Engineering

Engineering cycle

Synthetic biology is a branch of engineering, as it applies the same principles to develop biological systems that perform specific functions. Much like the process an engineer uses to design a product, synthetic biology follows a series of iterative steps known as the Engineering Cycle. This cycle repeats until the system meets the desired specifications. In synthetic biology, the cycle continues until a biological system capable of achieving the desired function is identified.

The Engineering Cycle consists of four key phases: design, build, test, and learn. In the design phase, the target biological system and its intended function are specified. Literature research and modelling tools are especially useful here, as they can accelerate the design process by learning from previous results and simulating different constructs, saving time and resources in the lab. During the build phase, these designs are implemented in a target organism, or “chassis”, which is typically a strain of bacteria or yeast. Next is the test phase, where the system is evaluated to ensure it performs the desired function. Finally, the learn phase involves analysing the results: how much does the performance deviate from the expected outcomes, and what can be done next time to improve. Insights gained during this phase are then used to refine the initial design, and the cycle begins again.

Hemicellulose

The goal of our iGEM project of 2024, named KlothY, is to produce a unique, customizable and environmental-friendly kind of textile. This textile is going to be based on bacterial cellulose. But with the usage of bacterial cellulose, there will be a few challenges to face, due to its different properties in comparison to established textiles or cloths. Bacterial cellulose alone simply does not have the characteristics we imagine, when thinking about cloth in general. When creating mats of bacterial cellulose, it is easy to spot that they are very stiff and brittle [1]. Therefore, they cannot function as a comparable textile substitute. So, the logical consequence, to tackle these insufficient properties of a pure bacterial cellulose mat, is to increase their elasticity, flexibility and within also their endurance. In short, that is the main task of the hemicellulose subgroup and the problem our engineering cycle is trying to find a solution for. To achieve these property changes we want to increase the amount of hemicellulose, which is incorporated in our bacterial cellulose mats. The increased amount of hemicellulose within the mats will give the cloth more of the needed elasticity and flexibility [2, 3, 4]. So, in the end, the produced cellulose mats should, at least property-wise, get a lot closer to already established textiles or cloths.

Our Hemicellulose cycle:

Design

In order to achieve the goal of increasing the amount of expressed hemicellulose, we will boost the produced mass of xyloglucan, one kind of hemicellulose. Our whole plan consists of multiple steps. At first, we want to produce the xyloglucan. Therefore, we imagined a 5-gene strain, which is responsible for producing xyloglucan inside of the golgi-apparatus.

The named 5 gene-strain includes the following genes:

AtCSLC4 cellulose synthase like fam. C -Golgi membrane; xyloglucan glycosyltransferase; constructs glucose backbone

AtXXT3 xyloglucan 6-xylotransferase -Golgi membrane; xylose transferase; directs inside the membrane and associates with AtCSCL4

AtUGDH1 UDP-glucose 6-dehydrogenase -Cytosolic UDP-glucose to UDP-glucuronic acid

AtUXS3 UDP-glucuronic acid decarboxylase 3 -Cytosolic UDP-glucuronic acid to UDP-xylose

AtUXT1 UDP-xylose transporter 1 -Golgi membrane; UDP-xylose transporter

A model of the xyloglucan synthesis:

(Made with BioRender®)

The product will be built inside of the golgi. Its structure consists of a glucose backbone with the added xylose, forming xyloglucan. The assembled xyloglucan strings will intercalate with the bacterial cellulose mats, allowing them to change their properties.

(Made with BioRender®)

That part of our project would not have been coming to life without the work of Ronja Immelmann, Alexander Schultink and Balakumaran Chandrasekar. We received a plasmid from them containing the 5 genes needed for the xylose production inside of the golgi (CD108). Their work as well as their willingness to share a part of it with us, made it possible to work on solutions for our project.

The inducibility of the xyloglucan production represents the second step. To solve this problem, we want to test multiple promotors, because we need to be able to induce the xyloglucan production before growing bacterial cellulose mats.

In total there are three different variants:

1. Sc-pCYC1, a constitutive promoter for all 5 genes

2. pREV1-Z3EV, an orthologous inducible promoter

3. Sc-pGal1, a galactose inducible promoter for all 5 genes

About the orthologous inducible promoter (pREV1-Z3EV):

The inducible promoter pREV1-Z3EV allows us to induce our genes independently from our colouring-system via ß-estradiol. The Z3EV transcription factor is encoded by a gene coupled to a constitutive promotor.

The last step is going to be adding a leucine auxotrophy. Including this property to our plasmids will make us able to tell, which of our cells incorporated the plasmid. After finishing the specific plasmids of all needed constructs, the next goal is the expression of the complete xyloglucan pathway in Saccharomyces cerevisiae, an eucaryotic model organism. Reaching that goal, also consists of a few sub-steps.

Beginning with an assembly of a modular cloning set (MoClo). A MoClo is an efficient way to assemble many DNA parts into one functional plasmid. Since we planned 3 final plasmids with the 5 genes, promotors, homology regions, markers and a few more components, a MoClo portrays the best choice. These plasmids will be transformed and amplified in Escherichia coli.

