The Engineering Cycle
Our core idea centers on modifying a fungus to act as a "Fungilyzer", a biological fertilizing agent, that can buffer soil phosphate and release it to crops as needed. Enhancing sustainable agriculture with Fungilyzer in the pursuit of sustainable and resource-efficient agriculture, we followed the principles of the engineering cycle: Design, build, test, and learn. Our whole engineering strategy aimed at delivering a mechanism for controlled cell death in yeast depending on different phosphate concentrations. Fungilyzer is designed to efficiently manage soil phosphate levels, providing crops with essential nutrients while minimizing the environmental impact of excessive fertilizer use. This design ensures both plant nutrient access and controlled fungal death, promoting resilience in harsh conditions. Our strategy aims to maximize efficiency in phosphate uptake while ensuring controlled fungal apoptosis when phosphate levels fall below critical thresholds, releasing stored nutrients in the plant root’s mineral depletion zone to benefit crops. This promotes rapid uptake of the mineral by the fungus, improving nutrient availability for plants. As a part of the design phase, we considered the choice of a suitable model organism for testing. Cloning of our constructs was done in Escherichia coli, to then be transformed into yeast. Yeast (Saccharomyces cerevisiae), provided a suitable platform for genetic modification and proof of concept testing. The ideal plant species for our experiments was zucchini (Cucurbita pepo subsp. pepo convar. giromontiina) due to its sensitivity to nutrient deficiencies.
Gel Ladder reference
For all gel electrophoresis we used 1 kb Plus DNA Ladder from NEB1:
Cycle 1 - Custom S. cerevisae YIp pSB1C30YIpHR-HO/BBa_K4706000
The initial step in the assembly of all constructs is the addition of the corresponding homologous regions (ScHR3p-HO/BBa_J435241 and ScHR5p-HO/BBa_J435242 homology region) into the plasmid backbone pSB1C30. Thus, the build phase in this cycle we traversed twice as we needed to perform two insertions and confirm that each of those was sucessful.
The main goal is to seamlessly integrate this homology into pSB1C30, allowing for the creation of various functional genetic constructs. There were originally four different constructs needed to assemble the planned composite parts BBa_K4706012, BBa_K4706013 and BBa_K4706015 into pSB1C30YIpHR-HO/BBa_K4706000 to facilitate reliable genome integration.
Design of custom S. cerevisiae YIp
Yeast Integrating plasmids (YIp) are non autonomously replicating plasmids that do not contain an ORI or similar replication modes like CEN/ARS containing plasmids that transform the DNA by homologous recombination. Once integrated into the genome those strains are usually relatively stable.2 3 As it has been shown that linear DNA fragments have a much higher recombination frequency than circular plasmids, we decided that we wanted to linearize our backbone via PCR in our design, but preferrably without restriction enzymes, as to not degrade potential cloning capabilities of the backbone. 4
We further decided to use NEBuilder to facilitate scarless integration of the homology sites into the vector.5 The main reason for not using the already existing cloning site of pSB1C30 was to avoid occupying it for genes of interests - this allows the user to utilize the BioBrick site for inclusion of their desired part.
Build of custom S. cerevisae YIp (HO5')
As we were conducting homology based cloning, linearizing pSB1C30 and ScHR5p-HO/BBa_J435242 was necessary first and foremost.
Each of the two fragments has a 40 bp homology to each other and was assembled as previously shown via homology based cloning (NEBuilder).
This is the resulting intermediate plasmid:
After yielding transformants, we performed a PCR to confirm that the insert was present. We also loaded the plasmid itself onto a gel to confirm the approximate total size via gel electrophoresis.
Gel electrophoresis of PCR and insert check. In our internal documentation we referred to pSB1C30YIpHR-HO5' as CD006.
We also obtained sequencing data which further confirmed that our insert was sucessful:
Sequencing results in PDF format are available in the references.6
Build of custom S. cerevisiae YIp (HO5' + HO3')
As we were conducting homology based cloning, we again needed to linearize our intermediate plasmid and ScHR3p-HO/BBa_J435241.
Each of the two fragments has a 40 bp homology to each other and was assembled via homology based cloning (NEBuilder).
This is the resulting intermediate plasmid:
After yielding transformants we performed a PCR to confirm that our insert was present. We also loaded the plasmid itself on the gel to confirm the approximate total size.
Gel electrophoresis of PCR and insert check. In our internal documentation we referred to pSB1C30YIpHR-HO as CD007.
Sequencing results in PDF format are available in the references7.
