Team:Toulouse INSA-UPS/Results

Overview

In this section, we present all the experimental works performed during our summer. We start here with the cultivation and co-culture set up which have been done in parallel to the cloning efforts, integration in yeast and analytical analysis, presented thereafter.

Cultivation and co-culture

This part describes our efforts to characterize and optimize the growth of our two microorganisms and the preliminary work to set-up a co-culture. It is divided into a yeast growth, cyanobacterium growth and co-culture sub-parts.

Yeast growth on sucrose

The first step towards the synthetic microbial consortium we have designed was to study and characterize the growth of our parent yeast strain LycoYeast on sucrose.

Growth of LycoYeast in the co-culture medium using exogenous sucrose

We first wanted to see if we could grow LycoYeast in a modified version of the cyanobacteria medium (BG11) and by adding exogenous sucrose as a carbon source. Different co-culture media formulations are described in the literature to grow cyanobacteria and yeast together (Hays et al. 2017; Li et al. 2017). We used BG11 media supplemented with 1.2 g/L yeast nitrogen base without amino acids (YNB w/o aa), and 100 mM NaCl (denoted BG-11[CoY]). We added 5 g/L of exogenous sucrose, a concentration that should be reachable by our sucrose-secreting cyanobacterium strains (Lin et al. 2020). The growth was followed by measuring the OD at 600 nm. The results are presented in Figure 1.

Figure 1: Growth of LycoYeast in BG-11[CoY] supplemented with 5 g/L exogenous sucrose.

After a long lag phase, it appears that our yeast is able to grow, although quite slowly, in the co-culture medium with sucrose as its sole carbon source which is encouraging for the continuation.

Determination of the maximal sucrose uptake rate of LycoYeast

To calibrate our dynamic co-culture model, we then chose to measure the maximum uptake rate of sucrose in our LycoYeast strain. It has indeed been shown that sucrose consumption by S. cerevisiae is a key point for the stabilization of the consortium with S. elongatus (Hays et al. 2017). The value of this parameter has also been shown to be highly strain-dependent (Rodrigues et al. 2021). This approach furthermore allowed us to test whether our model could predict the growth rate of S. cerevisiae on sucrose.

To do so, S. cerevisae was grown in YP+Sucrose medium which contained 10 g/L of sucrose. The biomass concentration was quantified over time by measuring the optical density (OD) at 600 nm and converted to g_DCW/L using a previously established correlation factor of 0.53g_DW/ODunit (Sonderegger and Sauer 2003). At every OD measurement, a sample of culture medium was also taken and analyzed after centrifugation by HPLC to quantify the sucrose concentration. Time course concentrations of biomass and glucose were then fitted using PhysioFit (Peiro et al. 2019) to estimate the maximum sucrose uptake rate of our yeast. The results obtained for the best fit are presented in Figure 2.

Figure 2: Raw and fitted data of the growth of LycoYeast using sucrose as a carbon source. The biomass concentration (left) is expressed in gDCW/L while the sucrose concentration (right) is in mM. The time unit is in hour.

We obtained the following values (confidence intervals at 95%): μ_max=0.33 ± 0.02 h^-1 q_sucrose,max=-3.0 ± 0.3 mmol.g_DCW^-1.h^-1 The values we previously used for our model were those determined by Rodrigues et al. (2021) for S. cerevisiae CEN.PK113-7D who had measured a qsucrose,max of -8.2 ± 0.4 mmol.g_DCW^-1.h^-1. Our experimental sucrose uptake value was used to constrain our yeast genome scale model and predict the growth rate in this condition. The predicted value (0.44 h^-1) was in agreement with the experimental growth rate (0.33 h^-1) measured in our condition. Updating this parameter has significantly improved the predictive capabilities of our model (predicted growth rate of 1.3 h^-1 with the initial uptake rate taken from the literature vs 0.44 h^-1 with the updated parameter). This demonstrates the value of this experiment to calibrate our model and improves its predictive power.

Cyanobacteria growth and sucrose secretion

Growth

This was the first time that an iGEM Toulouse team worked with cyanobacteria. We did not have any equipment specifically dedicated to this and had to set up everything, which was done step by step. Throughout the summer, we conducted liquid cultures of cyanobacteria with the objective of attempting to reproduce the rapid growth phenotype of the Syn UTEX 2973 strain as well as the high sucrose secretion of the strains provided to us (Yu et al. 2015; Lin et al. 2020). This goal proved to be much more difficult than expected, we most often obtained growth profiles that were linear rather than exponential, and with growth rates several orders of magnitude lower than expected. Numerous tests were performed in an attempt to optimize the culture parameters of these cyanobacteria. The results of these different approaches are presented below.

Culture configuration & CO₂ transfer

The cultures were routinely grown in small cell culture flasks which have the advantage of providing optically thin cultures with good light penetration as well as avoiding contamination. Since the growth was not very satisfying in these conditions (Figure 3), different cultures in erlen flasks of different volumes and different configurations were tested (with or without baffles, different types of lids). These modifications did not improve the growth of our strains. The cultures at that time were carried out with a light intensity of approximately 250 μmoles photon.m^-2.s^-1 and at atmospheric CO₂ (0.04%). In these conditions, Yu et al. (2015) obtained a doubling time of 6.3 hours which is way below our estimated doubling time (which was around 29 hours). Our idea was then to carry out cultures in bioreactors, in which we could control the pH during the culture and enrich in CO₂. Using 5% CO₂, and a light intensity of 850 μmoles photon.m^-2.s^-1, we were able to improve the growth rate relatively with a maximum doubling time of 10 hours but it was still extremely low compared to what would be expected under these conditions (doubling time of between 1.5 hours to 2 hours for Yu et al. (2015) and Ungerer et al. (2018)). As it can be seen in Figure 3, the saturation besides occur at an OD_750nm of approximately 1 which is very low. After discussing with experts met at the Phototrophs Meetups, it appeared that this was not normal and that something else was limiting the growth since Syn UTEX 2973 usually easily reach OD_750nm 10 to 20.

Figure 3: Presentation of the different cultivation configurations tested with examples of results A) Cell culture flasks cultures of 20 mL (Temperature 38°C, Light 250 μmoles photon.m^-2.s^-1, CO2 (0,04%) B) Culture in erlen flasks (Temperature 38°C, Light 250 μmoles photon.m^-2.s^-1, CO2 0,04%) C) Bioreactor culture (Temperature 38°C, Light 850 μmoles photon.m^-2.s^-1, CO2 5 %)

BG11 medium composition & pH of the cultures

Our first hypothesis was that the composition of the BG11 medium we were using at that time was suboptimal. We had bought this medium in the form of a ready-made powder sold by Phytotech and did not really have any information about its composition. This lack of standardization of the composition of the BG11 medium was a subject of discussion at the Phototrophs Meetups and also concerned the regulation of the pH in the medium since it appeared for example that our commercial medium was not buffered.

We therefore decided to try to prepare this medium ourselves. Since iGEM Marburg 2019 had proposed an optimized version named BGM medium for this strain, we followed their protocol. While with this medium the saturation phase occurred at a higher OD_750nm (approximately 5 rather than 1 in the bioreactor), the growth was still very slow.

Figure 4: Growth of WT Syn UTEX 2973 in BGM medium (Temperature 38°C, Light 250 μmoles photon.m-2.s-1, CO2 0,04%)

Light quantity and quality

After discussing this with Dr. Julie Zedler, we came to question the quality of the light spectrum provided to our cultures. The spectrum of the LED lights we were using (Aquabeauty 70W) seemed indeed to be very blue (Figure 5). This type of light is not optimal because unlike plants, cyanobacteria have in addition to chlorophyll (which absorbs in the blue) phycocyanins that absorb in the red (Ungerer et al. 2018). So, we decided to get a new LED panel (Greenception GC4) with a spectrum more adapted to our growths (we chose the reference provided by the iGEM Marburg 2019 team with which they were able to obtain a fast growth phenotype). The results obtained with these new lamps were again disappointing, not significantly improving the growths.

