Team:HokkaidoU Japan/ag43
Ag43 Secretion System
Overview
Antimicrobial-peptides (AMPs) have a wide range of toxicity to microbes, gram-negative / positive bacteria, archaea, fungi and viruses. However, even if we hope that E. coli would produce host-toxic AMPs, it would have harmful results for the host cells owing to their broad spectrum of toxicity. It would be difficult task for E. coli to produce AMPs. If we introduce host-toxic genes into E. coli, they would die of their toxicity. So, in order to yield host-toxic protein without killing host cells, we at iGEM HokkaidoU Japan designed a safe secretion system of AMPs, and using this system, we made E. coli produce thanatin, one of the AMPs from a shield bug, Podisus maculiventris.
Tandem-multimerization of thanatin
A broad-spectrumed antimicrobial-peptide from a shield bug, thanatin, is a short polypeptide composed of 21 amino acids. Thanatin contains 2 cysteine residues and a disulfide bond forms between these 2 residues (Cys11 and Cys18). The disulfide bond between 2 cysteine residues is the core of active thanatin (1). This disulfide bridge stabilizes the anti-parallel β-sheets (Fig. 1). Thanatin is a very stable short polypeptide and has a wide range of activity against bacteria, archaea and fungi. The detailed activation mechanism of thanatin is still unknown. However, it is known that thanatin attacks the membrane of target cells and causes pore-forming.
Fig. 1 Schematic of thanatin structure. Thanatin is a 21-residue short polypeptide. It has a disulfide bond between 2 cysteine residues and anti-parallel β-sheets.
C-terminal residues of thanatin are crucial to its activity rather than N-terminus. Thanatin N-terminus is comparatively tolerant towards modifying; adding or deletion of residues. However, in the case of C-terminus, newly added or removed amino acid residues might be catastrophic, perhaps causing loss or a remarkable reduction of activity. Consequently, ligating some thanatin tandem would make it possible to yield detoxicated form of thanatin without killing the host cells (Fig. 2).
Fig. 2 Image of thanatin multimer. We designed our thanatin fragment with aspartic acid at its C-terminus. N-terminal of aspartic acids are cleaved by endoproteinase AspN. Multimerized thanatin has less toxicity, but, when it is cleaved to monomers, each monomer regains wide-spectrum toxicity.
Then, we need to recover the toxicity of thanatin after expressing detoxicated tandem-repeated thanatin. In order to realize regaining in toxicity, we utilized endoproteinase AspN, which specifically cleaves the peptide bonds N-terminal to aspartic acid residues (purchased at New England Biolabs. website). We added aspartic acid at the carboxy-terminal side of thanatin. Thanatin-Asp-thanatin-Asp-thanatin… multimer is cleavable and monomerizable with AspN. Monomerized thanatin then regains original toxicity. We appreciate Dr. Seiichi Taguchi, a professor of Biosystems Engineering, Hokkaido University, for offering us much information about thanatin function.
Secretion of Thanatin by auto-transporter; Antigen43
An easy and rapid collecting method of expressed protein is desired. It would be laborious task to homogenize cells in order to collect inner protein or polypeptide. To collect proteins and polypeptides easily and rapidly, we utilized Antigen43 (Ag43), which is an auto-transporter protein of Escherichia coli. Usually, Ag43 is surface-displayed and used for autoaggregation of E. coli (2). Autotranspoter proteins, including Ag43, are commonly composed of N-terminal signal peptide following passenger domain (α-domain), and C-terminal β-barrel translocator domain (β-domain) (Fig. 3). Translocator domain is inserted in outer membrane (OM) and forms pathway across the OM. This pathway is necessary for translocation of N-terminal passenger domain from periplasmic space to bacterial surface across the outer membrane. Therefore, passenger domain (α-domain) are displayed on the surface of bacterial cell. Signal peptide is indispensable for translocation of unfolded Ag43 into periplasmic space across the inner membrane. In addition, β-domain of Ag43 contains autochaperon domain, which mediates α-domain folding. After the secretion of passenger domain, passenger is autocatalytically cleft from β-domain, but then non-covalent bond between α-domain and β-domain allows passenger remain on the surface of bacterial cells. Passenger domain is non-covalently bound to translocator domain with β-barrel, so easily released from β-domain with heat treating (60℃) (3).
Fig. 3 Structure of Antigen43. Ag43 contains N-terminal signal peptide, α-domain (passenger), autochaperon domain and β-barrel translocator domain.
