by PhilipJ on 14 January 2006
Whenever my lab gets a visitor (usually a physicist), after giving them the tour and showing off the optical tweezers instrument, I always get asked the same question: “so how do you attach DNA to beads, anyway?” We physicists are a clever bunch, but (bio)chemistry isn’t often in our bag of tricks. Since many of our readers are physicists too, here’s how we end-label DNA for force-extension measurements, using λ DNA as our DNA of choice, partially because you can buy it commercially, and also because its the DNA I’ve been labeling all day across the hall in our wet lab (using )!
A bit of background information first, unrelated to labeling: λ DNA is so-named because it is the genome for the λ phage, a virus that attacks E. coli. Shown on the right, it is particularly sinister-looking creature! The capsid (the round bulb at the top) holds and protects the DNA until it is ready to get injected into a bacterium. Once injected, the virus integrates into the host’s genome, tagging along into each new bacteria upon cell division and causing virtually no harm (called the prophage pathway). When the time is right (for example, when the host is signaling that it will die), the virus exits the host’s genome and uses the host’s replication machinery to create new copies of itself (the lytic pathway). Once all resources inside the cell are depleted, the cell lyses, freeing the λ phages to infect others.
Back to labeling! There are a couple of pairs of molecules that are commonly used when trying to make molecular handles, and I’ve mentioned the ones we use before, namely the antibody/antigen pair of digoxigenin and anti-digoxigenin, and streptavidin/biotin. They are so common that you can buy anti-dig and streptavidin (and many other) coated beads! Now we just need to find a way to attach biotins and digoxigenins to either end of our DNA. It turns out this isn’t as tricky as a physicist might imagine either — the demand is large enough that you can buy biotinylated and digoxigenined dNTPs! Now, lets see how we add them to our λ DNA.
Prior to injection in an E. coli host, λ DNA is a linear, double-stranded molecule with overhanging sticky (that is, having unpaired bases) ends, of total length 48,502 bases. It is in this form that you can buy it, and at fairly high concentrations too. Let’s take a closer look at the overhanging sticky ends:
Both sides have all four bases overhanging, but the order isn’t the same on either side! You’ll also note that there is a single free A on either side. If we are clever, we can do each side individually, and end up with a dig-dUTP on one end, and bio-dUTP on the other (U, uracil, is another
amino acidbase (thanks Neil!) that will bind with A, adenosine). If we put λ DNA in a test tube with dATP, dGTP, bio-dUTP and a DNA polymerase (experts will know that we often use klenow exo-), the DNA polymerase will start encorporating the complementary bases on either end up until we need a C, cysteincytosine (note to self: don’t do this at so late anymore, thanks UndergradChemist!), which is currently being withheld. Letting that reaction occur, we get DNA which looks like:
Alright, halfway there! At this point we need to get rid of any excess bio-dUTP bases, and so one normally ethanol precipitates the DNA and resuspends the DNA in a buffer free of nucleotides. We’re now free to label the other end, so adding dATP, dGTP, dCTP, and dig-dUTP along with a DNA polymerase will fill in the rest. We end up with:
This is exactly what we need to now attach the DNA and start pulling. The way I’ve been doing the attachments is to incubate DNA labeled on either end with one of the sets of beads so that they are covered in many (up to ~1000) DNA molecules each, and “fishing” for the other tether using the micropipettes and the other complementary beads, eventually making a tether between the two like so:
So, there you have it! This same technique can be used to label more or less all DNA of interest, and in some situations it is an even easier single-step process (for example, when one end has only As and Ts overhanging, while the other has only Gs and Cs). Now, if only my yields weren’t so poor after the ethanol precipitation step…..
 R. M. Zimmermann & E. C. Cox, DNA stretching on functionalized gold surfaces, Nucleic Acids Research 22 492-497 (1994).