Team:Beijing 4ELEVEN/Results
According to our project design, the experiments mainly consist of two parts: Part I: Verification of AMP production and antimicrobial potency Part II: Production of adhesive and cohesive proteins
Part I：Verification of AMP antimicrobial potency and production Through article research, we selected 5 Antimicrobial peptides as potential ingredients of our product：GDP20 (RWRRWYRKFCR), CEN1HC-Br (FKKTFHKVSHAVKSGIHAGQRGCSA LGF), snake cathelicidin-BF (KFFRKLKKSVKKRAKEFFKKPRVIGVSIPF), tridecapepetide (CFRVCYRGICYRC), human cathelicidin LL-37 (LLGDFFRKSKEKIGKEFKRIVQRIKDFLRNLVPRTE). We purchased these peptides from synthetic companies, tested their antimicrobial potencies on E.coli MG1655 and P.acnes, then expressed them in Pichia pastoris GS115.
We tested the antimicrobial potencies of the 5 AMPs through two methods: adding drops of AMP solutions onto plates inoculated with E.coli MG1655 and P.acnes and observing bacteria growth inhibition zones; and adding AMP solutions into liquid culture inoculated with the bacteria and measuring OD600.
1.1 Plate verification To begin with, we tested the antimicrobial potencies of CEN1HC-Br and Snake cathelicidin-BF on E.coli MG1655 because they were the first AMPs synthesized.
Figure 1. CEN1HC-Br efficiency to E.coli MG1655 (plate).
Figure 2. Snake cathelicidin-BF efficiency to E.coli MG1655 (plate).
CEN1HC-Br and Snake cathelicidin-BF's effects of killing E. coli MG1655 are weak as barely any difference can be seen between the plates with AMPs and those with water (Figures 1&2). However, in higher concentrations, significant inhibition zones of bacteria growth indicate the strong antimicrobial potencies of the two AMPs. When pieces of paper soaked with the AMP solutions are placed onto the plates, inhibition zones did not exceed the area covered by the paper. Therefore, the AMPs' range of bacteriacidal effect is limited within areas of direct contact with bacteria, which is good because it means that our product would not eliminate harmless microbes outside of acne infected regions.
Based on previous results, we tested the antimicrobial potencies of GDP20, tridecapepetide, and human cathelicidin LL-37 on E. coli MG1655 only in high concentrations.
Figure 3. GDP20, Tridecapepetide, Human cathelicidin LL-37 efficiency to E.coli MG1655 (plate).
All GDP20, CEN1HC-Br, and human cathelicidin LL-37 demonstrated significant E. coli killing effects within areas of direct contact with the bacteria (Figure 3). Plates with tridecapepetide and human cathelicidin LL-37 had E. coli growing on the edges of the inhibition zones, indicating the effects of these AMPs are weaker.
We then verified the 5 AMPs' effects of killing P. acnes with the same methods.
Figure 4. AMPs efficiency to P.acnes (plate).
All 5 AMPs proved effective within direct contact with P. acnes. Determining from the size of inhibition zones, CEN1HC-Br is least effective in killing P. Acnes. The inhibition zone of tridecapepetide has a white background because the amino acids of the peptide's composition are strongly hydrophobic.
1.2 OD600 Verification We first verified the antimicrobial concentrations of CEN1HC-Br and Snake cathelicidin-BF as they are synthesized earliest. The AMP solution concentration gradient was initially set as 0, 4, 8, 12, 16mg/L, and tested their abilities to kill E. coli MG1655. After discovering significant difference between the results of the 2 AMPs, we adjusted the gradients of their concentrations. In later verification, we collected samples once in every 2 hours and measured OD600 for identification of antimicrobial potency.
Figure 5. CEN1HC-Br efficiency to E.coli MG1655 (OD600).
Figure 6. Snake cathelicidin-BF efficiency to E.coli MG1655 (OD600).