Subsequently we want to transform S. cerevisiae and select specific colonies. That selection is going to be based on the leucine auxotrophy we included earlier. At last, and if everything else worked as desired, we want to measure and localize xyloglucan. By doing so, we will be able to make statements about the produced quantity of the xyloglucan and following also about the quality of our resulting mats.

Build

Our work for the subgroup started with the creation of all needed plasmids for our 5-gene-strain, the inducibility, as well as all of the fitting promoters, terminators and homology regions. The first plasmids were re-transformed from the iGEM distribution kits, containing multiple of our needed parts, and prepared for use at a later time. The general retransformation consists of the transformation, setting up liquid cultures, plating them and finally preparing the plasmids. A few of our needed parts, for example the 5 gene strain, were obtained from the institute of our PI Markus Pauly. Also, these plasmids were prepared and stored.

Test

Our work made it possible to test a few things. We were able to check and rate the quality, or more precisely the DNA concentrations, of our plasmid retransformations via the spectrometer. Our stock of L0 plasmids contained:

From the distribution kits.

The noted concentration of CD023 was measured before the fix.

The plasmid of the 5 gene strain (CD108) was prepared and shows the following concentrations:

While preparing these plasmids we got interrupted by one plasmid. CD023, which contained the OYC backbone. It was sequenced and did not fit the expectations. Multiple PCRs show the progress closer to our desired results.

The first gel electrophoresis:

(Its PCR-annealing happened at 60°C)

The second gel electrophoresis:

(Its PCR-annealing happened at 65°C and 70°C)

The third gel electrophoresis:

(Its PCR-annealing happened at 65°C and 72°C)

All of these gels refer to CD023a.

After fixing CD023, the samples were digested using SapI restriction enzymes and their concentrations were measured.

Through the digest of CD023 we obtained a few copies of CD024, which were sequenced and showed the expected sequences. To finalize CD024, we repeated its creation. Due to this repetition, we were sure, that CD024, as well as its basis CD023, would not contain any mistakes. The mentioned repetition also included a PCR, a gel electrophoresis and a DpnI digest.

The results of the gel electrophoresis:

It showed the desired results and its replication was stopped. To finish CD024, it was transformed and stored.

Learn

We stumbled over a few problems we were able to learn from. Primarily it is often useful to double check important parts of the construct. By doing so the occurrence of mistakes, especially in late phases of experiments or studies, can be minimized. Secondly, we learned, that it can be useful to test multiple PCR-annealing temperatures. Even little variance can lead to different results in quality

Sources: [1] Chanliaud E., Gidley M.J., (2002) In vitro synthesis and properties of pectin/Acetobacter xylinus cellulose composites https://doi.org/10.1046/j.1365-313X.1999.00571.x [2] Chanliaud E., et al., (2004) Mechanical effects of plant cell wall enzymes on cellulose/xyloglucan composites https://doi.org/10.1111/j.1365-313X.2004.02018.x [3] Whitney S.E.C., et al. (1999) Roles of Cellulose and Xyloglucan in Determining the Mechanical Properties of Primary Plant Cell Walls https://doi.org/10.1104/pp.121.2.657 [4] Chanliaud E., et al., (2002) Mechanical properties of primary plant cell wall analogues https://doi.org/10.1007/s00425-002-0783-8 Note: All pictures and tables above are made by team members and are saved in our elab-protocolls.

Inducible BC

To achieve our goal to have controlled activation of the bacterial cellulose synthesis apparatus in Komagataeibacter xylinus. We went through multiple design cycles to achieve a genetically engineered strain where the bacterial cellulose production can only be activated by the presence of arabinose sugar. Furthermore, by knocking out specific parts of the bacterial cellulose gene complex we can potentially investigate the metabolic cost by measuring the growth rate of K. xylinus and comparing it to the wildtype.

Knockout Cycle 1

Design

We initially researched the enzymatic pathway for bacterial cellulose in K. xylinus. In literature, we found that the responsible genes bcsA, bcsB, bcsC and bcsD are expressing for the main enzymes involved in the synthesis of bacterial cellulose. Sequentially other genes like bcIX, bcsZ and bcsH express for enzymes that are also essential to ensure stable bacterial cellulose production as those are part of the regulatory process and the secretory machine in bacterial in K. xylinus(McNarma JT, et al.,2015; Römling U., et al.,2015).

Figure 1: Enzymatic pathway map highlighting the enzymes responsible in the bacterial cellulose synthesis, secretion and regulation apparatus. Graphic originally from McNarma JT, et al., 2015. A molecular description of cellulose biosynthesis. Annual review of biochemistry, 84, 895–921. https://doi.org/10.1146/annurev-biochem-060614-033930.

