Team:BOKU-Vienna/Engineering
Design & Engineering
Our project was quite complex with many different facets and potential pitfalls involved. As a result, our progressions during the iGEM engineering cycle were numerous and of quite different nature.
Thus, we structured this section - as we structured all our lab work - into the topics “Cloning” and “Scaffold”.
Assembly of our Type-II-S backbone
Design
As we planned to assemble our constructs in E.coli and then switch to L.plantarum for the actual expression, we needed a fitting vector. The vector needed to work in both host organisms and match the Type-II-S assembly system. We started by using the plasmid pSIP403, which contained five sequences that would be identified as illegal restriction sites in the Type-II-S system, which we referred to as "illegal sites" in the following paragraphs.
Our approach included the decomposition of the plasmid pSIP403 into five pieces, because of the five illegal sites in the original plasmid. Hence, we designed 10 primers: one forward and one reverse primer for each of the five illegal sites. These primers would anneal a few basepairs (bp) downstream of the illegal site. The design of the primers includes a 5’ overhang upstream of the annealing site which does not match the template sequence. With this we excluded the plasmid section in which the illegal site resides. When running a PCR, the primers with the overlap will get incorporated in the newly synthesized DNA, resulting in a growing amount of DNA constructs devoid of the illegal sites. To obtain the respective five constructs, an Agarose gel purification of the PCR products was performed including further purification. For the re-assembly of the single constructs, a Golden Gate assembly approach was taken, yielding the full backbone 3 (BB3). The overhang of the primers was constructed to carry a BsaI restriction site as well as a fusion site, ensuring the correct order of constructs during the assembling.
This sophisticated primer design allowed us to use the pSIP403 to create our novel BB3. Additionally, the reporter gene GuS was removed and instead, we decided to use amilCP, a blue chromoprotein, as a reporter.
pSIP403 served as template for the constructs one to four. For construct five, we needed to use a vector carrying the amilCP gene, and we decided for the pET 30a amilCP plasmid.
Build and test
In the first run, a classical PCR was performed, followed by a preparative gel. The gel yielded good results for construct one, construct two and construct four. After gel purification and DNA sequencing, the identity of the respective constructs was confirmed. Construct three and construct five did not work under the originally defined conditions.
Learn and re-design
We decided to test if we had more success by slightly deviating the primer annealing temperature. We faced the problem of having very faint bands on the stained Agarose gels and multiple bands not matching the desired size. We troubleshot our problem and decided to use a touchdown PCR for our next approach since this kind of PCR regimen can reduce the number of unspecific PCR products. During the annealing step of each cycle, the temperature was decreased by 1°C per cycle. We started with a higher temperature (72 °C) than the calculated melting temperature (Tm) and gradually decreased the temperature.
Build and test
The touchdown PCR yielded a positive result for construct five, but the amount of PCR product was not sufficient for further processing. Construct three still refused to be synthesized. We tried further PCR runs with small modifications (different elongation temperatures and different amounts of template DNA) but still, no positive result could be achieved.
Learn and re-design
Because of the unique overhang-containing structure of our primers, we came up with the idea to split the PCR program into two different stages: the first 10 cycles were done using a lower annealing temperature well adapted for the Tm of the primers allowing for more specific binding since the overhang of the designed primer will not anneal when the original template is used. After five to 10 cycles one can increase the annealing temperature since the newly synthesized DNA will act as a new template, enabling the primers to anneal fully. The remaining 20-35 cycles are performed with the higher annealing temperature matching the new Tm.
Build and test
With this method we were able to increase the signal and yield a sufficient amount of construct five.
Learn and re-design
Nevertheless, construct three still failed. This time we decided to try a two-step PCR. This means the annealing temperature is set at the same value as the elongation temperature of 72 °C resulting in two steps - denaturation, and elongation. The idea of this approach is to get the maximum specificity. For this to work, we had to design new primers with a Tm matching the elongation temperature. We achieved this by increasing the length of the primers downstream.
Build and test
This method finally yielded a signal for construct three, enabling us to finally assemble the raw BB3.
The primers for construct five had another addition. Apart from the BsaI restriction site, they also included a BbsI restriction site. This enabled us to insert different linkers, including fusion sites, into the raw BB3 making it suitable for the insertion of up to four expression cassettes in one single BB3. (BBa_K3855401)
Backbone 2 expression cassette
Design
We designed the respective sequences for inserts, promoters, and terminators then codon-optimized them for the expression in Lactobacillus. In combination with the process of codon optimization, we cured our sequences from illegal sites. After finishing the design, we ordered the g-blocks from IDT (Integrated DNA Technologies) and cloned them into backbone 1 (BB1).