Figure 5: Comparison of the emitted light spectrum of the first lights we were using (A) and the second ones that were purchased (C) with the absorption spectra of our cyanobacterium strain Syn UTEX 2973 (B). The black arrow indicates the absorption peak related to phycobiliproteins, in red-orange for (B).

Overall, we were not able to find the growth rate-limiting element of our strains in the time available, despite all the tests conducted. These difficulties highlight the need for standardization in these fields and justify the interest of the Phototroph Guide to which we contributed. These autotrophic chassis are very promising for the future of industrial biotechnologies, but it can be difficult to work with them without specific equipment for the culture of phototrophic microorganisms.

Sucrose secretion

Consistent with what was described above, we were unable to reproduce the high sucrose productions obtained by Lin et al. (2020) with the strains we used. This seems logical in view of the differences in growth rates already detailed. Sucrose could still be produced and detected in Syn UTEX 2973 CscB and CscB SPS SPP strains. At the same time as the different tests to optimize growth, these mutants were also grown under the different conditions and sucrose production induced. The highest production was obtained with the CscB strain for which a titer of 1,36 g/L in 7 days (BGM medium with the new LED panels at light intensity 650 μmoles photon.m^-2.s^-1). For the CscB SPS SPP strain, the highest titer was 300 mg/L obtained after 8 days in the same conditions. In both cases, it remains too low considering these strains had reported productivities of 1.9 g.L⁻¹day⁻¹ for the CscB strain and 1.4 g.L⁻¹day⁻¹ for the CscB SPS SPP one (Lin et al. 2020).

Figure 6: Sucrose calibration curve (left) and example of HPLC analysis result (right). Both the WT and CscB strains were cultivated for 7 days in BG11 with 150 mM NaCl and 1 mM IPTG. The titer of sucrose was determined to be 1.36 g/L which corresponds to the highest production that we obtained. The peak at 7.2 minutes most likely is sucrose-6-phosphate which is the precursors of sucrose.

Tips

The transporter CscB is a sucrose/H+ symporter which implies that sucrose export is dependent upon culture pH. A pH of a least 7.8 is necessary to promote sucrose secretion (Ducat et al. 2012).

Co-culture

We demonstrated that the yeast was able to grow on sucrose in the co-culture medium. Our cyanobacteria strains, although very slowly, produce sucrose. The next step was to try to put the two organisms together and see if the sucrose produced by the cyanobacteria was sufficient to keep the yeast alive and growing.

To do this, we inoculated the BG-11[CoY] medium with the CscB strain by adding 1 mM IPTG to induce sucrose production. As with previous cultures, sucrose concentrations were extremely low, in the order of 100 mg/L after 4 days. We were not very hopeful that our yeast could grow with such low amounts of sugar. However, we still inoculated our LycoYeast strain, which had previously been grown in the sucrose-enriched co-culture medium, at a very low density (OD_600nm=0.03). The same procedure was applied with a culture of WT Syn UTEX 2973 as a control. The co-culture temperature was set at 35°C. Serial dilutions of these co-culture mixes were plated on YPD petri dishes. The plates were incubated 2 days at 30°C without light and the yeast colonies were counted.

No significant yeast growth could be observed in our experimental setup. Still, we observed that the yeast remained alive for at least 3 days, which shows that the two organisms can cohabit in the same compartment. This attempt has allowed the development of the protocols necessary for the establishment of the co-culture, which is a critical step forward. We are hopeful that we could establish co-culture, provided we could reproduce the sucrose productivity described by Lin et al. (2020), thus requiring us to identify the limiting parameter in our cyanobacteria cultures.

Conclusion

We showed both yeast and cyanobacterium strains to be compatible with growth in BG11[CoY] medium. The yeast strain is able to grow from sucrose on this medium. However, the sucrose production by the cyanobacterium and its growth remains to be optimized. From our preliminary results, well-controlled bioreactor growth seems the best solution.

Cloning

In this section, we will present you how we created the three plasmids to produce violet scent molecules in the yeast (pFLEUR, pFRAMBOISE and pVIOLETTE), and the plasmid to add a green leaf scent to the cyanobacterium (pCONCOMBRE).

In order to produce our molecules of interest, cloning was classically performed in two steps:

Amplification of the desired fragments (from IDT and Twist Bioscience blocks, available plasmids or genome sequences) by PCR.
Assembly of the PCR amplified fragments and vectors by InFusion. All our routine protocols can be found in the Protocol page. The list of all the primers used for PCR is on our Material page.

Saccharomyces cerevisiae

pFLEUR

pFLEUR (Fig.7) is a pCfB3038 XII-1 based vector (EasyClone-MarkerFree kit) designed to carry:

the DBR1 and LIS genes, for the production of linalool and dihydro-β-ionone.
the inducible promoters Z3eV and PCUP1, driving the expression of DBR1 and LIS
the resistance marker HygR to select yeast integrants

Figure 7: pFLEUR plasmid for production of linalool and dihydro-β-ionone. Two integrative sites flank the construction for integration into the yeast chromosome XII locus 1.

Cloning strategy

Fragments and vectors were amplified by PCR using the high fidelity Q5 polymerase and further assembled by In-Fusion. According to our advisors and the Takara kit handbook advice, we separated the cloning in two steps for an optimum of efficacy:

First, the Z3eV and HygR fragments were inserted into the pCfB3038 XII-1 vector, generating the intermediate construct pFLE.
Second, DBR1 and LIS fragments were inserted in pFLE linearized by inverse PCR, resulting in the final pFLEUR construct.

Vectors construction

pFLE:

Expected size of the amplicons:

Name of the amplicon	Expected size (pb)
Z3eV	2065
HygR	1860
pCfB3038 XII-1	3991

Click here for detailed results

Z3eV and HygR fragments were PCR amplified from IDT gblock while pCfB3038 XII-1 was linearized by PCR from Addgene template plasmid. Amplification products size was checked on EtBr stained agarose gel. Amplicons were further purified on gel when other amplification products were detected.
Figure 8abc shows amplification products matching with expected sizes.
Z3eV and HygR fragments were inserted into pCfB3038 XII-1 by InFusion. Transformants were selected on ampicillin plates.
14 transformants were screened by colony PCR with primer pairs flanking the inserted fragments (F Z3eV + R Z3eV and F HygR + P reverse screen pFleur).
2 positive transformants (clones 3 and 13) were detected.
Plasmids were extracted from clones 3 and 13 and digested by restriction enzymes (SfoI and ScaI) to check the correctness of the assembly.
Figure 8d shows correct restriction maps for clones 3 and 13.
Plasmids from clones 3 and 13 were sent to Eurofins Genomics to check the inserted sequences. A missense mutation was detected in the DBR1 coding sequence for clone 3, while a minor point mutation was detected in the HygR terminator sequence.

pFLE from clone 13 was then chosen to further insert DBR1 and LIS fragments.

Figure 8: pFLE assembly. PCR amplicon sizes of HygR (A), Z3eV (B) and pCfB3038 XII-1 (C) and the restriction profile of pFle (D) were checked with agarose electrophoresis gel and revealed with EtBr. A theoretical gel is presented on the right of each gel and the NEB 1 kb DNA ladder is on the right or the left of each of them (note that a different ladder is presented on the theoretical gel).

pFLEUR:

Expected size of the amplicons:

Name of the amplicon	Expected size (pb)
pFle	7916
LIS	2512
DBR1	1396

Click here for detailed results

LIS and DBR1 fragments were PCR amplified from IDT gblock while pFLE was linearized by inverse PCR on pFLE Miniprep. Amplification products size was checked on EtBr stained agarose gel. Amplicons were further purified on gel when other amplification products were detected.
Figure 9ab shows amplification products matching with expected sizes.
LIS and DBR1 fragments were inserted into pFLE by InFusion. Transformants were selected on ampicillin plates.
7 transformants were screened by colony PCR with primer pairs flanking the LIS block (F LIS+ R LIS).
1 positive transformant (clone 7) was detected.
Plasmid was extracted from clone 7 and digested by restriction enzymes (ScaI and SalI) to check the correctness of the assembly.
Figure 9 shows correct restriction maps for clones 7.
Plasmid from clone 7 was sent to Eurofins Genomics to check the inserted sequence. No mutation was detected for clone 7.

