Engineering Success

An outline of the engineering cycles we completed throughout ReneWool’s development

Project Description Future Directions Contribution
Engineering The Inducible URS Series pTac_RBS 1/2/3/4 Engineering Process for Spider Silk Engineering The Constitutive URS J23119_RBS 1 Engineering Process for The Keratinase KerDZ Engineering Process for Keratin-31

Engineering The Inducible URS Series pTac_RBS 1/2/3/4

Design

Our first task involved creating a library of plasmids to carry the genes required for ReneWool, as well as to build on the iGEM collection of knowledge. ReneWool includes a constitutive expression system for textile degrading enzymes, such as KerDZ, and an inducible system for spider silk production. This combination of constitutive and inducible plasmids allows us to increase spider silk production in periods of feedstock abundance and optimal biomass accumulation, while continuously producing the enzymes for degradation. With this in mind, we made two sets of plasmids, inducible and constitutive, which we compared in various strains to optimize protein expression. We tested different upstream regulatory sites in two plasmids from the pJUMP series and gathered data on protein expression and compatibility with our selected strains.

Our main experimental vectors were pJUMP24-1A, to carry the inducible promoters, and pJUMP28-1A, to carry the constitutive promoters. Apart from their origins of replication, these plasmids are similar in structure, allowing data gathered from one characterization experiment to be extrapolated to the other. They are also compatible with each other, owing to their different origins of replication, which allows us to design an expression system within a single organism capable of degrading textiles and producing spider silk!

Build

We began by assembling our inducible vectors. The IPTG inducible pTac promoter and one of four Ribosome Binding Sites (RBS1, RBS2, RBS3 or RBS4)1 were inserted into pJUMP24-1A via Gibson Assembly. Additionally, sfGFP was inserted downstream of the RBS as a reporter gene for expression under each pTac_RBS upstream regulatory site (URS). These plasmids were cloned into Escherichia coli strains DH5ɑ, K12, and the protein expression strains Rosetta-gami™(Hereby referred to as Rosetta-gami) and BL21. To select for isolated colonies containing our desired plasmid, the transformed strains were grown on LB agar containing 50 µg/ml Kanamycin.

Fig 1. | Our pJUMP24-1A_pTac_RBS(1-4)_sfGFP series of plasmids.
Fig 1. | Our pJUMP24-1A_pTac_RBS(1-4)_sfGFP series of plasmids.

Test

We carried out several characterization experiments to determine the optimal growth conditions for transformants carrying the plasmid, the metabolic burden of our inserts, and the strength of each pTac_RBS upstream regulatory site.

First, we conducted growth experiments on E. coli DH5ɑ, K12, Rosetta-gami, and BL21 strains transformed with pJUMP24-1A_pTac_RBS(1-4)_sfGFP. This was done to determine the optimal temperature for growth of transformed cultures. Additionally, we tested the metabolic burden of pTac_RBS(1-4)_sfGFP on each strain by comparing the biomass accumulation of our clones with that of the empty cell controls.

Fig 2. | Growth curves of DH5ɑ, BL21, K12 and Rosetta-gami strains carrying pJUMP24-1A_pTac_RBS(1-4)_sfGFP. Empty cell controls for each strain were grown in LB, and transformants were grown in LB Kanamycin (50µg/ml). All cultures were grown at 30oC for ten hours. The Optical Density (OD) at 600nm of each culture was measured every two hours.
Fig 2. | Growth curves of DH5ɑ, BL21, K12 and Rosetta-gami strains carrying pJUMP24-1A_pTac_RBS(1-4)_sfGFP. Empty cell controls for each strain were grown in LB, and transformants were grown in LB Kanamycin (50µg/ml). All cultures were grown at 30oC for ten hours. The Optical Density (OD) at 600nm of each culture was measured every two hours.
Fig 3. | Growth curves of DH5ɑ, BL21, K12, and Rosetta-gami strains carrying pJUMP24-1A_pTac_RBS(1-4)_sfGFP. Empty cell controls for each strain were grown in LB, and transformants were grown in LB Kanamycin (50µg/ml). All cultures were grown at 37oC for ten hours. The Optical Density (OD) at 600nm of each culture was measured every two hours.
Fig 3. | Growth curves of DH5ɑ, BL21, K12, and Rosetta-gami strains carrying pJUMP24-1A_pTac_RBS(1-4)_sfGFP. Empty cell controls for each strain were grown in LB, and transformants were grown in LB Kanamycin (50µg/ml). All cultures were grown at 37oC for ten hours. The Optical Density (OD) at 600nm of each culture was measured every two hours.

