Antibiotic-resistant bacteria pose a growing threat in developing and industrial regions alike. While antibiotics are widely available, it often remains unclear which antibiotics the germs at hand are resistant against. Currently, there are two approaches for the treatment of resistant bacteria: Treatment with a panel of antibiotics, or targeted treatment based on an antibiogram. Treating with multiple antibiotics may give rise to additional resistances in already dangerous bacteria. An antibiogram takes at least 3 days before a clear result is available because of the slow bacterial growth, limiting time available for treatment.1
We designed a point-of-care diagnostic method for detecting antibiotic resistances in patient samples. To accelerate diagnosis, we decided to directly detect the resistance genes instead of the bacteria’s response to antibiotics. A common challenge in synthetic biology is speed and adaptability of biosensors and genetic circuits. We aimed our solution to be adaptable to any resistance gene, and to reliably detect and report its presence in as little time as possible. To keep the degrees of uncertainty at a minimum, we decided to import our system into the bacteria in question, for example using phages. By regulating translation instead of transcription, we aimed to ensure adaptability to a wide panel of resistance genes, and rapid response to their presence. Recently, split ribozymes were published as a technology able to regulate translation in response to presence of a target RNA. Split ribozymes are programmable using a guide RNA to detect any RNA, which renders them a great fit for our project. We designed split ribozymes to detect mRNA of native and resistance genes. To report detected resistance mRNA, we designed multiple reporters: GFP, a fluorescent protein, eforRED, a chromoprotein and the enzyme horseradish peroxidase.
We built our constructs using GoldenGate Assembly (GGA). Our first design did not work because we tried to assemble too many small parts in one GGA. Therefore, we changed our strategy and ordered bigger parts to be synthesized that were much easier to assemble in one reaction. In order to get these synthesized, we had to change the promoters to be different but in a similar strength range. If the promoters were the same the region of similarity is too big to get synthesized.
As proof of concept, we introduced our system into E. coli DH5alpha by transformation. In the finished application, we propose using phages. We confirmed that our GGAs worked with a colony PCR that showed mostly positive results. We measured fluorescence for GFP at an emission of 485 nm and excitation 515 nm as well as absorbance assays for the chromoprotein eforRED at a wavelength of 582 nm.
Our system did not show any meaningful activity. We were only able to measure GFP in the inducible GFP positive control where GFP is controlled by a T7 and lacO promoter. We did not get a measurement for eforRED even for the positive control. The lack of signal can be explained by the relatively weak promoter strength of the chosen promoters BBa_J23115 and BBa_J23116. Our model showed that with stronger promoters like BBa_J23101 and BBa_J23104 the signal strength is much higher as well. We designed our system to have a low background signal but that resulted in us getting no signal at all. So in the next engineering cycle we will need to use the stronger promoters. For eforRED it is also important to mention that chromoproteins need time to get visible and we wanted to measure our samples promptly as our end goal is a point of care diagnostic device. So chromoproteins might not be the way to go for a quick and reliable reporter.