We managed the assembly of pFLEUR with the plasmid coming from clone 7. The pFLEUR construction for production of linalool and dihydro-β-ionone is ready to be integrated in the yeast genome

Figure 9: pFLEUR assembly. PCR amplicon sizes of LIS, DBR1 (A) and pFLE linearised (B) and pFLEUR restriction profile (D) were checked with agarose electrophoresis gel and revealed with EtBr. A theoretical gel is presented on the right of each gel and the NEB 1 kb DNA ladder is on the right or the left of each of them (note that a different ladder is presented on the theoretical gel).

pFRAMBOISE

pFRAMBOISE-fused (Fig.10) is a pCfB3034 X-3 based vector (EasyClone-MarkerFree kit) designed to carry:

the CrtY-phCCD1 genes fusion with a LGS linker, responsible for the production of both β-carotene and β-ionone.
the inducible promoters Teto7, driving the expression of CrtY-phCCD1.
the resistance marker NeoR to select for yeast integrants.

A backup plan, pFRAMBOISE-notfused, was also designed. In this case, CrtY is under the control of the Teto7 promoter, and the phCCD1 with a fynAnchor is constitutively expressed under a TEF1 promoter.

pFRAMBOISE-fused

pFRAMBOISE-notfused

Figure 10: pFRAMBOISE-fused and pFRAMBOISE-notfused plasmids for production of β-carotene and β-ionone.Two integrative sites flank the two constructions for integration in the chromosome X locus 3.

Cloning strategy

The cloning strategy is the same as pFLEUR:

First, the Teto7 promoter together with the NeoR fragment were inserted in the pCfB3034 X-3 vector, generating the intermediate construct pFRAMB.
Second, CrtY-phCCD1 and rtTA fragments were assembled with pFRAMB linearised by inverse PCR, resulting in the final pFRAMBOISE construct.

For pFRAMBOISE-notfused, CrtY and fynAnchor-phCCD1 fragments were assembled with pFRAMBOI linearized by inverse PCR from pFRAMBOISE-fused, resulting in the final pFRAMBOISE-notfused construct.

Vectors construction

pFRAMB:

Expected size of the amplicons:

Name of the amplicon	Expected size (pb)
Promoter Teto7	1008
NeoR	1293
pCfB3034 X-3	4031

Click here for detailed results

Promoter Teto7 was PCR amplified from the addgene plasmid #165976 and the NeoR fragment was PCR amplified from IDT gblock, while pCfB3034 X-3 was linearized by inverse PCR on Addgene template plasmid. Amplification products size was checked on EtBr stained agarose gel. Amplicons were further purified on gel when other amplification products were detected.
Figure 11abc shows amplification products matching with expected sizes. For Promoter Teto7, the right strip was cut for a gel purification.
Promoter Teto7 and NeoR fragments were inserted into pCfB3034 X-3 by InFusion. Transformants were selected on ampicillin plates.
9 transformants were screened by colony PCR with primer pairs on the promoter Teto7 and the NeoR fragments (F pTet + R pTet and FVCNeoR + R NeoR ).
6 positive transformants (clones 4A, 4B, 4C, 4E, 5A and 5C) were detected
Plasmids were extracted from the 6 clones and digested by restriction enzyme (EcoRV) to check the correctness of the assembly.
Figure 11d shows correct restriction maps for clone 4E.
Plasmid from clone 4E was sent to Eurofins Genomics to check the insert sequences. No mutation was detected in this clone.

pFRAMB from clone 4E was then chosen to further insert CrtY-phCCD1 and rtTA fragments.

Figure 11: pFRAMB assembly. PCR amplicon sizes of NeoR(A), Promoter Teto7(B) and pCfB3034 X-3 linearized (C) and pFRAMB restriction profile (D) were checked with agarose electrophoresis gel and revealed with EtBr. A theoretical gel is presented on the right of each gel and the NEB 1 kb DNA ladder is on the right or the left of each of them (note that a different ladder is presented on the theoretical gel).

pFRAMBOISE-fused:

Expected size of the amplicons:

Name of the amplicon	Expected size (pb)
CrtY-phCCD1	2862
rtTA	1565
pFramb	6332

Click here for detailed results

CrtY-phCCD1 and rtTA fragments were PCR amplified from IDT gblock while pFramb was linearized by inverse PCR on pFRAMB Miniprep. Amplification products size was checked on EtBr stained agarose gel. Amplicons were further purified on gel when other amplification products were detected.
Figure 12abc shows amplification products matching with expected sizes.
CrtY-phCCD1 and rtTA fragments were inserted into pFRAMB by InFusion. Transformants were selected on ampicillin plates.
12 transformants were screened by colony PCR with primer pairs on the CrtY-phCCD1 and rtTA fragments (F CrtY-phCCD1+ R CrtY-phCCD1 and F rtTA+ R rtTA ).
1 positive transformant (clone C) was detected.
Plasmid was extracted from clone C and digested by restriction enzymes (ScaI and SalI) to check the correctness of the assembly.
Figure 12d shows correct restriction maps for clones C.
Plasmid from clone C was sent to Eurofins Genomics to check the insert sequence. No mutation was detected for clone C.

We managed the assembly of pFRAMBOISE-fused with the plasmid coming from clone C. The pFRAMBOISE-fused construction for production of β-carotene and β-ionone is ready to be integrated in the yeast genome

Figure 12: pFRAMBOISE-fused assembly. PCR amplicon sizes of CrtY-phCCD1 (A), rtTA (B) and pFramb linearized (C) and pFRAMBOISE-fused restriction profile (D) were checked with agarose electrophoresis gel and revealed with EtBr. A theoretical gel is presented on the right of each gel and the NEB 1 kb DNA ladder is on the right or the left of each of them (note that a different ladder is presented on the theoretical gel).

pFramboise-notfused:

Expected size of the amplicons:

Name of the amplicon	Expected size (pb)
fyn-phCCD1	1795
CrtY	1796
pFramboi	7795

Click here for detailed results

CrtY and fyn-phCCD1 fragments were PCR amplified from Twist Bioscience gblock while pFRAMBOI was linearized by inverse PCR on pFRAMBOISE-fused Miniprep. Amplification products size was checked on EtBr stained agarose gel. Amplicons were further purified on gel when other amplification products were detected.
Figure 13ab shows amplification products matching with expected sizes.
CrtY and phCCD1 fragments were inserted into pFRAMBOI by InFusion. Transformants were selected on ampicillin plates.
5 transformants were screened by colony PCR with primer pairs on the CrtY and rtTA fragments (F CrtY+ R CrtY and F fyn-phCCD1+ R fyn-phCCD1 ).
2 positive transformants (clones 1 and 3) were detected.
Plasmids were extracted from the two clones and digested by restriction enzymes (NotI and SalI) to check the correctness of the assembly.
Figure 13c shows correct restriction maps for clone 3.
Plasmid from clone C was sent to Eurofins Genomics to check the insert sequences. No mutation was detected for clone 3.

We managed the assembly of pFRAMBOISE-notfused with the plasmid coming from clone 3. The pFRAMBOISE-notfused construction for production of β-carotene and β-ionone is ready to be integrated in the yeast genome

Figure 13: pFRAMBOISE-notfused assembly. PCR amplicon sizes of CrtY, fyn-phCCD1 (A),and pFRAMBOIi linearized (B) and pFRAMBOISE-notfused restriction profile (C) were checked with agarose electrophoresis gel and revealed with EtBr. A theoretical gel is presented on the right of each gel and the NEB 1 kb DNA ladder is on the right or the left of each of them (note that a different ladder is presented on the theoretical gel).

pVIOLETTE

pVIOLETTE (Fig.14) is a pCfB3040 XII-4 based vector (EasyClone-MarkerFree kit) designed to carry:

the LcyE gene fused to ofCCD1 with a LGS linker, for the production of both ε-carotene and α-ionone
the inducible promoters Gal 1, driving the expression of LcyE-phCCD1
the resistance marker NsrR to select for yeast integrants

Figure 14: pVIOLETTE plasmid for production of ε-carotene and α-ionone. Two integrative sites flank the construction for integration in the chromosome XII locus 4.