Learn

The growth curves demonstrated that 37oC is optimal for the propagation of pJUMP24-1A_pTac_RBS_sfGFP, as cultures at this temperature obtained higher biomass accumulation compared to cultures at 30oC. Although visually there appears to be variation in growth rate between transformants, no statistically significant difference was observed between any of the RBSs or between the RBSs and the empty cell control. From this data, we concluded that pTac_RBS1, 2, 3, and 4 do not impose a considerable metabolic burden on any of the strains tested. However, we still wanted to determine the relative strength of each RBS.

Test

In tandem with the above growth curve experiment, we tested the degree of protein expression under the pTac_RBS upstream regulatory sites. Our team conducted a fluorometric characterization experiment. We induced sfGFP expression in E. coli DH5ɑ carrying pJUMP24-1A_pTac_RBS1/2/3/4_sfGFP with IPTG, lysed the cells via French Press, and measured the fluorescence of the pellet after lysis.

Fig 4. | Fluorescence of E. coli DH5ɑ cultures transformed with pJUMP24-1A_pTaq_RBS(1-4)_sfGFP and induced with 3mM IPTG in liquid LB Kanamycin (50µg/ml). The reported fluorescence values have been background (fluorescence of empty DH5ɑ cells) subtracted to account for non-specific signals.
Fig 4. | Fluorescence of E. coli DH5ɑ cultures transformed with pJUMP24-1A_pTaq_RBS(1-4)_sfGFP and induced with 3mM IPTG in liquid LB Kanamycin (50µg/ml). The reported fluorescence values have been background (fluorescence of empty DH5ɑ cells) subtracted to account for non-specific signals.

Learn

RBS1 transformants reported the strongest fluorescence, indicating that pTac_RBS1 is the strongest URS of those tested. As such, pTac_RBS1 was selected as the URS for our spider silk gene, which needs to be inducible to maximize silk production. However, the fluorescence observed in RBS1 transformants was lower than is generally expected for sfGFP, implying that although RBS1 was the stronger RBS of our series, pJUMP24-1A_pTac_RBS1 is not an overly strong expression vector.

Engineering Process for Spider Silk

Design

Our spider silk gene, (A3I)3-A14, encodes a modified version of the recombinant spider silk NT2RepCT.2 We selected this protein due to its incredible physical properties, which match those of native dragline spider silk.3 We optimized this spider silk gene for expression in E. coli using the IDT Codon Optimization Tool, attached a 6xHis tag to the N-terminus, and added the standard BioBrick prefix and suffix around the gene.

Build

As pTac_RBS1 was the strongest inducible URS tested in our pJUMP24-1A_pTac_RBS1/2/3/4_sfGFP characterizations, it was selected as the URS for the expression of our spider silk gene. sfGFP was added in tandem to spider silk to create a translational fusion that allows for detection of spider silk production.

Fig 5. | Our pJUMP24-1A_pTac_RBS1_SpiderSilk_sfGFP plasmid.
Fig 5. | Our pJUMP24-1A_pTac_RBS1_SpiderSilk_sfGFP plasmid.

Test

To test whether our spider silk gene imposes a significant metabolic burden on the cell—which would impact the feasibility of producing spider silk on a commercial scale—we tested the growth of DH5ɑ transformed with pJUMP24-1A_pTac_RBS1_SpiderSilk_sfGFP.

