Difference between revisions of "Team:UiOslo Norway/HP/Silver"

Our project has had several issues that we have had to consider, some of which were:
- Patient confidentiality and availability of our product to the general public
- Optimization of our assay to achieve a clinically relevant detection range
- Proving that our set-up safely tests for ESBL-producing bacteria
- Creating and testing positive controls for PhoneLab that do not involve actual dangerous pathogens
- Addressing antibiotic resistance potential on a genetic level in addition to phenotypic resistance

All of the above problems have been points of thought and investigation to us. Even though there was not enough time to come up with working solutions for all of them, several of them have been addressed, some of them even borderline solved and integrated into our final design. For the investigation part of said issues, see below. For integrative design of solved problems, see “Gold”.

Patient confidentiality:
Along with the birth of our idea, to create a product that would have access to peoples personal devices, and inform them as to whether resistant bacteria are present in them or not, came the realization that the eventual generation and storage of data would have to operate on a high level of security. To investigate this, we brainstormed around the issue, and reached out to Gissle Hannemyr, an expert in software security at the University of Oslo. This is something that in general has been of great importance to us; to utilize the potential and the potent expertise that resides in the institution that we are lucky enough to be a part of.

However, he did not respond immediately, and this ended up being something that we kept in mind, and would have moved forward with if we were to continue with the project.

Detection limits:
The first thing we did in respect to clinically relevant detection limits, was to contact experts to find out the bacterial number one can expect in urinary tract infections. The numbers that returned (averaging around 10^5/10^6 bacteria/mL) did not correlate with the detection limits we saw in our initial experiments with nitrocefin.

Thus, we proceeded to better our set-up. To do so, we investigated different factors that could improve the detection limit. In the end, by using a high concentration of nitrocefin and lysing the bacteria, we achieved a detection limit of clinical significance.

Testing our set-up with ESBL-producing bacteria:
This was important to us; as a proof of concept, we formed our set-up using “regular” AmpR bacteria. But, eventually, we wanted to move on to the big bads; clinical isolates of ESBL-producing bacteria. To do this, several things had to be done. First of all, we had to get our hands on both the bacteria themselves and information about them. This was achieved largely through corresponding with Ørjan Samuelsen at the Norwegian Center of Competence on Antibiotic Resistance in Tromsø. Second, ESBL-producing bacteria classify as bio-safety level 2, meaning that we had to thoroughly think through and plan whatever we wanted to do with them. This was done by conferring with the experienced people working in our lab, walking them through our set-up and having them supervise us when we were handling the bacteria.

We did in the end attempt a run of our set-up with clinical isolates. This was not entirely successful, but we learned a lot from the experience, and we feel certain that we would have achieved clear results from this if we had time to do it again. For more information about the strains that were used, the set-up and the results, see under the “Lab”-section.

Positive controls:
The first positive control that came to mind when designing our project, was to use bacteria with known resistance. Naturally, this was not something we desired to do - we do not want our test to rely on distribution of resistant bacteria. To work around this, we decided to purify a representative protein, a β-lactamase. A purified enzyme could, in theory, serve as a very sensitive positive control, and would be safe to carry around.

This was achieved; we successfully purified a class A β-lactamase using a biobrick from the Calgary team of 2013 (see the “biobrick” section). We also moved on to prove that even very low concentrations of the protein produced a reliable signal with nitrocefin, that it was inhibited by clavulenic acid (CVA, the inhibitor for class A enzymes in the Penta Well test), and integrated this into our hardware and software.

We also want to stress the issue that we created the chassis, in form of biobricks, to also express class B and C ESBLs, but did not do this due to bio-safety concerns. Expressing those proteins would mean creating dangerous, resistant bacteria in our lab, which we did not want to do unnecessarily. Therefore, we settled on expressing and testing for class A, and could easily also have done so for the other classes.

Addressing resistance on a genetical level:
Upon the realization that the Penta Well test deals only with the enzymatic activity of antibiotic resistance, we decided to attempt detecting also the genes causing said enzymatic activity. This led to the CRISPR/Cas9 design described further under “Gold”.