As it is currently not known by how many operons regulate all genes, but all genes are located next to each other, we decided to focus on the knockout of specific genes and specific gene clusters that based on the literature, are regulated by the same operon in other bacterial cellulose producing strains: 1. bcIX; 2. bcsA, bcsB, bcsC, bcsD; 3. bcsZ, bcsH; and 4. the complete knockout of all afformentioned genes.

Figure 2: K. xylinus bacterial cellulose synthase and regulatory genome region.

Figure 3: in-silico costruct CD013 pSB1C30 ΔbcIX amp.

Figure 4: in-silico costruct CD014 pSB1C30 ΔbcsA-D amp.

Figure 5: in-silico costruct CD015 pSB1C30 ΔbcsZH amp.

Figure 6: in-silico costruct CD031 pSB1C30 ΔbcsABCDHZ ΔbcIX amp.

To achieve the planned knockouts in K. xylinus, we designed our constructs (Figure 3-6) with homology regions surrounding the targeted gene clusters. Furthermore we specifically went with the pSB1C30 backbone, because the ori sequence is compatible with K. xylinus. Thirdly, we also added an ampicillin ressistance cassete in addition to the chloramphenicol resistance casette already present in the backbone to increase the likely hood that selected colonies include the planned constructs. Finally, to assemble our constructs we decided to use a Golden Gate assembly aproach with PaqcI and therefore designed the primers used for amplification of all necessary fragments to have PaqcI restriction site overhangs.

Build

We assembled our constructs by firstly starting a genome DNA extraction in K. xylinus to have a DNA template to amplify all the necessary homology regions. Sequentially, we also amplified a ampicilin ressistance cassete and the pSB1C30 vector backbone that already has a chloramphenicol resistance cassette present. So that in the two antibiotics could be used for selection in E. coli DH5α K. xylinus.

Figure 7: Screenshot of Overview table used to document PCR amplification for all constructs planned for transformation in K. xylinus.

After PCR amplification of every fragment was complete, a Golden gate assembly with PaqcI and T4 ligase was done to assemble all our constructs. After completion, our presumed assembled constructs were then transformed in E. coli DH5α based on our DH5α transformation protocol and then plated on LB agar plates with both chloramphenicol and ampicillin for selection.

Figure 8: E. coli DH5α transformed with CD015 pSB1C30 ΔbcsZH amp. Overnight, visible colonies formed on LB agar plates with both ampicillin and chloramphenicol.

Unfortunately, some of the plates appeared to have dried out in the incubator after one day, they were placed in. Still, out of four planned knockout constructs, two of them were successful in growing colonies. Colony PCR with primers binding on the vector backbone and the ampicillin resistance cassete, showed visible bands in colonies transformed with CD015 pSB1C30 ΔbcsZH amp and colonies transformed with CD031 pSB1C30 ΔbcsABCDHZ ΔbcIX amp.

Figure 9: Gel picture of colony PCR results showing visible bands at predicted length for colonies transformed with assembled construct CD015 pSB1C30 ΔbcsZH amp(middle) and colonies transformed with assembled construct CD031 pSB1C30 ΔbcsABCDHZ ΔbcIX amp (right).

Before further transformation in K. xylinus, we wanted to verify the plasmid sequences. Therefore, we did whole plasmid sequencing through next generation nanopore sequencing by Microsynth.

Sequence results showed that only one of the knockout candidates CD015 pSB13C0 ΔbcsH ΔbcsZ, was successfully assembled. We are currently unsure why exactly. It was slightly unfortunate as both inducible constructs that were sequenced as well share the same affected homology region affected but due to time constraints a repetition of the transformation was not an option and we continued with CD015 pSB1C30 ΔbcsZH amp.

We transformed K. xylinus with CD015 pSB1C30 ΔbcsZH amp through electroporation. After 5 days, visible colonies formed on YPD agar plates with ampicillin and chloramphenicol.

Figure 10: K. xylinus CD015 pSB1C30 ΔbcsZH amp colonies on YPD agar plate with 0.2% cellulase as well as ampicillin and chloramphenicol for selection. After 5 days of cultivation visible colonies were marked on plate to use for cPCR to verify success of transformation.

Colonies for cPCR were picked and resulting PCR and gel electrophoresis indicate that knockout of the gene BcsH and BcsZ through homologous recombination was successfully integrated into the genome of K. xylinus

Figure 11: DNA electrophoresis gel pic. Layout of gel pic from left to right: ladder,2-7 replacing native constitutive promoter for the bcsABCD with paraBAD /inducible arabinose promoter and araC and araE genes through homologous recombination, 8-13 knockout of bcszH region through homologous recombination, 14-15 replacing native constitutive promoter for the bcsABCD with paraBAD /inducible arabinose promoter and araC genes through homologous recombination.

Test

After positive control we really wanted to immediately characterise our transformed strains. However, due to time constraints we were not able to achieve an in depth characterisation. However, we still managed to do an initial comparative inoculation test by preparing SOC media with 2% glucose and 1% arabinose as well as SOC media with just 2% glucose. We then added colonies from knockout strain CD015, inducible strain CD027 and WT for comparison.

