Abstract

The booming of the fashion industry has led to rising demand for leather. However, current industrial leather production causes environmental and ethical problems such as water pollution and animal cruelty. LINKS_China 2021 aims to create a sustainable and humane leather substitute using bacterial cellulose membrane (BCM) produced by co-culturing Komagataeibacter with Saccharomyces cerevisiae. By engineering S. cerevisiae, we induced transgenic expression of artificial spider-silk proteins fused with cellulose-binding matrixes, which increases the tensile strength and flexibility of BCM upon binding to it. Additionally, we enabled our engineered S. cerevisiae to synthesize ethyl acetate, giving our membrane a fruity fragrance. To endow our leather with more fashionable features, we used engineered Escherichia coli to synthesize different di-halogenated indigoid dyes for coloration. We expect our novel leather to be a transformative product that will continue the prosperity of fashion industry through a more humane and environmentally friendly technology.

Figure 1. illustration of the Neoleathic Age project workflow. Komagateaibacter sp. and S. cerevisiae BY4741 will produce bacteria cellulose membrane. E. coli BL21 (DE3) will produce spider silk proteins fused with cellulose binding matrixes. E. coli DH5α will produce natural pigment dyes. By combining these three components, we can produce NeoLeather.

Target 1：Producing Bacterial Cellulose Membrane

The base material for our project, bacterial cellulose membrane (BCM), is produced by a Symbiotic Co-culture Of Bacteria and Yeast (SCOBY) of Komagataeibacter intermedius/rhaeticus and Saccharomyces cerevisiae. SCOBY enables consistent and rapid BCM production. BCM formation is characterized by secretion and bundling of individual glucan fibers, which takes several days, forming a thick BCM matrix called pellicle. Before we begin our project, we had to acquire Komagataeibacter strains for our SCOBY.

Isolating Komagataeibacter strains

To acquire Komagataeibacter strains, we bought commercially available kombucha and isolated Komagataeibacters from the drink. We performed 16s rRNA sequencing and constructed a phylogenetic tree to identify our colonies of Komagataeibacter. We isolated four strains: B2, 12, 25, and 40. From 16sRNA analysis, we concluded that strain 40 is most similar to Komagataeibacter rhaeticus, and strains B2, 12, and 25 is most similar to Komagataeibacter intermedius (Fig. 2A). We also observed the single colony morphology for all four strains (Fig 2B). Interestingly, strain B2 and 40 have observable fiber production.

Figure 2. Characterization of strains 12, 25, 40, and B2. A) Phylogenetic tree of our strains (12, 25, 40, and B2) constructed using 16sRNA analysis. Due to size constraints, “rh” stands for rhaeticus, “xy" for “xylinus”, "in" for “intermedius”, “su” for

Characterizing BCM

We then measured the dry weight of BCM produced by each strain after 10-day-incubation(Fig. 3A & B). Strains B2 and 12 produced the most BCM, followed by strain 40 and25. B2 was chosen for the project since B2 displayed the greatest stability in SCOBY, consistently and rapidly producing BCM.

To determine the optimal culturing time for SCOBY, we measured the thickness of SCOBY with B2 BCM for ten days. We noticed that the BCM membrane generally grows faster during the first few days, and its growth rate gradually plateaus after seven to ten days (Fig. 3C). Thus, for the rest of the project, we grew strain B2 for 10 days for maximal BCM production unless otherwise stated.

Figure 3. Culturing and characterizing SCOBY. A) Schematic representation of how BCM is produced from SCOBY. The media used is YPD + 0.5% glucose (with 1% cellulase for culture tubes) B) Comparison between the dry weight after 10 days for strains B2, 12, 25, and 40. C) Representative curve of the thickness of BCM of SCOBY with B2 over ten days. All error bars represent two standard deviations from the mean.

We also experimented with different sized cultures of pure Komagataeibacter and SCOBY (Fig. 4). In a shaking culturing tube, pure Komagataeibacter forms furry balls (Fig. 4A). SCOBY BCM’s maximum thickness increases with bigger cultures (50mL centrifuge tube < 2L glass bowl < 4L glass tank; Fig 4B, C, D), but the growth rate of SCOBY BCM is generally the same between different cultures. We attempted multiple large cultures of SCOBY and acquired large BCMs (Fig. 4E) for demonstrating the viability of BCM as leather (visit our proof of concept page for more).