Fig 6. | Growth curve of DH5ɑ transformed with pJUMP24-1A_pTac_RBS1_SpiderSilk_sfGFP. Empty cell controls for each strain were grown in LB, and transformants were grown in LB Kanamycin (50µg/ml). Cultures were grown at 37oC for ten hours. The Optical Density (OD) at 600nm of each culture was measured every two hours.
Fig 6. | Growth curve of DH5ɑ transformed with pJUMP24-1A_pTac_RBS1_SpiderSilk_sfGFP. Empty cell controls for each strain were grown in LB, and transformants were grown in LB Kanamycin (50µg/ml). Cultures were grown at 37oC for ten hours. The Optical Density (OD) at 600nm of each culture was measured every two hours.

Learn

DH5ɑ pJUMP24-1A_pTac_RBS1_SpiderSilk_sfGFP transformants grew more slowly and reached lower overall biomass than empty DH5ɑ cells. This result indicated that the spider silk containing plasmid imposes a metabolic burden on cells. This may pose a problem for using this expression vector to produce spider silk on a commercial scale. See future directions Future Directions
for our proposed new designs and considerations.

Engineering The Constitutive URS J23119_RBS 1

Design

In addition to the inducible URS described above, the ReneWool system requires a constitutive expression system for textile degrading enzymes. In designing our constitutive system, we applied what we learned from the experiments outlined above, as well as research from the 2023 UAlberta iGEM team. The 2023 UAlberta iGEM team had success using the J23 series of promoters in pJUMP vectors.4 To expand on their knowledge, we chose to experiment with the constitutive promoter J23119 in pJUMP28-1A. pJUMP28-1A was chosen due to its similarity to pJUMP24-1A and their ability to be expressed together in a single organism. We chose RBS1 as the ribosomal binding site, as our previous results indicated this was the strongest RBS in our series

Fig 7. | Our constitutive pJUMP28-1A_J23119_RBS1 plasmid. This constitutive expression system is necessary for our textile-degrading enzymes.
Fig 7. | Our constitutive pJUMP28-1A_J23119_RBS1 plasmid. This constitutive expression system is necessary for our textile-degrading enzymes.

Build-Test

Following BioBrick standards, we cloned J23119_RBS1 into pJUMP28-1A via restriction cloning using EcoRI and PstI cut sites in the biobrick sites on both the pJUMP 28-1A vector and the J23119_RBS 1 URS. To test the metabolic burden of J23119_RBS1, we transformed DH5ɑ, Rosetta-gami, and BL21 with pJUMP28-1A_J23119_RBS1 and performed a growth experiment.

Fig 8. | Growth curve of DH5ɑ, BL21, and Rosetta-gami strains carrying pJUMP28-1A_J23119_RBS1. Empty cell controls for each strain were grown in LB, and transformants were grown in LB Kanamycin (50µg/ml). All cultures were grown at 37oC for 10 hours. The Optical Density (OD) at 600nm of each culture was measured every two hours.
Fig 8. | Growth curve of DH5ɑ, BL21, and Rosetta-gami strains carrying pJUMP28-1A_J23119_RBS1. Empty cell controls for each strain were grown in LB, and transformants were grown in LB Kanamycin (50µg/ml). All cultures were grown at 37oC for 10 hours. The Optical Density (OD) at 600nm of each culture was measured every two hours.

Learn

DH5ɑ transformants achieved the highest OD600 out of the tested strains, with Rosetta-gami transformants achieving the next highest. Interestingly, although BL21 transformants achieved the lowest biomass accumulation of the three transformed strains, transformed BL21 attained a greater OD than empty BL21 Taken together, these results suggest that J23119_RBS1 does not impose a large metabolic burden and that DH5ɑ is superior to BL21 and Rosetta-gami in terms of plasmid maintenance. Nevertheless, we still wanted to test whether DH5ɑ remained the ideal strain once kerDZ was incorporated into the plasmid, as DH5ɑ is not a protein expression strain.

