Intro

We further developed "Dr. E. coli": our project of iGEM 2010. Last year, we showed that Type 3 Secretion System (T3SS) works in E. coli by injecting GFP into RK13 cells. We thought this system can be applied to direct reprogramming of somatic cells among many other things.

See here for details.

This year we repeated previous competitions experiment, one more time showing that GFP can be really injected into a target cell with it. And then, as a second step, we tested T3SS performance and tried to make it more convenient. For this purpose we designed a plasmid backbone which can instantly produce ready-to-inject fusion proteins from ordinary biobrick part. Using it, we tried to further characterize this system by injecting characteristic proteins.

Type 3 Secretion System

T3SS is a system of pathogenic gram-negative bacterium such as Salmonella, Yersinia and EPEC (entero pathogenic E. coli). Using this system bacteria can inject whole protein molecules through a syringe like organelle named T3S Apparatus. During iGEM 2010 we found that E. coli with a part of Salmonella genome library expresses T3SS functionally. This presented opportunity to work with the amazing machinery without involving pathogenic bacteria. Read more

Plasmid Backbone for protein injection

We developed plasmid backbone which can attach tags needed for secretion and various other functions to a chosen protein biobrick. This can be used for big scale screening of various protein domains for their inject-ability. Backbone

Figure1. A backbone under constitutive promoter(pTetr). Has SlrP as a injection signal, GSK tag, Bsa I Cloning Site. Desired protein are inserted into the cloning site.

Last year we used T3SS to inject GFP to mammalian cells. This year we wanted to explore T3SS limits. We submitted it to a injectable protein screening. We made a small library of proteins with distinguished structures which were chosen from 2011 Biobrick distribution.

Assembling each protein, injection signal and tag would have been a laborious task which we didn't want to endure. So ready-to-inject backbones was proposed(Fig.1).

We dsigned Bsa I Cloning Site to facilitate quick assembly of proteins to inject. Thus it's name ready-to-inject backbone. All inserts must be PCRed with specific primer to remove stop codon. Of course all inserts will be in-frame.

See here for details.

Another problem was how check if protein was injected. To solve it we used a distinct property of Glycogen Synthase Kinase 3 β, it is phosphorylated only in eucaryotic cells. This phosphorylation can by detected with phospho-specific antibodies. Thus we now can easily detect if protein was injected in eucaryotic cells. This property is exhibited by first 13aa of this protein[1] which makes it very small tag. Because of it's small size the interference on tag proteing should be minimu.

See here for details.

We combining these futures and constructed ready-to-inject bakcbone. SlrP is an injection signal, without it the protein cannot be secreted. GSK is a tag, by detecting phosphorylation of it you can distinguish whether it has been it eukaryotic cell. For us it is an evidence of successful injection. Bsa I Cloning Site is used for inserting various BioBrick while retaining the whole constructs BioBrick properties. The whole proteins is under control of pTetr constitutive Promoter.

We used ready-to-inject backbone eight proteins representing different structures and functions to see which get injected. This was are shot to try and characterise T3SS a little bit further.

See here for details.

Bsa I Cloning Site

Bsa I Cloning site is unique in a sense that you can clone BioBrick into a middle of a construct and still retain the properties of biobrick. We used it to construct our backbones for T3SS characterization. Bsa I cloning site is valuable part when you need to screen vast libraries of proteins. It designed that inserted biobrick would be fused to preceding signals.

Bsa I restriction enzyme is in distinguish group of enzyme which cutting site is different from recognition site. Unlike EcoR I or Pst I, Bsa I regognizes GGTCTC sequence but cuts the sequence 1 base further ahead of it. Which results in a 5 prime 4 base overhang(Fig).

5'...GGTCTCN^.......3'
3'...CCAGAGNNNNN^...5'

You can manipulate the sequence of overhang as you like. By if you construct sequence GGTCTCNAATTN you can make it to ligate with EcoR I digested strand. Also long as NAATTN won't become GAATTG it wouldn't not be digested by EcoR I and that’s the beauty of it.

Of course there are other restriction endonucleases that exhibit same properties but Bsa I was the cheapest.

However there are some limitations Bsa I is not an official biobrick Restriction enzyme so you have to screen each part for Bsa I recognition sequences. pSB1A3 has one in Ar locus which requires silent mutation or avoiding using it. Thus fur we didn't encounter other BioBricks containing it.