Fungal contamination is a major problem with large-scale fermentation, which will damage the integrity of BCM. This inspired us to visit Classy-Kiss (visit our iHPs for more), where we gained insight into how they sterilized their equipment. Furthermore, they also provided helpful advice regarding optimal fermentation conditions. We designed and prototyped a NeoLeathic Tanner capable of growing and processing the leather in one place, without the need for humans (visit our hardware page for more).

Figure 4. BCM in different culture environments. A) Komagataeibacter sp. in 14mL culturing tube with 5mL YPD after shake culturing. B) Komagataeibacter sp. in 50mL centrifuge tube with 15mL YPD. C) SCOBY in 2L glass bowl with 600mL YPD. D) Raw BCM grown from SCOBY in 4L glass tank with 2L YPD. E) Raw BCM from SCOBY in 4L glass tank with 2L YPD curled up in hand. Growth medium for all is YPD + 0.5% glucose. Aerobic conditions are ensured for all cultures.

Target 2：Spider Silk modification of BCM’s properties

In order to modify BCM’s physical properties, we designed and expressed spider silk fibroins fused with cellulose binding matrixes (CBMs; learn more on our description page) to bind to BCM (Fig. 5). For our project, we experimented with CBM3 from Ruminiclostridium thermocellum (Protein Data Bank (PDB) accession: 1NBC; Fig. 5D) (2) and CBM2 from Cellulomonas fimi (PDB accession: 1EXG; Fig. 5C). For our spider silk protein, we chose to use the synthetic mini spider silk protein NT2RepCT (2Rep; first characterized by GreatBaySZ_2019). 2Rep is water-soluble due to hydrophilic interactions of protein N-terminal and C-terminal. When 2Rep is submerged in a coagulating bath and subjected to a shear force, the repetitive regions will uncoil, form beta-pleated sheet networks and solidify into silk fiber (Fig. 5A).

Figure 5. Structure and design of 2Rep and 2Rep fused with CBMs. A) Schematic representing how individual 2Rep proteins becomes silk fibers. B) Schematic drawing of CBM fused with 2Rep C) CBM2 structure from PDB. D) CBM3 structure from PDB

For adding CBM3 flanking to 2Rep, we synthesized CBM3-BsaI-CBM3 on a pET28a vector. Primers were then used to add BsaI restriction sites in 2Rep to fuse the respective domains together in the synthesized pET28a vector by Golden Gate assembly (Fig. 6A & 6B). After construction, the plasmids were transformed into E. coli BL21(DE3) for IPTG-inducible expression (Fig. 6C).

Comparing the solubility of CBM2/3 fused with 2Rep

2Rep was shown to possess good water solubility, but we’re uncertain whether our modification will change this desirable property. To test the solubility, we cultured the modified strains, induced the expression of the constructs, collected the cells and performed SDS-PAGE on cell lysate (Fig. 6). Expression of both constructs, CBM3-2Rep-CBM3 (72kDa) and CBM2-2Rep-CBM2 (65kDa), are observed. CBM2-2Rep-CBM2 was present in the whole cell sample, but absent in the cell lysate supernatant, indicating poor water solubility. In contrast, CBM3-2Rep-CBM3 was present in both whole cell and supernatant. Therefore, CBM3-2Rep-CBM3 was chosen for the rest of the project for its superiority in water solubility.

Figure 6. CBM-2Rep-CBM construction and expression. A) Schematic representing construction of CBM-2Rep-CBM plasmids using golden gate assembly. B) Gel electrophoresis of CBM-CBM and CBM-2Rep-CBM plasmids. C) Schematic representation of CBM2-2Rep-CBM2 and CBM3-2Rep-CBM3 constructs in E. coli DH5α D) SDS-PAGE analysis of whole cell (WC) and supernatant (S) samples of CBM3-2Rep-CBM3 and CBM2-2Rep-CBM2.