Fig. 4 shows the mechanism of Ag43 insertion into OM and secretion of thanatin on the surface of bacterial cell. First, N-terminal signal peptide is inserted into inner membrane (IM) and then, following polypeptide are translocated towards periplasmic space across IM. Next, β-domain is inserted into OM, forming β-barrel structure and finally, thanatin multimer is displayed through β-domain on the bacterial surface. Replacing α-domain with thanatin multimer, remaining signal peptide and β-barrel domain, would allow the display of thanatin multimer on the bacterial surface. So, we inserted thanatin as passenger between signal peptide and β-barrel domain. After secretion of multimerized thanatin, AspN treating cause monomerization of thanatin. Thanatin monomer would regain antimicrobial activity (Fig. 5).
Fig. 4 Image of secretion mechanism of thantin-Ag43 fusion protein from expression to surface-display.
Fig. 5 A method for thanatin-activation. On cleaving surface-displayed thanatin-multimer with AspN, thanatin-monomer will be active form.
Design
We have designed sequence of thanatin with C-terminal Aspartic acid (Asp). Thanatin fragment also contains BamHI restriction enzyme site (N-terminus) and BglII restriction enzyme site (C-terminus) (Fig. 6a). Auto-transporter biodevice (Fig.6b) is composed of PBAD promoter + araC BBa_I0500, N-terminal signal peptide, Ag43 translocator β-domain, and double terminator BBa_K346007. BglII restriction enzyme site is located between signal peptide and β-domain. We built a construct for thanatin detoxication and secretion by ligating thanatin fragment (BamHI / BglII cut) with BglII cut auto-transporter biodevice (Fig. 7).
Fig. 6 Schematics of fragment and plasmid that we used in this project. (a)Thanatin fragment. It contains BamHI and BglII restriction enzyme site. Aspartic acid at C-terminal side of thanatin is important for thanatin activation. (b)Auto-transporter biodevice. This plasmid contains the gene of Ag43, but lacks α-domain so we can insert any fragment between signal peptide (S.P.) and β-domain.
Fig. 7 Image of complete version of construct.
Thanatin secretion biodevice is under control of one of inducible promoters; PBAD promoter. In the absence of L-arabinose, AraC protein acts as repressor and binds to lac operator site in promoter region, therefore expression of downstream gene of PBAD is negatively controlled. However, on adding L-arabinose, this inducer interact with AraC and cause the conformation change of repressor. Finally, AraC is released from promoter region, leading recruiting of RNA polymerase and initiation of transcription. Thanatin - Ag43 fusion protein is inducible by adding L-arabinose. And then, AspN treating of surface displayed thanatin multimer would realize collection of antimicrobial active thanatin.
Experiments
Thanatin multimerization
How can we create thanatin multimer and β domain fusion protein correctly and securely? One of some methods to create thanatin multimer is inserting thanatin fragment into thanatin - Ag43 fusion protein. Once we have got thanatin monomer inserted auto-transporter biodevice, we can make thanatin multimer and Ag43 fusion protein from this plasmid. Thanatin-monomer-inserted auto-transporter biodevice contains BamHI/BglII scar (between signal peptide and thanatin) and BglII restriction enzyme site (between thanatin and β domain). So, after cutting thanatin-inserted auto-transporter with BglII, ligation this plasmid with thanatin fragment (BamHI/BglII cut) would make it possible to create thanatin multimer (Fig. 8). However, in this method, there is a possibility of reverse insertion of thanatin because C-terminal BglII of thanatin might be ligated with C-terminal BglII of auto-transporter.
Fig. 8 Insertion of thanatin into auto-transporter biodevice. (a) Forward insertion. It is desired pattern of insertion. (b)Reverse insertion. This insertion will happen by accident. We cannot insert thanatin only in forward direction selectively.
To prevent thanatin from reverse-insertion, we utilized another method. Fig. 9 shows the outline of alternative thanatin multimerization method.
Fig. 9 Flowchart of thanatin multimerization. (a)First, we amplified 2 types of fragment with 2 different primer sets. (b) Second, we digested 1st fragment (BamHI / SpeI cut) and 2nd fragment (BglII / SpeI cut). And finally, we ligated these 2 fragments. (c) Complete form of thanatin multimer secretion system.
Amplify the auto-transporter plasmid containing 1 thanatin by PCR with 2 primer sets below.
1st primer set: for making 1st fragment
Forward: a primer binding to PBAD region, which has an ability to regenerate BamHI restriction enzyme site.