Test of Custom S. cerevisae pSB1C30YIpHR-HO
We transformed what we thought at the time was the composite part BBa_K4706015/SrpR repressable PHO5 phosphate dependent cell death. Analysis of the assembly later revealed that we did not have a correct assembly. The linear DNA we transformed looked like this:
While this fragment entails the correct DNA sequence required for homology directed genomic DNA insertion, it contains no selective marker. As we prepared the uracil deficiency selective media, a mistake in the media preparation process took place. Uracil might have been accidentally added or was already present in our premix. The latter is more likely as we duplicated the transformation process. This time, no S. cerevisiae without the selective marker present grew on those plates.
We performed a colony PCR on the colonies grown in the transformation on non selecting plates, then searched for an insert flanked by BioBrick sites, which would look like this:
At first glance, it seems like the colony PCR results mostly in background noise, indicated by the insert bands being barely visible (we expected major intensity around 1kb), another possibility would be nonspecific amplification of our introduced insert. This would better explain why we still had 4 colonies without any amplification.
For this reason we also performed a reference colony PCR with 28 different colonies of S. cerevisiae BY4741 using the exact same protocoll and parameters to account for background amplifications of the S. cerevisiae genome. To our surprise, no background amplicons were present which suggests that the primers bound nonspecific to our introduced insert fragment.
This result suggests that our linear DNA, which was amplified from our backbone, has a suprisingly high transformation frequency. Out of 28 screened colonies we only observed four (4) colonies that tested negative on the insert. This would result in an approximate transformation frequency of 85.7%. We would have conducted additional tests but due to time and resource constraints we were not able to do so.
Learning from custom S. cerevisiae pSB1C30YIpHR-HO
This phase of our project has primarily taught us the critical importance of media preparations for transformations. It is worth mentioning that, without this error, we would not have had confirmation that our backbone was functioning as intended. We also came in contact with issues encountered whilst setting up a proper colony PCR, at first we tried amplifying fragments higher than 3 kb but we would recommend any team needing to perform colony PCRs to utilize for fragments of a size of <1 kb as otherwise degradation or nonspecific amplicons appear to be a problem. We would further advise any team using this backbone in the future to make additional tranformation frequency test on nonselective media as well as testing transformation efficency for inserts of different lenght and if the transformation efficency on selective media might be dependent on the marker used.
Reagrding our backbone: If we would have had the time to further improve our backbone we would have liked to introduce blunt cutters at the edges of our HO homologies to have a more stable way of storing them and having an easier way to yield higher concentrations that can be used for transformations.
Cycle 2 - Assembly of Phosphate regulatory mechanism
Composite parts BBa_K4706012/SrpR repressable PHO5 phosphate level sensor, BBa_K4706013/Gal1 induced RFP A2 AtBAG6 and BBa_K4706015/SrpR repressable PHO5 phosphate dependent cell death will be discussed in one cycle as we designed all of these Parts based on each other and also the cloning of these was done in parallel.
Design of a Phosphate-Regulated Mechanism
To establish a phosphate regulated mechanism, we need to induce a signal under either low or high phosphate conditions. In our project, our aim is to trigger lysis to release nutrients for plants in low phosphate conditions. Therefore, we selected a native S. cerevisiae promoter, BBa_K4706002, which is induced by low phosphate conditions.
To quantify the strength of this identified phosphate promoter, we initially designed BBa_K4706012/SrpR repressable PHO5 phosphate level sensor, a repressible PHO5 phosphate level sensor, by combining BBa_K4706002 with an RFP reporter BBa_K4706005. This composite part allows for the creation of a biological phosphate sensor that can be analyzed through fluorescence.
Our next step involves introducing the ability to induce controlled cell death in yeast. Before testing the cell death inducing gene in combination, we must assess the ability of BBa_K4706007 to induce cell death under the control of a well-characterized promoter. For this purpose, we adapted the Gal1 promoter to create BBa_K4706014. This adaptation enables us to characterize this part in a known system within our composite part BBa_K4706013/Gal1 induced RFP A2 AtBAG6.
Following separate tests on BBa_K4706002 and BBa_K4706007, we combined them in BBa_K4706015/SrpR repressable PHO5 phosphate dependent cell death to induce cell death under low phosphate conditions.
BBa_K4706012/SrpR repressable PHO5 phosphate level sensor and BBa_K4706015/SrpR repressable PHO5 phosphate dependent cell death are designed to be repressed by BBa_K4706016/CUP1-GFP 2A SrpR-SV40NLS. This design allows us to simultaneously test the creation of a well-regulated kill switch system based on phosphate concentration.
All of the cloning was planned wit NEBuilder as prevously described. Notable and the reason why we combined it in one cycle is that all of them share the same homology edges for our custom backbone pSB1C30YIpHR-HO/BBa_K4706000. These small homology regions also share sequences needed for restriction enzyme cloning to constitute as a backup plan if homology based cloning via NEBuilder or AQUA8 should fail.