Purification of CBM3-2Rep-CBM3

We purified CBM3-2Rep-CBM3 via the fused his-tag (Fig. 7A), and followed up with BCA assay to measure the yield of our modified spider silk protein. The concentration of CBM3-2Rep-CBM3 is approximately 115.78mg/L (Fig. 7B).

Figure 7. Purification results for CBM3-2Rep-CBM3. A) SDS-PAGE analysis of his-tag samples for CBM3-2Rep-CBM3. B) BCA test standard curve. The pink square represents CBM3-2Rep-CBM3.

We also constructed a 2Rep plasmid capable of expression in E. coli BL21(DE3) (Fig. 8A). We expressed 2Rep alongside CBM3-2Rep-CBM3 and purified both proteins (Fig. 8B & C). CBM3-2Rep-CBM3 had a titer of 133.08mg/L, which was higher than 2Rep’s titer of 98.44mg/L (Fig. 8D). Both proteins demonstrated great water solubility.

Figure 8. Expression and purification of 2Rep and CBM3-2Rep-CBM3. A) Schematic representation of 2Rep and CBM3-2Rep-CBM3 in E. coli BL21(DE3). B) SDS-PAGE analysis of 2Rep and CBM3-2Rep-CBM3 expression. C) SDS-PAGE analysis of his-tag purification of CBM3-2Rep-CBM3 and CBM2-2Rep-CBM2. D) BCA test of 2Rep and CBM3-2Rep-CBM3. The blue dot represents 2Rep while the pink dot represents CBM3-2Rep-CBM3.

Measuring physical properties of BCM mixed with 2Rep and CBM3-2Rep-CBM3

After purification of both 2Rep and CBM3-2Rep-CBM3, we went on to test whether the binding of additional spider silk layer can enhance the ability of BCM enduring pulling and the softness of our material. To bind our spider silk protein to BCM and to prevent rehydration of BCM, we added protein elute to dried BCM, soaked BCM in ethanol and subsequently soup water, and dried the BCM.

We compared the maximal force of BCM added with 2Rep or CBM3-2Rep-CBM3. We found statistical significant increase in maximal force endured when BCM mixed with CBM3-2Rep-CBM3. We then compared the effect of adding different quantities of 2Rep and CBM3-2Rep-CBM3 on maximal force of BCM (Fig. 9C & D), and found a one-fold increase from raw BCM to BCM with 5mg 2Rep and 1.5-fold increase from raw BCM to BCM with 5mg CBM3-2Rep-CBM3. We noticed that a statistically significant increase in maximal force endured will only be observed after adding at least 3.4mg of CBM3-2Rep-CBM3.

Figure 9. Measuring the maximal force of our new materials. A) Schematic representation of our tensile tester for testing the maximal force of BCM. The purple material is BCM. B) Comparison of the maximal force of BCM without protein, BCM with 2Rep, and BCM with CBM3-2Rep-CBM3. 5mg of spider silk protein were used. C) Comparison of the effect of 2Rep concentration on maximal force of BCM. D) Comparison of the effect of CBM3-2Rep-CBM3 concentration on maximal force of BCM. Approximately same weight of BCM were used. Error bars represent two standard deviations from the mean. Statistical test: two-sided student t-test. *: p < 0.05; **: p < 0.01

We then compared the softness of BCM mixed with different quantities of 2Rep and CBM3-2Rep-CBM3 (Fig. 10B & C). Both comparisons showed around 35% increase between raw BCM and BCM with 3.4mg of protein. Same amounts of 2Rep and CBM3-2Rep-CBM3 generally achieved similar results, indicating that softness is not affected by the addition of CBM3.

Overall, we have demonstrated that adding 2Rep and CBM3-2Rep-CBM3 can significantly improve the physical characteristics of BCM, with CBM3-2Rep-CBM3 outperforming 2Rep in improved maximal force. This demonstrates the viability of spider-silk-BCM composite material to be used as a leather substitute.