Figure 12: K. xylinus CD015 colonies were used to inoculate a 6-wellplate based on layout (left). A: shows the plate after initial inoculation. B: shows the well-plate after 3 days of standing cultivation in room temperature. C: shows the well plate after 9 days of standing cultivation in room temperature.

Although the test is qualitative in nature, it was mainly to assess if bacterial cellulose pellicle formation in the knockout is visually worse than in the wild type strain. After 3 and 9 days no bacterial cellulose pellilce was seen in the knockout strain, indicating that loss of function was succesfully engineered in K. xylinus CD015 pSB1C30 ΔbcsZH amp strain.

Learn

We were able to achieve a viable transformation and through initial inoculation testing we were able to show that the strain does indeed not produce a bacterial cellulose pellicle even after 9 days of cultivation. Furthermore, through growth rate experiments we were able to show that the growth rate of the knockout strain is higher. This matches the predictions made with our metabolic model, where bacterial cellulose production slows down growth in K. xylinus. Further experiments for future iGEM teams continuing our work may be to assemble and transform the remaining constructs targeting different or all regions in the bacterial cellulose synthase complex in K. xylinus to see if they prove viable for a complete knockout of the bacterial cellulose synthesis function or if potentially a knock down effect can be perceived. Finally, it would also be of interest to compare the growth rate in all of the growth rate in all of the knockout strains to better understand the metabolic cost of each gene in K. xylinus

Inducible Cycle 2

Design

We initially researched the enzymatic pathway for bacterial cellulose in K. xylinus. In literature, we found that the responsible genes bcsA, bcsB, bcsC and bcsD are expressing the main enzymes involved in the synthesis of bacterial cellulose. Sequentially, other genes like bcIX, bcsZ and bcsH express enzymes that are also essential to ensure stable bacterial cellulose production as those are part of the regulatory process and the secretory machine in bacterial in K. xylinus (McNarma JT, et al., 2015; Römling U., et al., 2015).

Figure 1: Enzymatic pathway map highlighting the enzymes responsible in the bacterial cellulose synthesis, secretion and regulation apparatus. Graphic originally from McNarma JT, et al., 2015. A molecular description of cellulose biosynthesis. Annual review of biochemistry, 84, 895–921. https://doi.org/10.1146/annurev-biochem-060614-033930.

As it is currently not known how many operons regulate all genes, but all genes are located next to each other, we decided to focus on replacing the constitutive promoter for specific genes and specific gene clusters that, based on the literature, are regulated by the same operon in other bacterial cellulose producing strains: 1. bcIX; 2. bcsA, bcsB, bcsC, bcsD; 3. bcsZ, bcsH.

Figure 2: K. xylinus bacterial cellulose synthase and regulatory genome region.

Figure 3: In-silico costruct CD016 pSB1C30 araC paraBAD bcIX amp

Figure 4 : In-silico costruct CD017 pSB1C30 araC paraBAD BcsA-D amp

Figure 5: In-silico costruct CD018 pSB1C30 araC paraBAD bcsZH amp

Figure 6: In-silico costruct CD026 pSB1C30 araC am araE paraBAD bcIX amp

Figure 7: In-silico costruct CD027 pSB1C30 araC am araE paraBAD BcsA-D amp

Figure 8: In-silico costruct CD028 pSB1C30 araC am araE paraBAD bcsZH amp

To achieve the planned exchange of the native constitutive promoters in K. xylinus with an inducible promoter, we designed our constructs (Figure 3-8) with homology regions surrounding the targeted gene clusters. Furthermore we specifically went with the pSB1C30 backbone, because the ori sequence is compatible with K. xylinus. Thirdly, we also added an ampicillin resistance cassette in addition to the chloramphenicol resistance cassette already present in the backbone to increase the likelyhood that selected colonies include the planned constructs. Finally, to assemble our constructs we decided to use a Golden Gate assembly approach with PaqcI and therefore designed the primers used for amplification of all necessary fragments to have PaqcI restriction site overhangs.

Build

We assembled our constructs by firstly by starting a genome DNA extraction in K. xylinus to have a DNA template to amplify the necessary homology regions for the constructs with PaqcI overhangs to later assemble.In addition we amplified the Level 0 backbone pSB1C30 with a Chloramphenicol resistance cassette and PaqcI to act as our backbone. And an ampicillin resistance cassette to select for later in final transformation in K. xylinus.

Figure 9: Screenshot of table used for overview of necessary PCRs for both knockout and inducible BC

After the amplification of every fragment was complete, a Golden gate assembly with PaqcI and T4 ligase was done to assemble our constructs. After completion, our constructs were transformed in E. coli DH5α based on our DH5α transformation Protocol and then plated on LB agar plates with both chloramphenicol and ampicillin for selection.

Figure 10: E. coli DH5α colonies transformed with CD027 pSB1C30 arac am araE paraBAD bcsA-D amp, plated on LB agar plates with both ampicilin and chloramphenicol.