Cloning strategy

The first cloning strategy was to directly integrate the 3 fragments LcyE, ofCCD1 and NsrR into pCfB3040 XII-4. Unfortunately, no positive transformants were selected. We then considered a new cloning strategy:

First, the ofCCD1 fragment was inserted together with the NsrR fragment in the pCfB3040 XII-4 vector, generating the intermediate construct pVIOLET.
Second, LcyE was assembled with pVIOLET linearised by inverse PCR, resulting in the final pVIOLETTE construct.

Vectors construction

pVIOLET:

Expected size of the amplicons:

Name of the amplicon	Expected size (pb)
ofCCD1	2101
NsrR	1196
pCfB3040 XII-4	3856

Click here for detailed results

NsrR and ofCCD1 fragments were PCR amplified from IDT gblocks and pCfB3040 XII-4 was linearized by inverse PCR. Amplification products size was checked on EtBr stained agarose gel. Amplicons were further purified on gel when other amplification products were detected.
Figure 15abc shows amplification products matching with expected sizes.
NsrR and ofCCD1 fragments were inserted into pCfB3040 XII-4 by InFusion. Transformants were selected on ampicillin plates.
10 transformants were screened by colony PCR with primer pairs flanking the integration of the fragments ofCCD1 and NsrR (P seq LcyE CCD1 (1) + P reverse screen LcyE CCD1).
3 positive transformants (clones 8, 21 and 22) were detected.
Plasmids were extracted from the 3 clones and digested by restriction enzymes (SacI + EcoRV) to check the correctness of the assembly.
Figure 15d shows correct restriction maps for clone 8.
Plasmid from clone 8 was sent to Eurofins Genomics to check the insert sequences. No mutation was detected in this clone.

pVIOLET from clone 8 was then chosen to further insert the LcyE fragment.

Figure 15: pViolet assembly. PCR amplicon sizes of ofCCD1 (A), NsrR (B),and pCfB3040 XII-4 linearized (C) and pVIOLET restriction profile (D) were checked with agarose electrophoresis gel and revealed with EtBr. A theoretical gel is presented on the right of each gel and the NEB 1 kb DNA ladder is on the right or the left of each of them (note that a different ladder is presented on the theoretical gel).

pVIOLETTE:

After many unsuccessful attempts to integrate LcyE, we decided to change the homology region for InFusion. A 500 bp fragment (NewLinker) flanking the linker was designed together with new LcyE (NewLcyE) and ofCCD1 (NewofCCD1) fragments for a 3-fragment InFusion cloning with the pVio vector (pCfB3040+NsrR) from the inverse PCR of pViolet. According to Takara kit handbook advice and our experiments, a GC-rich homology zone is less likely to work for inFusion cloning. We therefore advise future iGEM teams to have homology zones for InFusion that flank such sequences

Expected size of the amplicons:

Name of the amplicon	Expected size (pb)
NewLcyE	1735
NewofCCD1	1886
NewLinker	498
pVIO	7951

Click here for detailed results

NewLcyE, NewofCCD1 and NewLinker fragments were PCR amplified from IDT and Twist Bioscience gblocks and pVIO was linearized by inverse PCR from pVIOLET Miniprep. Amplification products size was checked on EtBr stained agarose gel. Amplicons were further purified on gel when other amplification products were detected.
Figure 16abc shows amplification products matching with expected sizes.
NewLcyE, NewofCCD1 and NewLinker fragments were inserted into pVIO by InFusion. Transformants were selected on ampicillin plates.
12 transformants were screened by colony PCR with primer pairs flanking the integration of the fragments NewofCCD1 and NewLcyE (P seq LcyE CCD1 (1) + R NsrR).
4 positive transformants (clones 1,2,3 and 4) were detected.
Plasmids were extracted from the 4 clones and digested by restriction enzymes (HindIII + XbaI) to check the correctness of the assembly.
Figure 16d shows correct restriction maps for clone 3.
Plasmid from clones 1,2,3 and 4 were sent to Eurofins Genomics to check the insert sequences. No mutation was detected in clone 3, mutations in junk DNA were present for clones 1 and 2, and a 5 codons deletion in the linker sequence was detected for clone 4.

We managed the assembly of pVIOLETTE with the plasmid coming from clone 3. The pVIOLETTE construction for production of ε-carotene and α-ionone is ready to be integrated in the yeast genome.

Figure 16: pViolette assembly. PCR amplicon sizes of NewLcyE, NofCCD1 (A), NewLinker (B),and pVio linearized (C) and pVIOLETTE restriction profile (D) were checked with agarose electrophoresis gel and revealed with EtBr. A theoretical gel is presented on the right of each gel and the NEB 1 kb DNA ladder is on the right or the left of each of them (note that a different ladder is presented on the theoretical gel).

Synechococcus elongatus

pCONCOMBRE

pCONCOMBRE (Fig.17) is a pAM4951 based vector designed to carry:

the Nb-9-LOX, Cm-9-HPL and LacI genes, for the production of 3,6-nonadienal and LacI repressor
the inducible promoters Trc-theoE-riboswitch, driving the expression of Nb-9-LOX and Cm-9-HPL
the resistance marker SpecR to select for Cyanobacteria integrants

Figure 17: pCONCOMBRE plasmid design for production of 2,6-nonadienal and 2,6-nonadienol. Two integrative sites flank the construction for an integration in the NSI locus.

Cloning strategy

The cloning strategy is the same as for the LycoYeast plasmids, but the fragments Nb-9-LOX, Cm-9-HPL and LacI were integrated into the pAM4951 vector in only one step.

Vectors construction

Expected size of the amplicons:

Name of the amplicon	Expected size (pb)
Nb-9-LOX	2796
Cm-9-HPL	1646
LacI	1276
pAM4951	4021

Click here for detailed results

Nb-9-LOX, Cm-9-HPL fragments were PCR amplified from IDT gblocks, LacI fragment was PCR amplified from the genome of Syn UTEX 2973 CscB strain and pAM4951 was linearized by inverse PCR. Amplification products size was checked on EtBr stained agarose gel. Amplicons were further purified on gel when other amplification products were detected.
Figure 18abc shows amplification products matching with expected sizes.

One run of PCR colony on Synechococcus elongatus genome to obtain an amplified fragment is not enough, we had to purify on gel our LacI amplicons for 10 more cycles of PCR
Nb-9-LOX, Cm-9-HPL and LacI fragments were inserted into pAM4951 by InFusion. Transformants were selected on spectinomycin plates.
14 transformants were screened by colony PCR with primer pairs on the fragments LacI, Nb-9-LOX and Cm-9-HPL (F LacI + R LacI, F LOX + R LOX and F HPL and R HPL).
3 positive transformants (clones A2, A5 and B6) were detected.
Plasmids were extracted from the 3 clones and digested by restriction enzymes (EcoRI + EcoRV) to check the correctness of the assembly.
Figure 18d shows correct restriction maps for clone A5.
The plasmid from clone A5 was sent to Eurofins Genomics to check the insert sequences. No mutation was detected.

We managed the assembly of pCONCOMBRE with the plasmid coming from clone A5. The pCONCOMBRE construction for production of 2,6-nonadienal and 2,6-nonadienol is ready to be integrated in the cyanobacterium genome.

Figure 18: pCONCOMBRE assembly. PCR amplicon sizes of HPL, LOX (A), LacI (B),and pAM4951 linearized (C) and pCONCOMBRE restriction profile (D) were checked with agarose electrophoresis gel and revealed with EtBr. A theoretical gel is presented on the right of each gel and the NEB 1 kb DNA ladder is on the right or the left of each of them (note that a different ladder is presented on the theoretical gel).

Conclusion

Even if some cloning required design modifications, we successfully built all of our constructions to engineer the yeast and the cyanobacterium.

Strain construction

Here we present our efforts to adapt our yeast chassis and then to introduce our constructions in the yeast and in the cyanobacterium genomes.