Figure 10. Measuring the softness of our new materials. A) Schematic representing our softness testing for testing the softness of BCM. The purple material is BCM. B) Comparison of softness of BCM with 0, 0.5, 1.5, and 3.4mg of CBM3-2Rep-CBM3. C) Comparison of softness of BCM with 0, 0.5, 1.5, and 3.4mg of 2Rep. BCM of approximately same weight were used. Error bars represent two standard deviations from the mean. Statistical test: two-sided student t-test. *: p < 0.05; **: p < 0.01

Target 3：Dyeing BCM with Natural Dyes

Tanning and dyeing animal leather is a very polluting process. In areas lacking regulations, such pollution will enter the natural ecosystem, significantly impacting its health. Therefore, we produced three natural indigo dyes — indigo, 6, 6’-di-bromo-indigo (tyrian purple), and 6, 6’-di-chloro-indigo (tyrian red) in E. coli from tryptophan (trp) to dye our BCM.

Figure 11. Production pathway of indigo and related dyes. A) Pathway representing the production of indigo and tyrian purple from tryptophan using Trp-6-halogenase, TnaA, and FMO. B) Schematic representation of Fre-SttH and the function(s) of each domain.

Production of indigoid dyes from trp requires three enzymes: Fre-SttH (a 6-trp-halogenase), TnaA (tryptophanase) and FMO (flavin-containing monooxygenase) (Fig. 11A). To produce tyrian red and tyrian purple, we need to express Fre-SttH in a different strain than the one with TnaA and FMO (learn about why at our descriptions page).

Improving solubility and expression of Fre-SttH

Lee et al designed and characterized the fusion enzyme Fre-SttH as a trp-6-halogenase. Fre-SttH is composed of two separate domains, Fre and SttH. SttH is a trp-6-haloganese that requires FADH2 as a cofactor to convert trp into 6-X-trp, and is highly insoluble in E. coli. Lee et al fused Fre, a highly-soluble flavin reductase from E. coli which reduces FAD to FADH2, with SttH as a N-terminal soluble tag, enabling the protein to become soluble and eliminating the need for costly FADH2 cofactors to be added (Fig. 11B). Thus, we chose to use Fre-SttH as our trp-6-halogenase.

We expressed Fre-SttH under T7 promoter in E. coli BL21(DE3) (Fig. 12A). SDS-PAGE of the sample showed decent Fre-SttH expression but suboptimal water solubility, since the supernatant sample has less intense target band compared to the whole cell sample (Fig. 12B).

For producing tyrian purple, Fre-SttH and TnaA has to be in two different strains, so we used a ΔTnaA E. coli strain, courtesy of Sha Zhou. However, our ΔTnaA E. coli was supplied as E. coli DH5α, which was incompatible with the T7 promoter.

Because of the need for a ΔTnaA E. coli strain, we decided to switch from the T7 system to the E. coli DH5α compatible ptac system. We constructed two ptac plasmids, ptac-Fre-SttH and ptac-histag-Fre-SttH, and transformed them into E. coli DH5α ΔTnaA (Fig. 12C).

Figure 12. Construction and expression of Fre-SttH proteins. A) Schematic representation of E. coli BL21 (DE3) transformed with histag-Fre-SttH plasmid. B) SDS-PAGE analysis of histag-Fre-SttH expressed by E. coli BL21 (DE3) described in A. C) Schematic representation of E. coli DH5a ΔTnaA transformed with histag-Fre-SttH and Fre-SttH plasmids. D) SDS-PAGE analysis of Fre-SttH and histag-Fre-SttH expressed by E. coli DH5a ΔTnaA described in C.

We then induced expression of both proteins and performed SDS-PAGE. Results show that histag-Fre-SttH expression and solubility were poor, but Fre-SttH had extremely high expression and solubility (Fig. 12D). Thus, ptac-Fre-SttH in E. coli DH5α ΔTnaA was used for all further experiments.

Measuring Fre-SttH activity and substrate-product concentrations

In order to model the enzymatic dynamics of Fre-SttH, we induced Fre-SttH expression and added the substrates. We then took samples of the culture in 6-hour-intervals for 24 hours for HPLC, and calculated the relative concentrations of trp and 6-X-trp (Fig. 13C & D). To learn about the detailed experimental setup, visit our experiment and measurement pages.

