See the slides for some brief info.
Express monomeric streptavidin on E. coli cell surface
Scaffold 1: GPI anchor
Scaffold 2: Beta autotransporter
Scaffold 3: Lpp-OmpA
- Where can we get the DNA expressing the scaffold protein. I need to look more into this, but I don't think we can just order it.
- Where do we get streptavidin DNA (smaller problem)
- What constraints do we have to worry about with streptavidin? In the original and later autotransporter papers, the passenger protein was mutated to lack cysteines in order to prevent disulfide bond formation in the periplasm: tertiary structure prevents passage of the protein through the beta barrel.
- How much time to get a first experiment up and running? This depends on how much work we have to do with the scaffold protein sequence and how fast we can get them (likely a long time).
- Will we be able to correspond with experts on this system? The founder of autodisplay, Joachim Jose, has not responded to some intial questions about autotransporter.
Demonstrate control of a nucleic acid interface between proteins and the cell surface
(Maybe Un-, but more likely..) fortunately, another group has done work like this. We'll expand on the "adaptamer" concept of Tahiri-Alaoui et al. (12000850). In this paper, the group evolved one aptamer which bound streptavidin, then spliced it to a CopA RNA. Then they spliced a complementary CopT RNA to a CD4 aptamer. Following CopA/CopT binding, the resulting construct could bind both CD4 and streptavidin. Neato. CopA and CopT bind each other by a kissing complex between two loops; mutations have been introduced before, so looking at those studies might be a good starting places for adjusting the dissociation rate (16199086).
The reason they go through the trouble of using CopA and CopT is that they maintain stable secondary structures; the group reports that when you just have any old complementary strands, the binding affinities of the aptamers goes down considerably. We may be able to design our own RNA-RNA interfaces to have well-defined secondary structure, enhancing our ability to control dissociation rate.
Additionally, we want to make one of these aptamers bind to the E. coli cell surface. There is a serious dearth of known aptamers that do this. The only report about one is 15541352; their aptamer binds LPS, a major component of the E. coli outer membrane, so it has a lot of targets. They don't give a Kd value, maybe because it didn't matter too much with so much LPS. Hope we can get our hands on the sequence..
One thing to consider: in the absence of aptamers that bind to the cell surface, we might as well try it with aptamers that bind proteins in vitro as done in 12000850. Why worry about cells for now? We're interested in a proof of principle.
We'll attempt to build an adaptamer that can bind thrombin and streptavidin simultaneously. More details to come. Oligos to be used here
- What aptamers exist that bind the E. coli cell surface??? (and have public sequences! The sequence of the aptamer that binds LPS is not reported in the paper.)
- Is there any chance that we'll be able to evolve our own aptamer this summer? Online resources point toward 'no' but see last question.
- What other secondary structures might work as interfaces? If we design our own, the online programs mfold and Vienna RNA may be useful. Otherwise, we'll have to go with known complementary structures. Do the faculty members have any suggestions for structures?
- Time and reliability: Hopefully short! Aptamer sequences are relatively short, so we'll likely be able to order them instead of asking groups for them. Then we'd try tons of intermediate interfaces, which is just a matter of ordering more oligos.
- Will people respond to email? I've emailed David Liu about evolving aptamers and Sulatha Dwarakanath (author of 15541352) about aptamers that target the E. coli cell surface to no response. It would be quite helpful to hear their thoughts.
A couple of related projects that we haven't really researched..
- Having E. coli export an aptamer outside the cell
- Using aptamers to actually stimulate a response on the cell surface
Streptavidin Fusion Protein
I have contacted Filiz Aslan at DFCI, and she has agreed to give me clones of her single-chain dimeric streptavidin (Aslan et al., 2005). I'm in the process of having the MTA (Material Transfer Agreement) approved by Harvard's Office of Technology Development. Once the approval has gone through, we can pick up the clone!
I've also gotten a reply from Dr. Takeshi Sano, also at DFCI. He has pTSA13-wt, which is wild-type streptavidin clone. Another MTA will have to be approved.
I have word from Dr. Charles Earhart that his colleague Dr. George Georgiou "has been distributing LppOmpA vector plasmids and has a large collection." I've emailed Dr. Georgiou several times without reply. I tried emailing his administrative assistant Donna, and she replied she would pass it along to someone in the lab. Closer? Maybe.