Yeast genetic engineering

Generate a kanMX - LycoYeast using Cre-Lox system

Integration of both pFRAMBOISE vectors in the LycoYeast genome relies on the neoR gene coding for kanamycin/neomycin/G418 resistance. Unluckily, our LycoYeast is a modified yeast strain containing several resistance markers, including the kanMX gene coding for kanamycin/neomycin/G418 resistance. This kanMX gene is flanked by two facing loxP sequences allowing its removal by the action of the Cre recombinase.

Figure 19: The loxP/Cre gene disruption and marker rescue procedure.

Cre-Lox recombination is a site specific recombination system used to carry out genetic modifications. The protein Cre recombinase recognizes 34 bp loxP sites and allows rearrangement of the in-between genetic material. Here, we used the Cre-loxP to remove the kanMX gene from the LycoYeast strain.

To do so, we transformed our yeast with the plasmid pSH65 which contains:

the cre gene coding for the Cre recombinase, under the control of a galactose inducible promoter.
a phleomycin resistance marker.

Transformants were selected on phleomycin (20 μg.mL-1) plates and further grown on YP+Galactose medium in order to induce cre expression and promote kanMX excision.

Tips

It is important to know that phleomycin loses its activity at acidic pH. Thus, the pH of the medium has to be adjusted to 8.

After 12 h of growth in presence of galactose, transformants were plated on phleomycin plates and growing colonies were replica-plated on three media (Figure 20):

YPDA (all clones should form colonies).
YPDA with phleomycin 20 μg.mL-1 pH 8 (only clones containing the pSH65 plasmid should grow).
YPDA with neomycin 250 μg.mL-1 (only the colonies for which the Cre-loxP has not worked should grow, i.e., kanMX+ clones)

Figure 20: Left: phleomycin medium. Center: no antibiotic medium. Right: neomycin medium. In each plate, 6 colonies from the clone B and 6 colonies from the clone D are grown.

Unexpectedly, all colonies grew on the plate containing neomycin. We initially thought that we had prepared the neomycin plates incorrectly. So, we conducted a test for neomycin selectivity with three different strains: the kanMX+ LycoYeast and two other strains, BY4741 and CEN-PK2-1C, that were kanMX-. In addition to neomycin, we conducted the same test with YPDA medium containing G418 (500 μg.mL-1), a neomycin analog.

Figure 21: Left: G418 dish. Right: neomycin dish. On the G418 dish, only the resistant LycoYeast strain grows. On the contrary, on the neomycin dish, all strains grow.

kanMX+ and kanMX- strains both grew on neomycin plates, while only kanMX+ strains grew on G418 plates, showing that the kanMX cassette is specific for G418 resistance. The selection of induced transformants was then repeated on plates containing G418.

Figure 22: Cre-lox induced transformants on YPDA medium supplemented with G418 (500 μg.mL-1). 3 colonies seem to have grown despite the presence of the antibiotic (B4, B7, D10). For these colonies, the Cre-loxP did not work since the selection marker is still present.

In Figure 22, we can see that most clones show G418 sensitivity profile so the kanMX resistance gene was indeed deleted in these colonies. This kanMX- strain was named LycoYeast-G418 and saved as glycerol stocks.

Integration into the kanMX- strain

All our constructs were designed to be integrated into the yeast genome by homologous recombination. They must therefore be linearized by PCR. The transformation technique we used was LiAc/ss-DNA/PEG following the TAKARA protocol (see Protocol section).

PEG is necessary for the transformation since it allows the DNA to bind to the yeast walls.
LiAc and heat shock promote the penetration of DNA into the cell.
Finally, the carrier ss-DNA increases the transformation frequency but its function is still unclear (Kawai et al. 2010).

In addition, to improve the transformation frequency, 5 μg of DNA was added to the transformation mix.

Tips

We tried many protocols before successfully integrating our constructs into the yeast genome. We then realized that the problem was the amount of DNA template we were adding to the mix. In order to successfully integrate a large construct into the genome of S. cerevisiae, large amount of DNA is necessary. This is what we observed with our 10 kb constructs using about 5μg of DNA.

For each construction, integration of our constructs was checked by PCR (yeast DNA extraction protocol can be found in the protocols section) with various primers pairs:

external primers: would give a product too long to be amplified
LA: one primer on the upstream genome region and one primer on the insert
RA: one primer on the downstream genome region and one primer on the insert

The three integration sites were chosen thanks to the advice of Sara Castano-Cerezo (see IHP and design) because they were the few left available in the LycoYeast strain she provided us.

pVIOLETTE → locus 4, chromosome XII
pFLEUR → locus 1, chromosome XII
pFRAMBOISE (fused and not-fused) → locus 3, chromosome X

The objective is first to obtain strains containing one construct each. Then, once the first integration is achieved, the other constructs are inserted into the yeast genome sequentially. We started with pVIOLETTE, then pFLEUR and finally pFRAMBOISE-fused or pFRAMBOISE-notfused.

LycoYeast-VIOLETTE

LycoYeast-G418 was transformed with pVIOLETTE following the TAKARA protocol

The recombination takes place at locus 4 of chromosome XII.
The integration of pVIOLETTE confers resistance to nourseothricin through the nsrR gene.

Transformants were selected on YPDA medium containing 250 μg/mL nourseothricin. About ten clones were obtained and integration was verified in all clones by PCR (Figure 23). In Figure 23, only two clones are shown; other results are not shown. It can be seen in Figure 23 that only clone 2 shows a correct integration profile, i.e. a band for both LA and RA at the expected size.

Figure 23: 0.8 % agarose gel stained with EtBr. Integration of pVIOLETTE in LycoYeast-G418 at chromosomic XII locus 4 was checked by PCR. At the bottom left are presented the result of genotyping of LycoYeast – VIOLETTE, clone 1 contains only the left arm (LA) while clone 2 contains the left and right arm (LA and RA). 4. At the bottom right are presented the theoretical results (LA: 627 bp ; RA: 644bp).

Integration of the pVIOLETTE insert at locus XII-4 was successful. The integrant strain was named LycoYeast - VIOLETTE and saved as glycerol stock.

LycoYeast–FLEUR

LycoYeast - G418 was transformed with pFLEUR.

The recombination takes place at locus 1 of chromosome XII.
The integration of pFLEUR confers resistance to hygromycin through the HygR gene.

Transformants were selected on YPDA medium containing 200 μg/mL hygromycin. About ten clones were obtained and integration was verified on all clones by PCR (Figure 24). In Figure 24, 3 clones are shown of which only one has the construction (the right and left homology arms show a band at the expected size).

Figure 24: Integration of the correct integration of pFLEUR in LycoYeast-G418 at chromosome XII locus 1 was checked by PCR. At the bottom left are presented the result of genotyping of Lycoyeast – FLEUR, one clone has both left arm (LA) and right arm (RA) i.e. the entire construction in the locus 1. At the bottom right are presented the theoretical results (LA: 2,709 bp ; RA: 2,757 bp).

Integration of the pFLEUR insert at locus XII-1 was successful. The integrant strain was named LycoYeast - FLEUR and saved as glycerol stock.

LycoYeast–FRAMBOISEfused

LycoYeast-G418 was transformed with pFRAMBOISE-fused.

The recombination takes place at locus 3 of chromosome X.
The integration of pFRAMBOISE-fused) confers resistance to G418 through the neoR gene.

Transformants were selected on YPDA medium containing 400 μg/mL G418. About twenty clones were obtained and integration was verified on 10 clones by PCR (Figure 25). In Figure 25, only one correct clone is shown.

Figure 25: Integration of pFRAMBOISE-fused in LycoYeast-G418 at chromosome X locus 3 was checked by PCR. At the bottom left are presented the result of genotyping LycoYeast – FRAMBOISE(fused), one clone has both left arm (LA) and right arm (RA) i.e. the entire construction. At the bottom right are presented the theoretical results (LA: 2 kbp ; RA: 2 kbp).

Integration of the pFRAMBOISE-fused insert at locus X-3 was successful. The integrant strain was named LycoYeast-FRAMBOISE-fused and saved as glycerol stock.

LycoYeast–FRAMBOISEnotfused

LycoYeast-G418 was transformed with pFRAMBOISE-notfused.

The recombination takes place at locus 3 of chromosome X.
The integration of pFRAMBOISE-notfused confers resistance to this G418 through the neoR gene.

