User:Anthony Salvagno/Notebook/Research/2010/03/08/New Mini Project
Quick Navigation: Pages by Category | List of All Pages | Protocols
Sorry for not posting anything in my notebook in over a week. It's not that I haven't been working, I've just been doing odds and ends that weren't really notebook material. I guess I should have put it in anyways, but I didn't think conversing on friendfeed was all that worth it both for your reading and for my effort. I did get a bunch done and I'll recap on a separate entry later.
Anyways I've got a new collaboration and I'd like to get that started. I need a codename until I can disclose the information. Remember since it's not my project I need to only detail the things I'm working on and try and keep everything else a secret. Sorry for being closed, but I'm working with closed people. Don't worry, they won't convert me.
Details
I need long DNA to have a fluorescent molecule bound to each end. I am going to use QD's because I don't know how visible a single Cy-3 or Cy-5 or other dye molecule would work here. So right now I'm thinking digesting with something (maybe two different cutters) and ligating hairpins to the ends. The hairpins would contain a dX-biotin (where X is any nucleotide) which could attach streptavidin QD's.
Lambda DNA is my starting point and for my hairpins I'll use the SapCap template but change the overhang and add my biotin-dX. Lemme do a little research.
New Plan: Klenow polymerization with incorporated labeled nucleotides! Read below.
SJK 16:42, 8 March 2010 (EST)
Agree that Klenow is good bet, especially since you found that exact protocol. Hairpin method may be easier for down the road when we want dig on one end and bio on the other end. A couple other answers here while in my head: (1) We want the klenow without exonuclease activity. This is "proofreading" and we don't need it or want it. (2) More than one label would be fine. I don't think more than one QD would be able to bind, so it would just increase the lifetime of the QD-DNA interaction which would be good. (3) We can debug / assess success using polystyrene spheres at first, and then move to QDs. (4) We'll probably want to use FITC-colored and TRITC-colored QDs (green and red). That's my first intuition.
What to do
First I will look into already manufactured biotinylated lamda phage DNA. Then I will learn about the Kelnow reaction to make it myself. BTW this page is highly unorganized, but will be better once I get a better scope of what I want and need to do.
Lambda DNA
<html><iframe height="200px" width="650px" src="http://www.ncbi.nlm.nih.gov/nuccore/9626243?report=genbank#sequence_9626243"></iframe></html>
On this page, you can find the sequence for the bacteriophage lambda. It is a huge linear piece of DNA around 48kb. It is useful to use because it has two 5' overhangs (one on each end) of 12 bases each. In the above sequence those sequences are:
- 5'-gggcggcgacct
- 5'-aggtcgccgccc
It would be nice to have exactly one biotin in a complementary piece of each sequence, but I don't think more than one is bad. Biocurious pointed out to me that there is exactly one A in each overhang allowing for binding of precisely one T or U. I'll have to do some reading to understand why they specifically use a U instead of a T, but this is a great starting point.
(Other: I spent a bunch of time analyzing the html sourcecode for that framed page looking to just frame sequence and nothing else. After much analysis I figured out the coding, and have decided that the information comes from some internal source like a database or something and need a access to it, so I give up. I'm sure my sentence isn't technically correct, because of my base understanding of web design, but basically I have no access to the information that everything is stored on.)
Klenow Polymerase
See here (Note: I won't frame that page for risk of overdoing it. Framing is awesome though.)
The Klenow fragment is a piece of DNA Polymerase I found in E. coli. It has two main and awesome functions. It can polymerize 5'->3' and it can remove nucleotides 3'->5' (known as an exonuclease, shouldn't it have been called a depolymerizer though?). It can do these things (stolen straight from the Wikipedia page):
- Synthesis of double-stranded DNA from single-stranded templates
- Filling in( meaning removal of overhangs to create blunt ends) recessed 3' ends of DNA fragments
- Digesting away protruding 3' overhangs
- Preparation of radioactive DNA probes
There is a change to the gene that encodes for the fragment that removes its exonuclease ability making the fragment known as exo- Klenow fragment. I have listed purchases of both fragments below (see: Shopping). I don't yet know why I'd want exo- over the regular one. With the exo- I could only polymerize, but the wt fragment could also cleave. For lambda DNA I wouldn't need to worry about the cleaving since I only have 5' overhangs and no 3' overhangs. This would also give me a viable option in the future in case I needed the ability to chew the fat (so to speak). Safest bet is the exo- though.