For sample added with trp and NaBr, the concentration of trp decreased from approximately 1.5 mM to 0.5mM, while the concentration of 6-Br-Trp increased from approximately 0mM to 1.0mM (Fig. 13A). We obtained similar result from sample with trp and NaCl, which the trp concentration went from approximately 1.7mM to 0.4mM and 6-Cl-Trp concentration increased from 0 to 1.3mM (Fig. 13B). For both samples, the yield of trp to 6-X-trp conversion by Fre-SttH was near 100%. As trp still remains after 24h in both cultures, a fermentation time of 36 or 48h might yield a higher concentration of 6-X-Trp, should it be desired. From this, we acquired several enzymatic constants for Michaelis-Menten equation, which was vital to our modeling team.

Figure 13. Quantifying the production of 6-X-Trp in Fre-SttH. A) Concentration-time graph of Trp and 6-Br-Trp in ptac-Fre-SttH in E. coli DH5a ΔTnaA with 1.5mM Trp and NaBr. B) Concentration-time graph of Trp and 6-Cl-Trp in a culture of ptac-Fre-SttH in E. coli DH5a ΔTnaA with 1.7mM Trp and NaCl. C) HPLC results, from top to bottom, of Trp, 6-Br-Trp, NaBr-supplemented culture at 0, 6, 12, 18, and 24h. D) HPLC results, from top to bottom, of Trp, 6-Cl-Trp, NaCl-supplemented culture at 0, 6, 12, 18, and 24h.

Fusing TnaA and FMO together

TnaA and FMO are two vital but separate enzymes for converting trp/6-Br-trp to our indigo and tyrian purple dye. To increase the overall reaction speed, we fused these two proteins together into TnaA-linker-FMO.

We designed and engineered three strains of E. coli DH5α ΔTnaA: ptac-TnaA-rbs-FMO (RBS; in this strain TnaA and FMO are expressed as separate proteins), ptac-TnaA-rigid linker-FMO (RL), and ptac-TnaA-Flexible linker-FMO (FL) (Fig. 14A & C). As TnaA is expressed as a tetramer and FMO a dimer, we put the TnaA tetramer at the center of the fused protein, with FMO forming two dimers to each side of the TnaA tetramer (Fig. 14B).

Figure 14. Construction and expression of TnaA-FMO proteins. A) Schematic representing TnaA-RBS-FMO, TnaA-Flexible linker-FMO, and TnaA-Rigid linker-FMO transformed into E. coli DH5a ΔTnaA. B. Schematic representing the structure of TnaA-linker-FMO. C) The sequences between TnaA and FMO, with the rbs and amino acids labelled. FL and RL stands for flexible linker and rigid linker respectively. D) SDS-PAGE analysis of TnaA-RBS-FMO, TnaA-RL-FMO, and TnaA-FL-FMO.

After culturing and inducing the expression, the three strains, the SDS-PAGE showed separate expression of TnaA (60kDa) and FMO (54kDa) for RBS, and expression of one fused protein at 114 kDa for RL and FL (Fig. 14D). This indicated expected expression of our fused proteins.

We then cultured and induced RBS, RL, and FL with IPTG. After 20 hours, 1mM of either trp, 6-Cl-trp, or 6-Br-trp was added as substrate, and the relative dye concentration produced by each strain was calculated by using a standard calibration curve (Fig. 15B). The comparison between RBS, RL, and FL shows that there is similar production of tyrian red and tyrian purple, and a significant difference between indigo production of RL and FL. Titers of indigo is approximately 0.30mM (60% yield) for FL and RBS, and 0.20mM (40% yield) for RL. Titers of tyrian red is approximately 0.30mM (60% yield) for RL, FL, and RBS. Titers of tyrian purple is approx. 0.25mM (50% yield) for FL and 0.20mM (40% yield) for RL and RBS.

Figure 15. Measuring dyes production of different TnaA-FMO strains from Trp, 6-Cl-Trp, and 6-Br-Trp. A) Pictures of TnaA-RL-FMO, TnaA-FL-FMO, and TnaA-RBS-FMO with Trp, 6-Cl-Trp, or 6-Br-Trp added. B) Comparison of production titers of TnaA-RL-FMO, TnaA-FL-FMO, and TnaA-RBS-FMO with Trp, 6-Cl-Trp, or 6-Br-Trp added, pictured in A. Error bars denote two standard deviations from the mean.