Transformants were selected on YPDA medium containing 400 μg/mL G418. Only one colony was obtained and integration was verified by PCR (Figure 26).

Figure 26: Integration of pFRAMBOISE-not fused in LycoYeast-G418 at chromosome X locus 3 was checked by PCR. At the bottom left are presented the result of genotyping of LycoYeast – FRAMBOISE(not fused), one clone has both left arm (LA) and right arm (RA) i.e. the entire construction. At the bottom right are presented the theoretical results (LA: 2 kbp ; RA: 2 kbp).

Integration of the pFRAMBOISE-notfused insert at locus X-3 was successful. The integrant strain was named LycoYeast-FRAMBOISE-notfused and saved as glycerol stock.

LycoYeast–VIOLETTE–FLEUR–FRAMBOISE

After having successfully transformed LycoYeast-G418 with pVIOLETTE, we sequentially transformed LycoYeast-VIOLETTE with pFLEUR, and then pFRAMBOISE-fused or pFRAMBOISE-notfused.

The final strains containing the three constructs are called:

LycoYeast–VIOLETTE–FLEUR–FRAMBOIS-fused
LycoYeast–VIOLETTE–FLEUR–FRAMBOISEnotfused

LycoYeast-VIOLETTE-FLEUR:

Transformants were selected on YPDA medium containing 250 μg/mL nourseothricin and 200 μg/mL hygromycin. About twenty clones were obtained and integration was verified on 10 clones by PCR (Figure 27). In Figure 27, only one correct clone is shown.

Figure 27: Integration of pFLEUR in LycoYeast – VIOLETTE at chromosome XII locus 1. At the bottom left are presented the result of genotyping of Lycoyeast – VIOLETTE – FLEUR, one clone has both left arm (LA) and right arm (RA) i.e. the entire construction in the locus 1. At the bottom right are presented the theoretical results (LA: 2,709 bp ; RA: 2,757 bp).

Integration of the pFLEUR insert at locus XII-1 was successful. The integrant strain was named LycoYeast-VIOLETTE- FLEUR and saved as glycerol stock.

LycoYeast-VIOLETE-FLEUR-FRAMBOISE-fused:

Transformants were selected on YPDA medium containing 250 μg/mL nourseothricin, 200 μgm/L hygromycin and 400 μg/mL G418. About twenty clones were obtained and integration was verified on 10 clones by PCR (Figure 28). In Figure 28, only one correct clone is shown.

Figure 28: Integration of pFRAMBOISE-fused in LycoYeast – VIOLETTE - FLEUR at chromosomic X locus 3. At the bottom left are presented the result of genotyping of ELIXYeats(fused) one clone has both left arm (LA) and right arm (RA) i.e. the entire construction. At the bottom right are presented the theoretical results (LA: 2 kbp ; RA: 2 kbp).

Integration of the pFRAMBOISE-fused insert at locus X-3 was successful. The integrant strain was named LycoYeast-VIOLETTE-FLEUR-FRAMBOISE-fused and saved as glycerol stock.

LycoYeast-VIOLETTE-FLEUR-FRAMBOISE-not fused:

Transformants were selected on YPDA medium containing 250 μg/mL nourseothricin, 200 μgm/L hygromycin and 400 μg/mL G418. About twenty clones were obtained and integration was verified on 10 clones by PCR (Figure 29). In Figure 29, two clones are shown but only clone 1 has the correct integration profil, i.e. both left and right arms. We can see that clone 2 gives no bands for the LA and RA tracks but gives a band for the track using both primers on the yeast genome, so the construct is not present in its genome.

Figure 29: Integration of pFRAMBOISE-notfused in LycoYeast – VIOLETTE - FLEUR at chromosome X locus 3. At the bottom left are presented the result of genotyping of ELIXYeast(not-fused). The clone 2 has both left arm (LA) and right arm (RA) i.e. the entire construction. On the contrary, clone 1 still has the integration insert so the construct is not present in clone 1. At the bottom right are presented the theoretical results (LA: 2 kbp ; RA: 2 kbp).

Integration of the pFRAMBOISE-notfused insert at locus X-3 was successful. The integrant strain was named LycoYeast-VIOLETTE-FLEUR-FRAMBOISE-fused and saved as glycerol stock.

Success 1

The four inserts were independently integrated into the LycoYeast genome.

Success 2

The final strain with concomitant integration of pVIOLETTE, pFLEUR and pFRAMBOISE inserts was successfully constructed.

Cyanobacterium genetic engineering

Transformation of S. elongatus through triparental conjugation

After successfully cloning pCONCOMBRE, it was time to use it to engineer Syn UTEX 2973. Conjugal transfer of DNA between bacterial cells is a well-described process and has been used previously to engineer cyanobacterial species, particularly those that are not naturally competent, such as Syn UTEX 2973 (Elhai and Wolk 1988; Gale et al. 2019). We therefore relied on this method to introduce pCONCOMBRE in our cyanobacterial strains.

From a practical standpoint, we followed the protocol found in the well-illustrated video article provided by Gale et al. (2019) (for more information, see our Methods page). The details of this technique are presented in Figure 30. In brief, cyanobacterial cultures are incubated in presence of two different E. coli strains called respectively the conjugal strain and the helper strain which allow the transfer of the plasmid of interest (cargo plasmid) to the cyanobacterium in three steps.

Figure 30: Transformation of S. elongatus through triparental conjugation. 1) The helper strain carries two plasmids: the vector to be transferred to the cyanobacteria, called cargo vector, and a helper plasmid. The latter encodes for specific DNA-methylases that act on the cargo vector so that it is not degraded after entry into the cyanobacterium cellular space. (2) The conjugal plasmid carried by the conjugal strain contains the many genes required to encode the conjugal apparatus. In the first step of conjugation, the conjugal plasmid is transferred from the conjugal strain to the helper strain. (3) This helper strain is then able to transfer the methylated cargo vector into the cyanobacterium.

After incubating our E. coli strains with the cyanobacterium strains, the conjugation mixes were plated on sterile filters placed on petri dishes containing a mixture of BG11 and LB media. The protocol was applied to the WT strain as well as to the two different sucrose-secreting strains CscB and CscB SPS SPP. After 24 hours on non-selective medium, these filters were transferred on plates containing spectinomycin, the antibiotic used to select the integration of pCONCOMBRE. Isolated conjugants were obtained after 4 days of incubation of these plates (Figure 31).

Figure 31: Evolution of the visual aspect of the triparental conjugation. After 24h of incubation on non-selective medium, the cyanobacteria cover the filter which becomes green (left). Following the passage of these filters on a selective medium, the mutants that have integrated the spectinomycin resistance cassette appear isolated on the filter (center). The isolated conjugants are restreaked successively on BG11 plates to eliminate E. coli contamination and genotyped by PCR (right).

Tips

The conjugal plasmid used for triparental mating can be very large. For example, the conjugal plasmid we used, pRL443, is 57 kb long. To transform it into E. coli, do not hesitate to highly increase the quantity of DNA you add. In our case, we went as high as 400 ng of DNA to obtain transformants.

Verification of the correct integration of pCONCOMBRE

Integration of our construct was checked by PCR with various primers pairs:

LA: one primer on the upstream genome region and one primer on the insert
RA: one primer on the downstream genome region and one primer on the insert
external primers: would give a product too long to be amplify in case of insert integration

Almost all the clones tested had the same profile for the different tests performed (Figure 32). Amplification products did not however always correlate with expected sizes:

insert product at 2.3 kb when expecting no signal: It is known that S. elongatus UTEX 2973 carries multiple copies of its chromosome (Yu et al. 2015). The conjugants were perhaps not entirely segregated which means that there was still at least one chromosome copy that did not have the insert.
RA additional product at 4.8 kb: Probably corresponds to non-specific amplification.
LA product at 4 kb that was not observed for a WT strain (data not shown) while the expected band was supposed to be 1.5 kb: Our hypothesis is that the cloning vector we used (pAM4951) was poorly sequenced/annotated and that there was a sequence between the LA and the spectinomycin resistance gene (it was noted that several primers designed in this region were inexplicably non-functional).

