Team:DTU-Denmark/Demonstrate
Demonstrate
Introduction
The goal for our project was to create a piece of software that could predict synthetic promoters that could be utilized by both future iGEM teams and by industry. After deciding on an organism and consulting key stakeholders, it became apparent that one of the most important criteria for industry-relevant Aspergilli promoters were that they were functional and reliable in multiple scales.
To show the scalability of our promoters, we wanted to demonstrate that the synthetic promoters from our library not only worked under many different conditions but also had the same relative gene expression from each other, regardless of which scale we tested them in. We, therefore, performed experiments in four different scales to test the strength, and more importantly, the consistency of our promoters.
Click the figure below to see the different scales:
The experiments
For these experiments, the two synthetic promoters PLEAPglaA_2 and PLEAPsonB_1, were inserted in a USER plasmid in front of the fluorescent reporter gene mCherry. The plasmids were then transformed into a protease deficient variation of the Aspergillus niger ATCC 1015 strain, which was used to test the promoter activity at multiple scales.
The model hypothesizes that both promoters are the most active in the exponential growth phase. Additionally, the model predicted PLEAPsonB_1 to be a, relatively speaking, weak-medium strength promoter. The PLEAPglaA_2 promoter is a weakened variation (as described in the Model page) of the PLEAPglaA_1 promoter. This weakened promoter was included to validate the noise injection procedure of the model. For verification of the promoter activity and measurement of fluorescence, we used an engineered A. nidulans strain producing mCherry as a positive control, and an A. niger ATCC 1015 without mCherry production, as a negative control.
The details regarding the choice of this specific A. niger strain, and fluorescent marker can be found on the Design page.
It should be noted that since the positive control is an mCherry producing A. nidulans strain, and not an A. niger strain, the positive control cannot be compared quantitatively to our other data, as we are working with a different organism entirely. However, the strain is still useful in comparing mCherry production to no mCherry production.
The production of mCherry from the two promoters was evaluated in a confocal microscope based on fungi grown on agar plates and although we were unable to determine the relative promoter strength quantitatively, it was evident that both promoters produced mCherry.
Microtiter scale
From figure XX, the following becomes evident: 1) The two promoters are both different from the negative control, meaning that they work. 2) Like the model predicted, the promoters are displaying the highest activity in their exponential growth phase. 3) The two promoters behave very similarly. However, PLEAPglaA_2 appears to have slightly higher expression than PLEAPsonB_1 after 60 hours. Given more time, a greater difference in expression levels might have been revealed.
However, working with filamentous fungi on this scale could very well lead to measurement issues over time, as the fungi quickly overgrow the wells, and thereby blocks the light required for measurement. To avoid this, and to investigate the scalability of the promoters, we decided to run additional tests on the promoters in larger scale.
Shake flask scale
For a larger scale testing of the promoters, fungal cultures were grown in 500 mL shake flasks where biomass samples were collected continuously and stored at -20 °C for further testing. During the run, the media was also sampled and analyzed to determine the remaining glucose, as a way of estimating the growth phase of the fungi. The media sampling was discontinued after the glucose concentration fell close to zero.
After completion of the run, protein extraction was performed on the biomass samples and subsequently measured for both fluorescence and total protein content. In this way, we hoped to produce an accurate picture of the promoters’ relative strengths, regardless of biomass generation and minor variations over time. We measured the relative promoter strength in millimoles TexasRed equivalent per µg protein, resulting in figure XX.
From figure XX we see that both promoters are generally different from the negative control and that the expression of mCherry follows the exponential growth phase, as seen in the remaining glucose concentrations. We also see that the relative strength of PLEAPglaA_2 and PLEAPsonB_1 is different across the two biological duplicates. Furthermore, as overgrowth is not an issue in shake flasks the experiment could continue for longer than 60 hours before the nutrients ran out and thus, we were able to gain more information on the promoter activity compared to the BioLector. This means that we can see that PLEAPglaA_2 seems to increase in strength is the later phases of growth compared to PLEAPsonB_1 which is a very interesting characteristic for a synthetic promoter (see Integrated Human Practises) and confirmed our expectations that it would be active in the exponential phase.
Bioreactor scale
The largest scale, in which our promoters were tested was in a 1 L bioreactor where industrial applications could be emulated. The reactor setup also allowed for measurement of the CO2 production in the off-gas produced during cultivation, enabling the monitoring of the correlation between biomass growth and promoter expression in greater detail than at the smaller scales. Again, the biomass was sampled and after the run was completed, the proteins were purified and fluorescence and protein content was analyzed.
From the bioreactor data, we observe multiple things of great interest. First, by comparing relative fluorescence with the accumulated CO2 data, we can confirm that the promoters are in fact their most active in the exponential growth phase. This is consistent with our modeling data. Secondly, the PLEAPglaA_2 promoter has higher expression of mCherry per protein than the PLEAPsonB_1 promoter, a behavior which here in the bioreactor is more pronounced than it was in the shake flask experiment, where the biological replicates yielded conflicting results as to which promoter was the strongest. However, based on this experiment, we can see that even though the PLEAPglaA_2 promoter is stronger than PLEAPsonB_1, they are within the same order of magnitude with regards to TexasRed equivalent production.
Conclusion
By testing our two promoters in these different scales (1.5 mL, 150 mL, 1L), we have demonstrated that we can; 1) Design synthetic promoters of varying strengths based on consensus sequences from several Aspergillus spp. genomes. 2) Produce heterologous proteins in different scales using these synthetic promoters. 3) Use our model to predict the behavior of said synthetic promoters. 4) Validate the noise injection procedure of the model by observing that the PleAPglaA_2 promoter works in vivo. 5) Last but not least, we have demonstrated that these two promoters behave fairly evenly, but with PLEAPglaA_2 being slightly stronger than PLEAPsonB_1 in the later growth phases, based on multiple scales of cultivation. This means that we have not only created synthetic promoters that produce significantly more product than the negative control but we have also made promoters that are both stable and scalable, living up to the industrial standards.
