Team:UFRGS Brazil/Results
Results
Overview
1. Glyphosate in silico measurement
2. Construction of promoter cassettes and transformation
3. Construction of phnEG correcting cassette and transformation
4. Growth Curves
5. Glyphosate quantification
6. Alginate Spheres
7. Biobrick construction
8. Biobrick characterization
9. Challenges
1. Glyphosate in silico measurements
We collected data from Agro Census 2017 - a census about agriculture made by Brazilian Institute of Geography and Statistics (IBGE) - about soy, the most planted grain in Brazil. About 95% of all soy planted here is transgenic, with glyphosate resistance.
Data about type of soil in five cities inside this regions were gathered from the database of climatic data of Brazil were applied to the software Araquá, a tool developed for ambiental risk assessment of compounds used as agrochemicals. This software uses the input data to calculate the amounts of a determined agrotoxic in superficial waters. We considered the glyphosate dose equal to 1920 $g/ha$ as indicated by the software. Also we tested the doses indicated by RoundUp leaflet. Below you can see the table of this five cities and the concentrations of glyphosate found in freshwater evaluated by Araquá software.
As the soil type was the same in all five locals evaluated, we can observe the same amounts of glyphosate carried superficially in all cities analyzed.
The brazilian legislation allows 500 µg/L, showing that even with large amounts of glyphosate per hectare led to values admitted by law. This result shows that only superficial carriage is not enough to generate the high concentrations of glyphosate measured in Brazilian rivers, and there must be other factor influencing these measures, like misuse by agricultors and wrong discard of glyphosate packaging, as pointed by some agricultors and specialists in our human practices survey.
The intensive use of glyphosate-based herbicides are shown to be present in water bodies even long after its application, mainly in areas nearby, and this is not a recent reality, as there are older studies that already demonstrate the fact. In a study related to the quantification of glyphosate and AMPA in surface waters in Passo do Pilão river, Brazil, “The analysis revealed the presence of the herbicide in superficial areas of this watershed, even in the samples after 30 days of Glyphosate applicatication. […] High concentrations (above 100 ppb) were detected, mainly in points near to intense cultivation areas”. (SILVA et al. 2003)
2. Construction of promoter cassettes and transformation
Achievements:
1. Added homologous flanking sequence to promoter parts
2. Added homologous flanking sequences to CmR_Cassette
3. Construction of three promoter cassettes and confirmation by gel electrophoresis
4. Transformation and confirmation in Escherichia coli K-12
To fuse each of the three promoters with the CmR resistance, we first PCR amplified each promoter with a chimeric primer in order to add a flanking sequence of 30 bp homologous to the CmR_Cassette. Size was confirmed by gel electrophoresis. DNA was purified from gel.
The CmR_Cassette was also PCR amplified in three individual reactions with different chimeric primers in order to add 20 bp flanking sequences homologous to each promoter part. Size was confirmed by gel electrophoresis. DNA was purified from gel.
We individually fused promoter fragments derived from the homologous flanking sequence PCR with the CmR_Cassete + flanking sequences, by Single-Joint PCR (primerless PCR followed by a second-round PCR with external primers) generating the composite parts: BBa_K3215007, BBa_K3215009 and BBa_K3215010. Fragments were confirmed by gel electrophoresis and purified from gel.
We then used the purified DNA to transform E. coli K12 harboring a PKD46 plasmid, by arabinose-induced Lambda-Red Recombinase method. Transformation was plated on LB + chloramphenicol plates. Colonies were screened by Colony PCR (cPCR). Colonies that showed promising amplification had its DNA extracted by alkaline lysis and PCR was done to confirm them.
3. Construction of phnEG correcting cassette and transformation
Achievements:
1. Added partial flanking homologous sequences in the phnEG correcting cassette
2. Transformed mutants from the promoter transformation
E. coli K-12 have an endogenous phn operon with a frameshift at phnE, inactivating it, and phnF which is a repressor. In order to correct the frameshift in phnE and remove the phnF repressor, we designed a phnEG correcting cassette, as phnG is the cistron next to phnF. This cassette was codon optimized to allow synthesis by IDT. However, by doing that, we changed the homology of the cassette with the endogenous operon. To correct that, we designed 70 nt chimeric primers (20 nt annealing with the cassette, 50 nt to generate the homologous flanking sequences).
