with BtgZI to that new definition.
However, this method has drawbacks. First off, BtgZI is extremely common in the E coli genome. In fact, it is over 20xs as common as BsaI is. This is unacceptable for my genome engineering work. In addition, redefining a definition is far easier using a technique like BASIC which doesn’t have the inconvenience of having to remove BtgZI sites.
In the past I have investigated using BtgZI methylase to block random BtgZI sites from cutting, but all in all it became a Rube Goldberg way of going about this problem. Therefore, I am working more on characterizing BASIC and related techniques than this inefficient method.