Our detection system is based on the mechanism of toehold switches. Toehold switches are synthetic mRNA molecules, with riboswitch functionality. The hairpin-like structure at the 5’ end regulates the translation of a reporter protein coded by the 3’ end of the mRNA molecule. Reporter translation is only allowed upon the unfolding of the hairpin, which is triggered by the binding of a specific RNA sequence to a complementary site of the switch.
Cell-Free Paper-Based Biosensor
The simplicity of the design and the operating principle of toehold switches renders this mechanism a highly specific and modular tool with a vast variety of applications. Following the work of Pardee et al., the toehold switches are incorporated in the biosensor as their DNA predecessors, solving the problem of the lack of stability RNA molecules present. Incorporating this mechanism in a cell-free transcription and translation system and freeze-drying the construct on paper generates a robust, genetically engineered biosensor. The system is activated upon rehydration from the processed patient sample, maximizing its storage time and stability.
Having the foundation of our mechanism in our hands, we needed to translate our idea into actual DNA sequences and fix every step of the process
Our sequence design was based on determining the MERS-CoV genome sequences that would be highly specific for the virus detection. Those sequences would serve as our trigger RNAs. Complementary to that, the corresponding toehold switches created for detecting these viral sequences needed to present a certain thermodynamic profile, in order to favor the best possible functionality of the mechanism. Through this process, every possible pair of toehold switch - trigger RNA from MERS-CoV was ranked and finally the 4 most promising out of them were chosen to be laboratory tested.
Perhaps the most important point in our mechanism design was deciding on the reporter protein. The enzyme chosen was trehalase (TreA). The first reason for this choice is the simplification and accuracy of signal indication provided by this enzyme. Trehalase hydrolyzes trehalose to glucose, which can be easily measured by a commercial strip glucotest for diabetes mellitus, providing the diagnosis.
The second reason for this choice was the prospect of minimizing the waiting time to diagnosis. The most common reporter used in toehold switch diagnostic systems is β-galactosidase, which is a 1023aa protein. A larger reporter tends to delay the final signal, increasing the transcription / translation load, as well as inhibiting stereochemically the approach of the trigger RNA to the toehold switch. Trehalase is a 535aa protein, thus is expected to give a faster diagnosis.
Attempting to further minimize the reporter coding sequence on the switch, we decided to use trehalase as a split reporter molecule, dividing it into two fragments, based on the mechanism of protein-fragment complementation assays (PCAs). The smaller fragment (TreA-a, 66aa long) was added in the toehold switch structure as the protein coding sequence, while the larger one (TreA-b, 456aa long) was expressed and purified. The two separate fragments do not present catalytic activity. Combining the two fragments would create TreA as a functional enzyme. However, these fragments do not have the ability of self re-assembling, therefore they were fused with a pair of leucine zippers. The zippers consist of two intertwining alpha-helices, with a certain complementary motif of hydrophobic / hydrophilic amino acids, which forces them to bind strongly. Subsequently, the presence of the trigger RNA activates the expression of TreA-a, which binds to TreA-b through zipper interactions, assembling in order to create the functional enzyme.
Assembling The Pieces Of The Puzzle
After designing the individual parts of our synthetic biological circuit, the next step was choosing an assembly method. This choice was based on the fact that we demanded no “scars” between our parts, in order to create a logical flow in our sequences without redundancies. Thus, we chose the method of Gibson Assembly. The operating principle of this method is based on sequence homology, connecting fragments that present overlaps on both ends. Gibson Assembly also has the advantage of speeding up the whole process, since it performs a practically one-pot, isothermal cloning in maximum 4 hours, regardless of the number or size of the parts to be assembled.
Toehold Switch Testing
The next step in our project design was planning a way to test the functionality of the toehold switches that had been designed. We decided to use a cell-free in-vitro transcription / translation system for this testing. In this system, each toehold switch tested would be transcribed by default, but it would only be translated if the trigger RNA added in the system could actually activate the switch. Following up, samples from this system, containing trehalase in case of a successful toehold switch, would be assayed with trehalose, leading to glucose production. In case of a switch with the split reporter, an intermediate step of assembling TreA-a and TreA-b into TreA would be demanded.
The final step of the process is the glucose detection. It was chosen to be performed through two different assays (DNS, commercial glucose meter). We generally wanted to use a method with high sensitivity for our lab testings, since being able to detect low glucose concentrations implies the ability to detect low amounts of trehalase, meaning less time needed for its expression. The DNS method has a very low threshold, thus it appears as the most convenient for a rapid response. The threshold of the commercial glucose meter is much higher, which would not be practical for day-to-day lab testings, but it offers a greater certainty in the final diagnosis, since surpassing a greater glucose concentration threshold signifies an undoubtedly functional switch and most importantly, an unquestionable diagnosis.