Because of discrepancies between obtained and expected genotyping profiles, we decided to also check for the presence of all genes from our insert with additional primers: LOX, HPL, LacI and SpecR. Figure 32 shows amplification products for all genes of our insert.

Figure 32: Integration of pCONCOMBRE in S. elongatus at chromosomic neutral site I (NSI)was checked by PCR. At the bottom left are presented the result of genotyping of WT pCONCOMBRE clone 4 (0.8% agarose gel stained with EtBr for visualisation) and at the bottom right are presented the theoretical results.

Overall, it appears that we have successfully inserted pCONCOMBRE into the genome of S. elongatus. The next step is to test our mutant strain!

Tips

Efficent cell lysis ensuring optimal genomic DNA release is critical for the success of colony PCR. The traditionals protocols used for E. coli colony lysis proved to be inefficient in our hands. We advise future iGEM teams to use two methodologies advised during the Phototrop Meetups (see our Methods page):

1-freezing-thawing (-80°C / 60°C cycles)

2-colony lysis in DMSO with 95°C heating for 10 minutes

Conclusion

All our constructions have been successfully integrated in the yeast and cyanobacterium genomes. The strains are ready to be challenged for their capacity to produce violet scent molecules.

Strain characterization for the production of the molecules of interest

Here we explore the functionality of our constructions. We assessed their capacity to transform substrates from the terpene pathway and to produce the violet scent.

Color change in the modified LycoYeasts strains

After verifying the correct integration of our constructs by PCR, our engineered LycoYeast strains were placed on YPD plates containing the inducers with the aim to:

Detect color changes due to the conversion of lycopenes (red) to carotenes (orange), as indicated on our Design page.
Detect smell changes due to production of ionones.

For all the strains that are supposed to produce ionones (LycoYeast-VIOLETTE, LycoYeast–FRAMBOISEfused and LycoYeast–FRAMBOISEnotfused), the expected color change was obvious (Figure 33).

Figure 33: Color change in the modified Lyco Yeast strains. The mutants seem to change from red (lycopene) to orange (carotene) which was the expected result.

For the VIOLETTE construction, it appears (upright corner of Figure 33) that without inducer, the colonies remain red. This is coherent with the literature since P_GAL1is supposed to be a tight promoter (Adams 1972). Conversely, for both pFRAMBOISE constructs, the color change is observed in the presence and absence of doxycycline. Again, this is consistent since P_tetO7 is well known to suffer from significant leakage (Tominaga et al. 2021).

SPOILER: motivated by this encouraging visual evidence, we smelled the Petri plates to maybe detect a difference between the WT and induced mutant strains (knowing this has to be safe as explained in our Safety page). We were delighted to find that a characteristic ionone sweet smell could be noticed! We asked various people from the lab to blindly smell the plates, they also recognized the ionone aroma. We then continued the demonstration by investigating the profile of carotenoids in the cells.

Carotenoid profile analysis

The carotenoids contained in the cells (precursors of the ionones) were extracted using the method described by López et al. (2020). Briefly, the yeast cells were lysed in acetone using glass beads and the supernatant obtained after this lysis was analyzed by RP-HPLC using a C18 column.

pVIOLETTE

In the VIOLETTE strains, lycopene is indeed converted into a new product with a higher retention time upon induction (Figure 34). We did not have a standard for ε-carotene, but this peak had a retention time close to our β-carotene standard. Considering the yellow color of pVIOLETTE strains, as well as the in-line following α ionone production results, this new peak most likely corresponds to ε-carotene, the expected precursor. This confirms that at least the lycopene ε-cyclase activity of our fusion enzyme LcyE-OfCCD1 is present. As observed with the color of the strains, it appears that P_GAL1 does not have any detectable leak with our system.

Figure 34: Carotenoid analysis of the strain engineered with pVIOLETTE. The observed color change in the galactose-induced mutant strain LycoYeast-VIOLETTE is confirmed in HPLC by the apparition of a new peak that most likely corresponds to ε-carotene.

pFRAMBOISE

Using the same method, we were able to detect upon doxycycline induction a peak corresponding to the β-carotene retention time. Our Lyco Yeast pFRAMBOISE-fused and pFRAMBOISE-notfused produce β-carotene (Figure 35)!

The previous observations on the significant leakage of P_tetO7 are confirmed with these carotenoid profiles since no significant difference between induced and non-induced strains are observed. Two hypotheses can be formulated: either the rtTA promoter activator is inactive, or the promoter leakage is sufficient to cause the conversion of all lycopene present in our strains.

Figure 35: Carotenoid analysis of the strains engineered with pFRAMBOISE-fused (left) and pFRAMBOISE-notfused (right) The observed color change in the mutant strains LycoYeast-FRAMBOISEfused and LycoYeast-FRAMBOISEnotfused are confirmed in HPLC by the apparition of a new peak that corresponds to β-carotene.

Tips

Carotenoids are sensitive to light and O₂ and can degrade quite rapidly. This can be seen for example on Figure 35 B) where the samples were left at room temperature for too long before being injected causing the appearance of degradation products that can disturb the analyses. It is necessary to try keeping the samples protected from light and in the freezer (-20°C).

Production of the molecules of interest in vivo

With these exciting results confirming lycopene conversion, it was finally time to test the production of our molecules of interest in vivo. The molecules we aim to produce, as all fragrance compounds, are very volatile. A common strategy to avoid losing these molecules during the culture is to grow the engineered microorganisms in a culture medium supplemented with an organic phase to trap the molecules of interest. This method has particularly been applied to the production of terpenes (Chen et al. 2019; López et al. 2020). Since no other method had been described for the violet leaf aldehydes (2E,6Z)-nonadienal and (2E,6Z)-nonadienol, the same strategy was also applied with the mutant strains S. elongatus.

The most common organic solvent used for ionones is dodecane Chen et al. 2019; López et al. 2020), we therefore relied on this technique for the cultivation of all of our strains. Briefly, yeast and cyanobacterial cultures were performed in a total volume of 20 mL respectively in YPD supplemented with dodecane (20% v/v) for the yeast and in BG11 supplemented with dodecane (20% v/v) for the cyanobacteria. When the cultures reached the stationary phase, samples of the organic phases of the different cultures were collected for GC-MS analysis.

α-ionone

The results of GC-MS for the LycoYeast-VIOLETTE strain are presented in Figure 36. We were delighted to observe the apparition of a peak at the same retention time as the α-ionone standard for the induced LycoYeast-VIOLETTE strain, correlating with the violet smell detected by human noses from the petri dishes. The mass spectra associated with this peak matched with the one obtained with the analytical standard. The α-ionone attribution was further confirmed by the NIST mass spectral library (National Institute of Standards and Technology). Overall, this demonstrated the functionality of our LcyE-OfCCD1 fusion enzyme in S. cerevisiae.

Figure 36: GC-MS analysis of the dodecane layer of the LycoYeast-VIOLETTE cultures. α-ionone is produced in vivo by our strain when it is induced by galactose. On the right are presented the matching mass spectra between the standard and the observed peak. (Note: the standard is here a little too concentrated but it still allows to realize the attribution)

β-ionone

Similarly, we were also able to detect β-ionone in the cultures of LycoYeast-FRAMBOISE-notfused (Figure 37).

Figure 37: GC-MS analysis of the dodecane layer of the LycoYeast-FRAMBOISE-notfused cultures. β-ionone is produced in vivo by our strain with or without induction with doxycycline due to promoter leakage. On the right is presented the mass spectra of β-ionone that matched between the standard and the observed peak (not shown).

We unfortunately did not have time to analyze by GC-MS the cultures of LycoYeast-FRAMBOISE-fused. This strain characterization work was indeed done during the last three weeks of the summer with only two students still in the lab. Due to a malfunction of the GC-MS we were supposed to use initially, we could only perform one round of analysis which was kindly conducted by Toulouse White Biotechnology, one of our sponsors.

The LycoYeast-FRAMBOISE-fused strain was successfully constructed a few days after this round of GC-MS analysis. We were however able to show in HPLC that at least the CrtY activity was present (production of β-carotene) with our CrtY-PhCCD1 fusion enzyme. Because of the odor of this strain, which is similar to the one of previous strains for which we have detected ionones, we are confident that the fusion enzyme we have designed is functional and can produce β-ionone.