And because only backbone has to be digested by Bsa I you don't have to worry about inserts having Bsa I sites.

We designed a cloning site which when digested with Bsa I will produce Not I like overhang and Spe I like overhang (Fig). Which will ligate to Not I and Spe I but won't be digested after.

         Bsa I    Not I'           Spe I'   Bsa I
  
5'...GG GGTCTC A^GGCC ….........^CTAG A GAGACC...3'
3'...CC CCAGAG T CCGG^TCCGGCCGCT GATC^T CTCTGG...5'

5'...GG GGTCTC A                 CTAG A GAGACC...3'
3'...CC CCAGAG T CCGG                 T CTCTGG...5'

We dealt with TAG stop codon at Xba I site by inserting a mutation and destroying it. Relatively easy step using just primers and PCR.

New BioBrick standard would not by realistic as making several cloning sites with the same restriction site would defeat the purpose. Using Bsa I Cloning Site as a BioBrick will open the door to insert new BioBricks in various places of the construct. Using other Bsa I like restriction sites you can create multiple insertion sites in the construct.

GSK tag

Glycogen Synthase Kinase 3β (Fig) is known to be phosphorylated by several enzymes in eukaryotic cell. We used first 13 amino acid sequence to construct a tag (GSK tag)^[1] which phosphorylation state could be detected. 9th amino acid, serine is phosphorylated in eukaryotic cell(Fig). There are antibodies which bind to only phosphorylated GSK tag using them it is possible to distinguish whether it has been it eukaryotic cell. So you can see proteins which were injected into cell and which were not. This was a vital ingredient in our experiments.

GSK tag was constructed by Julie Torruellas Garcia, Gregory V. Plano et al. We removed present Spe I site in the sequence by silent mutation.

Translation: M   S   G   R   P   R   T   T   S-p  F   A   E   S
Original   :ATG AGT GGT CGC CCT CGC ACT ACT  AGT TTC GCT GAA AGT
rm Spe I    :ATG AGT GGT CGC CCT CGC ACT ACA* AGT TTC GCT GAA AGT 

Phosphorylated Serine is shown as S-p.

GSK tag can be added to N terminus, C terminus^[1] and anywhere in between^[2] of the protein. We opted to insert it between SlrP secretion tag and the protein we wonted to inject. To SlrP is required to be on CorN terminus. And insertin on C trminus would have required some costly primers.

Using non-phosphospecific antibodies it is possible to check the total amount of created tag protein. Comparing it with the injected protein you can determine the efficiency of the injection.

By comparing the mass of the protein with GSK tag it is also possible to see if it had been modified in eucaryotic cell. It can be used alongside of TEV site. And can be used as proof TEV protease activity.

Investigation of T3SS-injectable proteins

Here we will discuss the structure of proteins which are injected and which are not. We tried eight different proteins: mnt repressor, Gal4, RFP, GFP, Cre DNA recombinase, (CCR5) transmembrane, LacI and Luciferase. All were chosen from biobrick distribution which shows their significant importance for iGEM.

Our main concern was not with the size the protein but its stability. Previous research show that proteins like Zinc-Finger are were stable and couldn't be injected. Stability prevents unfolding by T3SS chaperons. Our asortment includes Gal4 which is representative of stable proteins.

We showed that GFP can be injected into eucaryotic cells by confocal laser microscope imaging. Thus in can serve as a control. Next is RFG, a fluorescent protein bu with different structure from GFP. MNT represor

References

1. Julie Torruellas Garcia, Franco Ferracci, Michael W. Jackson,1 Sabrina S. Joseph, Isabelle Pattis, Lisa R. W. Plano, Wolfgang Fischer, and Gregory V. Plano. 2006. Measurement of Effector Protein Injection by Type III and Type IV Secretion Systems by Using a 13-Residue Phosphorylatable Glycogen Synthase Kinase Tag. Infect Immun.Vol.74:5645-57. [http://www.ncbi.nlm.nih.gov/pubmed/16988240 PubMed]
2. JWensheng Luo and Michael S. Donnenberg. 2011. Interactions and Predicted Host Membrane Topology of the Enteropathogenic Escherichia coli Translocator Protein EspB. J. Bacteriol.Vol.193:2972–80. [http://www.ncbi.nlm.nih.gov/pubmed/21498649 PubMed]