Build of Phosphate Regulatory Mechanism
Homology based aproach
At first we decided to continue with our first approach as we had massive success at our first cycle. For this to work we again needed to linearize our backbone pSB1C30YIpHR-HO/BBa_K4706000 and the URA3 selection marker BBa_J435253. Additionally the functional part of the three composite parts that was synthesized was already linear and had these 40 bp homologous regions.
BBa_K4706012/SrpR repressable PHO5 phosphate level sensor
We first tried to assemble BBa_K4706012/SrpR repressable PHO5 phosphate level sensor using NEBuilder.
This approach did not yield any colonies we could further analyse.
Cosequently, we tried to use AQUA8 cloning to assemble our composite part which this time resulted in colonies. After further analysis none of the analyzed colonies had an insert of our desired composite part present.
The resulting plasmid would have looked like this:
BBa_K4706013/Gal1 induced RFP A2 AtBAG6
We tried to assemble BBa_K4706013/Gal1 induced RFP A2 AtBAG6 using NEBuilder.
This approach did not yield any colonies we could further analyse.
Then we tried to use AQUA8 cloning to assemble our composite part which this time resulted in colonies. But after further analysis none of the analyzed colonies showed an insert of our desired composite part present. The resulting plasmid would have looked like this:
BBa_K4706015/SrpR repressable PHO5 phosphate dependent cell death
We consequently tried to assemble BBa_K4706015/SrpR repressable PHO5 phosphate dependent cell death using NEBuilder.
This approach did not yield any colonies we could further analyse.
After that we tried to use AQUA8 cloning to assemble our composite part which this time resulted in colonies. But after further analysis all analyzed colonies did not have an insert of our desired composite part present.
The resulting plasmid would have looked like this:
Restriction-Ligation Based Cloning
After our homology based cloning strategy failed, we decided to move on to our backup plan, which consisted of a combination of blunt end cloning and restriction based cloning.
BBa_K4706012/SrpR repressable PHO5 phosphate level sensor
Firstly, we again needed to linearize all fragments, but we also needed to amplify the fragments ordered via PCR which proved quite difficult due to nonspecific amplification for this part. Additionally we did a restriction digest with PstI+EcoRI for our backbone pSB1C30YIpHR-HO/BBa_K4706000, URA3 was digested by PstI and the ordered fragment we amplified and added 5' phosphate groups to facilitate blunt end joining were cut with EcoRI. The final overlaps/cloned ends looked like this:
We did obtain a few colonies but sadly none of them had the correct insert.
BBa_K4706013/Gal1 induced RFP A2 AtBAG6
To start over, we again linearized all fragments, but we also need to amplify the fragments ordered via PCR which proved quite difficult due to nonspecific amplification for this part. Additionally, we performed a restriction digest with PstI+EcoRI for our backbone pSB1C30YIpHR-HO/BBa_K4706000, URA3 was digested by PstI and the ordered fragment we amplified and added 5' phosphate groups to facilitate blunt end joining were cut with EcoRI. The final overlaps/cloned ends looked like this:
We did obtain a few colonies but sadly none of them showed the correct insert.
BBa_K4706015/SrpR repressable PHO5 phosphate dependent cell death
To move on, we again had to linearize everything, but we also had to amplify the fragments ordered via PCR which turned out successful for this part. Additionally we did an restriction digest with PstI+EcoRI for our backbone pSB1C30YIpHR-HO/BBa_K4706000, URA3 was digested by PstI and the ordered fragment we amplified and added 5' phosphate groups to facilitate blunt end joining were cut with EcoRI. The final overlaps/cloned ends looked like this:
The resulting plasmid was used for transformation in S. cerevisiae which resulted in us finding out that cycle 1 is a success but further analysis revealed that we do not have the right insert (in regard of the composite) present in the plasmid. We did obtain a few colonies but sadly none of them had the correct insert.
Test of Phosphate Regulatory Assembly
As we sadly were not able to generate S. cerevisiae transformants we can only outline how our tests would have looked.
BBa_K4706012/SrpR repressable PHO5 phosphate level sensor
Our initial plan was to test these parts using fluorometric measurements with a plate reader. Different phosphate concentrations would lead to distinct expression levels. A high emission in the 560 nm range indicates great levels of expression of our sensor, thus indicating low phosphate in the surrounding media.
BBa_K4706013/Gal1 induced RFP A2 AtBAG6
We planned on testing these parts with fluorometric measurements with a plate reader as outlined for BBa_K4706012 but this time we would have also needed to conduct living cell counts after a certain amount of time and change the testing media from phosphate to galactose. Based on the time it would take for the fluorescence to decrease we would have chosen appropriate timepoints for living cell count measurements.