We may be able to make Dr. Georgiou's construct ourselves. It is a fusion protein N - Lpp aa's 1-9 - OmpA aa's 46-159 - surface protein - C. E. coli K12 genome and BioBrick primers for the full OmpA gene (start through stop) have been ordered.
Perry 14:45, 30 June 2006 (EDT)
I got a reply from Mark Howarth (Howarth et al, 2006) from the Ting Lab at MIT. He kindly offered to give me three clones of streptavidin: wild-type, wild-type + His6 tag, non-biotin-binding mutant (N23A, S27D, S45A), in pET21a plasmid; I went to MIT and picked them up.
Perry 13:25, 3 July 2006 (EDT)
GPI anchors and autotransporter are almost dead. No group ever got back to us about supplying plasmids. The autotransporter vector was derived from an enteropathic E. coli (strain 2787 [O126:H27]). Genomic DNA of another enteropathic E. coli, O157:H7 str. Sakai, is available for purchase. However, the AIDA-I gene differs considerably between these two strains, so it may not be worth attempting to PCR out.
We've received the short DNA sequences that we'll be using. We are waiting for purified streptavidin protein to come in. The main question is what we will use in our gel shift assay. We cannot use 32P, so our alternatives are:
Invitrogen EMSA kit (cat. E33075) $224. enough for 10 minigel assays.
Roche DIG gel shift (cat. 03353591910) $453. enough for 20 labeling rxns, 200 gel shift rxns, 20 blots, and 20 control rxns. Don't know what 'gel shift reactions' are. Also need nylon membranes, DIG wash and buffer set.
LightShift Chemiluminescent EMSA Kit (prod. # 20148) $349. Enough for 100 binding reactions (??), detection reagants for 800cm^2 membrane.
The first kit uses fluorescent dyes. We would need to buy a $600 filter for the gel imager to use it. The second two involve blotting membranes. Since this project will probably involve trying out a few types of aptamer sequences and many types of intermediate sequences between the two aptamer ends, the latter two, while supposedly more reliable, will likely be too costly. Licor also has a system based on fluorescent dyes but it looks like you buy the parts separately...
Alain suggested that all adaptamers that we create have a sequence that can bind a complementary short sequence; we'd order a lot of labeled short complementary sequence. Sounds like a good idea, as long as those short sequences do not interfere with the aptamer binding.
Shawn and William suggested that for an initial assay, it'd be easy to just stain protein with Coomassie blue. So that's what we attempted.
We ran two 10-20% native polyacrylamide gels at 15 V overnight; the first used some lanes from the Nano group's gel. Visualization to come tomorrow.
The composition of what we added to each line is below; materials were added and incubated for 30 minutes in 2 uL total volume for lanes 1-11, 1-12 prior to dilution to 10 uL including addition of loading dye. Lanes 2-1 to 2-9 were incubated for an hour; longer incubation carried out since DNA basepairing reactions for lanes 2-10 to 2-12 were carried out for 30 minutes using results from 30-minute incubations of lanes 2-1 to 2-9. This incubation time should not be significant.
Materials used: Thrombin: a 2 uM solution of Thrombin. Streptavidin: a 2 uM solution of Streptavidin. Thromb5, Thromb20, Thromb35, Thromb50, Strep5, Strep20, Strep35, Strep50 are described in Plan section above. Bock's buffer is a reaction buffer for DNA-protein binding, used in several papers (see Nano group's page). 10X dye is a few flakes of bromophenol blue + 500 uL Bock's + 500 uL glycerol.
All quantities in uL.
|4X Bock's Buffer (see Nanostructure page)||.5||.5||.5||1||1||1||.5||.5||1||1||1|
|Bock's 1X buffer||7||7||7||7||7||7||7||7||7||7||7||5||5||5|
Ran 2 10-20% polyacrylamide gels along with nano group.
The gels came out completely blank. Bummer. Potential problems were protein concentration, pH of reaction buffer, and voltage. Different concentrations were tried out by the nanogroup to no success.
For now, we will
1) debug SDS-PAGE with the nanos. 2) Search for possible cell surface protein targets for which aptamers have been developed (and have published sequences).