Though both are effective in inhibiting E. coli growth, CEN1HC-Br has weaker efficiency than Snake cathelicidin-BF. (Figures 5&6). E. coli reproduction nearly stopped when Snake cathelicidin-BF solution reached a concentration of 4 mg/L. However, it was almost the same as that in control group in CEN1HC-Br solution with concentration 16 mg/L. Therefore, in later verification we rearranged the concentration gradients of CEN1HC-Br and Snake cathelicidin-BF to 20, 24, 28, 32 mg/L and 0.5, 1, 2, 3 mg/L. Snake cathelicidin-BF proved to be greatly effective at 1mg/L while CEN1HC-Br's effect remains insignificant even at 32mg/L. Additionally, the efficient concentration of AMP solutions are lower in liquid culture because bacteria suspended in the solutions have better contact with AMPs.
We set the concentration gradient as 2, 4, 8, 12，16 mg/L for verification of GDP20, tridecapepetide, and human cathelicidin LL-37's antimicrobial potency on E. coli MG1655.
Figure 7. GDP20, tridecapepetide, human cathelicidin LL-37 efficiency to E.coli MG1655 (OD600)
The effects of all 3 AMPs peaked at 8 hours after cultivation. GDP20 and tridecapepetide proved to be most efficient at 16mg/L while human cathelicidin LL-37 displayed strongest inhibition at 4mg/L (Figure 7). It can be inferred that tridecapepetide has the strongest effect among the 3 AMPs.
We also tested the AMPs' abilities to kill P. acnes in liquid culture. Since P. acnes has a slow growth rate, we only measured OD600 after 48 hours of cultivation.
Figure 8. AMPs efficiency to P.acnes (OD600)
Tridecapeptide proved to be most efficient in killing P. acnes, followed by GDP20 and CEN1Hc-Br.
Thus, all 5 AMPs are effective in eliminating E. coli and P. acnes.
We selected Pichia pastoris GS115 as chassis, pPIC9K as vector, and AOX1 as induced promoter for AMP synthesis. First, five antimicrobial peptides were inserted into the pPIC9K vector by restriction enzyme ligation, then, the recombinant plasmid was transformed into P. Pastoris GS115, finally, the antimicrobial peptides were produced by methanol induction. The product of this AMP expression system is secreted through α-factor secretion signal. AMP antimicrobial potency is measured by the effect of bacteria growth inhibition of their fermentation products.
2.1 pPIC9K-AMPs plasmid construction We selected EcoR I and Not I on pPIC9K as acting sites and acquired the AMP sequences through primer-annealing. After digestion of vector and sequences, T4 ligase is used to connect the segments. The resulting plasmid is transformed into E. coli DH5α, which is then verified through PCR and sequencing.
Figure 9. PCR verification results of pPIC9K-AMPs
We transformed the constructed plasmids into Pichia pastoris GS115. In this process, the exogenous genes would integrate itself into the genome of Pichia pastoris. We then extracted the recombinant genome and verified it by PCR using the original genome as template. Finally, we sequenced the recombinant genome.
Figure 10. PCR verification results of recombinant yeast
2.2 AMP fermentation production and verification The fermentation of Pichia pastoris that are capable of expressing AMPs is induced by methanol (method of fermentation can be found in Methods page). The supernatant of the fermentation product is the AMP solutions. We tested the solutions' antimicrobial potency through plate and OD600 verifications.
2.2.1 Plate verification We added drops of fermentation products onto plates inoculated withE. coli MG1655 and P. acnes and observed inhibition zones. Pieces of paper soaked with these products were not placed onto plates because prior experiment already proved the range of function of AMP solutions.
Figure 11. Fermentation products efficiency to E.coli MG1655 (plate).
Figure 12. Fermentation products efficiency to P. acnes (plate).
It can be inferred from the results that the fermentation products' effects of killing E. coli MG1655 and P. acnes are not significant. Only a certain amount of bacteria within contact of the products were eliminated and the inhibition zones are quite unclear due to low volume of AMPs dissolved in to solutions.
