Introduction: A Cell-free, Paper-based Biosensor

We think drug testing kits should be safe, affordable, and readily available to those at risk. Where tests are affordable and understandable to the end-user, accuracy, and thus safety, is compromised. It is clear that the current methods for checking drugs are not adequate. To change this we envisioned a test designed and modeled around the fundamental needs of the users. This is a test built from their personal input and experience, to create the most useful and appropriate device possible.

The current methods available for home testing of drug samples consist of chemical-based tests. One of the main suppliers of these kinds of tests is EZ Test Kits, who produce everything from tests for cocaine purity to testing for the presence of certain cannabinoids. While these tests are a very important resource, they have two main flaws. Firstly, the detection relies on interpreting a subjective colour change by eye, and secondly that each test is specific for one compound only. There are other very accurate methods used to test drug samples. These include mass spectrometry and other tests which use gas chromatography and infrared spectroscopy. Since these testing methods are expensive and not portable they cannot be used as a home testing method, and are instead incorporated in the police and border agencies enforcement.

The utilization of a cell-free system was the most appropriate choice for this project, for it was conducive to the biosensor being used outside of the laboratory in a safe and effective manner. Indeed, it is paramount that users of the biosensor are not exposed to potential pathogens, and that that the environment is protected against the accidental release of genetically modified organisms. Since our device is intended to be used primarily by people with little or no training, the best way to provide this level of safety and security was to ensure that no live organisms were incorporated in the test.

Avoiding the use of live organisms also allows for a degree of portability, for regulation would have otherwise hindered the implementation in non-laboratory environments, including the environments in which we intend to employ the device (see p&p for more - link to our story). Furthermore, there would be a long and costly legal process to get our biosensor the green light if it were to utilize genetically modified organisms.

The choice of a paper-based biosensor was also fundamental to its ability to compete on the free market, as this reduced the estimated cost to no more than 10 pence per test paper1.

Enzymes or Genes?

In the earliest stages of design, we contemplated using gene networks to operate our detection system; however, our human practices research quickly led us to re-evaluate this decision. Through conversations with our end users, as well as professionals working directly with the users, we quickly learnt that users would be seriously deterred from using the sensor if they were subjected to long waiting-times for results. Thus, it became clear that minimising time-to-detection was a priority for our design, and that using an enzyme system, with its increased speed of detection and stability compared to gene networks, was the best method to meet this criterion.

Producing the Enzymes

In order to decrease the cost of the biosensor, we decided not to purify the enzymes, because this is both a tedious and expensive process. Instead, we will place crude cell lysate on each of the bioactive zones. By transforming the plasmids in the chassis Escherichia coli BL21 (DE3), we can increase enzyme production. This is a result of it being a robust and non-pathogenic E. coli B strain able to grow vigorously in minimal media in comparison to the Top10s we completed our cloning in. It is also deficient in ompT and lon; two proteases that could cause problems with protein isolation2.

All protein expression is under the control of the lactose induced expression cassette (BBa_K314103) produced by iGEM10_Washington. This expression cassette includes a LacI generator and a lactose inducible promoter with the Elowitz standard RBS. By putting the lac operon in front of the enzyme coding sequence, we should be able to regulate transcription of the gene via production of the LacI repressor protein. IPTG, an analogue of lactose, can be used to induce protein expression3. Having a repressible promoter is key in situations where the proteins can be toxic to the cells in high concentrations4.

Immobilising the Enzymes

After considering the large amount of diffusion of the crude cell lysate through the paper, we thought about methods to immobilise the enzymes so they remain in the intended bioactive zones. This led us to fusing our enzymes to Cellulose Binding Domains (CBDs) to ensure the enzymes would stick to the paper.

CBDs are proteins that recognise and bind specifically to polysaccharides, such as cellulose5. Their advantages include: the fact that they are independent folding units, having inexpensive and abundant attachment matrices and, most importantly, having controllable binding affinities6. This means that we can make sure enzymes that work at different speeds are fused to the CBD with an appropriate binding affinity in order to give the optimal concentration of enzyme on paper. This also allows an inexpensive purification method by washing the paper with PBS after being incubated with the crude lysate as only the CBD fusions will remain on the paper.