Unfortunately some of the plates appeared to have dried out in the incubator they were placed in. Still out of 6 planned inducible constructs, 3 of them were successful in growing colonies and initial colony PCR result showed that part of plasmid full plasmid was inside them

Figure 11: DNA electrophoresis gel picture of cPCR results showing visible bands at predicted length for assembled constructs CD017, CD018 and CD027. Fragment corresponding to CD028 was smaller than predicted and therefore not taken into further consideration

To verify the plasmid sequence, we did full plasmid sequencing through next generation nanopore sequencing by Microsynth.

Sequence results showed that two of the inducible BC constructs showed almost identical sequences for the planned plasmids CD017 and CD027. However, due to time constraints, a repetition of the transformation was not a feasible option and we continued with the verified plasmid constructs.

After verification through sequencing, we then transformed the previously prepared electrocompetent K. xylinus dsm 2325 strains with the plasmid CD017 and CD027. After 5 days visible colonies formed on plates

Colonies for cPCR results were then picked and resulting gel indicate that knockout through homologous recombination was probably successfully integrated into the genome of K. xylinus

Figure 12: DNA electrophoresis gel picture Layout of gel pic from left to right: ladder, 2-7 replacing native constitutive promoter for the bcsABCD with paraBAD inducible arabinose promoter and araC and araE genes through homologous recombination, 8-13 knockout of bcszH region through homologous recombination, 14-15 replacing native constitutive promoter for the bcsABCD with paraBAD inducible arabinose promoter and araC genes through homologous recombination.

Test

After positive control we wanted to immediately characterise our transformed strains. However, due to time constraints we were not able to achieve an in depth characterisation. Still, we managed to do an initial comparative inoculation test by preparing SOC media with 2% glucose and 1% arabinose as well as SOC media with just 2% glucose added. We then added colonies from knockout strain CD015, inducible strain CD027 and WT for comparison.

Figure 13: K. xylinus CD027 colonies were used to inoculate a 6-wellplate based on layout (left). A: shows the plate after initial inoculation. B: shows the well-plate after 3 days. C: shows the well plate after 9 days of standing cultivation in room temperature.

Although the test is qualitative in nature it was mainly to assess if bacterial cellulose pellicle formation in the inducible strain is visually different in media with and without inducer present and if the pellicle formation is visually similar than the wild type strain. After 3 and 9 days bacterial cellulose pellicles were forming in the media with arabinose present, indicating that the inducible bacterial cellulose production was succesfully engineered in K. xylinus CD027 strain. However, the visible pellicle looked visually thinner and less dense than the wild type formed pellicle, indicating some kind of perfomance reduction, though quantitatve experiments need to occur to be conclusive.

Learn

While we were ultimately successful in engineering a viable inducible bacterial cellulose producing Komagataeibacter strain. There were several points during the in silico cloning as well as the wet lab stage where improvement is indeed possible. Furthermore, the whole plasmid sequencing showed that in all transformed E. coli with the inducible constructs, the RBS sequence was missing.

Figure 14: Part of sequence alignment of CD017 showing missing RBS sequence in assembled construct.

Figure 15: Part of sequence alignment of plasmid CD027 showing missing RBS sequence in assembled construct.

Sequencing revealed that the ribosome binding site is missing in all of the inducible constructs. A possible explanation for that could be due to increased stress from abundance of transcriptional factor AraC in E. coli, resulting in the consistent removal of the ribosome binding site of all plasmid containing thearaC by E. coli through recombination.

While we were lucky that our plasmids were still able to function, future iGEM Teams should definitely take this into consideration if they want to continue our work with the constructs.

Updates

After Wiki freeze

As final update experiment, we were interested in further investigating differences in cell viability between all engineered strains. To achieve this, we measured the OD Growth of both the knockout and inducible BC K. xylinus strains in a microplate reader.

Outlook

Outlook

With our results, we showed that engineering the bacterial cellulose production in K. xylinus is possible and we hope that future iGEM teams go even furter. Eventually reducing the need for cellulase necessary to increase viable Komagataeibacter cell concentration as well as optimise the carbon intake by targeting knockouts, overexpression and clone an inducible promoter in front of different genes we want to up or down regulate. Giving us all in the end even more control over bacterial cellulose synthesis. Through this, we hope to one day push the economic scale in our favour and achieve a true sustainable production method of bacterial cellulose.

Sources:

1. McNamara, Joshua T et al. “A molecular description of cellulose biosynthesis.” Annual review of biochemistry vol. 84 (2015): 895-921. https://doi.org/10.1146/annurev-biochem-060614-033930
2. Römling, U., & Galperin, M. Y. (2015). Bacterial cellulose biosynthesis: diversity of operons, subunits, products, and functions. Trends in microbiology, 23(9), 545–557. https://doi.org/10.1016/j.tim.2015.05.005
3. Mangayil, R., Rajala, S., Pammo, A., Sarlin, E., Luo, J., Santala, V., Karp, M., & Tuukkanen, S. (2017). Engineering and Characterization of Bacterial Nanocellulose Films as Low Cost and Flexible Sensor Material. ACS applied materials & interfaces, 9(22), 19048–19056. https://doi.org/10.1021/acsami.7b04927

Co-culture

To achieve our goal of a stable co-culture that is viable to produce consistent mats with potentially homogoneus properties. We went through multiple synbio and non synbio engineering cycles to improve cultivation methods and results.