This PCR proved very difficult, as chimeric primer Tm were as high as 80-90 ℃. To overcome that, we tried many different strategies, including different primer combination, different annealing temperatures, different annealing times and Touchdown PCR. The only strategy that showed positive results was to use phnEG as a template for two individual PCRs. One of them used the chimeric forward primer together with a conventional reverse primer, while the other was the opposite: conventional forward primer together with reverse chimeric primer. This way we managed to amplify the phnEG Cassette and add partial flanking sequences, homologous to the endogenous E. coli K-12 phn operon. PCR products were size confirmed in gel electrophoresis and purified from gel.
The purified DNA from both reactions (i.e. phnEG_Forward and phnEG_Reverse) was mixed in 1:1 molar ratio and heated to 95 ℃ in order to separated DNA strands. We allowed it to cool at room temperature for reannealing of strands. The product of this reaction was used to transform E. coli K-12 mutants from the first transformation (i.e. promoter transformation) by the same lambda-red recombinase method. However, as we needed the phnEG cassette to insert in the middle of the phn operon, it was not possible to add a antibiotic resistance gene for selection. The transformation was plated in different volumes (i. e. 25, 50, 100, 150 $\mu L$) in LB + Chloramphenicol plates for selection. Even in the 25 $\mu L$ plate, there was too many colonies, as the antibiotic resistance was the same from before transformation. We then streak plated transformants from the 25 $\mu L$ plate in a M9 medium plate with $1 \over 5$ of phosphorus sources, addition of 50 ppm of glyphosate and chloramphenicol for selection. We screened 20 colonies of each of the three transformations (one for each promoter) without any positive results.
To overcome that, we streak plated transformants again, but now in a Tris-glucose medium, without any phosphorus source besides 50 ppm of glyphosate. Isolated colonies were screened again, around 20 colonies for each promoter, with only a few showing putative amplification. The phnEG_Cassette from synthesis was used as positive control.
4. Growth Curves
Growth curves are representative of each microorganism capability of metabolize different nutrients from different culture media and use it for their growth. For our final objective is growing our synthetic E. coli in fresh water contaminated with glyphosate, we evaluated the growth behavior of wild type E. coli in nutrient depleted media.
In this assay we performed the growth curves in two different growth media:
M9 broth: salt depleted media that is used to simulate marine environment. We tested it with three different phosphate concentrations (1x, 0.5x and 0.2x) to see the ability of E. coli to survive without phosphate. Also, we tested its growth using glyphosate as a phosphate source on 0.2x media.
Tris-glucose broth: it is a buffer with glucose supplementation. This broth challenges our E. coli in a nutrient-depleted environment. With it, our aim was to access the capability of our engineered bacteria to survive with solely glyphosate as a phosphorous source.
As we can observe in figures 7 and 8, the E. coli wt was able to survive in all conditions. However, after 24 hours, only the complete M9 media supported bacterial growth.
For the curves shown in figure 9, is possible to observe that all mutants and wt had diminished growth in TRIS-glucose (TG) media and TG supplemented with glyphosate. However, it is possible to notice that wt and four mutants were able to grow in TG with phosphorus addition, where the latter grew up to 250% more than the wt. This result showed that our mutants may have some improvement in phosphorus metabolism, but they are still not able to use glyphosate as sole phosphorus source. Besides, TG media without supplementation was shown not to be a good choice for our experimental design.
We also made a colony formation unity assay to access the viability of bacterial cells after 24 hours of broth exposure. We diluted the samples of M9 1x and M9 0.5x in order 1:20 and plated 50 microliters. At the other broth concentrations, we plated 50 microliters without the dilution step.
