Contact In the news Acknowledgements References Links Downloads Protocols Brainstorm Meetings

Conclusions & Recommendations

Conclusions

After a great summer of hard work, we've learned quite a few things. The most important are listed below.

Temperature sensitivity - Protein measurements

Due to the fact it is impossible to state that lysis has a constant efficiency, it is not possible to take the cell density before lysis as a value for correcting luciferase activity. Therefore it is important to measure something in the soluble fraction of your sample correlating to actual biomass . We decided to use a total protein content (Bicinchoninic acid assay) for the value to correct luciferase activity by. However, the lysis buffer (of unknown composition) in the luciferase assay kit interfered with our protein measurements. To prevent this interference we tried precipitating the protein in the samples using 4 different methods to get rid of the (unknown) interfering agent in the buffer. Although calibration curves which were first treated with lysis buffer and then precipitated varied from quite linear (R²=0.983) to not linear at all (R²<0), samples containing the soluble fraction of lysed cells, usually gave protein contents of below 0 mg/ml.

A different approach to solve this problem of lysis buffer interference was trying various different lysis methods, including sonication, using glass beads in a fastprep or using lysozyme and glass sand in a bead beater. Although it looked like all these methods destroy some of the activity of luciferase in the sample, time was lacking to optimize each method. The sonication protocol, although time consuming, seemed most reliable in our experiments. Still, this method is not optimal, as samples lysed with lysis buffer show a lot higher raw luciferase output, although it is not known what the total protein content of these samples is.

Temperature sensitivity - Results with sonication

Using sonication of our samples, we obtained our most reliable results. These indicated, as displayed here, that our strain K115035 shows the temperature induced switch-like behavior which was the aim of our project.

Color Pathway

It proved to be too much work to PCR 15 genes out of Escherichia coli and Saccharomyces cerevisiae in addition to making the temperature sensitive input system work (with only 2 wetlab guys). It was a good choice to split the project into two paths, as advised at the teachers workshop in Paris, as this allowed us to choose for the most promising path, should one of them not work out as expected. When it became obvious the actual color producing genes we aimed to get from the registry were of low quality (bad sequencing or did not grow at all) we gave priority to the temperature sensitive input system. In the end we did make some progress towards constructing our envisioned color pathway by PCR-ing three of the genes out of E. coli, which have been sent to the registry and to a sequencing company.

Modeling

Future work

In this project we worked towards a biothermometer, existing out of an RNA thermometer and a coupled pathway to produce color molecules. To measure the functioning of an RNA thermometer in general we used luciferase assays. In the end this provided us with unexpected problems as the lysis buffer provided with the luciferase assay kit interfered with our protein content measurements. Other lysis methods destroyed luciferase activity.

Continuing this part of the project we would suggest using a different enzyme. This would have to be a non Escherichia coli enzyme which is easily measured, preferably in whole cells. It is tempting to use the general protein expression indicator GFP for this analysis, but this is not reliably quantifiable. Quantification is very likely to be important for these thermometer RNAs, as expression as a function of temperature will likely be a sigmoid curve. These type of thermometers have been shown to function in previous research [1][2][3] and we are convinced that they can work in a biobrick environment as well. If the temperature dependent expression curves are not like an on-off switch, other systems in the registry such as the Schmitt trigger might be used for making on-off behavior sharper.

To work further on the color pathway, first of all the genes still missing must be obtained by either DNA synthesis or by PCR on S. cerevisiae or other suitable organisms. All enzymes should preferably be tested for individual activity, which has not yet been done on the E. coli genes we provided to the registry. If enzyme activity has been confirmed, this pathway can be implemented as described by Martin et al.[4].

If both systems are functional, they can be easily coupled because of the biobrick standardisation. This would deliver the biothermometer as we designed it.

References

1. ^ F. Narberhaus, T. Waldminghaus & S. Chowdhury. RNA thermometers. FEMS Microbiol Rev, 30(1):3-16, 2006. PMID:16438677
2. ^ Saheli Chowdhury, Curdin Ragaz, Emma Kreuger, and Franz Narberhaus. Temperature-controlled Structural Alterations of an RNA Thermometer. The Journal of Biological Chemistry, 278(48):47915-47921, 2003. PMID:12963744
3. ^ Torsten Waldminghaus, Nadja Heidrich, Sabine Brantl, and Franz Narberhaus. FourU: a novel type of RNA thermometer in Salmonella. Molecular Microbiology, 65(2):413-424, 2007. PMID:17630972
4. ^ V. Martin, D. Pitera, S. Withers, J. Newman and J. Keasling. Engineering a mevalonate pathway in Escherichia coli for production of terpenoids. Nature Biotechnology. 21(7):796-801, 2003. PMID:12778056