BBa_K4706015/SrpR repressable PHO5 phosphate dependent cell death
We planned on testing these parts with fluorometric measurements with a plate reader as outlined for BBa_K4706012 but this time we would also need to conduct living cell counts after a certain amount of time. Based on the time it would take for the fluorescence to decrease we would have chosen appropriate timepoints for living cell count measurements.
Learning from Phosphate Regulatory Assembly
Our main takeaways from this cycle are that, again, specific primer design and the avoidance of multiple PCR products have a major role in cloning preparations and can, if not done properly, prolong the time needed immensly. Also, we had issues in our cloning process as it might be possible that the products we originally designed for S. cerevisiae expression are for some unknown reason producing toxic byproducts when expressed in E. coli. But sadly due to time constraints we were not able to confirm if this is the case.
Cycle 3 - Assembly of Copper-Induced Repression System
Design of BBa_K4706016/CUP1-GFP 2A SrpR-SV40NLS
During our design process, we explored the option of incorporating a repressor system that, if needed, could halt the initiation of cell death in the components BBa_K4706012/SrpR repressable PHO5 phosphate level sensor and BBa_K4706015/SrpR repressable PHO5 phosphate dependent cell death.
To achieve this, we opted for CUP1 BBa_J435201, a copper-inducible promoter for S. cerevisiae. This promoter can be succinctly described as an "On/Off" switch. When combined downstream with BBa_K4706021 to produce a DNA binding protein, SrpR (commonly used in E. coli), it can effectively repress protein expression in S. cerevisiae as well.
Build of BBa_K4706016/CUP1-GFP 2A SrpR-SV40NLS
For BBa_K4706016/CUP1-GFP 2A SrpR-SV40NLS we also decided for a homology based approach. The assembly would have been conducted into the plasmid (http://parts.igem.org/Part:BBa_K4706020) from pSB1C30, BBa_K4706011 our ordered insert fragment and HIS3 BBa_J435273. The following fragments should have looked like this:
Sadly we were not able to amplify BBa_K4706011 from the 2022 distribution kit and a sample we sent off for sequencing indicated that BBa_K4706011 contained the wrong insert. We were luckily offered a CEN/ARS plasmids from the Institute of Microbiology, which hosted our lab. But due to time constraints we were not able to finish our assembly.
Test of BBa_K4706016/CUP1-GFP 2A SrpR-SV40NLS
In order to test the functionality of the copper-induced promoter in an environment close to the target environment, we would have started several testing series with C. pepo in order to confirm that it works under any given circumstances. The soil should include different nutrient conditions, especially important in this case would be an example of average copper concentrations, and an example of a strongly increased phosphate concentration, as it would be found after the application of a combined phosphate-copper fertilizer, as well as a test group with a copper concentration starting on a high level that would not be renewed over the course of the experiments. This case would ideally show a final result of Fungilyzer dying off after a while when the copper concentration has significantly dropped through use of nutrients by the plants.
Learn From the Assembly of Copper-Induced Repression System
One of the most important learnings we had, trying to work with the copper induced repression system, was the importance of sequence verification. The failure to amplify BBa_K4706011 from the 2022 distribution kit and the discovery that the sequence contained the wrong insert emphasizes the importance of verifying the authenticity of DNA sequences prior to assembly. Furthermore, it shows the significance of having backup plans and alternative sources when planning projects and staying flexible throughout the whole engineering cycle. Also, having valuable partners like the Institute of Microbiology goes a long way, as it helps overcoming challenges and getting expertise from different people.
https://international.neb.com/products/n0559-1-kb-plus-dna-ladder-for-safe-stains↩
Parent, S.A., Fenimore, C.M. and Bostian, K.A. (1985), Vector systems for the expression, analysis and cloning of DNA sequence in S. cerevisiae. Yeast, 1: 83-138. https://doi.org/10.1002/yea.320010202↩
Hinnen, A., Hicks, J. B. and Fink, G. R. (1978). Transformation of yeast. Proceedings of the National Academy of Sciences U.S.A.75,1929↩
Orr-Weaver, T. L.; Szostak, J. W.; Rothstein, R. J. (1981). Yeast transformation: a model system for the study of recombination.. Proceedings of the National Academy of Sciences, 78(10), 6354–6358. doi:10.1073/pnas.78.10.6354 ↩
https://www.neb.com/en/applications/cloning-and-synthetic-biology/dna-assembly-and-cloning/nebuilder-hifi-dna-assembly↩
Beyer HM, Gonschorek P, Samodelov SL, Meier M, Weber W, et al. (2015) AQUA Cloning: A Versatile and Simple Enzyme-Free Cloning Approach. PLOS ONE 10(9): e0137652. https://doi.org/10.1371/journal.pone.0137652↩