Team:EPFL/Results
Results
Introduction
Once our copper retrieval system was designed, lab experiments were performed with the aim of expressing one or two copies of our protein of interest, CUP1, at the outer membrane of the yeast. First, we performed a series of experiments to check our protein expression level and localisation. Next, we tested our transformed organisms by conducting copper removal tests that consist of monitoring medium copper levels over time in presence of the engineered yeast.
Overview
In order to solve the problem of copper pollution in the field, we came up with the idea of using a microorganism capable of absorbing copper in water, by expressing the copper intracellular yeast metallothionein CUP11 at its surface and improving the capacity of the cell to retrieve copper, a phenomenon called bio-adsorption2.
To express the protein extracellularly, we cloned its coding sequence into the pCTcon2-V5 plasmid3 in frame with Aga24, a protein that links by disulfide bonds to Aga1, a yeast membrane protein. In order to check its extracellular expression, we first performed a Western Blot with a primary antibody targeting the V5 tag and then performed immunocytochemistry in order to visually check the presence of CUP1 on the outer membrane of our cells.
Once CUP1 extracellular expression was validated in our transformed yeast cells, we performed tests monitoring yeast's ability to remove copper in different conditions in comparison with non-transformed yeast. We then completed the same workflow with the CUP1 dimers, namely, DNA cloning, yeast transformation, Western Blot, Immunocytochemistry assay, yeast growth curves and finally, copper absorption assays.
DNA cloning
CUP1 coding DNA fragment was obtained by PCR-amplification of yeast genomic DNA since CUP1 genomic and coding sequence are the same. Amplicon size was checked by gel electrophoresis before proceeding to a PCR clean-up. We then digested the plasmid with two restriction enzymes, NheI and BamHI, and also checked the result by gel electrophoresis. Next, we cloned our sequence into the plasmid by Gibson Assembly, transformed E.coli with the assembled product and amplified the recombinant plasmid. The pCTcon2-V5 plasmid containing an ampicillin resistance gene, transformants were selected by plating the bacteria on ampicillin LB plates and amplifying clones in liquid medium supplemented with ampicillin. Plasmid purification was performed by Miniprep. Next DNA integrity and sequence were checked by absorbance measurements and Sanger sequencing respectively. Eventually, we transformed yeast competent cells with the right plasmids and selected them on trp- plates. As our plasmid contains a trp+ gene and the transformed yeast strain EBY100 is trp-, it allowed the auxotrophic selection of transformants.
Protein expression characterization
Aga2-CUP1-V5 protein expression quantification by immunoblot allowed the selection of the best clones. Protein extraction was achieved by cell lysis and expression levels quantified by Western Blot. Chemiluminescence-based protein detection was based on the binding of a mouse antibody to the C-terminal V5-epitope or a mouse monoclonal antibody to a tubulin-specific epitope, both subsequently recognized by an HRP-conjugated goat secondary antibody, binding the constant region of the primary antibodies. As expected, non-transformed yeast strains (EBY100) or transformed strains (pCTcon2-V5) cultured in medium without induction (galactose -) do not show any protein expression, while transformed strains (pCTcon2-V5/plasmid backbone, pCTcon2-CUP1-V5/CUP1 monomer and pCTcon2-CUP1-linker-CUP1/CUP1 dimer) cultured in medium with induction (galactose +) show elevated protein levels, Figure 1.
System membrane expression was tested by immunocytochemistry. Yeasts membranes were stained with a mouse primary antibody binding the V5 tag and recognized by a second antibody harboring a fluorophore, the Alexa Fluor 4885, while nuclei were marked by DAPI staining. Results validated that CUP1 surface expression, CUP1 monomer in Figure 2a and CUP1 dimer in Figure 2b, has similar expression to a CUP1-less system (plasmid backbone positive control), Figure 2d, and that is specific to the cell membranes. As these are non permeabilized cells, the antibodies bind only the extracellular proteins. In order to check whether the protein was not also present intracellularly, we permeabilized the cells before staining them and confirmed that our protein was only expressed at the membrane. Background ectopic staining was checked by non-transformed yeast cells (EBY100) staining which displayed poor signal, validating our results.
(a)
(b)
(a)
(b)
(c)Growth assay
We also wanted to check whether our expression system would affect the growth of the yeast in any way. The wild type EBY100 yeast could normally only grow in a complete yeast medium, such as YPD which is yeast extract peptone dextrose. Whereas the yeasts transformed with pCTcon2V5 vector were selected in SGCAA, which is a reduced medium containing galactose but lacking tryptophan that was used for auxotrophic selection of yeast transformants.
For that purpose, we monitored the growth over time of the transformed yeasts in both rich media (YPD) and selection media (SGCAA), as well as the growth of wild type EBY100 yeast in rich media, Figure 3 by measuring the cultures OD600 every 15 minutes. With no surprise, the transformed yeasts with the backbone vector grew faster in rich media, similarly to wild type yeasts. It took longer for the engineered yeasts (transformed with backbone vector or CUP1 construct) to grow in selection media, but the growth curves still reached a plateau. Moreover, we also performed some growth assay with the wildtype EBY100 yeast in YPD media supplemented with various concentrations of copper, Figure 4. Our yeast strain demonstrated good resistance to growing concentration of copper (up to 1000 mg/l), validating our choice to use it as a chassis for copper remediation.
