Team:Toulouse INSA-UPS/Results
Results
Information
In order to supplement the astronauts' food with tasty nutritious yeast, we had to do several clonings to insert key genes into the yeast genome. However, our expectations were high and we did not have the time to complete all of our clonings. We made a cloning progress chart here below at the time when we left the lab. We managed to enhance the pool of precursors of β-carotene as you can see below and in our Engineering success page. We are also proud to prove with our encouraging coculture results that our project has a promising future.
Cloning process chart
Yeast modification results
Introducing new metabolic pathway in engineered Saccharomyces cerevisiae to constitutively synthetize vitamin and flavors in a controlled manner.
The goal of this section is to provide engineered yeast in order to constitutivly produce β-carotene and geraniol and limonene in an optogenetic controlled system. Thus, all cloning for gene of interest were performed in Escherichia coli as described in the Experiments section. Our strategy was to assemble several blocks (some of our clonings with 4 to 5 blocks) in a single cloning step by using the In-Fusion kit (ligation-independent cloning) from TaKaRa Bio. However, due to the limitation of the technology, we soon found out that assembling 4 or 5 blocks in a single reaction was not conclusive. Thus, we subcloned the genes of interest with no more than two blocks at the same time, which increasing dramatically the cloning steps.
Materials and methods:
Most of our blocks were synthesized as gBlocks by IDT (Integrated DNA Technologies). They were all given a number to facilitate their identification and our organization. The sequences that were not synthesized by IDT are coming from plasmids obtained from different laboratories.
All the protocols used in this cloning section can be found here (Experiments page).
Molecular engineering for flavors production
Cloning for geraniol and brazzein production
Summary and cloning strategy (BBa_K3570001):
In our project, we aimed to give to the astronauts the opportunity to choose the flavor of the yeast. Thus, we explored the possibility to use geraniol and brazein to make sweet rose flavored yeast. For geraniol and brazzein, we aimed to construct two different vectors (figure 1) and to fuse them together into a single plasmid:
Figure 1: Cloning strategy for geraniol and brazzein production
The first intermediary plasmid is composed of blocks B1, B2 and B3 in the pUC19 chassis to create pUC19-B1B2B3, so it is composed of the upstream integration site (BBa_K3570014), the brazzein gene and its FLAG tag, the geraniol synthase (BBa_K3570012) and their common bidirectional promoter GAL1/10. The other intermediary plasmid is composed of blocks B4 and B5 in the pUC19 to create pUC19-B4B5 (so including the leucine selection marker for the yeast S. cerevisiae BBa_K1430001 and the downstream integration site BBa_K3570015).
Results and discussion:
- Assembly of pUC19-B1B2B3 (block 1,2 and 3):
The gBlocks B1, B2 and B3 have been amplified by PCR then purified. At the same time, digestion of our vector pUC19 by SbfI and XbaI was performed and purified on gel with the same kit. Before we proceed to cloning, we verified all our purified PCR fragments (Figure 2).
Figure 2: Agarose electrophoresis in order to verify gBlocks B1, B2, B3 and pUC19 digested with SbfI and XbaI. On the right the expected results and on the left the obtained gel. All gBlocks and the pUC19 present the expected sizes.
The purified DNA fragments had the expected length and were subsequently used for cloning. In-Fusion, transformation of Stellar cells, selection on ampicillin, miniprep of a few clones were done. Then, the plasmid construction was checked out with the restriction enzymes SbfI and SacI in order to release the expected fragment of 3748 bp and 2672 bp. After several attempts, we accepted that this cloning strategy was very unlikely to yield proper clones. As previously proposed, a new design strategy consisting of two steps instead of one step was adopted (Figure 3).
Figure 3: Revisited cloning strategy for pUC19-B1B2B3. First, the B1 block is cloned into pUC19 vector and then B2 and B3 blocks are cloned in the resulted vector from step 1.
First, the pUC19-B1 was built by the In-Fusion kit and three of the obtained colonies were verified by digestion with SalI and SbfI (figure 4).
Figure 4: Agarose electrophoresis of DNA fragments in order to validate the cloning experiments. DNA plasmids from 3 colonies were double digested with SbfI and SalI.
Colonies 2 and 3 have the expected digested profile on the gel: pUC19-B1 was successfully obtained. It was digested by SalI and XbaI to insert the B2 and B3 fragments. After the In-Fusion reaction and transformation, the colonies were verified by PCR and we expected a fragment at 872 bp (Figure 5).
Figure 5: Agarose electrophoresis of PCR amplification on ampicilin selected clones after cloning gBlocks B1, B2 and B3 into pUC19 by the In-Fusion kit. On the right the expected results and on the left the obtained gel. The expected size is 872 bp.
As it can be seen, only clone #4 presented a strand at 872 bp. The plasmid was sent for sequencing. Unfortunately, it presented a deletion in the ORF of the geraniol synthase (B2) involving a shift in the reading frame. We tried again this cloning step without success. At this point, we decided to stop linearizing the vector by digestion and switch to preparing the linear vector by PCR. This would prevent getting a majority of colonies presenting the empty vector. This PCR was performed using different concentrations of the template plasmid (figure 6).
