Toehold switches were shown to be an effective construct for regulating RNA translation. Recently, Pardee et al1 also showed they could be used to detect viral RNA in cell-free paper based expression systems, effectively developing a low-cost, easy to use, and portable diagnostic test. Inspired by this result, we aimed to develop biosensors in the same vein as that developed by Pardee, but this time targeting proteins. We ended up reworking a classic ELISA with aptamers and a cell-free expression system.
In more details
An ELISA test uses a pair of antibodies specific to a target protein. The first antibody is immobilized on the bottom of a well. Then the test solution is incubated in the well along with the second antibody, which is conjugated to a reporter molecule such as a fluorophore. The target protein, if present in the test solution, will then bind both antibodies, forming an immobilized “sandwich” complex. Thus, after washing the well, this complex will persist and will be detectable because of the reporter molecule conjugated to the second antibody.
The problem is that antibodies are difficult to engineer for specific proteins. Current methods are lengthy and afford no guarantess on the affinities of the resulting antibodies. DNA aptamers, on the other hand, can be selected from large libraries of candidates through multiple rounds of SELEX. Thus by replacing antibodies with DNA aptamers, it is possible to retool our biosensor to detect virtually any protein.
A problem with biosensors is the usually fluorescent read-out. Indeed some parts of the world may not have access to fluorescence readers. Therefore we wanted our sensor to produce a clear and simple output. It could still be quantitative when using sophisticated machinery to measure the output, but it could also be qualitative, which is often good enough for diagnosis.
To adress this challenge we integrated toehold switches into our project. These switches can control the expression of a gene. In the presence of a trigger sequence expression is on, and otherwise it is off. So instead of attaching a fluorophore to our second aptamer, as in an ELISA, we attached a trigger sequence. Then in the case of a positive test, instead of measuring a fluorescent signal the trigger sequence would switch on the transcription and translation of lacZ, producing beta-galactosidase. This enzyme would then process its substrate into a colored product, and this color change would be visible by eye.
This scheme requires a way to transcribe and translate genes into proteins outside of cells. Just as Pardee did for his Zika virus sensor1, we used a cell-free expression system that is still able to process DNA into proteins, but does away with the complexity and biosafety hazards of working with live cells. Then because we wanted our sensor to be affordable, we improved on Pardee’s work by using crude E.Coli lysates instead of the highly expensive PURE system.
Finally, we realized that there was no convenient software available to design toehold switches, although Pardee did outline his design process in his paper. So in order to make the use of toehold switches easier for the iGEM community and for our own project, we automated Pardee’s design pipeline to generate toehold switch sequences for a desired trigger sequence.
Implications and applications
The main application of this biosensor will be in diagnostics. While our starting point was the Zika virus sensor developped by Pardee, we realized that RNA or DNA detection is not always the best way to diagnose diseases. Sometimes the viral RNA load in a sample might be much too low, or it might be rapidly degraded. For other bacterial or fungal pathogens there may be no DNA or RNA to detect at all. For this reason many diseases are diagnosed by looking for protein antigens, with the most common test being ELISA. But these laboratory protocols are not well suited for certain regions of the world. Our biosensor is meant to meet the need for a cheap and easy-to-use ELISA substitute and could thus be used for biomarker detection for example, as the methods used nowadays are quite expensive.2
Diagnostics need not be restricted to human health either. Crop diseases for example are a serious issue around the globe, and farmers with less means might not be able to catch these diseases in time without access to a lab. An easily distributed biosensor could allow these farmers to quickly identify whether their crops or harvests are affected and react accordingly.
Having more farmers able to report on the health of their crops would also help monitor the spread of crop diseases within a whole region. This could help avoid serious food shortages3. Additionally, although our sensor currently uses a single toehold switch to produce its colored output, adding a network of these switches could allow for more sophisticated signal processing. For example the sensor could target multiple proteins. Then depending on the combinations of proteins detected it could produce different outputs. So the sensor could discriminate between bacterial strains by looking at a multitude of antigens, allowing for more precise diagnosis.
Based on the inspiration we got from a workshop on including synthetic biology in high school curricula, we created a small toolbox that would enable teachers to give a live demonstration of various biomolecular processes. An easy to handle, high-impact collection of experiments that poses no biohazard as it doesn’t contain living cells. It is perfect for explaining concepts such as transcription, translation or gene regulation to high school kids. They can experience this all in a hands-on way, while at the same time developing an intuition for biology that students cannot get from books.
1 Pardee, Keith, et al. "Rapid, low-cost detection of Zika virus using programmable biomolecular components." Cell 165.5 (2016): 1255-1266.
2. Martinez, Andres W., et al. "Diagnostics for the developing world: microfluidic paper-based analytical devices." (2009): 3-10.
3. Fang, Yi, and Ramaraja P. Ramasamy. "Current and prospective methods for plant disease detection." Biosensors 5.3 (2015): 537-561.