Copper absorption assay
The ability of the engineered microorganism to remove copper from a solution was assessed by monitoring the copper concentration of medium in contact with yeast over time. To have a complete understanding of the assay technicalities, see the Copper assay design. Different yeast cultures were prepared, each one harboring a different condition. The conditions varied between yeast strain and yeast concentration. We then collected samples from the cultures at different time points, spinned the cells down by centrifugation and measured the copper concentration in the supernatant using a colorimetry test with Cuprizone and spectrophotometer.
The first strain we tested was the non-transformed EBY100, referred to as “wild type” (WT) thereafter. We tested three different concentrations of yeast in media supplemented with 4 mg/l copper, a concentration in the same order of magnitude as the pollution found in vineyards and the results are summarized on Figure 5.
First we could conclude that the wild type strain absorbs copper, as expected because they express endogenous intracellular CUP1. Also, the denser the yeast population, the greater the ability to absorb and remove copper, Figure 5. To illustrate, the sample with the highest yeast density absorbed three times more copper than the culture with the lowest yeast concentration after 120 minutes of incubation (p value < 0.0001).
We next tested our engineered yeasts that are: (1) one single CUP1 protein expressed at its surface and (2) CUP1 dimers. We tested 7 different dimers constructions as described in the Design page. These differences essentially rely on linker properties used for the two CUP1 proteins fusion. In a first batch of experiments, we compared CUP1 monomer surface expression and three different CUP1 dimers (dimer-1, dimer-3 and dimer-7) surface expression ability to remove copper compared with the wild type strain. The results are depicted in Figure 6.
Unfortunately, we observed that our newly engineered yeasts, CUP1 monomer or CUP1 dimer were not only not performing better than wild type but did not even absorb copper in solution. Disappointed by this finding but not less determined to find out why, we continued digging. We had three hypotheses to explain this results as there were three variables changing from wild type to engineered yeast, the membrane expression of CUP1, the medium and the fact that the yeast is genetically modified or not:
-
The extracellular membrane expression of CUP1 interferes with the endogenous bio-accumulation property of yeast by, for example, blocking the entry of copper into the cell
-
Transformation of yeast with pCTcon2V5 vector (with or without CUP1 insert) interferes with the endogenous bio-accumulation property of yeast
-
The reduced medium is the problem and prevents the yeast from bio-accumulate copper.
To test if the problem was coming from the extracellular CUP1 expression we tested our transformed yeast with backbone plasmid in SGCAA medium (+ galactose, induced expression) and in SDCAA (- galactose, no expression induction) and compared the results with the wild type strain EBY100 in YPD:
We can clearly see on this figure that neither the backbone system in SGCAA nor in SDCAA was retrieving copper from the solution. As the system backbone in SDCAA was not expressing the surface display, we could eliminate the first hypothesis. Indeed, the only differences between the wild type and backbone system strains were in the medium and in the fact that the backbone system had been genetically modified. Therefore, we decided to check if the problem was coming from the medium by testing our engineered yeast in YPD + galactose (to induce the system) as well as the WT strain in SGCAA + tryptophan (because it would not survive without tryptophan). The results are shown below in Figure 7.
Seeing that wild type yeast in SGCAA supplemented with tryptophan was behaving like the engineered version in SDCAA and that engineered yeast in YPD supplemented with galactose was behaving like WT in YPD made us eventually conclude that the problem was in fact coming from the medium.
Unfortunately we ultimately found that our engineered yeast, even when cultured in YPD, was not better than the wild type strain in absorbing and removing copper, Figure 8. To conclude, our strategy based on CUP1 surface expression did not improve the endogenous bio-accumulation ability of yeast.
Copper recovery
Since the wild type yeast could remove copper in solution and seemed to be the best option to do it compared to the engineered strains we made a last experiment to test whether we could retrieve the captured copper from the yeast in order to eventually further re-use it. Our literature research led us to consider EDTA as a candidate for copper retrieval6. We thus tested this chemical on loaded (1 mg/l) wild type yeast with different times of reaction as a function of EDTA concentration. The results are shown in the following figure:
From this, we found that the use of EDTA for copper retrieval to be effective with the best profitability using a 20 min reaction time with 20 mM. During our second meeting with Ludovic Vincent, the CEO of Biomede, we shared these results with him. It seemed that EDTA was indeed a great option for what we were trying to do. However, it is difficult to subsequently remove EDTA from the copper containing water. We must take this into account when thinking about the future uses of this copper.
Thus a rather complicated chemical question arises : we would need to find an effective chelating ligand (a compound that triggers complexe formation), which is not too expensive and not too harmful to the environment.
In our case, if we consider this copper solution to be reusable by wine growers, EDTA at a low concentration is not in itself a major problem as it is already used as a fertilizer. However, other options should be considered, such as separating EDTA from the copper-filled water to avoid further contamination of plants and soils.