Test

As Kanamycin is susceptible to high salt concentrations, we also conducted this experiment in Lennox LB (43.1 mM NaCl) and regular LB (86.2 mM NaCl) to determine the optimal growing conditions for maintaining pJUMP28-1A_J23119_RBS1 in DH5ɑ transformants.

Fig 9. | Growth curve of DH5ɑ carrying pJUMP28-1A_J23119_RBS1 grown in LB (86.2 mM NaCl) Kanamycin (50µg/ml) or Lennox LB (43.1 mM NaCl) Kanamycin (50µg/ml). Cultures were grown at 37oC over 6 hours. The Optical Density (OD) at 600nm of each culture was measured every two hours.
Fig 9. | Growth curve of DH5ɑ carrying pJUMP28-1A_J23119_RBS1 grown in LB (86.2 mM NaCl) Kanamycin (50µg/ml) or Lennox LB (43.1 mM NaCl) Kanamycin (50µg/ml). Cultures were grown at 37oC over 6 hours. The Optical Density (OD) at 600nm of each culture was measured every two hours.

Learn

We observed no significant difference between DH5ɑ growth at high and low salt concentrations. If a higher rate of cell growth were achieved in the higher salt media, that would imply Kanamycin was being deactivated by the salt, causing our plasmids to be lost and cells to grow faster as a result. Based on our data, Kanamycin is not affected by salt concentration, and either condition is likely suitable for maintaining the pJUMP28_J23119_RBS1 plasmid in DH5ɑ, Rosetta-gami, or BL21 strains. The data gathered on the constitutive pJUMP28-1A_J23119_RBS1 plasmid informed the design of our KerDZ part. As pJUMP28_J23119_RBS1 did not impose a large metabolic burden on cells, it was selected as the plasmid for KerDZ.

Engineering Process for The Keratinase KerDZ

Design

To design our keratin-degrading enzyme, we searched the literature for a keratinase that would function efficiently in our system. KerDZ was selected based on its relatively high thermostability, ability to function at a wide range of pHs, and its known ability to degrade wool when expressed in E. coli.5, 6 We optimized kerDZ, originally from Actinomadura viridilutea, for expression in E. coli using the IDT Codon Optimization tool, attached a 6xHis tag to the N-terminus, and added the standard BioBrick prefix and suffix around the gene.

Build

Following BioBrick standards, we cloned kerDZ, the constitutive promoter J23119, and RBS1 into pJUMP28-1A via restriction cloning using the SpeI and PstI cut sites for the J23119_RBS 1_pJUMP 28 vector and the XbaI and PstI cut sites for the KerDZ gene fragment. We transformed DH5ɑ with this plasmid for propagation and later transformed it into BL21 and Rosetta-gami to achieve constitutive expression of KerDZ in our system.

Fig 10. | Our constitutive pJUMP28-1A_J23119_RBS1_KerDZ plasmid.
Fig 10. | Our constitutive pJUMP28-1A_J23119_RBS1_KerDZ plasmid.

This construct incorporated data from the fluorometric characterization of pJUMP24-1A_pTac_RBS1/2/3/4_sfGFP, which revealed that RBS1 was the strongest of our RBS series. Additionally, the pJUMP28-1A_J23119_RBS1 growth curve (Figure 8) confirmed that J23119_RBS1 does not impose a significant metabolic burden on DH5ɑ, Rosetta-gami, or BL21.

Test

To test the metabolic burden of J23119_RBS1_KerDZ, we transformed DH5ɑ, Rosetta-gami, and BL21 with pJUMP28-1A_J23119_RBS1_KerDZ and performed the following growth experiments.