Reverse: a primer binding to 200 bp-downstream region of suffix.
2nd primer set: for making 2nd fragment
Forward: a primer binding to 100 bp-upstream region of suffix.
Reverse: a primer binding to 200 bp-downstream region of β domain.
Digest the amplified 2 fragments. Cut the 1st fragment with BamHI and SpeI, and cut the 2nd fragment with SpeI and BglII.
Ligate 1st and 2nd fragment and complete the creation of thanatin-multimer.
Repeat these experimental operation to increase the number of inserted thanatin.
Result
Even if we hoped that E. coli produce thanatin, it is painful task for them to do so owing to its broad-spectrum toxicity (1). Produced active thanatin would kill its host cells. In order to yield thanatin, we designed auto-transporter biodevice containing thanatin multimer (monomer (BBa_K1714000), dimer (BBa_K1714005), trimer (BBa_K1714006), tetramer (and BBa_K1714007)) and introduced them into E. coli. After enough cultivation (10 hours) in liquid LB medium, we resuspended cells in LB medium to optical density at 600 nm (OD600) =0.1 (total volume of LB: 2ml). Before initiating cultivation, 200 µL of L-arabinose (final concentration; 0.1%) was added for PBAD induction. Then, we measured temporal change of OD600 every hour over 10 hours (until cultured cell reached stationary phase). We also determined OD600 of E. coli containing auto-transporter biodevice alone as negative control. This control is for examination of the influence of expression of Ag43. In addition, we prepared 5 types of E. coli above without L-arabinose adding in order to examine whether gene expression is correctly induced. (Fig.10).
(a)Growth curve of transformant expressing thanatin-monomer fused with Ag43
(b)Growth curve of transformant expressing thanatin-dimer fused with Ag43
(c)Growth curve of transformant expressing thanatin-trimer fused with Ag43
(d)Growth curve of transformant expressing thanatin-tetramer fused with Ag43
(e)Growth curves of 5 types of transformants expressing thanatin fused with Ag43
Fig. 10 Growth curves of E. coli expressing different number of thanatin tandem multimer. We measured temporal change of OD600 and draw growth curve. (a) Thanatin-monomer, (b) Thanatin-dimer, (c) Thanatin-trimer, (d) Thanatin-tetramer, (e) Growth curves of 5 transformants. As negative control, we also draw growth curve of E. coli containing only Ag43 β domain.
Fig. 10 shows that as the number of thanatin repeat increases, growth of E. coli is more inhibited. Growth inhibition is monitored 2 hours after induction on every transformants. Theransformant expressing thanatin-monomer reaches stationary phase about 11 hours after initiation of cultivation. As length of thanatin polymer increased, time to reach stationary phase were shortened. Thanatin-dimer, trimer, tetramer take 10, 7, 3 hours to be grown to stationary phase respectively.
In order to make sure that transformed E. coli secrete thanatin towards outside of the cell, we carried out an operation to collect thanatin. We needed to produce thanatin as much as possible for determination of thanatin’s activity. So, we cultivated transformant with thanatin-tetramer and Ag43 β subunit fusion protein. Collection method is as below;
Resuspend the cells in LB medium to OD600 =0.1 (total volume: 2 mL) after 10-hour-cultivation. At the same time, induce the gene expression by adding L-arabinose.
3 or 6 hours after induction, centrifuge at 5000 rpm for 1 minute and remove the cells.
Remove the supernatant, add 5 µL of AspN with 100 µL of reaction buffer and 100 µL of DW, resuspend it and then, incuvate the mixture at 37℃ for 30 minutes. As negative control, we prepared mixture of 1 µL of AspN and 20 µL of reaction buffer. If AspN alone had toxicity, the result of negative control would indicate the toxicity of endoproteinase AspN.
Centrifuge the suspension at 5000 rpm for 1 minute again. If E. coli secreted thanatin, we would be able to collect active thanatin monomer from the second supernatant.
After collecting operation, we needed to examine that if collected supernatant really include thanatin and have toxicity. In order to determine its toxicity, we carried out Minimum Inhibition Concentration (MIC) test. MIC test is one of the most general examination of an antimicrobial activity against microorganism. The lowest concentration of antimicrobial agent that inhibits visible growth of microorganisms is defined as MIC. In order to evaluate thanatin’s activity, we concentrated supernatant containing thanatin 5-fold with centrifugal concentrator and prepared dilution range of the supernatant including thanatin from 1/2 to 1/32 (total volume is 20 µL each) with 96-well microtiter plates. Wild-type E. coli (DH5α) was cultivated in LB medium until OD600 reached 0.4 and diluted 100,000 folds. After that, 80 µL of the suspension was added to dilution range of supernatant. We incubated the cells at 120 rpm, 30℃ for 18 hours. After incubation, the OD600 was determined with a microtiter plate reader. The result is shown in Fig. 12.