Sticky Yeast Cycle 1

Design

We consulted with Prof Dr Daniel C. Ducat. During our meeting with himn he proposed several ideas including to engineer our yeast strain so that the outer cell wall has an increased affinity towards bacterial cellulose. Based on that, we did some research on yeast surface display systems and were also provided with Saccharomyces cerevisiae EBY100, an engineered strain that is commonly used for yeast surface display purposes through an engineered expression vector pYD1 with multiple cloning sites to add a gene of interest. The gene of interest for our purposes is the cellulose binding domain. A binding protein with . Fortunately, the dye group already started working with cellulose binding domain sequence and already cloned it into an Level O construct. So we just needed to design primers to amplify the CBD with EcorI and XhoI overhangs to enable insertion at multiple cloning site in pYD1 later on.

Figure 1: in silico construct of expression vector pYD1 with cellulose binding domain gene sequence insert.

Build

We started, by amplifying the CBD with EcorI and XhoI restriction site overhangs. Then to create sticky ends on both the backbone pYD1 and the CBD, enzyme digests has been done. Then followed a DNA purification through gel extraction and finally an assembly with T4 ligase. Afterwards we digested our backbone pYD1 and the CBD used T4 ligase to assemble our construct and finally transformed our construct in E. Coli DH5α and initial verification the presence of our insert through colony PCR.

Figure 2: Colony PCR verification if CBD is in assembled construct.

We then verified the results through sequencing the pYD1 plasmid with CBD insert. After sequencing results verified that sequence and plasmid sequence are almost identical to in silico construct, we then transformed our construct in S. cerevisiae EBY100. After 2 days initial colonies were visible and cPCR results show succesful transformation of S. Cerevisae EBY100.

Figure 3: S. cerevisiae EBY100 on drop out media -uracil -threonine agar plates after transformation with CD105 pYD1 CBD "Stcky yeast"(right).

Figure 4: Colony PCR verification if plasmid with CBD insert is in S. cerevisiae EBY100.

Test

To test Sticky yeast strain we attempted to perform a cell viability drop assay after induction of the expression vector. However due to time constraints we were not able to evaluate our results. In addition plans by using cellulose strips to compare the "stickiness" after induction in a sticky yeast culture were also considered but were not feasible in the scope of our project.

Learn

In the end we were not able to proof that Sticky yeast is a viable concept during our iGEM run. However we implore other iGEM Teams to continue our work here and proof if the idea of a yeast that is binding to a specific surface is feasible. If proven viable, the engineered yeast could have multiple applications far beyond textile purposes.

Stadardised BC mat protocol Cycle 1

Design

We initially designed all our in silico constructs for the expression of our chromoprotein constructs in E.coli with the terminator EF_TZ (BBa_J435371 )for our L1 constructs.

Build

To develop the Protocol we were researching different cultivation methods as well as investigating some of the initial BC mats cultivated by the other subgroups. By combining data from literature with our attempts we were able to write an initial protocol that covers the step from initial cultivation of pre-culture to final washing and post treatment

Test

To verify the feasibility in producing any BC mat succesfully with the standardised protocol occured through a mini workshop to give members outside the Co-culture subgroup the chance to create a BC mat. The mats resulting from the test were later used in meetings with the fashion college, University of applied science in Niederrhein for further testing as well as dyeing experiments by the dye subgroup. However we were not able to meaningfully test if the mats were comparable due to time constraints.

Learn

During the proeject run we made several mistakes that resulted in Bacterial cellulose mats that were not really comparable and initial stages of the project we produced mats by inoculating rods. We also learned that the Cellulase amount could be reduced in half and still result in satisfactory results.

Dye Group

To achieve our goal to have coloured bacterial cellulose mats, we went through multiple design cycles in order to come closer to achieving a modular colouring system.

Dyeing of bacterial cellulose mats Cycle 1

Design

Due to chromoproteins being not heat-stable, we wanted to try dyeing our mats with different (substances). To achieve this, we settled on dried algae powder, beetroot, curcurmin and activated charcoal.

Build

We mixed the powdered pigments into the growth media (YPD) and inoculated it with K.xylinus from a plate. They were then set to incubation for 7 days at 30 degrees.

Test

All mats were autoclaved in dH2O and washed on an aggregator for 50 min in water and then dried on a drying rack. All mats except for the curcumin mat lost their colour.

Learn

Curcumin is suited as a natural pigment to use for staining bacterial cellulose. The bacterial cellulose mats are in this state not under the perfect condition of being dyed due to the intrinsic brownish colour.