5. Glyphosate Quantification
Glyphosate measurements were performed in partnership with Federal Laboratory of Food and Agriculture (LFDA) through liquid chromatography-mass spectrometry with parameters previously established by LFDA. In order to standardize their method and test whether it is applicable to our experiments, we prepared samples with borderline amounts of glyphosate and tested if the high degradation rates of this compound could influence our analysis.
Briefly, we prepared a standard glyphosate (Sigma-Aldrich) solution at 10 ppb. This solution was used as control and to quantify the samples. Also, we made solutions containing M9 media with glyphosate at the concentrations allowed by Brazilian legislations in rivers (50 ppb and 500 ppb). This was used to test the the possibility of glyphosate detection in complexes samples by this method. Besides, we tested the glyphosate self degradation by heat exposing the samples to bacteria growth temperature, 37 ℃, overnight at 200 rpm.
The results shown in table 2 reinforce the sensibility of the method and its applicability. Besides, we can observe that glyphosate is not degraded under prolonged exposure to heat conditions and apart from human errors, the sample recovery was very similar to that added in the beginning.
This table put in evidence the human error applied to our samples, where different amounts of glyphosate were retrieved. This makes noticeable that our approach must be more analytical, making use of more accurate instruments in order to prevent duplicate errors.
6. Alginate Spheres
Alginate spheres were prepared in for different concentrations - 0.5, 1, 1.5 and 2% -, three different sizes - - and four different reticulation times.
Briefly, to produce these alginate spheres, alginate was dropped into calcium chloride 2% solution with p1000 micropipette, plastic pasteur pipettes and cutted plastic pasteur pipettes to form the different sizes. To test the reticulation, we let the alginate spheres in solution during 15 minutes, 30 minutes, 1 hour and 24 hours.
To test the ability of these spheres to absorb molecules, as glyphosate, we incubated them in solution with lab colouring Coomassie Blue overnight and the absorbance was measured.
As observed, all alginate concentrations were able to absorb comassie blue and the reticulation time did not influence the absorption. We chose the smallest size to the next experiments for its easy preparation, high rigidity and less deformation when touching the calcium chloride solution. This is corroborated by the results presented in figure 11, where the high deformability made the coomassie absorption inhomogeneous.
For electron microscopy analysis, we produced the alginate spheres in the same manner as described before. After 15 minutes of reticulation, we poured the alginate spheres in liquid nitrogen for 15 minutes and then lyophilized overnight. The samples were precipitated with silver ions and images were made in partnership with Microscopy and Microanalysis Center (CMM) of Federal University of Rio Grande do Sul.
We could observe pores in only one of the imaged samples. This could be due the sample preparation method that provoked shrinkage and collapse of our samples. A representative image of this event is shown in figure 12a. We were able to measure the pore size of the alginate spheres prepared with 1% of calcium alginate, and the medium size measured was 22.5 $\mu m$.
7. Biobrick construction
1. Existing Biobrick
To test an already existent biobrick, we combined the part BBa_K259007, which consists of an AraC Promoter fused with RBS (1), with a mRFP translation unit fused with a terminator (RBS + CDS + Terminator) (7). To do so, we digested the fragment 1 with EcoRI and SpeI and the fragment 7 with XbaI and PstI in an overnight cleavage reaction.
After this, we made a ligation reaction overnight, to unite these generating the part BBa_K3215011. This construction was inserted into pSB1K3, that was previously cleaved overnight with PstI and EcoRI, in another overnight ligation reaction. The colony PCR is shown below.
2. New parts
With the aim to construct an arabinose death control, we ligated a cI Repressor Cassette (AraC promoter fused with RBS + cI CDS + Terminator + cI Repressed promoter) (2) with Tse2 translation unit fused with a terminator (RBS + CDS + Terminator) (8). Briefly, we digested in an overnight reaction the former with SpeI and EcoRI and the latter with XbaI and PstI. Both were ligated in an overnight reaction with T4 ligase, generating the BBa_K3215013. This new part was also inserted into the pSB1K3 plasmid.