2.2.2 OD600 verification We added fermentation products into liquid culture and verified their effects on E. coli MG1655 and P. acnes growth.
Figure 13. AMPs Fermentation products efficiency to E. coli MG1655(2.5% & 25%)
The results show that all fermentation products are effective in killing E. coli MG1655. The inhibition is not significant at concentration of 2.5%, but becomes strong at 25%. CEN1HC-Br and human cathelicidin LL-37 are most efficient probably because larger amounts of them are dissolved in their fermentation products.
Figure 14. AMPs Fermentation products efficiency to P.acnes(25%)
We verified the fermentation products' efficiency in eliminating P. acnes at 25%. Tridecapepetide and human cathelicidin LL-37 proved most effective.
Part II: Ahesive and cohesive protein production Since AMP solutions have strong fluidity, they cannot remain at a certain region for long durations to function. Therefore we decided to help the AMPs of our product stay on the skin with protein adhesives. These adhesives consist of two parts：adhesive proteins, which make them sticky, and cohesive proteins, which maintain their form. Going through the project, SticKit, of team 2019 Greatbay_SCIE, we selected fp1-mfp5-fp1, CsgA-mfp5, and CsgA-mfp5-mfp5 recombinant proteins.
However, the adhesive proteins' function require DOPA modification by tyrosinase, which can be achieved through two means: expressing adhesive proteins in Picha pastoris and tyrosinase intracellularly, then conduct modification extracellularly; and expressing both adhesive proteins and tyrosinase in E. coli, so that the secretion would be able to function right away. We explored both processes.
We selected E. coli BL21 as chassis for tyrosinase production and improved 2019 Greatbay_SCIE's tyrosinase expression system by constructing a RBS library. At the same time, we transformed the adhesive and cohesive protein sequences into Pichia pastoris GS115.
1.1 Establishment of RBS library to optimize the yield of tyrosinase We improved 2019 Greatbay_SCIE's tyrosinase expression system by modifying its 6 base pairs of its RBS sequence TAAGTATAAGNNNNNNATAT. 15 RBSs are taken into our RBS library through RBS calculator (https://salislab.net/software/login) verification of the modified RBSs.
After construction of our RBS library, we verified strains with sfGFP sequences. The results demonstrate the increase of florescence which follows the increase in inducer concentrations and peaks at IPTG concentration of 0.5 mM. It can also be inferred that our RBS library contains a wide range of RBSs with distinct potency.
Figure 15. Fluorescence analysis of sfGFP expression with RBS library
We also verified the expression of mTyr-CNK with different RBSs at inducer concentration 0.5mM, 25℃, and induction time for 20 hours. Verified bacteria was then put through ultrasonication and protein purification, and SDS-PAGE and BCA verification was done on the purified proteins. The results indicate correct bands of purified proteins in the SDS-PAGE gel, and that the production of proteins controlled by different RBSs reached 2.94 mg/mL, which was more than 4 times higher than the control group of 0.70 mg/mL.
Figure 16. SDS-PAGE gel analysis of mTyr-CNK protein expression with RBS library
Figure 17. BCA protein assay results ofmTyr-CNK protein expression with RBS library
1.2 Production of adhesive and cohesive proteins in Pichia pastoris We chemically transformed plasmids containing fp1-mfp5-fp1 and CsgA-mfp5-mfp5 into Pichia pastoris, and controlled the synthesis of protein production with methanol. Adhesive and cohesive proteins in the fermentation supernatant was verified by SDS-PAGE electrophoresis.
However, we did not observe any correct bands in gel electrophoresis results--no bands appeared in the results of fp1-mfp5-fp1, and a few large bands appeared in that of CsgA-mfp5-mfp5. This might be because that plasmids are integrated into the yeast genome with multiple copies, producing proteins that fold together.