Figure 6: Agarose electrophoresis of PCR amplification of pUC19-B1 with different template quantities (from 10 ng to 100 ng of DNA). On the right the expected results and on the left the obtained gel.
This showed that the linearized pUC19-B1 could be amplified with template concentrations as low as 10 ng. In-Fusion was then performed to integrate B2 and B3. The selected colonies were checked by PCR and we expected a 1205 bp fragment (figure 7).
Figure 7: Agarose electrophoresis of PCR amplification on ampicilin selected clones after cloning gBlocks B2, B3 into pUC19-B1 with the In-Fusion kit. On the right, the expected results and on the left, the obtained gel. The expected band should be at 1.2 kb.
According to the results, none of the clones had the expected size that would correspond to a correct insertion to pUC19-B1 of B2 and B3. Unfortunately at this time, a re-emergence of coronavirus restricted our work (Impact of COVID-19). We had to choose which cloning we would carry on. As this part was not our first priority, its progress simply stopped here.
- Assembly of the pUC19-B4B5 (Block 4 and 5)
In parallel with pUC19-B1 construction, we tried to build pUC19-B4B5 which contains the leucine auxotrophic marker and the downstream integration site.
The expected sizes of the amplified DNA fragments were obtained for both fragments (data not shown). These were integrated by the In-Fusion kit into the pUC19 plasmid. Four clones were verified by digestion with EcoRI which should reveal fragments at 1366 bp and 3807 bp if correct (figure 8).
Figure 8: Agarose electrophoresis in order to verify ampicilin selected clones after cloning pUC19-B4B5 with the In-Fusion kit and digested with EcoRI. On the right, the expected results and on the left, the obtained gel. The expected sizes are 1.3 kb and 3.8 kb.
The three first colonies presented correct profiles and their plasmids were sent for sequencing. Clone M28 presented the correct sequence.
The next step should have been to include fragment B1B2B3 in this pUC19-B4B5 but it had to be abandoned because of the re-emergence of the coronavirus.
Cloning for limonene production
Summary and cloning strategy:
The final strategy for the cloning of limonene synthase is presented Figure 9.
Figure 9: Final cloning strategy for limonene production. To build our construction for limonene production, the four DNA fragments needed were inserted one by one in the pUC19 plasmid.
Results and discussion:
The block B10, B11 and B13 were synthesized as gBlocks by IDT, amplified by PCR with CloneAmp HiFi PCR, and purified by NucleoSpin Gel or NEB Monarch Gel extraction kits. The auxotrophic marker (block B12) was amplified by PCR with CloneAmp HiFi PCR using S. cerevisiae genomic DNA as template (data not shown).
So we started by inserting block B10 in the pUC19 plasmid. Four colonies presented plasmids with the expected digestion profiles (Figure 10).
Figure 10: Agarose electrophoresis in order to verify ampicilin selected clones after cloning pUC19-B10 with the In-Fusion kit and digesting with SbfI and SacI. On the right the expected results and on the left the obtained gel. The expected sizes are 2.6 kb and 1.2 kb, the non-digested plasmid should be 3.8 kb.
The first four tracks of this gel correspond to four different clones tested by digestion. The last line contains undigested plasmid. The sequences proved to be valid. The next step was to integrate fragment B11. The obtained colonies were assessed by PCR but unfortunately, none presented the expected amplicon size (Figure 11).
Figure 11: Agarose electrophoresis in order to verify ampicilin selected clones after cloning pUC19-B10B11 with the In-Fusion kit. On the right, the expected results and on the left, the obtained gel. The expected size is 2kb. None of the clones presented the expected size.
At this step, we had to stop the experiments: we still had three fragments to insert in our vector before being able to integrate it in the yeast, and we lacked time and workers to undertake it with the COVID-19 lurking around.
Molecular engineering for Optogenetic Systems
Our aim is to give the choice to astronauts between either raw yeast, lemon or sweet rose taste. This is why we thought to use optogenetic systems to regulate the production of limonene, geraniol and brazzein using light. The red optogenetic system required two fused proteins encoded by GBD-PhyA and GAD-FHY1 genes. For the blue ontogenetic system, it required only the NLS-EL22-vp16 gene.
Cloning of the red optogenetic system
Summary and cloning strategy (BBa_K3570004):
The insert containing genes for the PhyA-FHY1 optogenetic system was designed to be cloned into the limonene production plasmid, or the brazzein/geraniol production one (if you have read the previous cloning descriptions, you can guess how this will end up!). The following blocks B6, B7, B8 and B9 (figure 12) were designed to be cloned by the In-Fusion kit into a pUC19.
Figure 12: Cloning strategy for the red optogenetic system
Results and discussion:
As for the previous clonings, the first strategy consisted in inserting 4 blocks all together. Since all the colonies we obtained contained empty pUC19 vectors, we tried to avoid the problem by preparing the vector from PCR amplification instead of digestion using restriction enzymes.