Fig 11a. | Growth curve of DH5ɑ, BL21, and Rosetta-gami strains carrying pJUMP28-1A_J23119_RBS1_KerDZ in Lennox LB (43.1 mM NaCl). Empty cell controls for each strain were grown in LB, and transformants were grown in LB Kanamycin (50µg/ml). Cultures were grown at 37oC for 10 hours. The Optical Density (OD) at 600nm of each culture was measured every two hours.
Fig 11a. | Growth curve of DH5ɑ, BL21, and Rosetta-gami strains carrying pJUMP28-1A_J23119_RBS1_KerDZ in Lennox LB (43.1 mM NaCl). Empty cell controls for each strain were grown in LB, and transformants were grown in LB Kanamycin (50µg/ml). Cultures were grown at 37oC for 10 hours. The Optical Density (OD) at 600nm of each culture was measured every two hours.
Fig 11b. | Growth curve of DH5ɑ, BL21, and Rosetta-gami strains carrying pJUMP28-1A_J23119_RBS1_KerDZ in LB (86.2 mM NaCl). Empty cell controls for each strain were grown in LB, and transformants were grown in LB Kanamycin (50µg/ml). Cultures were grown at 37oC for 10 hours. The Optical Density (OD) at 600nm of each culture was measured every two hours.
Fig 11b. | Growth curve of DH5ɑ, BL21, and Rosetta-gami strains carrying pJUMP28-1A_J23119_RBS1_KerDZ in LB (86.2 mM NaCl). Empty cell controls for each strain were grown in LB, and transformants were grown in LB Kanamycin (50µg/ml). Cultures were grown at 37oC for 10 hours. The Optical Density (OD) at 600nm of each culture was measured every two hours.

Learn

We observed that DH5ɑ and BL21 transformants achieved higher biomass accumulation than Rosetta-gami transformants, indicating that these strains are optimal for pJUMP28-1A_J23119_RBS1_KerDZ plasmid propagation. Once again, BL21 transformants grew faster and achieved a higher overall biomass than empty BL21 cells, emphasizing that the pJUMP28-1A_J23119_RBS1_KerDZ plasmid does not impose a significant metabolic burden on BL21 cells. Although DH5ɑ transformants did not reach the same optical density as empty DH5ɑ cells, this difference is not significant, implying that either of these strains would be suitable for pJUMP28-1A_J23119_RBS1_KerDZ plasmid propagation. At a glance, transformant growth in Lennox LB (Figure 22a) appears slower, and transformants in this media reached lower overall biomass compared to those in regular LB. While there may be a link between salt concentration and the activity of Kanamycin, this relationship remains unclear as there is not a statistically significant difference in growth rate.

Engineering Process for Keratin-31

Fig 12. | Model of Keratin-31 interaction with KerDZ. Created using HDOCK HDOCK
by inputting AlphaFold2 AlphaFold2
models of Keratin-31 and KerDZ. The interaction was then visualized with SWISS-MODEL SWISS-MODEL
.
Fig 12. | Model of Krt31 interaction with KerDZ. Created using HDOCK HDOCK
by inputting AlphaFold2 AlphaFold2
models of Krt31 and KerDZ. The interaction was then visualized with SWISS-MODEL SWISS-MODEL
.

Learn

To test the efficiency of our keratinolytic enzyme, we aimed to express keratin in our system. krt31 was optimized for expression in E. coli using the IDT Codon Optimization tool. We attached a 6xHis tag to the C-terminus to allow for downstream Immunoaffinity Chromatography (IMAC) and added the standard BioBrick prefix and suffix around the gene. In parallel to the above experiments, we used Gibson Assembly to clone krt31 downstream of the inducible T7 promoter in the pET22-b(+) backbone. This design was chosen because of the reliability of the pET vector and the T7 promoter. Additionally, pET22-b(+) was an ideal vector because it is compatible with the pJUMP series and contains a pelB sequence, which aids in protein purification.

Fig 13. | Our inducible pET22-b(+)_T7_Kert31 plasmid.
Fig 13. | Our inducible pET22-b(+)_T7_Kert31 plasmid.