After confirming the efficacy of Fre-SttH and TnaA-FMO, we wanted to compare TnaA-rbs-FMO with our TnaA-FL-FMO. Therefore, drawing inspiration from GreatBaySZ 2019’s TnaA-rbs-FMO expression system using a constitutive promoter system(TALEsp2), we designed and constructed TALEsp2-TnaA-FL-FMO.

Instead of using 1mM standard samples of trp or 6-X-trp, we attempted to produce dyes from trp and NaX salts. We induced Fre-SttH expression, took the sample supernatant and added it to the ptac-TnaA-FMO and TALEsp2-TnaA-FMO cultures, and compared the titers of the all using supernatant from Fre-SttH cultures as substrate (Fig. 16). There is no significant difference between any of the samples. Both TALEsp2 cultures produced titers of approx. 0.09mM for tyrian purple and 0.23mM for 6, 6’di-chloro-indigo. RL, FL, and RBS both achieved titers of 0.12 and 0.28mM for tyrian purple and tyrian red respectively.

Figure 16. Measuring dye production of different TnaA-FMO strains from Fre-SttH + Trp + NaCl/NaBr. A) Pictures of different TnaA-FMO strains with supernatant of Fre-SttH + Trp + NaCl/NaBr added. B) Comparison of dye production titers of TALEsp2-TnaA-RBS-FMO and TALEsp2-TnaA-FL-FMO. C) Comparison of dye production titers of ptac-TnaA-RL-FMO, ptac-TnaA-FL-FMO, and ptac-TnaA-RBS-FMO.

To dye clothe and our BCM, we scaled up production of our dyes to 400mL cultures (Fig. 17A). We then dyed silk handkerchiefs and our BCM (Fig. 17C & D). Visit our proof of concept page for more!

Figure 17. Production and dyeing of natural indigoid dyes. A) Indigo, 6, 6’-di-chloro-indigo, and 6, 6’-di-bromo-indigo produced in shake flasks. B) Extracted indigoid dyes. C) Cloth dyed with tyrian purple (left) and indigo (right). D) BCM dyed with indigo (left) and tyrian purple (right).

Target 4: Engineering yeast to achieve modification of BCM

In SCOBY, yeast is a common synbio chassis. Previously, Ellis et al has achieved functionalization of BCM in yeast by engineering yeast to secrete proteins. Inspired by their attempts, we want to engineer our yeast and modify our BCM for two different properties: in situ spider silk binding and fragrance production.

Engineering yeast to express and secrete CBM3-2Rep-CBM3 for BCM binding in situ

As the addition of CBM3-2Rep-CBM3 was shown to increase the softness and tensile strength of BCM, making it more like traditional leather, we wanted to engineer yeast to produce and secrete CBM3-2Rep-CBM3 to bind to the BCM as it is forming in SCOBY. This way, the labor-intensive process of protein purification can be skipped.

To enable secretion of production in yeast, we attached a short signal peptide called maturation factor alpha (Mα) to our protein. Aza et al have already characterized Mα and mutated it to become more efficient at protein secretion. For our project, we selected Mα E86T; A87N (referred to as Mα) as it had one of the highest secretion rates. We constructed two plasmids, Mα-sfGFP-CBM3 and Mα-CBM3-2Rep-CBM3 (Fig. 18B), and transformed both plasmids into yeast (Fig. 18A).

Figure 18. Expression and measurement of Ma proteins in Saccharomyces cerevisiae. A). Schematic of Ma plasmids transformed into S. cerevisiae BY4741 B) Gel electrophoresis results of Ma-sfGFP-CBM3 and Ma-CBM3-2Rep-CBM3 C) SDS-PAGE analysis of yeast whole cell and YPD media for Ma-sfGFP-CBM3. D) Pictures of single colonies of Ma-sfGFP-CBM3. E) Liquid cultures of Ma-sfGFP-CBM3 and Ma-CBM3-2Rep-CBM3. F) BCM produced using Ma-sfGFP-CBM3 S. cerevisiae BY4741 SCOBY.