At this time, however, because of the coronavirus re-emergence restricting our work ( COVID-19 impact ), we were not able to finish our construction and decided to stop this cloning (not very surprising!).
Clonings for the blue optogenetic system
Summary and cloning strategy (BBa_K3570005):
Figure 13: Cloning strategy for the blue optogenetic system
Results and discussion:
Since the B27 PCR amplification was not conclusive because of its small length (232 bp), it was suppressed from the design and we decided to build only pUC19-B28B30 by integrating sequentially B28 and B30.
Both blocks (B28 and B30) have been amplified by PCR. First, the pUC19-B28 construction was built by the In-Fusion kit and the plasmids from the obtained colonies were verified by PCR (figure 14).
Figure 14: Agarose electrophoresis of PCR amplification of pUC19-B28. On the right the expected results and on the left the obtained gel. The expected band size is at 1.9 kb.
As can be seen, all the plasmids presented the correct profile. They were sent for sequencing and it appeared that plasmid #3 had no mutation. It was then used as a chassis to integrate B30 with the the In-Fusion kit. The colonies obtained were checked by PCR, expecting a fragment at 1419 bp (figure 15).
Figure 15: Agarose electrophoresis of PCR amplification of pUC19-B28B30. On the right, the expected results and on the left, the obtained gel. The expected size should be at 1.4 kb.
The results showed that tracks 1, 3, 4, 5, 11 and 12 presented the correct profile. Thus, pUC19-B28B30 was obtained.
As you can guess, the optogenetic parts were put in standby at this step because we were late in our cloning program and too few to perform all the experimentations because of the COVID-19. Besides, the blue optogenetic system was designed to be integrated in the limonene or the geraniol constructions that we failed to produce.
Molecular engineering for β-carotene precursors
Cloning of tHMG1 and CrtE
Summary and cloning strategy (BBa_K3570000):
All the compounds we want to produce with the yeast derived from the mevalonate pathway, this is why the improvement of this pathway was required to reach sufficient yields.
The cloning strategy was to clone the blocks into two vectors before bringing them together in a unique plasmid. The blocks B14, B15 and B16 would be cloned in a pUC19 using the In-Fusion kit to form pUC19-B14B15B16 containing the DPP1 upstream sequence, tHMG1 and TDH3-TEF1. The other blocks B17, B18 and B19 containing CrtE, the histidine auxtrotrophic marker and the DPP downstream sequence would be cloned in parallel in another pUC19 using the same cloning method to form pUC19-B17B18B19. pUC19-B17B18B19 would be used as a template vector to insert the sequence of B14B15B16 from pUC19-B14B15B16 (Figure 16)
Figure 16: Cloning strategy for β-carotene precursors production
Results and discussion:
Assembly of pUC19-B14B15B16:
The gBlocks corresponding to B14, B15 and B16 have been amplified by PCR, purified and verified on agarose gel (Figure 17).
Figure 17: Agarose electrophoresis verification of PCR amplified gBlocks B14, B15, B16 and digested pUC19 with SbfI and BamHI. On the right, the expected results and on the left, the obtained gel. The expected sizes were 2.6 kb, 0.4 kb, 1.8 kb and 1.0 kb respectively. All have the expected size.
pUC19 was double digested by SbfI and BamHI and prepared to receive the PCR products B14, B15 and B16 by the In-Fusion kit. After transformation of Stellar cells, selection on ampicillin, and minipreps of eight clones, we checked the restriction profiles of the constructions. The resulted plasmids containing two SacI restriction sites were then verified by digestion with the enzyme SacI (Figure 18).
Figure 18: Agarose electrophoresis of PCR amplification of pUC19-B14B15B16 cloned with the In-Fusion kit and digested using SacI. On the right, the expected results and on the left, the obtained gel.
We succeeded to isolate six clones displaying a plasmid with the expected profile. Sequencing of the plasmids confirmed that we obtained the correct pUC19-B14B15B16 construction also known as the first step for tHmg1-CrtE construction.
Assembly of the pUC19-B17B18B19
The gBlocks B17, B18 and B19 have been amplified by PCR with CloneAmp HiFi PCR and then purified by NucleoSpin Gel and PCR Clean-up (Figure 19).
Figure 19: Agarose electrophoresis of PCR amplification of the pUC19 digested using BamHI and EcoRI, and gBlocks B17, B18 and B19. On the right, the expected results and on the left, the obtained gel. The expected sizes were 2.6 kb, 1.4 kb, 1.4 kb and 0.4 kb respectively. All have the expected size.
pUC19 vector was double digested using BamHI and EcoRI and prepared to receive the PCR products B17, B18 and B19 with the In-Fusion kit. After transformation of Stellar cells, selection on ampicillin, and minipreps of 6 clones, we checked the restriction profiles of the constructions. The resulted plasmids were verified by digestion with BamHI and EcoRI (Figure 20).