Improvement of the dye assembly approach Cycle 2

Design

We initially designed all our in silico constructs for the expression of our chromoprotein constructs in E.coli with the terminator EF_TZ (BBa_J435371 )for our L1 constructs.

Build

We started our golden gate assembly with BBa_J435300 as the acceptor vector, our own designed constructs BBa_K5146000 and BBa_K5146020, a chromoprotein like aeBlue ( (BBa_K864401)) and the beforehand mentioned BBa_J435371.

Test

Running cPCRs with the white colonies that formed on xGal plates showed that either the E.coli had taken up multiple plasmids or only an uncut BBa_J435371 because it also contained an ampicillin resistance.

Learn

We learned that it extremely complicates the cloning and screening process and success if one of the L0 parts has the same resistance marker as the L1 you are aiming for. With that knowledge in mind, for all future reactions we switched to BBa_J428092 as a terminator because it has a chloramphenicol resistance and it does not pose the same problems as the previously mentioned terminator.

Dyeing of bacterial cellulose mats utilizing pigments Cycle 3

Design

Due to the first dyeing test with natural pigments being unsuccessful, we wanted to try a new approach, where the bacterial cellulose mat itself is better suited for dyeing to the colour. For the new dyeing approach, we wanted to try to utilise beta carotene as a pigment. From our experience gathered over the course of the project, we wanted to use a bacterial cellulose mat that is better suited for dyeing and has a lighter colour than incubated on YPD.

Build

We extracted a bacterial cellulose mat grown on YPG media which is way lighter in colour. The mat was washed in distilled water at 20 rpm for 50 minutes afterward, it was washed twice in 1% NaOH solution for 40 minutes each. The mat was then shaken in a beta-carotene solution for 2 hours and dried on parchment paper for 2 days.

Test

The bacterial cellulose mat did not desaturate throughout the drying.

Learn

Beta carotene is suited to stain bacterial cellulose mats furthermore due to the added washing steps the bacterial cellulose mat absorbs the colour better.

Property test

To cultivate our bacteria, K. xylinus, and thus our bacterial cellulose as a product, we first designed the composition of our culture media. For the media, we decided to set two variable parameters: the concentration of glycerol as our alternative sustainable carbon source and the concentration of xyloglucan as hemicellulose. The xyloglucan used in this project was partly obtained by extraction from tamarind seeds in our lab.

Mechanical Tests Cycle 1: Visual evaluation of the effects of hemicellulose and glycerol on the BC mat

Design

We planned to grow K. xylinus in “Hestrin-Schramm” (HS) media with varying hemicellulose and glycerol concentrations to test its effects on the properties of the bacterial cellulose produced.

In other experiments, we wanted to test how well Bacterial Cellulose mats would grow in “Yeast Peptone Sucrose” (YPS) and “Yeast Peptone Dextrose” (YPD) media.

Build

K. xylinus was grown in the media. After six days, the bacterial cellulose mats were harvested, washed and lastly dried.

Test

The mats were evaluated visually, where mats with higher Xyloglucan concentrations seemed to perform better relatively, as they retained their shape. Some of the mats in lower concentrations were also not extractable due to their flimsy attributes. Further, mats grown in YPD or YPS seemed to grow faster, thicker and more stable in general

Learn

All samples shrunk significantly after the drying process, and therefore an additional post-treatment or a change in the drying process needed to be devised. We learned that mats grown with xyloglucan were easier to extract and showed a more rigid structure, which supports that xyloglucan may enhance the physical properties of BC mats. The variation of glycerol concentration only resulted in a negligible effect on the pellicle formation.

As mats tended to grow better in YPD and YPS media, changing the media, we added our xyloglucan to would be the next step.

Mechanical Tests Cycle 2: Mechanical evaluation of the effects of Xyloglucan

Design

“Yeast Peptone Sucrose” (YPS) media with varying Xyloglucan concentrations were prepared. We also planned to test washing the BC mat with a 1% (w/v) NaOH solution hoping to remove media stains better than with ddH2O only and to maybe change properties. We also wanted to evaluate the possible property-enhancing effects of different drying methods like parchment drying and drying the mats in the drying cabinet.

Build

We grew mats with “Yeast Peptone Sucrose” (YPS) media of varying Xyloglucan concentrations. Different washing and drying methods were applied.

Test

Tensile strength tests were conducted on each sample to determine the effects of different xyloglucan concentrations, washing methods, as well as drying methods.

Learn

BC mat samples grown in YPS medium with 0.25% (w/v) xyloglucan concentration displayed the greatest tensile strength compared to other BC mat samples but showed reduced elasticity compared to mats without xyloglucan. Samples with xyloglucan concentrations showed reduced tensile strength as well as elasticity compared to 0% xyloglucan mats. This shows that other ways to increase the tensile strength and elasticity should be tested, as the addition of xyloglucan seems unviable. These could be things like reinforcing the textile by letting them grow into a net added to the growth media beforehand.