Build and test
This first step of our cloning process did not pose a challenge and the building blocks for our expression cassette were ready. But when we tried to assemble them into backbone 2 (BB2), we failed to get positive results. There was hardly any growth observable and the colonies picked showed no insert when sequenced.
Learn and re-design
Since promotor, the gene of interest (our enzymes), and terminator are assembled into a single plasmid during this step of the cloning process, the result was a functional expression cassette. As a consequence, the enzymes are expressed and translated into the catalytically active enzymes, causing various effects in the expression host including a high metabolic burden. We assumed that our promotors are too strong and thus pose a very high metabolic burden on the transformed clones. Therefore, we stepped back and used weaker promotors.
Build and test
The weaker promotors chosen are P4 and P5 and have been reported to be up to 10 times weaker than P1, P2, and P3. We repeated the assembly as well as the transformation. As a result, we were rewarded with several clones which were processed, purified, and sent to sequencing.
Learn and re-design
All clones turned out to contain a frameshift mutation. One example is shown below. The left-hand side shows the sequences of our three clones aligned with the expected sequence. A frameshift is seen at the first base pair of the enzyme sequence. On the upper right, the original sequence and the according translated protein sequences are shown, with the clone version containing a one-bp frameshift mutation in the coding sequence below. The mutation significantly alters the protein sequence, as it causes a shift in the open reading frame. During the first engineering cycle, the few clones we obtained, also showed mutations rendering the enzyme unfunctional. This indicated a reason, other than the metabolic burden, causing the complications.
Even though we found no hint towards the origin of this problem in the literature, we had to assume that our enzymes are toxic to E.coli in some way, as only clones with a functioning mutation survived.
Since the translated enzymes contained a secretion signal for Lactobacillus, we expected fewer complications as soon as we would reach Lactobacillus as an expression host. Therefore, we designed an approach that would allow us to circumvent cloning the BB2 in E.coli.
Build and test
Cloning the plasmid into E.coli after the assembly has two main purposes: Selection of functioning plasmids and multiplication of the plasmid DNA. In order to avoid E.coli during the cloning process, we designed a way to perform these steps artificially.
We did so by running a PCR to replicate the plasmid DNA. Then, we isolated the right plasmids via a quantitative gel. Further, the constructs were purified and the DNA content was quantified.
Learn and re-design
Apparently, this approach did not yield sufficient amounts of plasmid DNA to perform a successful transformation in Lactobacillus since the results of the transformation were all negative. We compared the documented amount of plasmid DNA needed for a successful transformation with our yield and were off by more than one power of ten.
Thus, we went back to the drawing board and decided to use an inducible promoter.
Build and test
Using the inducible promotor ORF (BBa_K3855006) yielded positive clones in E.coli with no mutation. Thus, we were able to progress further with our project. The assembly of the BB3 constructs containing the ORF promoter succeeded in Lactobacillus at the first try for almost all constructs. Finally, we achieved our goal of cloning the fructan-degrading enzyme cassette into Lactobacillus.
Synthesis of pepNTT
Design
One should be prepared that IDT will not be able to synthesize certain DNA sequences with a too high complexity value. This was the case for our second terminator: pepNTT. We came up with two different approaches to synthesize the desired sequence. The vector pSIP403 carries the pepNTT terminator. Therefore, we designed two primers to amplify pepNTT using pSIP403 as a template sequence. Since we could not guarantee the success of this approach, our second idea was to assemble primers into the full pepNTT sequence. In our case, the sequence was divided into five pieces of approximately 55 bp in size. For the 3’ -> 5’ direction, the five pieces were shifted approximately 20 bp to create large enough overlapping sequences to ensure a reasonable Tm for the annealing of the single pieces to each other. The terminal area of the construct not covered by the overlapping sequence is covered by a smaller primer. The graphic shown illustrates the arrangement of the primers forming the pepNTT terminator.
Build and test
To get the full sequence, we tested different assembly approaches. The assembly is performed in a classical thermocycler and the following program was the most effective:
98 °C for 20 seconds followed by a decrease of 1 °C per 15 seconds until 59 °C. Then follows a quick cooldown to 8 °C for two minutes.
Subsequently, a preparative gel including the respective purification steps was performed to obtain the purified construct. Both approaches yielded not only correct pepNTT terminator constructs but also a sufficient DNA concentration.