Figure 20: Agarose electrophoresis of PCR amplification of the pUC19-B17B18B19 cloned by In-Fusion kit and digested using BamHI and EcoRI. On the right, the expected results and on the left, the obtained gel. The expected sizes were 4.8 kb and 2.6 kb.
Only one clone had the expected profile (figure 20, lane 4). This clone has been sequenced by Eurofins and it was fortunately valid. We had successfully obtained the second plasmid pUC19-B17B18B19 of our tHMG1-CrtE construction.
Assembly of tHMG1-CrtE:
The next step was to combine both plasmids by subcloning the fragment B14B15B16 into plasmid pUC19-B17B18B19.
To do this, we first extracted the DNA with the QIAGEN Plasmid Plus Midi Kit. Then, we digested both plasmids with SbfI and BamHI and purified with the Monarch Genomic DNA Purification Kit by NEB. The fragments were ligated together with T4 DNA ligase by NEB followed by a transformation into Stellar cells (ampicillin selection). Over the eight assessed colonies, two colonies presented the expected restriction profile when digested with SbfI and EcoRI (Figure 21).
Figure 21: Agarose electrophoresis of ampicilin selected clones of pUC19-B14B15B16B17B18B19 ligated with T4 DNA ligase digested with SbfI and EcoRI. On the right, the expected results and on the left, the obtained gel. The expected sizes were 6.8 kb and 2.6 kb. Two of them present the expected size (with some residual undigested clone left at 9.4 kb).
Yeast transformation:
Since the construction was successful, we proceeded to the yeast transformation by integration in the DPP1 locus. The plasmid was digested with enzymes SbfI and EcoRI and the fragment containing our construction was purified to transform the Saccharomyces cerevisiae strain BY4741 . The resulted cells were grown on YNB LEU+, URA+, MET+ HIS- with 2% glucose for 3 days. At the third try, we were able to observe around 20 colonies in our yeast transformation, about the same amount of colonies on the positive control and none on the negative control plate.
Verification of integration of BBa_K3570000 using the DPP1 homology sequence (BBa_K3570006 and BBa_K3570007) was performed by a genomic PCR using the TaKaRa PCR amplification Kit and the following primers: primer 1 and primer 2.
Primer 1: ATCAGGATTTGCGCCTTT
Primer 2: GCCGCCGAGGGTATTTTACTTCCG
We randomly chose eight clones from our transformation and one from the positive control plate (Figure 22).
Figure 22: Verification of DPP1 integration. On the right, the expected results and on the left, the obtained gel. The expected size is 1.2 kb.
All clones have the expected size (1.2 kb), and the PCR negative control where we inserted pRS313 did not show any band, which means that the BBa_K3570000 part is well integrated in the yeast genome. The integration in the yeast genome is a success that means that the parts BBa_K3570008 (HIS3 selective marker), BBa_K3570006 and BBa_K3570007 (for DPP1 homologous sequence) all work. Our modified strain is BY4741 DPP1::tHMG1-crtE.Analysis and Validation of the constructed strains
The integration of the BBa_K3570000 part in the yeast genome should enhance the flow of the mevalonate pathway. This modification should result in an increase of the concentration of GGPP, which is the precursor of beta-carotene. GGPP have been extracted from BY4741 DPP::tHMG1-crtE and analysed by LC-MS (see our experiments here ).Our strain BY4741 DPP::tHMG1-crtE showed a five-fold increase in GGPP concentration in comparison with the wild type yeast (figure 23). This important result demonstrates that the part BBa_K3570000 works.
Figure 23: Results of the GGPP analysis from culture on YNB with 2% Glucose by LC-MS
Data from the literature [2] suggest a stable turn-over of mevalonate pathway intermediates in native and engineered S. cerevisiae strains, i.e. a stable ratio between the GGPP production flux and its concentration. This would indicate that our strain optimization strategy successfully enhances by five-fold the capability of GGPP production, and thus the potential of β-carotene production. Moreover, the increased flux through the mevalonate pathway is expected to increase not only β-carotene production, but also the production of limonene and geraniol (a provitamin A, a flavor of lemon and of rose) synthesized from other mevalonate intermediates.
Molecular engineering for β-carotene production
In order to make yeast produce β-carotene, the precursor of vitamin A, we also needed to insert a fusion protein: CrtYB(ek)I.
Cloning of the fusion protein CrtYB(ek)I
Summary and cloning strategy (BBa_K3570002):
Our strategy was to divide our part into two plasmids: one composed of blocks B20 (HO up), B221 (TDH1) and B22 (CrtYB(ek)I) and tCYC) in pUC19 vector and forms pUC19-B20B21B22; and another composed of B23 (URA3 auxotrophic marker) and B24 (HO down) also in pUC19 vector leading to pUC19-B23B24. Then we would use pUC19-B23B24 as a template vector to insert B20B21B22.