Test

To test whether the keratin-containing plasmid imposes a significant metabolic burden on the cell, we performed growth experiments on DH5ɑ, Rosetta-gami and BL21 transformed with pET22-b(+)_T7_krt31.

Fig 14. | Growth curves of DH5ɑ, BL21, and Rosetta-gami strains carrying pET22-b(+)_T7_krt31. Empty cell controls for each strain were grown in LB, and transformants were grown in LB Ampicillin (100µg/ml). Cultures were grown at 37oC for 10 hours. The Optical Density (OD) at 600nm of each culture was measured every two hours.
Fig 14. | Growth curves of DH5ɑ, BL21, and Rosetta-gami strains carrying pET22-b(+)_T7_krt31. Empty cell controls for each strain were grown in LB, and transformants were grown in LB Ampicillin (100µg/ml). Cultures were grown at 37oC for 10 hours. The Optical Density (OD) at 600nm of each culture was measured every two hours.

Learn

We observed that DH5ɑ transformants achieved the highest biomass accumulation, followed by Rosetta-gami transformants. This result indicates that these strains are optimal for pET22-b(+)_T7_krt31 plasmid propagation. Although BL21 transformants grew slower than DH5ɑ and Rosetta-gami transformants, their growth is very similar to that of empty BL21 cells. Combined with the relatively high biomass achieved by Rosetta-gami and DH5ɑ transformants, this implies that pET22-b(+)_T7_krt31 does not impose a significant metabolic burden on these strains and that they are suitable for further experimentation.

Test

To test for Krt31 expression in this vector we transformed pET22-b(+)_T7_krt31 into the E. coli Rosetta-gami and BL21 protein expression strains. The plasmid was purified and sequenced and was confirmed to contain Krt31 in the correct frame. To characterize protein expression, we induced the strains with 1 mM IPTG overnight and compared the protein levels with uninduced strains. After induction, we lysed the cells using either French Press or Bacterial-Protein Extraction Reagent (B-PER) and collected the lysate for further purification. We attempted to purify Krt31 from the lysate using Immobilized Metal Affinity Chromatography (IMAC). To verify for successfully purified keratin, we ran the purified lysates on 12% SDS-PAGE gels and visualized using silver staining

Learn

Krt31, which is approximately 46kDa, did not appear on our silver-stained gel. In response, we did further research on Krt31 and discovered it is highly hydrophobic and is unlikely to be found in the soluble lysate. Instead, it was likely to be left in the pellet after lysis. Additionally, the low amount of protein visualized on our silver-stained gel led us to western blotting, which is more sensitive and specific.

Test

We set up the next protein purification in a similar manner. However, after lysis using French Press or B-PER, we collected the remaining pellet. We again purified the soluble lysate using IMAC but ran both the resuspended pellet and the lysate on a 12% SDS-PAGE gel. We then visualized the soluble and insoluble fractions using Western blotting with a MouseɑHis primary antibody and a GoatɑMouse secondary antibody.

Learn

No 46kDa bands appeared in either the Rosetta-gami or BL21 induced and uninduced pellets. However, two bands approximately 36 and 34 kDa appeared in all Rosetta-gami and BL21 strains, including the untransformed controls. This is indicative of either contamination with a 6xHis tagged protein, or the non-specific binding of our antibodies to proteins native to both Rosetta-gami and BL21. Additionally, the presence of a faint band around 20kDa in the Rosetta-gami strain induced with 1mM IPTG lane likely suggests non-specific antibody binding to a protein unique to Rosetta-gami, and potentially inducible by IPTG. In any case, the numerous bands that appeared on this blot indicated that 6xHis antibodies were not specific enough to Krt31. The absence of any band corresponding to Krt31 indicates that our Krt31 was not purified from either the lysate or the pellet. Further research into the purification of insoluble proteins yielded a new protocol for the isolation of insoluble proteins from inclusion bodies; this was likely where Krt31 was stored, which explains why our antibodies were not capable of binding its 6xHis tag.

