Collaborations

Viability Assay with Azurin

This year’s iGEM team from ETH Zurich is working on a tumor microenvironment based approach to fight cancer as well. Their approach is responding to the lactate secreted by tumor cells and should result in the secretion of azurin by E. coli. Azurin has the ability to induce apoptosis by binding to p53 which results in intracellular accumulation of p53 and BAX. Elevated levels of p53 and BAX lead to the release of cytochrome c into the cytosol which initiates the caspase cascade process resulting in apoptosis (Punj et al., 2004).

In order to show the cytotoxic effect of azurin on tumor cells we, the iGEM team Freiburg, performed a killing curve experiment with azurin and HEK293T cells. A fixed amount of HEK cells were seeded into 12-well plates, with one plate for every counting day. Four different conditions were applied in this experiment, no azurin treatment (negative control), 1 µg/ml azurin, 10 µg/ml and 100 µg/ml azurin. Since azurin is not stable at 37°C, 2 % of glycerol was added to the medium for stabilization. Every second day, the medium was changed without washing the cells.

Figure 1: Viability curve with HEK293T cells.
Course of the used azurin concentrations in the killing experiment. The negative control was not treated with azurin. All experiments were performed in triplicates.

Figure 1 shows the results of the killing experiment. After 8 days, the untreated cells show constant viability of around 80 %, while azurin treated cells show decreased viability. For 1 µg/ml and 10 µg/ml azurin in medium, the cells show a loss of viability of more than 50 %. The highest azurin concentration (100 µg/ml) has a stronger killing effect on HEK cells. Hereby, viability goes down to 18 %, which indicates a possible working concentration for further experiments.

During our Human Practice efforts, we often faced the question “How far can a T cell migrate after being activated in a tumor microenvironment?”. Finding an answer to this question proved to be challenging as the prediction of T cells migration is not trivial. Looking for a possible collaboration with the iGEM team from ETH Zürich, they offered us to answer this question with their modeling expertise.

As a basis for their modeling literature values for the migration behavior of T cells were used (Miller et al., 2003).

Animated simulation of CARTEL^TM cell behavior in- and outside the tumor microenvironment.

At the beginning of the simulation, CARTEL^TM cells, depicted by dots, are randomly distributed inside the tumor microenvironment (grey circle). Blue indicates a fully activated cell and yellow a cell close to deactivation. It can be observed that after leaving the tumor microenvironment fast deactivation of the the CARTEL^TM cells takes place. This basic model indicates that cells migrate on the average 6 mm from the tumour before getting inactivated. This on the one hand underlines the increased safety of our CARTEL^TM cells and on the other hand this model provides a perfect basis for further refinements.

Mutagenesis rate

To improve current directed evolution approaches, iGEM team Heidelberg set out to make use of their newly designed method called PREDCEL (phage-related discontinuous evolution). For the collaboration with Heidelberg, we analyzed the mutagenesis level in E. coli containing different mutagenesis plasmids (MP1, MP2 and MP3). The plasmids enhance the mutation rate by the inhibition of DNA repair mechanisms (Badran & Liu, 2015).
It was assumed by team Heidelberg that the mutation rate, which is several magnitudes higher in E. coli containing these mutagenesis plasmids, might vary between laboratories.

According to the data provided by team Heidelberg and literature (Badran & Liu, 2015), MP3 was supposed to have the highest mutagenesis rate, followed by MP2 and MP1. Due to these plasmids, E. coli is able to rapidly develop antibiotic resistances. To test this ability, E. coli cells were transformed with one of each of the mutagenesis plasmids and plated onto 2xYT (Yeast Extract Tryptone) plates, containing various antibiotics. After 21 hours of incubation at 37°C the number of colonies were counted.

Protocol

Plating of the bacteria onto agar plates containing 100 mM glucose and 34 µg/ml of chloramphenicol.
After 12 hours of incubation at 37°C single colonies were picked and inoculated overnight in 4 ml 2x YT medium (with 200 mM arabinose and 4 µl of 34 mg/ml chloramphenicol) at 37°C for 21 hours. For each E. coli with a different plasmid negative controls were inoculated with 200 mM arabinose instead of glucose.
2x YT agar plates with 100 mM glucose and different antibiotics were prepared:

a. 9 plates with streptomycin (50 µg/ml)
b. 9 plates with carbenicillin (50 µg/ml)
c. 9 plates with kanamycin (30 µg/ml)
d. 9 plates with tetracyclin (10 µg/ml)
e. 9 plates without antibiotics for the negative control

50 µl were plated onto the matching plate undiluted, 10x diluted and 100x diluted, all done in triplicates.
The plates were incubated for 21 hours at 37°C and the number of colonies counted.

The protocol was kindly provided by the Heidelberg iGEM 2017 team.

Results

As expected, we could observe that E. coli gained resistances and grew on almost all of our plates. Only on those containing kanamycin, there could not be any or very few E. coli colonies observed.
Due to the fact that most of our plates were overgrown, we could not see any difference between E. coli containing one of the three plasmids (Tab. 1).
To see the final results, check it out on the homepage of Heidelberg.

Table 1: Counted E. coli colonies on plates treated with different antibiotics.

Cloning of constitutively active HIF1A

Under normoxic conditions HIF1A cannot interact with HIF1B to form the transcription factor HIF1, because its hydroxylated and marked by the E3 ubiquitin ligase for degradation by the proteasome (Ziello et al., 2007).

Therefore the substitution of the two hydroxylated amino acids, that are prolines at the positions 402 and 564 with alanines, would generate a HIF1A (P402A,P564A), that is stabilized and interacts with HIF1B under normoxic conditions (Ziello et al., 2007) to activate the hypoxia response element (HRE).