Figure 24: Cloning strategy for pUC19-B21B22 and pUC19-B23B24. First, the B21 and B22 blocks are cloned into pUC19 vector, B23 and B24 blocks are cloned in pUC19, then the two constructions are ligated together.
Results and discussion:
Assembly of pUC19-B20B21B22:
The blocks B20, B21 and B22 have been amplified by PCR, purified and verified on agarose gel (data not shown). pUC19 was digested by SbfI - XmaI and purified on gel. We proceeded to the In-Fusion reaction, transformation of Stellar cells, selection on ampicillin, and miniprep of clones. Unfortunately, this cloning repeatedly failed. So we decided to first clone B21 and B22 in the pUC19. The resulting colonies were checked by PCR (Figure 24).
Figure 24: Agarose electrophoresis of PCR amplification verification of the B21 insertion in pUC19 using the In-Fusion kit. On the right, the expected results and on the left, the obtained gel. The two Crt - columns are negative controls.
To verify our cloning, we tested two different PCRs on our four clones: one is expected to provide a 4300 bp fragment and another one, a 300 bp fragment. We observed that 2 clones seemed to match our expectations. Both were sequenced and appeared to us as valid (or so we felt…). After following the same protocol to insert the last block B20, we obtained many colonies and tested the restriction profiles of six of them (figure 25).
Figure 25: Agarose electrophoresis of the ampicilin selected clones after the insertion of B20 into pUC19-B21 digested by SbfI and XbaI. On the right, the expected results and on the left, the obtained gel.
The six columns contain different DNA plasmids of 6 different colonies that have been digested by SbfI and XbaI. Among the 6 clones, only two presented the expected size (1300 bp). We sent them to be sequenced by Eurofins, and the sequence again appeared as valid to us…
- Assembly of pUC19-B23B24:
The blocks B23, and B24 have been amplified by PCR with CloneAmp HiFi PCR and then purified by NucleoSpin Gel and PCR Clean-up (data not shown). The pUC19 vector was digested by EcoRI andXmaI and purified on gel. We proceeded as usual to the In-Fusion reaction, transformation of Stellar cells, selection on ampicillin, and minipreps of a few, …well…, of the unique clone. Its plasmid was checked by several restriction profiling (figure 26).
Figure 26: Agarose electrophoresis of the pUC19-B23B24 digested with SbfI and XmaI, XmaI and EcoRI, and EcoRI and EcoRV respectively on each lane. On the right, the expected results and on the left, the obtained gel.
Luckily enough (for a change!), this single clone got the right digestion profile so we sent it to be sequenced and it was fine.
- Assembly of pUC19-CrtYB-CrtI:
The next step was to integrate B20B21B22 into pUC19-B23B24 as a vector with a classical cloning method using restrictions enzymes and ligase. Plasmids purified using QIAGEN Plasmid Plus Mini Kit were double digested by SbfI and XmaI. The vector was purified and insert B20B21B22 was ligated into linearized pUC19-B23B24 using the T4 DNA ligase from NEB. Selected transformant E. coli cells were recovered from solid selective media and the presence of the gene of interest was assessed using a colony PCR followed by a transformation into Stellar cells (ampicillin selection). The next day, we obtained hundreds of colonies so we chose six of them for a DNA extraction on the following day. To verify our colonies, we performed a PCR and a digestion. From the six clones tested by PCR, four were positive (data not shown). We tested two of these clones by a series of digestion to validate our cloning (Figure 27).
Figure 27: Agarose electrophoresis of the pUC19-B23B24B20B21B22 digested with SbfI and EcoRI (D1), SbfI and XmaI (D2), and EcoRI and EcoRV (D3).
Our clones presented the expected digestion profile, which allowed us to proceed to the transformation step in the yeast that had previously integrated tHMG1-CrtE.
- Yeast transformation:
As we managed to boost the precursors of Vitamin A, the final step to make yeast produce β-carotene was to insert the CrtYB-CrtI construction.
We tried a first round of transformation in the previously obtained tHMG1-CrtE yeast strain. We expected orange-colored yeast but our clones were desperately white. At this step, we had another careful look at our sequencing results. Knowing that something was wrong, we identified a mutation at the very beginning of block B22. Unfortunately, this mutation triggered a shift in the reading frame, preventing the expression of CrtYB-CrtI.
Because of the COVID-19 regulations and the university courses that were starting, we did not have time to continue our experiments. Further experiments will be necessary to correct the mutation by site-directed mutagenesis in order to produce β-carotene.
OVERALL CONCLUSION:
Fermentation experiments for the coculture system
All the clonings we have tried to do aim to build a nutritious yeast that will be used as food supplement for the astronauts. We have designed a coculture between Clostridium ljungdahlii and Saccharomyces cerevisiae to produce this nutritious yeast from CO2 and H2.
In this part of our results, we are presenting the experiments and the results to prove that a such coculture is feasible. These experiments can be broken down into three main areas. Firstly, the definition of a culture medium common to both microorganisms. We tested this medium on the yeast and on the bacteria individually in order to find out whether it is suitable for both. According to our model, coculture is feasible. However, we need to simulate the different possible solutions proposed by the model in the yeast and the bacteria individually. These experiments have allowed us to refine the model and collect parameters to size the coculture reactor. The latest experiments are the coculture experiments.