Other molecules of interest

As explained before, we only were able to do one set of GC-MS analysis. Unfortunately, we found that linalool, (2E,6Z)-nonadienal and (2E,6Z)-nonadienol were released at the same retention time as dodecane, our solvent. It was therefore not possible to confirm that these molecules were produced. It would be necessary to modify the temperature gradients used in the GC-MS method, which we were unable to do in the allocated time, or to use another extraction solvent.

Enzymatic activity assay of DBR1

The last molecule we were interested in is dihydro-β-ionone. This molecule is produced from β-ionone as a precursor. Since we were not sure that β-ionone was produced by our strains when we prepared our samples for our single GC-MS analysis, we decided to try investigating if the enzyme catalyzing this step of conversion, DBR1, was active.

To do so, we cultivated the strain LycoYeast-FLEUR and induced the expression of DBR1 by adding copper to the culture medium in late exponential phase. After waiting for the enzymes to be produced, the yeast cells were lysed in TRIS-HCl pH 6.5 buffer using glass beads. The activity was studied by GC-MS using β-ionone as substrate, adapting the method described by Zhang et al. (2018). Assays contained 50 mM TRIS-HCl (pH 6.5), 1 mM NADPH, 2 mM dithiothreitol (DTT), 1 mM β-ionone and 500 μL of the yeast lysate in a total volume of 1 mL at 45 °C for 1.5 h. At the beginning and end of the incubation, 200 μL of the mix was collected and then a liquid-liquid extraction was carried out by vortexing the sample with 200 μL of dodecane. The dodecane phase was then analyzed by GC-MS. The results are presented in Figure 38.

Figure 38: Activity assay of DBR1. A portion of the β-ionone is converted into dihydro- β-ionone by DBR1.

At the initial state t=0, we can only see β-ionone which is the substrate that was added to the reaction mix. After 1.5 hours, dihydro-β-ionone is detected in the reaction media, thus demonstrating that the enzyme DBR1 is indeed active.

Conclusion

The results are beyond our hopes. We managed to smell yeast actually smelling like violet. We observed the change of color of our strain induced to produce our molecules. Eventually, we analytically detected all three ionones. However, we have not been able to assess the production of linalool, (2E,6Z)-nonadienal and (2E,6Z)-nonadienol due to our concentration strategy with dodecane and lack of time.

Final conclusion

We engaged in this project like all iGEMers do, full of hope and great ideas. Our instructors kept telling us that most experiments are often prone to fail. Here our main dream was to create a violet-smelling yeast and we definitely succeeded. Our second ambition was to prove the viability of the co-culture to sustainably produce this scent. This will require more time and energy since the growing Synechococcus elongatus was a lot more challenging than expected. Anyway, we have demonstrated that a small team of student in a limited time can engineer a microorganism to produce a defined synthetic scent, and hence, the power and accessibility of synthetic biology for application in the perfume industry.

References

Adams BG. 1972. Induction of galactokinase in Saccharomyces cerevisiae: kinetics of induction and glucose effects. J Bacteriol. 111(2):308–315. doi:10.1128/jb.111.2.308-315.1972.

Chen X, Shukal S, Zhang C. 2019. Integrating Enzyme and Metabolic Engineering Tools for Enhanced α-Ionone Production. J Agric Food Chem. 67(49):13451–13459. doi:10.1021/acs.jafc.9b00860.

Ducat DC, Avelar-Rivas JA, Way JC, Silver PA. 2012. Rerouting carbon flux to enhance photosynthetic productivity. Appl Environ Microbiol. 78(8):2660–2668. doi:10.1128/AEM.07901-11

Elhai J, Wolk CP. 1988. Conjugal transfer of DNA to cyanobacteria. Methods Enzymol. 167:747–754. doi:10.1016/0076-6879(88)67086-8.

Gale GAR, Osorio AAS, Puzorjov A, Wang B, McCormick AJ. 2019. Genetic Modification of Cyanobacteria by Conjugation Using the CyanoGate Modular Cloning Toolkit. JoVE (Journal of Visualized Experiments).(152):e60451. doi:10.3791/60451.

Gueldener U, Heinisch J, Koehler GJ, Voss D, Hegemann JH. 2002. A second set of loxP marker cassettes for Cre-mediated multiple gene knockouts in budding yeast. Nucleic Acids Research. 30(6):23e–223. doi:10.1093/nar/30.6.e23. Hays SG, Yan LLW, Silver PA, Ducat DC. 2017. Synthetic photosynthetic consortia define interactions leading to robustness and photoproduction. J Biol Eng. 11(1):4. doi:10.1186/s13036-017-0048-5.

Kawai S, Hashimoto W, Murata K. 2010. Transformation of Saccharomyces cerevisiae and other fungi: Methods and possible underlying mechanism. Bioengineered Bugs. 1(6):395–403. doi:10.4161/bbug.1.6.13257.

Li T, Li C-T, Butler K, Hays S, Guarnieri M, Oyler G, Betenbaugh M. 2017. Mimicking lichens: Incorporation of yeast strains together with sucrose-secreting cyanobacteria improves survival, growth, ROS removal, and lipid production in a stable mutualistic co-culture production platform. Biotechnology for Biofuels. 10. doi:10.1186/s13068-017-0736-x.

Lin P-C, Zhang F, Pakrasi HB. 2020. Enhanced production of sucrose in the fast-growing cyanobacterium Synechococcus elongatus UTEX 2973. Scientific Reports. 10(1):390. doi:10.1038/s41598-019-57319-5.

López J, Bustos D, Camilo C, Arenas N, Saa PA, Agosin E. 2020. Engineering Saccharomyces cerevisiae for the Overproduction of β-Ionone and Its Precursor β-Carotene. Front Bioeng Biotechnol. 8. doi:10.3389/fbioe.2020.578793. [accessed 2021 Jan 30]. https://www.frontiersin.org/articles/10.3389/fbioe.2020.578793/full.

Peiro C, Millard P, de Simone A, Cahoreau E, Peyriga L, Enjalbert B, Heux S. 2019. Chemical and Metabolic Controls on Dihydroxyacetone Metabolism Lead to Suboptimal Growth of Escherichia coli. Appl Environ Microbiol. 85(15):e00768-19. doi:10.1128/AEM.00768-19.

Rodrigues C, Wahl A, Gombert A. 2021. Aerobic growth physiology of Saccharomyces cerevisiae on sucrose is strain-dependent.

Sonderegger M, Sauer U. 2003. Evolutionary engineering of Saccharomyces cerevisiae for anaerobic growth on xylose. Appl Environ Microbiol. 69(4):1990–1998. doi:10.1128/AEM.69.4.1990-1998.2003.

Tominaga M, Nozaki K, Umeno D, Ishii J, Kondo A. 2021. Robust and flexible platform for directed evolution of yeast genetic switches. Nature Communications. 12. doi:10.1038/s41467-021-22134-y.

Ungerer J, Lin P-C, Chen H-Y, Pakrasi HB. 2018. Adjustments to Photosystem Stoichiometry and Electron Transfer Proteins Are Key to the Remarkably Fast Growth of the Cyanobacterium Synechococcus elongatus UTEX 2973. mBio. 9(1). doi:10.1128/mBio.02327-17. [accessed 2021 Mar 31]. https://mbio.asm.org/content/9/1/e02327-17.

Yu J, Liberton M, Cliften P, Head R, Jacobs J, Smith R, Koppenaal D, Brand J, Pakrasi H. 2015. Synechococcus elongatus UTEX 2973, a fast growing cyanobacterial chassis for biosynthesis using light and CO2. Scientific reports. 5:8132. doi:10.1038/srep08132.

Zhang X, Liao S, Cao F, Zhao L, Pei J, Tang F. 2018. Cloning and characterization of enoate reductase with high β-ionone to dihydro-β-ionone bioconversion productivity. BMC Biotechnol. 18(1):26. doi:10.1186/s12896-018-0438-x.

Go to Engineering success