On YPD plates, secretion of Mα-sfGFP-CBM3 can be seen by the green halo surrounding each individual colony (Fig. 18D). After culturing Mα-sfGFP-CBM3 and Mα-CBM3-2Rep-CBM3 yeast in liquid YPD (Fig. 18E), we performed SDS-PAGE on both the whole cell and growth media (Fig. 18C). We discovered that only a small portion of the expressed proteins were found in the media for Mα-sfGFP-CBM3 and no secreted proteins were found in the Mα-CBM3-2Rep-CBM3 media. We speculate that little CBM3-2Rep-CBM3 will be secreted, not enough to significantly alter the characteristics of the resulting BCM. Even though there was minimal protein secretion for both constructs, we still attempted SCOBY with Mα-sfGFP-CBM3. In the resulting SCOBY, BCM did display fluorescent characteristics, showing that some sfGFP-CBM3 proteins have bound to the BCM. In the future, we hope to increase the secretion rate of Mα-sfGFP-CBM3 and Mα-CBM3-2Rep-CBM3 to achieve in situ modification of BCM.

Synthesizing ethyl-acetate from the byproducts of fermentation

We observed an accumulation of acetate in SCOBY, which contains an odorous smell and can interfere with secreted proteins such as Mα-CBM3-2Rep-CBM3. Therefore, in order to utilize the accumulated acetate and to convert it into something useful, we decided on a simple metabolic pathway to turn ethanol (also produced by yeast fermentation in SCOBY) and acetate into ethyl acetate, which contains a fruity smell. This pathway can be completed with two enzymes: SaACS2 (acetyl-CoA synthase) from Salmonella enterica and AeAT9 (acyltransferase) from Actinidia eriantha (kiwifruit). SaACS2 converts acetate into acetyl-CoA under anaerobic condition. AeAT9 will transfer the acetyl group from acetyl-CoA to ethanol to form ethyl acetate (Fig. 19A).

To construct plasmids capable of expression in yeast, we utilized a yeast genetic toolkit first characterized by Lee et al. to construct three plasmids: pTEF1-SaACS2-tADH1, pRPL8B-AeAT9-tSSA1, and pTEF1-SaACS2-tADH1-pRPL8B-AeAT9-tSSA1 (Fig. 19C). The whole construction process was done in close contact and collaboration with AISSU_Union (learn more about the yeast toolkit and our collaboration on our partnership page). After our final plasmid (pTEF1-SaACS2-tADH1-pRPL8B-AeAT9-tSSA1) was constructed, we transformed it into yeast (Fig. 19B).

After transformation, we attempted to do fermentation but we have yet to receive optimal fermentation results due to time constraints.

Figure 19. Production of ethyl-acetate using engineered S. cerevisiae BY4741. A). Production pathway of ethyl-acetate from acetate and ethanol using ACS2 and AT9. B) Schematic representing engineeredS. cerevisiae BY4741 expressing ACS2 and AT9. C) Gel electrophoresis results of pTEF1-SaACS2-tADH1-pRPL8B-AeAT9-tSSA1.

Conclusion and Discussion

Bacterial cellulose membrane (BCM) is biocompatible, cheap material made from common microorganisms already in use in the food industry. Our project characterized BCM and modified the material in three main ways and proved the viability of BCM to be used as a more sustainable leather substitute. First, binding fused spider silk-cellulose binding matrix proteins increases the tensile strength and softness of BCM. Second, natural indigoid production counters the environmental damages of the traditional dyeing process. Third, in situ modification of BCM can be achieved by engineering yeast.

Further research is needed to achieve price-competitive industrial production of BCM compared to traditional leather (check out our entrepreneurship page for more). We also hope to achieve one-cell production of natural indigoid dyes by expressing the different proteins at different times (check our our model page). Additionally, the possibility of modifying Komagataeibacter to produce dyes by itself is attractive. We also hope to further our research in SCOBY, by increasing the amount of secreted proteins or achieve in situ modification of BCM in an alternative way.

The possibilities opened up by the Neoleathic age are endless. As our quality of life increases and we demand more leather products, it is vital to consider the hidden costs to our planet. The Neoleathic age hopes to guide the industry towards a more sustainable future, and we aspire to the day when we see BCM leather products enters the market.

Team:LINKS China/Engineering