Figure 28: Scheme of the coculture between C. ljungdahlii and S. cerevisiae
Definition of a coculture medium
Purpose of the experiments
Coculture between microorganisms requires the design of a common environment, especially if the microorganisms have very different needs. This is why we have developed a coculture medium and we have carried out growth experiments on the yeast and then on the bacteria in order to find out whether certain compounds that are absolutely necessary for one would be toxic to the other.
Experiments
The range of temperature for the yeast is 25-37 °C and the range of temperature for the bacteria is 30-40 °C. This is why we fixed the coculture temperature at 33 °C. The optimal pH of S. cerevisiae is 5.5 and C. ljungdahlii can grow at this pH.
We elaborated a medium for the coculture based on the media used for Clostridium ljungdahlii and Saccharomyces cerevisiae [1], [2]. To design it, we received help from Jason M. Whitham. All the compounds used in S. cerevisiae media are approximately in the same quantity in the media used for C. ljungdahlii. The medium is therefore composed of YNB (what the yeast needs) and we added everything else the bacteria needs. Here is how we elaborated the first coculture medium:
Dissolve ingredients of the pre-medium, warm the solution if necessary to dissolve YNB. Add 10 ml of the trace elements solution. Sparge medium with 80% N2 and 20% CO2 gas mixture for several minutes to make it anoxic and adjust pH to 5.5. Autoclave at 121 °C for 15 min. Add 10 ml of vitamins (sterilized by filtration) from sterile stock solutions prepared under 100% N2 gas atmosphere. Adjust pH of complete medium to 5.5. Before yeast inoculation, be sure to have a pO2 at 5%.
Results and discussion
- Growth experiments with S. cerevisiae
We grew CENPK1137 cells at 33°C, pH 5.5 and 150 rpm as followed:
-
pre-medium without Na2S + trace elements solution + 0.38% ethanol + 0.5% acetate
-
pre-medium + trace elements solution + 0.38% ethanol + 0.5% acetate
-
pre-medium without Na2S + trace elements solution + vitamins solution + 0.38% ethanol + 0.5% acetate
-
pre-medium without Na2S + trace elements solution + vitamins solution + 2% glucose
Figure 29: Growth rate of S. cerevisiae on various media to test the coculture medium. pre-medium. 1: without Na2S + trace elements solution + 0.38% ethanol + 0.5% acetate; 2: pre-medium + trace elements solution + 0.38% ethanol + 0.5% acetate; 3: pre-medium without Na2S + trace elements solution + vitamins solution + 0.38% ethanol + 0.5% acetate; 4: pre-medium without Na2S + trace elements solution + vitamins solution + 2% glucose
CENPK1137 grew on pre-medium without Na2S, with trace elements solution, vitamins solution and 2% glucose. Nevertheless, the yeast did not grow in the other conditions. Na2S and L-cysteine HCl are used to complex dioxygen and keep the medium anaerobic. S. cerevisiae need dioxygene to grow on acetate and ethanol. The properties of Na2S and L-cysteine HCl could be harmful in these conditions. Metals contained in the trace elements solution could also be harmful for the yeast on acetate and ethanol.
- Growth experiments with C. ljungdahlii
We made all our C. ljungdahlii growths on the coculture medium as we described it in the experiments section and bacteria grew well. See in the paragraphe “Growth experiments with Clostridium ljungdahlii ‘’ below.
Perspectives
-
Tests must be carried out independently on the pre-medium, the trace element solution and the vitamin solution to find out which compounds are toxic to the growth of yeast on ethanol and acetate.
-
Once a suitable medium for the yeast on the basis of the initial medium is found, then growth tests on the bacteria will have to be repeated.
Growth experiments with Clostridium ljungdahlii on CO2 and H2
Purpose of the experiments
In our coculture system, Clostridium ljungdahlii catalyses the transformation of CO2 and H2 into ethanol and acetate. Growth rate and specific production rate can be found in the literature but it is a growth rate in optimal condition. We had to experiment the capacity of C. ljungdahlii in our experimental reactor before going through coculture experiments because our reactor is not optimal and we needed parameters for dimensioning the coculture system.
Experiments
We obtained the strain from DSMZ as a fresh culture that we inoculated in a vial containing 30 ml of coculture medium. After reaching an OD600 nm of 0.354 over two days, we inoculated with 10 mL of this preculture a 200 mL reactor for the growth on CO2 and H2. We used the set-up which is on the figure 2. The reactor had been made gas-tight and it has two heads in order to have the entry of CO2 and H2, the entry of N2, the output of gases and the output/input of medium for the medium flow in the coculture. The OD and a supernatant sample were taken three times per day. We ran the fermentation for two days.
We received help from Jason M. Whitham (interview Jason M. Whitman) to know how to manipulate C. ljungdahlii, see on our experiment page how we did.