Fig. 11 Antimicrobial activity assay of thanatin. Growth inhibition of E. coli DH5α was assayed by adding 20 µl of serial dilutes containing thanatin into 80 µL of PB containing bacterial cells (cultivated to OD600 values of 0.4 and diluted 100,000 folds).
The minimum inhibitory concentration (MIC) of collected thanatin lies between 5-folds and 2.5-folds dilution range (Table 1). Take the MIC value of purified thanatin (between 1.6 and 3.1 µM) into account, it is indicated that the concentration of collected thanatin would be about 7.8 µM.
Table1 Antimicrobial activity assay.
Conclusion
Although 2 growth curves of transformants (arabinose induced or not) slightly differs, it is concluded that the expression of Ag43 does not affect the growth of transformants (Fig. 10). In the absence of L-arabinose, the growth curves of different thanatin-copy-number variants resemble that of transformant with Ag43 alone. We succeeded in inhibiting the expression of thanatin-Ag43 β subunit fusion protein with PBAD promotor. Thus, we concluded that we succeeded in inactivation of thanatin by multimerizing and fusing with Ag43. It is indicated that we will be able to cultivate trsnsformants under safe condition and make them produce thanatin at the most appropriate point for thanatin mass-production.
As shown in Fig. 10, as the number of thanatin-repeat increases, the growth of E. coli is more inhibited. At first, we expected that tandem-multimerization would detoxicate thanatin because one thanatin or Ag43 β subunit connected to C-terminus of neighboring thanatin inactivate its toxicity. However, this result indicates that tandem-multimerized thanatin is host-toxic. In addition, the more the number of thanatin is increased, the more its toxicity is strengthened.
There is a possible reason why toxicity is strengthened as copy-number of thanatin is increased. It is possible that modifying of C-terminus of thanatin does not completely detoxicate it, so each fragment of thanatin has toxicity to some extent. Because of increase of thanatin’s copy number, surface-displayed part of fusion protein is lengthened. Consequently, the probability that thanatin will interact with outer membrane is increased.
Although surface-displayed thanatin remains to have host-toxicity in some extent, it is possible that transformant is grown without intensive host-killing, compared to transformant with thanatin without modification.
The result of MIC test indicates that we succeeded in cleaving thanatin by AspN, collecting thanatin from bacterial suspension, and to determining the antimicrobial activity of collected thanatin.
Our conclusion is that expression and secretion of thanatin to bacterial surface, cleavage and activation of thanatin by endoproteinase AspN, and determination of thanatin’s antimicrobial activity succeeded. Transformants secreting thanatin grew up without extinction. Multimerization and fusion with Ag43 is one of the useful methods to produce host-toxic antimicrobial peptide. We revealed that thanatin that is once inactivated certainly regain toxicity against microbes after treated by AspN. We estimated the concentration of thanantin using the result of purified thanatin’s MIC test. A possibility of mass-production of thanatin is implied in our experiment. What we have to do next is to improve the yield and purify the thanatin. There would be plenty of room for improvement. We will try harder to establish methods of thanatin mass-production. Perhaps, thanatin will be used as an alternative of antibiotics in the future.
Reference
- Seiichi Taguchi, Kanako Kuwasako, Atsushi Suenaga, Miyuki Okada and Haruo Momose, Functional Mapping against Escherichia coli for the Broad-Spectrum Antimicrobial peptide, Thanatin, Based on an In Vivo Monitoring Assay System, J. Biochem. 2000, 128:745-754
- van der Woude, Marjan W Henderson, Ian R. Regulation and function of Ag43 (flu). Annu. Rev. Microbiol. 2008, 62:153–69
- Glen C. Ulett, Jaione Valle, Christophe Beloin, Orla Sherlock, Jean-Marc Ghigo and Mark A. Schembri. Functional Analysis of Antigen 43 in Uropathogenic Escherichia coli Reveals a Role in Long-Term Persistence in the Urinary Tract Infect. Immun. 2007, 75(7):3233. DOI: 10.1128/IAI.01952-06.