RFC10 vs RFC25

One of the goals of our biosensor is to make it a proof of concept for future biosensors with interchangeable parts. This requires us to think about how every enzyme and CBD can be exchanged for a different enzyme or CBD. This standardisation was started with the BioBrick assembly method, such that every part has complementary terminals. The original BioBrick assembly method (RFC10), however, does not allow for the construction of protein fusions, as there is an 8-nucleotide scar between the two genes leaving a frameshift mutation7.

This led us to use the Freiburg format of RFC25 since everything we did relied on protein fusions. RFC25 is compatible with all existing parts in RFC10 and only leaves a 6-nucleotide scar, which allows the genes to stay in frame. The RFC25 prefix and suffix required certain restriction sites to not be in the gene to allow for fusions8. It also meant that our protein-coding genes could be in frame in a single plasmid so they will produce fused, functional proteins when transcribed.

Freeze drying

The final step in creating our biosensor is ensuring its mobility. This is done by freeze-drying all of the enzymes on paper. This process increases stability and shelf-life of enzymes so that freshness, and therefore reliability, of the biosensors is not compromised. It further involves freezing the sample, as well as two subsequent stages of drying9. Freeze-drying with 20% trehalose is a method of further increasing the stability of the enzymes10.

An adaptable design

Whilst Edinburgh iGEM 2015’s biosensor may leave a great legacy in improving drug safety, we hope the impact it has does not end there. Our biosensor is a fluid and adaptable model that can be easily manipulated to serve a range of analytical purposes, bringing synthetic biology to the forefront of detection sciences.

Through exchange of standard biological parts, the device can easily be produced with different enzymes by switching the genes in our backbone. This allows our biosensor to be important not only as a harm reduction tool for drug users, but as a proof of concept with much wider applications. Exchanging the enzymes in the biosensor could allow for our device to test for a multitude of compounds, be they drugs, contaminants or counterfeit-pharmaceuticals. (Find out more in Future Applications- link to future applications).

Our device is far more than a tool to improve the safety of a small group of user; its user-targeted design and enhanced biosafety allows our biosensor to bring synthetic biology out of the lab and into society in ways that have not been possible before.


1 Pelton, R. (2009) Bioactive Paper Provides Low Cost Platform for Diagnostics. Trends in Analytical Chemistry, 28(8), 925-942.
2 Chart, H., Smith, H. R., La Ragione, R. M., & Woodward, M. J. (2000). An investigation into the pathogenic properties of Escherichia coli strains BLR, BL21, DH5α and EQ1. Journal of applied microbiology, 89(6), 1048-1058.
3 K.S. Matthews, J.C. Nichols (1998). Lactose repressor protein: functional properties and structure. Prog Nucleic Acid Res Mol Biol, 58, 127–164
4 O'Connor, C. D., & Timmis, K. N. (1987). Highly repressible expression system for cloning genes that specify potentially toxic proteins. Journal of bacteriology, 169(10), 4457-4462.
5 Boraston, A. B., Bolam, D., Gilbert, H., & Davies, G. J. G. J. (2004). Carbohydrate-binding modules: fine-tuning polysaccharide recognition. Biochem. J, 382, 769-781.
6 Bayer, E. A., Morag, E., & Lamed, R. (1994). The cellulosome—a treasure-trove for biotechnology. Trends in biotechnology, 12(9), 379-386.
7 Anderson, J., Dueber, J. E., Leguia, M., Wu, G. C., Goler, J. A., Arkin, A. P., & Keasling, J. D. (2010). BglBricks: A flexible standard for biological part assembly. Journal of biological engineering, 4(1), 1-12.
8 Grünberg, R., Ferrar, T. S., van der Sloot, A. M., Constante, M., & Serrano, L. (2010). Building blocks for protein interaction devices. Nucleic acids research, gkq152.
9 Roy, I., & Gupta, M. N. (2004). Freeze‐drying of proteins: some emerging concerns. Biotechnology and applied biochemistry, 39(2), 165-177.
10 Mazzobre, M. F., & Buera, M. D. P. (1999). Combined effects of trehalose and cations on the thermal resistance of β-galactosidase in freeze-dried systems. Biochimica Et Biophysica Acta (BBA)-General Subjects, 1473(2), 337-344.