The data have been analyzed using PhysioFit [3].
Figure 30: Experimental set-up for Clostridium ljungdahlii growth on CO2 and H2. A: the reactor of Clostridium ljungdahlii. The output of gases was connected to the outside of the building for further security. The reactor was into a water bath at 37°C. B: the electrolyzer of water whose H2 output was connected to the reactor.
Results and discussion
- Growth
We had some difficulties getting information about the growth because the medium tends to precipitate and we had to learn how to carefully measure the ODs. Nevertheless, we managed to have enough data to express a growth rate. It was evaluated to 0.022 h-1 which is below the data described in the literature (0.058 h-1) [4] and below the prediction of our model (0.07 h-1) but we assumed that the experimental set up can be improved.
Figure 31: C. ljungdahlii growth on CO2 and H2.
- Acetate and ethanol production
In accordance with the literature, the Clostridium ljungdahlii growth should be coupled with the production of acetate and/or ethanol [5]. The sample supernatant at 47 h (figure 31) was analysed by NMR. We detected the classic triplet of ethanol at 1.2 ppm (figure 32). We also detected the singulet of acetate at 1.92 ppm.
Thanks to an internal standard, we measured the concentration of acetate and ethanol. Acetate concentration was 0.3 mmol/L and ethanol concentration was 3 mmol/L. We approximated the specific acetate production rates to 0.038 mmol/OD/h, and the ethanol one to 0.38 mmol/OD/h. These results are lower than expected. Since we did not control the input of H2 and CO2, and since the membranes quickly became wet, it can be argued that our reactor needs to be optimized to avoid substrate limitation in order to obtain better productivity.
Figure 32: 1D H NMR analysis of the supernatant of C. ljungdahlii growth on CO2 and H2 after 47h. The singlet at 1.92 ppm is the one of acetate. The triplet at 1.20 ppm is the one of ethanol.
Figure 33: 2D TOCSY RMN analysis of the supernatant of C. ljungdahlii growth on CO2 and H2 after 47 h. This graph confirms that the triplet at 1.2 ppm is indeed that of ethanol, where we observe the expected correlation for ethanol between the methyl protons (1.2 ppm in the F2 dimension) and those related to COH (at 3.65 ppm in F1).
Perspectives
-
The experimental set up could be improved for a better growth. The flow of gases could be too low and we observed that the gas output filter got wet rapidly. The gas-tightness could therefore be better.
-
The growth of C. ljungdahlii could be also experimented at a pO2 of 5-8% before testing the coculture experiments because S. cerevisiae would need dioxygen to grow on ethanol and acetate.
Growth of Saccharomyces cerevisiae on minimal medium with acetate and/or ethanol as sole carbon source
Purpose of the experiments
In order to design our coculture between Clostridium ljungdahlii and Saccharomyces cerevisiae, we needed to know the maximum growth rate of the yeast on ethanol and acetate, as well as to find the acetate/ethanol mix on which the yeast present the best growth rate. We also needed to know the quantity of carbon source necessary and sufficient to quickly start the growth of the yeast. We would also need to know if ethanol and acetate are consumed at the same time. If this is not the case, the accumulation of one of the two compounds could upset the balance of the coculture.
Experiments
The yeast strain used was the CENPK1137 because it is a resistant and efficient prototrophic strain. All experiments were conducted at 33 °C, the defined coculture temperature. All cultures were performed as three replicates and were followed up over day with sampling approximately every hour. The carbon source concentration was always 2% (g/100 mL).
The data have been analyzed using PhysioFit [3]. It is a software designed to quantify exchange (production and consumption) fluxes and cell growth rate during (batch) cultivations of microorganisms.
Results and discussion
- Correlation gDW/OD
We tried to measure a correlation gDW/OD during the growth of our cultures but our measures were noisy because of our balance. We used therefore a correlation from literature [7] which is 0.53 gDW/OD.
- Growth rate on various carbon source
Tab.1: Growth rate of CENPK1137 on various carbon source
Figure 34: Growth rate of S. cerevisiae of various mix of ethanol and acetate. The upper chart represents our results of growth rates. The chart below represents the quantity of carbon sources of each experiment.
Tab.2: Average growth rate of CENPK1137 on various mix of ethanol and acetate
First we observed that the initial concentration of acetate or ethanol did not significantly impact the growth rate (Tab. 2). This means that we can start the growth of the yeast in the coculture with a substrate concentration as low as 0.1% acetate or 0.1% ethanol.
The strain CENPK122 is described to have a growth rate of 0.17 h-1 on YNB with 1% acetate and 0.12 h-1 on YNB with 1% ethanol [6]. The strain CENPK1137 is very close to this CENPK122 strain and therefore should present a close maximal growth rate on these substrates. Nevertheless, the growth rate of yeast on acetate, ethanol or both in various ratios was slower, according to our results (0.11 ± 0.019 h-1). We can conclude that whatever the ethanol/acetate ratio produced by Clostridium ljungdahlii in coculture, the growth of our strain won’t be impacted.