Previously, the washing steps were carried out with cold water but not NaOH. Treatment with NaOH enhanced both the elasticity and tensile strength of our BC mats.

In addition, drying in the drying cabinet increased the tensile strength and elasticity of the mats compared to parchment drying.

Chemical Tests Cycle 1: First HPAEC

Design

K. xylinus was grown in Hestrin Schramm (HS) media with varying hemicellulose concentrations (xyloglucan - 0.1~0.5%), to see if and how much of it was integrated into the mat. The amount of glycerol in the used HS media was also varied, to evaluate possible effects on the mats’ composition.

Build

K. xylinus was grown in the media. After six days, the bacterial cellulose mats were harvested, washed and lastly dried.

Test

The sugar composition of the produced BC mats was evaluated by HPAEC. Ribose was used as the internal standard.

Learn

During the HPAEC evaluation, we found that our samples already contained ribose. Another sugar needed to replace Ribose as the internal standard.

Chemical Tests Cycle 2: Second and Third HPAEC

Design

K. xylinus was again grown in HS media with hemicellulose but this time only with 0.5% and 0% xyloglucan which served as a contrast, as we found in the first engineering cycle of the mechanical tests, that mats grown with higher concentrations of Xyloglucan seem to be more rigid.

Build

After six days, the bacterial cellulose mats were harvested, washed and lastly dried. In addition, BC mats grown in 0% xyloglucan but later washed in a 0.5% xyloglucan solution were prepared.

Test

The sugar composition of the produced BC mats was evaluated by HPAEC. Ribose was used as the internal standard.

Learn

Xylose was absent in the BC mat grown without Xyloglucan (0% XG). A considerable amount of xylose was found in the mat washed with 0.5% (w/v) XG solution (0% XG washed in 0.5% XG) and a relatively high amount in the mat grown with 0.5% Xyloglucan (0.5% XG), which means, that Xyloglucan was incorporated in the Bacterial Cellulose mat.

Hardware Cycle 1: Designing the first machine

Design

We designed a property testing machine to be able to run our own mechanical tests on our produced textiles.

Build

We built the designed machine with wood as our main building material.

Test

We tested if tear tests would be possible with the construction by ripping apart pieces of cardboard, held by self-built clamps, with the upper one being connected to a string which was pulled upwards by hand.

Learn

We learned that our designed test could work if improved further, but when we showed the machine to one of our PIs, Prof. Markus Pauly, he made us aware of the problem, that we wouldn’t be able to properly clean the machine if something happened in the lab, meaning we had to go for a non-porous material, to build our machine with.

Hardware Cycle 2: Designing the improved machine and first winding mechanism

Design

We redesigned the initial property testing machine’s design, with galvanised sheet metal as our main building material, as it could be cleaned without problems. We also designed a winding mechanism to substitute for pulling the string/cable by hand.

Build

We built the frame for the machine as well as the first winding mechanism made of LEGO to substitute for pulling by hand (see picture). For the clamps, we reused the ones from the previous machine.

Test

We tested the winding mechanism and realised that it wasn’t able to properly wind up the string/cable. The stability of the frame was evaluated in the same way as the wooden machine before.

Learn

We learned that we would have to redo the winding mechanism with more area to wind the string on. The frame worked as intended.

Hardware Cycle 3: Improving the winding mechanism

Design

Based on a construction by the iGEM team from GreatBay_SZ 2019, we redesigned the winding mechanism.

Build

We built the new winding mechanism (see picture) and also connected it to the frame for the tests this time.

Test

We tested the winding mechanism by ripping apart cardboard with it, which worked well, but faced problems when we tried to go for textiles, which we couldn’t rip apart. To quantify the force output, we held against the pull of the motor with a Newtonmeter and found that the force output was only about 6 N.

Learn

We learned that our winding mechanism worked, but that our motor was too weak, meaning we had to take a stronger one.

Hardware Cycle 4: Increasing the force output

Design

We redesigned our winding mechanism to fit two stronger motors, which could then move simultaneously, increasing the force output.

Build

We built the new winding mechanism with two bigger LEGO motors (see picture). As this couldn’t be used with the regular control panel, we also wrote code to be able to make it usable (See software)

Test

We tested the winding mechanism and code by ripping apart the textile which we had failed to rip apart with the previous construction. As this worked, we measured the force output which was above what we could measure with our Newtonmeter which could only measure up to 30 N.

Learn

We learned that our winding mechanism worked, but that our measuring range was too small, which would have to be increased with stronger Newtonmeters for example. We also learned from one of the tests, where the upper clamp flew away when the textile ripped, that this was a safety hazard which had to be dealt with.

Hardware Cycle 5: Integrating Safety Mechanism

Design

We designed a safety mechanism so that our upper clamp wouldn’t fly off when a textile rips apart.

Build

We integrated the safety mechanism by adding a WAGO clamp to the Newtonmeter’s hook.

Test

We tested the safety mechanism which worked at lower forces but less above the detection limit.

Learn

We learned that our safety mechanism would have to be improved further to allow for higher force outputs.