To validate this new part, we also developed another new part by fusing the cI Repressor Cassette with the mRFP translation unit fused with a terminator. This was also inserted into pSB1K3.
All plasmids were inserted into E. coli K-12 by quimioporation.
8. Biobrick characterization
1. Existing Biobrick
To add new data to this part, we made an arabinose curve according to data showed in our modelling. The concentrations used were of 0.01, 0.05, 0.1, 0.2 and 1% of arabinose. The mRFP expression was measured overtime using the imaging system SpectraMax i3, with excitation in 558 nm and excitation at 607 nm, and the measures were compared with WT E. coli K-12.
As observed in the graph, E. coli with BBa_K3215011 part was able to produce fluorescence signals stronger than the observed in the wt (background signal), although concentrations of arabinose did not show impact in mRFP expression, going in contrast with the predictions made by our in silico kinetic study of the promoter induction.
To show mRFP fluorescence, we plated the cells with arabinose and made images of microscopy with an inverted microscope of fluorescence Axiovert 200
As we can observe, only colonies with the BBa_K3215011 plated in arabinose media were able to express mRFP, evidencing that our constructions were correct.
New parts
To test the new created parts, bacteria of the strain E. coli k-12 were quimioporated and plated. The colonies formed were inoculated in two halfs: one in arabinose 0,5% supplemented LB media and one without arabinose. All media were supplemented with Kanamycin. After 8 hours, optical density and red fluorescence were measured.
Our goal was to create a biobrick that could slowly kill our engineered bacteria and also guaranteed that it is incapable of surviving at outside of the filtering system. However, as we can see by fig. 16 a, the construct that was supposedly to work by arabinose control, that in its presence permits the production of the inhibitor protein of TSE2 gene, actually was inducing the expression of the toxic protein, killing most of the colonies. The result for verification, showed in 16 B, also follows the same pattern. Investing these phenomena, we can tell that overexpression of inhibitory protein could induce dimers that acts as activators of TSE2 gene. However more results are necessary to prove this hypothesis and to guarantee the safety of our system.
9. Challenges
As mentioned, E. coli K12 have an endogenous phn Operon, with a frameshift at phnE, inactivating it. We decided to correct this operon, change the promoter to a constitutive one, and excise the repressor, in order to develop a glyphosate degrading bacteria. With the time we had, the locus dependency of our design proved to be a challenge harder than expected. The first transformation (promoter changing) had an antibiotic resistance gene in the construction, allowing for effective selection of positive colonies. However, as mentioned, we needed a loci-specific insertion (upstream of the phn operon) for our design to work. We did found positive colonies by growth in LB + chloramphenicol, and confirmation made by PCR (amplifying the resistance + promoter sequences). But any cell that had the construction inserted in its genome would have resistance to chloramphenicol, almost independent of locus. We had to order new primers to confirm locus insertion, and colony screening by colony PCR was not very effective.
For our second transformation (operon correcting and repressor excising) it was not possible to add any selection gene, as the insertion was in the middle of the operon. We first thought that using minimal media with glyphosate as sole phosphorus source would be enough to allow growth of positive colonies, and confirmation would be made only by PCR. But, as for the first transformation, we needed a second event of insertion in a specific loci, on top of not loci-confirmed mutants. The challenge to screen colonies and screen locus insertion at the same time took too long, and more time would be needed to effectively select positive colonies.
References
Yu JH, Hamari Z, Han KH, Seo JA, Reyes-Dominguez Y, et al. (2004) Double-joint PCR: a PCR-based molecular tool for gene manipulations in filamentous fungi. Fungal Genet Biol 41: 973-981
DATSENKO, Kirill A.; WANNER, Barry L (2000). One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products.
SILVA, Marcelo Dutra et al. Determinação De Glifosato E Ácido Aminometilfosfônico Em Águas Superficiais Do Arroio Passo Do Pilão. Pesticidas: Revista de Ecotoxicologia e Meio Ambiente, [s.l.], v. 13, p.19-28, 12/31 2003. Universidade Federal do Rio Grande.
