Our saxitoxin biosensor is based on an AMP biosensor that joins an AMP aptamer domain and a DNAzyme domain by a few bases, called linkers. These linkers are extremely important for the biosensor to success since they are responsible for the correct folding of a functional 3D structure. A functional 3D structure includes a correct sequestration of the catalytic domain when there's no ligand present, and the most important feature of these type of biosensors: the free energy released when the aptamer binds to the ligand must be close enough to the free energy needed to separate the linker from the DNAzyme domain (un-sequestration) and let it catalyze the REDOX reaction.
Designing the linker region for the Aptazyme is a very difficult task. We could even say that it is impossible for a human to design proper linkers without computer help in a reasonable time. That's why we needed an automated process that could help us with our linker region design.
Heidelberg's 2015 team, with their project "CATCH IT IF YOU CAN", developed a software called JAWS (Joining Aptamers Without Selection) that basically automated the random process of selecting DNA bases that accomplish all the requirements mentioned above (see more at http://2015.igem.org/Team:Heidelberg/software/jaws). We used this software to model our linker region, joining the saxitoxin Aptamer with the DNAzyme domain. JAWS gave us 4 linker regions that satisfied the requirements with high probability.
Above are shown the sequences provided by JAWS, bold letters are the linkers modeled the software.
Did it work?
Apart from the linkers modeled by JAWS, we designed a linker ourselves. For this, we analyzed the linker used in the AMP aptazyme biosensor and we simply changed the bases of the linker that hybridized with the AMP aptamer domain, for bases that should hybridize with the saxitoxin aptamer domain. We contrasted the performance of the saxitoxin aptazymes with our likers versus the aptazymes with the linkers provided by JAWS. The results indicated that the models made by the software were significantly better than the one designed by us (see details at 'RESULTS' section).
If we really wanted to design an aptazyme that can correctly respond to the exposition to different levels of saxitoxin, varying the colorimetric response in function of the toxin concentration, it was absolutely necessary that the linker was designed by this specific software. Even though the response wasn't exactly what we wanted, it is clearly shown that we are on the right path. In other words, the modeling provided by this software was a key part in the development of our project and future improvement of it.
We are highly interested in working with Heidelberg's team on improving both their software and our aptazyme with this new experimental results we had based on their previous work. See you next year, Heidelberg!