Putting it all together
So the goal is to get biotin on the ends of lambda DNA for QD attachment. For this I would non-selectively polymerize with all dNTPs. If I wanted different labels and thus different attachments, I would want to polymerize each end separately. More knowledge is needed in this regard.
Protocols
<html><iframe height="500px" width="800px" src="http://biofizika.aok.pte.hu/en/research/nanobiotech/protocol/DNA_Biotinylation.pdf"></iframe></html>
From here. This one allows for lots of biotinylated nucleotides to end up on the DNA, and is biotin on each end.
<html><iframe height="500px" width="800px" src="http://biocurious.com/2006/01/14/labeling-l-dna"></iframe></html>
Not really a protocol, but close enough to understand what I need to do. This method allows me to have different labels on each end.
What if I didn't care about every single DNA molecule having one biotin and one anti-dig? If I just mixed the dNTP's (dCTP, dGTP, dATP, and biotin-dTTP and anti-dig-dTTP) molecules would end up having two biotin, two anti-dig, or one of each. If we get two different color dots each with a separate linker surrounding it we could selectively look at only the different color DNA's.
If it is absolutely needed that we know which molecule is on which end and always want one of each, then selectivity is an issue.
SJK 16:46, 8 March 2010 (EST)
For now, we just want biotin on both ends, and we'll deal with only 50% of them being 2-colored. When we go to dig on one end, biotin on the other, I'm guessing hairpin method will be easier. Dig-labeling w/klenow is super-expensive (unless the patent has run out from Roche).
Shopping
- Lambda DNA
- <html><iframe width="700px" src="http://www.neb.com/nebecomm/products/productN3011.asp"></iframe></html>
- Polymerase:
- <html><iframe width="700px" src="http://www.neb.com/nebecomm/products/productM0212.asp"></iframe></html>
- <html><iframe width="700px" src="http://www.neb.com/nebecomm/products/productM0210.asp"></iframe></html>
- dNTP's - we have this but it may be worth getting a separate stock for this reaction.
SJK 16:48, 8 March 2010 (EST)
Definitely purchase new nucleotides. Also, you'll need to pick which one to be biotinylated. I'm leaning towards having more than one biotin. I seem to remember biotin-11-dCTP being available 10 years ago? (The 11- refers to an 11-carbon linker on the biotin.)
Finalizing
{{#widget:Google Spreadsheet |key=t8Q1jJ0eJ_yZc4Odq5ofbjw |width=500 |height=300 }} I am going with:
- Klenow exo-
- biotin-dCTP (because there are 6 dCs on one end and 4 on the other)
- QD's from Invitrogen at wavelengths 525 and 625 (green and red respectively)
- Also I want some free streptavidin (NeutrAvidin) and I found this. Is what I want on that page?
SJK 17:58, 8 March 2010 (EST)
A couple things: (1) Make sure that the amount of dCTP will be enough for your reaction. Note how low the concentration is! (you'll probably want to reduce concentration of other nucleotides as well? check protocol.) (2) make sure dNTPs set is not premixed. (3) Do you need columns or other things for the cleanup? I've never had good luck with centricon. I wonder if size-exclusion column would work well. If S-400's (or similar) work for this, I'd buy some. We don't want to waste our dime later. (4) Buy some streptavidin from Pierce Biochem (actually neutravidin, I think). You can use this on an agarose gel to verify that lambda DNA at least has some biotin. (Gel mobility shift.)