In brief
To sum, we were able to express different CUP1 configurations on the extracellular membrane of the yeast S.cerevisiae. However we did not succeed in improving the endogenous ability for bio-accumulation of the yeast with our engineered strain. Instead we found out that the medium in which the yeast is growing is a major factor to take into account when performing bioremediation experiments. Indeed, our engineered yeast were not able to retrieve copper when growing in SGCAA, a reduced medium used for transformants selection, but was acting like the wild type when put in YPD, a rich medium used to grow wild type yeast.
In conclusion, WT yeast already performs well in removing copper by bioaccumulation from a solution in a 4 mg/l range, relevant to the fungicide contaminated water copper concentration. Surface expression of CUP1 does not improve yeast ability to remove copper from a solution. Nonetheless, since autogenous bioaccumulation seems to be an effective process on its own, we suspect that increasing it further by rising endogenous intracellular CUP1 expression would make the yeast a very efficient genetically modified engineered machine for copper removal.
Discussion
In this project we identified an underappreciated problem related to organic or ‘bio’ based horticulture in particular of wine grape vines and tea plants. Treatment with copper based fungicides, while not detrimental to the food produced (e.g. wine), is deleterious to the soil these plants are cultivated in as copper accumulates to toxic levels from rain runoff. These copper fungicide treatments have a long history, they are approved for organic use and their use is unlikely to decline in the near future given the additional environmental issues with alternate fungicides, and public sentiment against novel compounds (see our survey). Copper ground contamination inhibits cultivation of new plants in the polluted soil. Given the already limited soil types and regions suitable for wine and tea production, which are becoming even further limited due to global warming, finding a solution to stop this problem becomes even more pressing.
Our solution, CuRe, aimed to employ bioremediation to remove copper from fungicide contaminated rainwater. Our strategy was to express the endogenous yeast copper-binding protein CUP1 on the surface of yeast to bind to and allow the removal of copper from water. We generated constructs which allow the expression of single and multiple copies of CUP1 on the surface of yeast separated by linker protein regions of varying flexibility. We confirmed that these proteins are produced and present on the surface of yeast using both western blotting and immunofluorescence.
To couple with these novel yeast strains, we designed a bioreactor to allow yeast suspended in beads to interact with rainwater and remove pollutants such as copper. We designed, manufactured and built a device easy to use in the field by implementing a graphical user interface on an associated tablet that allows control of the bioreactor activities.
Evaluation of the efficiency of CUP1 expressing strains was more circumspect. We established a copper removal chemical assay and growth conditions for our yeast strains suitable for introduction to copper containing solutions. A key unexpected variable however was the potent ability of our background yeast carrier strain to bioaccumulate copper from solution. Comparison of our CUP1 surface expressing strains to our chassis strain revealed little additional removal of copper from solution compared to the control background yeast strain. From these experiments we can conclude:
- Control yeast strains have a potent ability to absorb copper from solution.
- CUP1 monomers or multimers could be expressed on the surface of yeast though retention of their copper binding activity could not be verified.
- Addition of single or multiple copies of CUP1 provided no significant additional removal of copper from solution.
In sum, we discovered yeast can remove copper efficiently from solution, though our efforts to further enhance this process were not fruitful. We also provide a strategy whereby copper can be released from yeast, for return back into fungicides for example, forming a closed loop for the recycling of this metal.
Our central goal of depleting copper from solution has been achieved, albeit not enhanced by our synthetic biology approach. Ironically, this finding could make real-world application of our technology easier to achieve. The release of genetically modified organisms is highly restricted in Europe and other countries. The finding that wild type yeast strains allow efficient removal of copper removes an important stumbling block to exploitation of our bioremediation strategy. As it stands, wildtype yeast strains in our bioreactor prototype could be quickly made ready for field testing, with little or no regulatory impedance.
From our data and experiments, we can envision the next steps upon which to iterate and improve.
Firstly, we would suggest other ‘wild type’ yeast strains used in laboratory experiments or even food or beer production be tested for their ability to bioaccumulate copper. Existing strains may have an even higher ability to absorb copper from solution and bring the advantages for direct application described above.
Second, in a synthetic biology strategy, our data suggests that bioaccumulation rather than biosorption should be the cellular process upon which to aim to enhance. Yeast's natural ability to bioaccumulate copper was far more potent than our attempt to promote biosorption. One can envision strategies to further enhance copper accumulation through the modification of the copper import and detoxification machinery, such as increasing intracellular CUP1 levels or activity, increasing the copper ions influx by enhancing transporters activity or raising intracellular copper metabolism. A combination of these approaches could further enhance the existing effective ability of yeast to deplete copper in solution.
Through our outreach activities, we have sought to encourage the next generation of bioengineers to rise to this challenge by both promoting synthetic biology in schools through a children's book, by encouraging EPFL students to get involved in iGEM and by outreach the public through surveys, podcasts and lay public articles. We believe our results have laid the first foundations upon which a solution can be generated by future investigators to prevent copper pollution in wine and tea growing areas and preserve both the productivity and beauty of these regions including in the world heritage Lavaux wine growing site immediately adjacent to our institute.