Test

Unsuccessful attempts at purifying the expressed Krt31 protein led us to design a new protocol for the purification of Krt31. A key change to the previous protocols included the addition of a modified solubilization buffer [25 mM Tris-HCl, 9.5 M Urea, 5 mM DTT, 2.5 mM EDTA] which was previously shown to be effective in the purification of Keratin 35 and 857 in a 1:1.5 ratio to the pelleted cells. Additionally, rather than attempting cell lysis with French Press or B-PER, we used sonication to lyse the cells by giving them 10-second shocks, followed by 10-second rests, for thirteen cycles. From here, we performed IMAC with modified buffers to optimally and selectively purify our protein.

Fig 15. | Coomassie-stained 12% SDS PAGE gel showing Krt31 and KerDZ from soluble and insoluble fractions of sonicated Rosetta-gami and BL21. The signal in the first lane at approximately 36 kDa may include Krt31.
Fig 15. | Coomassie-stained 12% SDS PAGE gel showing Krt31 and KerDZ from soluble and insoluble fractions of sonicated Rosetta-gami and BL21. The signal in the first lane at approximately 36 kDa may include Krt31.

Learn

After several attempts at redesigning our Krt31 purification protocol we finally began to see progress. The protein purified from the induced Rosetta-gami pellet showed a strong signal at roughly 36 kDa, slightly smaller than the expected size of Krt31. Still, that the strong expression was found only in the induced strains implies that this may be the isolated keratin. While this is not a definitive result, it clearly shows progress in the right direction. Future studies would involve further refinement of this protocol and could utilize RT-PCR to uncover the exact nature of krt31 expression. See future directions Future Directions
for more information.

References

  1. Ryu, M.-H., and Gomelsky, M. (2014) Near-infrared light responsive synthetic C-di-GMP module for Optogenetic Applications. ACS Synthetic Biology 3, 802–810.
  2. Andersson M, Jia Q, Abella A, Lee X-Y, Landreh M, Purhonen P, Hebert H, Tenje M, Robinson CV, Meng Q, et al. 2017. Biomimetic spinning of artificial spider silk from a chimeric minispidroin. Nat Chem Biol. 13(3):262–264. doi: https://doi.org/10.1038/nchembio.2269
  3. Arndt T, Greco G, Schmuck B, Bunz J, Shilkova O, Francis J, Pugno NM, Jaudzems K, Barth A, Johansson J, et al. 2022. Engineered Spider Silk Proteins for Biomimetic Spinning of Fibers with Toughness Equal to Dragline Silks. Adv Funct Materials. 32(23):2200986. doi: https://doi.org/10.1002/adfm.202200986
  4. UAlberta iGEM. 2023. Parts | UAlberta - iGEM 2023. https://2023.igem.wiki/ualberta/parts (accessed 2024-09-24).
  5. Ben Elhoul M, Zaraî Jaouadi N, Rekik H, Omrane Benmrad M, Mechri S, Moujehed E, Kourdali S, El Hattab M, Badis A, Bejar S, et al. 2016. Biochemical and molecular characterization of new keratinoytic protease from Actinomadura viridilutea DZ50. Int. J. Biol. Macromol.. 92:299–315. doi: https://doi.org/10.1016/j.ijbiomac.2016.07.009
  6. Ben Elhoul M, Zaraî Jaouadi N, Bouacem K, Allala F, Rekik H, Mechri S, Khemir Ezzine H, Miled N, Jaouadi B. 2021. Heterologous expression and purification of keratinase from Actinomadura viridilutea DZ50: feather biodegradation and animal hide dehairing bioprocesses. Environ Sci Pollut Res. 28(8):9921–9934. doi: https://doi.org/10.1007/s11356-020-11371-1
  7. Ishii, D., Abe, R., Watanabe, S., Tsuchiya, M., Nöcker, B., and Tsumoto, K. (2011) Stepwise characterization of the thermodynamics of trichocyte intermediate filament protein supramolecular assembly. Journal of Molecular Biology 408, 832–838.