- Consumption of ethanol and/or acetate
We needed to know if acetate and ethanol are consumed together or one after another because depending on the answer the ethanol and acetate concentration in the coculture system will be impacted and the accumulation of a product could be detrimental to the coculture. To answer this question, we measured the specific consumption rates of ethanol and acetate using HPLC analysis.
Tab. 3: Specific consumption rates of ethanol or acetate of CENPK1137 on various carbon source
We observed significant standard deviation which means that our data just gave us an order of magnitude of each specific consumption rate. The specific ethanol consumption rate was -4.93 ± 1.7 mmol/gDW/h and the specific acetate consumption rate was -4.97 ± 2.51 mmol/gDW/h. These data were used to feed our model. It will next serve us for the dimensioning of our experimental coculture system and of our prototype coculture system.
Figure 35: Consumption of ethanol and acetate of CENPK1137 on YNB + 2% Acetate/Ethanol 50/50.
According to figure 35, ethanol and acetate are consumed at the same time and not one after the other. This information ensured us that the variation of ratio of ethanol and acetate produced by C. ljungdahlii will not lead to the accumulation of one product in the media.
Conclusion on coculture
The results of C. ljungdahlii growth are conclusive even if the experimental setup must be improved. We have to optimize the gases supply to obtain higher growth rates and higher ethanol and acetate production rates.
The results with S. cerevisiae showed that it is able to co-consume acetate and ethanol. This information assures us that the variation of ratio of ethanol and acetate produced by C. ljungdahlii will not lead to the accumulation of one product in the media. Now, tests would have to be performed on oxygen limitation too to get closer to the coculture conditions.
Our results are encouraging. The next step will be to run the coculture with both microorganisms. We have set up the system for coculture experiments (Figure 36). These experiments will prove the feasibility of the system for space application for which we made a 3D model (Figure 37). This system for space application have been designed according to the specifications (Implementation) based on feedbacks from experts (Integrated human practices).
Figure 36: Experimental set-up for coculture between S. cerevisiae and C. ljungdahlii. The reactor at the left is fed by gases (CO2 and H2) and it has an entry of N2 and an output of gases. The reactor at the right is connected to a monitor in order to control the pH, the temperature and the dissolved O2. The two reactors are connected via two pumps in opposite directions which are working on alternatively (see the design for further information). The water electrolyzer is in the middle and brings H2 for the C. ljungdahlii reactor and O2 for the S. cerevisiae reactor.
Figure 37: 3D model of the system for space application
References
https://www.dsmz.de/microorganisms/medium/pdf/DSMZ_Medium879.pdf
[2] BD Difco™ Yeast Nitrogen Base without Amino Acids
[3] Camille Peiro, Pierre Millard, Alessandro de Simone, Edern Cahoreau, Lindsay Peyriga, Brice Enjalbert, and Stéphanie Heux, 2019. Chemical and Metabolic Controls on Dihydroxyacetone Metabolism Lead to Suboptimal Growth of Escherichia coli. Appl Environ Microbiol. 2019 Aug 1; 85(15): e00768-19. doi: 10.1128/AEM.00768-19
http://github.com/MetaSys-LISBP/PhysioFit/
[4] Michael E. Martin, Hanno Richter, Surya Saha, Largus T. Angenent, 2015. Traits of Selected Clostridium Strains for Syngas Fermentation to Ethanol. Biotechnology and Bioengineering Volume 3, Issue 3 p531-539; DOI 10.1002/bit.25827
[5] Michael Köpke, Claudia Held, Sandra Hujer, Heiko Liesegang, Arnim Wiezer, Antje Wollherr, Armin Ehrenreich, Wolfgang Liebl, Gerhard Gottschalk, and Peter Dürre, 2010. Clostridium ljungdahlii represents a microbial production platform based on syngas. PNAS July 20, 2010 107 (29) 13087-13092; https://doi.org/10.1073/pnas.1004716107
[6] J.P. van Dijken, J. Bauer, L. Brambill, P. Dubo, J.M. Francois, C. Gancedo, M.L.F. Giuseppin, J.J. Heijnen, M. Hoare, H.C. Lange, E.A. Madden, P. Niederberger, J. Nielsen, J.L. Parrou, T. Petit, D. Porro, M. Reuss, N. van Riel, M. Rizzi, H.Y. Steensma, C.T. Verrips, J. Vindeløv, J.T. Pronk, 2020. An interlaboratory comparison of physiological and genetic properties of four Saccharomyces cerevisiae strains. Enzyme and Microbial Technology 26 (2000) 706–714.
https://doi.org/10.1016/S0141-0229(00)00162-9
[7] Sonderegger M, Sauer U. Evolutionary engineering of Saccharomyces cerevisiae for anaerobic growth on xylose. Appl Environ Microbiol. 2003 Apr69(4):1990-8. p.1991. DOI: 10.1128/AEM.69.4.1990–1998.2003
