From OpenWetWare

Jump to: navigation, search

EchiDNA 2014

Design of Cooperative Beacon


To design the two key components of our cooperative molecular biosensor: a barrel composed of 24 helices and the molecular beacons mounted upon it.


We wanted maximal flexibility possible from a single design so we wouldn't need to re-design or re-order parts, saving us time and money. To achieve this we opted for a modular design where the components could be interchanged easily and independently without changing the rest of the design. In particular we needed the following features to be modular:

  • The sequence targeted by our biosensor
  • The strength of the clip that holds the molecular beacons closed
  • The number of molecular beacons in the cooperative biosensor
  • The degree of cooperativity between molecular beacons
The first two design parameters were addressed in the single switch design, the third would be determined by the design of the barrel at the core of our biosensor, and the last by the design of mounted molecular beacons.
Fig. 1. Modular aspects of our design


Our model demonstrated that the number of switches in our cooperative biosensor would affect its overall behaviour. Therefore we wanted to design a barrel that could support a flexible number of switches, allowing us to explore and optimise the behaviour of our cooperative biosensor. We chose to create a barrel with 24 helices so that we could alter the number of sub-units by systematically including or excluding switches on the various helices (factors of 24 include 1, 2 , 4, 6, 8, 12, and 24).

The height of the barrel needed to be similar to the length of the molecular beacon (18-30nt). Combined with the number of helices, this constrained the size of the entire barrel. This necessitated a custom scaffold strand as the standard scaffold in DNA origami, M13, is far too long (around 7000nt). Therefore we developed a method for creating ssDNA scaffolds of any length. Since this scaffold could be of any arbitrary sequence, we decided to pay homage to the Bacterial Flagella Motor, by choosing as the scaffold for our barrel a section of the coding region of FliG, the protein involved in the switching mechanism of the BFM.

We took two different approaches to designing the staples necessary to fold our DNA scaffold into the barrel at the core of our biosensor.
Fig. 2. 3D model of spatially designed barrel

Spatial Approach

We created a 3D representation of DNA from first principles, which unlike CaDNAno's 'rules of thumb' pays no heed to approximations such as axial symmetry of DNA and includes fine structural details such as major and minor grooves. After placing helices in the arrangement we desired for a 24mer barrel we manually altered the positions and orientations of each helix to minimise the length of crossover links and converted this 3D spatial model into a CaDNAno File for computation of the scaffold and staple sequences.

Theoretical Approach

Fig. 3. Angles generated between adjacent helices by staples [from Castro, et al. (2011)]
We wanted to make a barrel, which meant that neighbouring helices had to align with specific angles between them. Neighbouring helices are connected by staple strands that move around a curve before connecting neighbouring helices. This means that the number of nucleotides in a staple is directly related to the angle between neighbouring helices, and thus to whether or not the helices fold naturally into a barrel.
Fig. 4. CanDo output of the theoretically designed Core Barrel.

We calculated that a 15˚ change in angle was required between neighbours, which allowed us to approximate the barrel as a 12mer with angles of 30˚. Based on this we generated a bunch of CaDNAno models with skips and inserts on different helices and modelled them with CanDo to select the design with minimal twisting.

We anticipate that both of these designs would form an adequate barrel to mount our molecular beacons, however, we chose the proceed with the spatial design as we wanted to test this novel method of generating staples.

Mounted Molecular Beacons

Now HERE'S the tricky part: we wanted to arrange our molecular beacons on the barrel and mechanically link them together. Importantly, we wanted the link between neighbours to be variable so that we could completely characterise the effect of cooperativity on the behaviour of the biosensor. To achieve this we used two symmetrical four-way junctions for each molecular beacon. The four-way junctions allowed us to connect each molecular beacon with its neighbours and with the barrel.
Fig. 4. Design and specifications for cooperatively-linked subunit

We wanted to be able to vary the strength of the band connecting molecular beacons, so we ordered a range of bands with 2 and 8 single polyT in their middle, as well as a set of 'null' bands that should prevent any cooperativity between neighbours by removing the physical connection.

Given all these constraints, we used a script for NUPACK design to generate all the remaining undefined sequences in each four-way junction. We then used NUPACK analysis to determine the theoretical yield of each junction in isolation for equimolar rations of oligos. We re-iterated this design-analysis loop until we had a sequence set that had high theoretical yields.

With 12 sub-units included, the whole sensor should look like this:

Click here to go back to the Lab Book Overview

Personal tools