Cellulose sulfate
Design
For the scaffold, which in our case is a polyelectrolyte complex, we needed two strongly opposingly charged polymers. The positively charged polymer is polyDADMAC which is commercially available. The negatively charged component is water-soluble cellulose sulfate which is not commercially available and thus needed to be synthesized.
Build
The synthesis of the cellulose sulfate was carried out under the supervision of Dr. Hubert Hetteger. With his help, we sulfated microcrystalline cellulose wit the modification of a lower amount of SO3-Pyridin (1 mol equivalent) according to his method (See Protocol SSP in PolyDADMAC).
Test
After the successful synthesis of cellulose sulfate, we tried to form a polyelectrolyte complex capsule by dropping the cellulose sulfate solution into the polyDADMAC solution using a syringe. Unfortunately, the capsule did not form since the cellulose sulfate solution started to spread in the solution and precipitated.
Learn
We discovered that the cellulose sulfate synthesized from the microcrystalline cellulose was of too low molecular weight to form a capsule. Thus, we needed cellulose of much higher molecular weight.
Design
The candidates were spruce sulfite pulp (short: SSP) and a cotton linter, which have a higher molecular weight than the microcrystalline cellulose.
Build
The same sulfation method as described above was used for the synthesis of the respective sulfared, negatively charged polyelectrolyte. Only the raw material, the unmodified cellulose, has been changed.
Test
We discovered that the cellulose sulfate we tried to synthesize using different raw materials was not water-soluble. There were also a few undissolved cellulose fibers in the otherwise clear solution.
Learn
The problem was that the synthesized cellulose was neither completely dissolved in the DMA (N,N-Dimethylacetamid) solution nor homogeneously distributed. Thus, we added more LiCl, since the molecular weight of the cellulose derived from raw materials than microcrystalline cellulose is higher. A further problem we encountered, were the hygroscopic properties of SO3-pyridine and LiCl. Since the sulfating reaction needs to take place in the total absence of water, we needed to work with dry substances. The water would react with SO3 and form sulfuric acid.
Design
We now knew that we had to work water-free as much as possible and tried to completely dissolve the cellulose fibers.
Build
To keep our substances dry, we used drying beads and a vacuum dryer. Fresh chemicals were closed with parafilm to protect them from humidity. We also dissolved the celluloses in an already prepared DMA/LiCl solution, the rest of the method remained unchanged.
Test
There were no longer any undissolved fibers, but the synthesized cellulose sulfate was still not water-soluble when we tried to solve it in distilled water and diluted NaOH.(4)
Learn
Water-soluble cellulose is achieved when the substitution degree with sulfate is high enough and homogeneous.
Design
To achieve a higher degree of substitution with sulfate, higher concentration of SO3-Pyridin was used.
Build
The amount of SO3-Pyridin was increased. Instead of 1mol equivalent, three to eight equivalents were used for further experiments.
Test
The cellulose sulfate made from spruce sulfite pulp and cotton linters using more SO3-Pyridin and working as water-free as possible was water-soluble. We dissolved the cellulose sulfate and dripped it in the polyDADMAC solution. Capsules were formed, and the cellulose sulfate synthesized from spruce sulfite pulp and cotton linters showed similar results in the FTIR spectrum (see Notebook in the Scaffolds section). But since some material loss of the cotton linters at the filtration step happened, we decided to work mostly with the spruce sulfite pulp.
The capsule stability was tested by first placing the empty capsules in distilled water overnight. At the next day, it was still intact and the bottle containing the capsule was put on a slow-moving shaker for five hours (see Notebook in the Scaffolds section). The formed capsules were left in distilled water over several days and the stable ones were also shaken slowly for a few hours on a shaker.
Learn
Water-soluble anionic cellulose sulfate which formed scaffolds when mixed carefully with cationic polyDADMAC was produced. But there was not enough time left to fine-tune solution concentrations, thereby also adjusting the pore size, and other modifications to enhance the properties.
Thiolation and Mucoadhesivity
Design
One of the main functions of our scaffold - besides retaining the cells - is its mucoadhesivity. In order to achieve it, we decided to use a method called thiolation - meaning the addition of thiol groups. These can then form disulfide bridges with cysteine residues of proteins lining the gut mucosa and thus, would allow our scaffold to adhere to the mucosal surface.
We initially planned to confer mucoadhesivity via direct thiolation of cellulose sulfate. Since cellulose sulfate is not commercially available, we wanted to carry out a synthesis of our cellulose sulfate using microcrystalline cellulose as raw material and then in a second step, thiolation would be achieved using thiourea and a microwave (see protocol section).