Figure 18. SDS-PAGE gel analysis of supernatant samples during fp1-mfp5-fp1 and CsgA-mfp5-mfp5 fermentation
We constructed the adhesive/cohesive protein and tyrosinase co-expression system in E. coli BL21. Analyzing the project of 2019 Greatbay_SCIE, we found that their intracellularly modified co-expression system involved only the inducible promoter T7-LACI, which made the expression of the two proteins impossible to be regulated separately. Therefore, we aimed to change the induction expression system of tyrosinase and reconstruct the co-expression system.
2.1 Construction of a new induced system During article research we found in the work of Thomas et al an inducible expression system, EilR-PJExD, controlled by cationic dyes. We then inserted sequences of this system into pSC101 vectors and measured its induction with sfGFP verification.
Figure 19. Fluorescence analysis of sfGFP expression with EilR-PJExD inducible system
We set the inducer concentration gradient of cationic dyes AO (Acridine orange), MG (Malachite green)，PB (Pyronin Y), MB (Methylene blue), VB (Victoria blue R), and CV (Crystal violet) as 0, 0.05, 0.1, 0.5, 1, 5, 10, 20, 50 μM. Since the article stated that NR (Neutral red) requires higher concentration, we set its concentration gradient as 0, 0.1, 0.5, 1, 5,10, 20, 50, 80, 100 μM. It can be inferred from the results that CV and AO induced strong sfGFP florescence at lowest concentrations. Therefore, we selected CV as the inducer of our expression system.
Then, we inserted mTyr-CNK sequences into EilR-PJExD, and induced expression with CV. The system displayed great efficiency as 1.44mg/mL tyrosinase is produced with the induction of only 1μM CV.
Figure 20. SDS-PAGE gel analysis and BCA assay of mTyr-CNK expression with EilR-PJExD inducible system
2.2 Construction of the co-expression system We constructed our co-expression system by transforming the mTyr-CNK and adhesive/cohesive protein expression plasmids into E. coli BL21. Induction verification is done on bacteria with the co-expression system as 2 inducers are added. However, cell growth was severely supressed: large amount of cells with single plasmids can be collected in 20 hours after induction, but 5 days are required for cells with the co-expression system to reproduce to an amount capable of sufficient protein purification, and little adhesive/cohesive proteins can be found in the products. Therefore, we rearranged the induction time of the two induction systems through modelling so that our cells could have some time to grow before induction. According to our model, sufficient amount of cells can be collected in 24 hours after induction, and both proteins would be expressed with similar efficiency.
Figure 21. SDS-PAGE gel analysis of co-expression system
Reference 1. Choi, Y. S., et al. (2012). In vivo modification of tyrosine residues in recombinant mussel adhesive protein by tyrosinase co-expression in Escherichia coli. Microb Cell Fact 11:139. 2. Do, H., et al. (2017). A tyrosinase, mTyr-CNK, that is functionally available as a monophenol monooxygenase. Sci Rep 7(1):17267. 3. Hwang, D. S., et al. (2007). Practical recombinant hybrid mussel bioadhesive fp-151. Biomaterials 28(24):3560-8. 4. Kim, E., et al. (2018). Microbially Synthesized Repeats of Mussel Foot Protein Display Enhanced Underwater Adhesion. ACS Appl Mater Interfaces 10(49):43003-43012. 5. Liang, C., et al. (2015). Protein Aggregation Formed by Recombinant cp19k Homologue of Balanus albicostatus Combined with an 18 kDa N-Terminus Encoded by pET-32a(+) Plasmid Having Adhesion Strength Comparable to Several Commercial Glues. PLoS One 10(8):e0136493. 6. Thomas L. Ruegg, Jose H. Pereira, et al. (2018). Jungle Express is a versatile repressor system for tight transcriptional control. NATURE COMMUNICATIONS, 9:3617 7. Zhong, C., et al. (2014). Strong underwater adhesives made by self-assembling multi-protein nanofibres. Nat Nanotechnol 9(10):858-66.