Build
The thiolation was carried out in the lab of Prof. Andreas Bernkop-Schnürch at the University of Innsbruck. He was kind enough to share his methods and expertise on thiolated polymers with us. Thiolation of our own cellulose sulfate derived from microcrystalline cellulose was done according to a published method (1) using a microwave, but we implemented changes to the original protocol.(2)
Test and learn
We tested the success of the thiolation and, thereby mucoadhesion, performing two tests called "Ellman’s test" and "disulfide bridge test" with 5,5-di-thio-bis-(nitro-benzoic acid).(3) Unfortunately, we saw that something was wrong with our cellulose sulfate. After the thiolation treatment, it had become insoluble in water.
We learned that microwave irradiation led to the desulfation of cellulose sulfate. This was very unfortunate, since the negative charges of the sulfate are needed to form the polyelectrolyte complex that holds our bio container together.
Re-design
So instead of thiolating cellulose sulfate directly, we had to come up with an alternative solution. We needed a material that would not disturb the polyelectrolyte complex formation while being biocompatible and susceptible to our thiolation method.
We chose to add another polysaccharide, chitosan, which we wanted to thiolate prior to adding it. Chitosan is the second most prevalent naturally occurring polysaccharide after cellulose and it is present in the shells of crustaceans, insect exoskeletons, and in some fungi. Our idea was to add the thiolated chitosan directly into the polyDADMAC (polydiallyldimethylammoniumchlorid) solution that could then be used to encapsulate the cells suspended in a cellulose sulfate solution. Thanks to its positive charge, chitosan should be well compatible with the cationic polyDADMAC solution and not interfere with its ability to form the polyelectrolyte complex. Also, with this approach, the chitosan should be located on the outside of the capsule and thus, be able to easily interact with the surface of the gut mucosa and form the desired disulfide bridges.
Build and test
We thiolated chitosan in the laboratory of Prof. Bernkop-Schnürch using the same microwaving method as described above for the initial experiement and confirmed the successful thiolation using the Ellman’s test and disulfide bridge tests.
Back home in Vienna, we started our first encapsulation assays. In the beginning, we focused on assessing capsule stability using empty capsules without cells. There we saw that despite the capsules being slightly opaquer than without chitosan addition, they showed good stability over several days.
Next, we went on to do experiments using living E.coli cells and tested whether they could escape the bio container. There we saw that capsule formation was successful but over time more cells inside the capsule died than in capsules only composed of cellulose sulfate and polyDADMAC.
Learn
We figured out that this might be due to the fact, that chitosan has certain anti-bacterial properties. This pointed into the direction that it still influenced the cells inside the container, although most of the chitosan must be located on the outside of the capsule since we need the sulfide of the thiolated chitosan outside to form disulfide bridges with the cysteine of the mucosa.
After this, we unfortunately ran out of time and were not able to conduct further experiments.
Outlook
To optimize the capsule properties and allow better cell survival, further experiments using different concentrations of thiolated chitosan, polyDADMAC, and cell densities for encapsulation could be done in order to fine-tune the conditions and achieve optimal results.
References
(1) Grosso, R., & de-Paz, M. V. (2021). Thiolated-Polymer-Based Nanoparticles as an Avant-Garde Approach for Anticancer Therapies-Reviewing Thiomers from Chitosan and Hyaluronic Acid. Pharmaceutics, 13(6), 854. https://doi.org/10.3390/pharmaceutics13060854
(2) Hussain Asim, M., Nazir, I., Jalil, A., Matuszczak, B., & Bernkop-Schnürch, A. (2020). Tetradeca-thiolated cyclodextrins: Highly mucoadhesive and in-situ gelling oligomers with prolonged mucosal adhesion. International journal of pharmaceutics, 577, 119040. https://doi.org/10.1016/j.ijpharm.2020.119040
(3) Hussain Asim, M., Nazir, I., Jalil, A., Matuszczak, B., & Bernkop-Schnürch, A. (2020). Tetradeca-thiolated cyclodextrins: Highly mucoadhesive and in-situ gelling oligomers with prolonged mucosal adhesion. International journal of pharmaceutics, 577, 119040. https://doi.org/10.1016/j.ijpharm.2020.119040
(4) Muhitdinov, B., Heinze, T., Turaev, A., Koschella, A., & Normakhamatov, N., (2019). Homogenous synthesis of sodium cellulose sulfates with regulable low and high degree of substitutions with SO3/Py in N,N-dimethylacetamide/LiCl. European Polymer Journal, 119, 181-188. https://doi.org/10.1016/j.eurpolymj.2019.07.030