Shattering DNA

[amethyst73]

As many of you probably know, my job consists primarily of taking long pieces of DNA (many hundreds of thousands of base pairs long) and breaking it into much smaller pieces (3000-4000 base pairs).  This is done by a process called shearing.  Dissolved DNA is passed several times through a hole that's pretty small (maybe 0.2 mm, probably smaller).  The shear forces in the liquid that are created by being passed through a tiny hole break the DNA into the smaller pieces in a remarkably predictable fashion, based on the size of the hole and how fast the solution is forced through the hole.  As a result, making sheared libraries is pretty straightforward.

For the last month or so, I've been working with smaller pieces of DNA (3000-8000 bp) and breaking them into very small pieces (200-400 bp).  This process is considerably more complicated. 

To start with, instead of working with 6-8 samples of DNA at a time, I'm working with (effectively) 192 samples - two 96-well blocks, with a different piece of DNA in each well.  Preparing the samples is done all in a 96-well format, so that's straightforward.  BUT!  Complication #1!  Often you don't get sufficient DNA for a given sample, so you have to spend several days streaking out the bacteria for the individual samples, growing up little liquid cultures, prepping DNA, seeing if you have enough, potentially growing and prepping again.

Once I have sufficient DNA for all my samples (or have tried sufficiently hard without success to declare them dead), then I pool the samples: for each column of 8 samples (the 96-well plate is laid out in an 8x12 format), I take some from each well and put them all into a single tube.  Repeat for all 24 columns of samples (12 pools for each plate).  Not difficult, merely tedious; it takes about 2 hours to do this.

Now comes the interesting part.  Instead of passing the samples through a tiny hole to break them up, we sonicate them.  (Yup, we screech at them and this makes them break down.)  High-frequency soundwaves make teeny tiny bubbles in solution, and those bubbles provide shear forces to break up the DNA.  This is called shattering the DNA.  While it ought to be a predictable process, for reasons which remain obscure to me it is not.  The last set of 24 samples I did this to mostly behaved very well - only two samples required further sonication.  The set I ran last night... had one successfully shattered sample.  All the others obviously required more, and variable amounts more between samples.  So I'm spending the day running diagnostic gels and further sonicating the samples. 

They're not done yet. 

I'm clearly not going home early today.

Comments

[meepodeekin.livejournal.com]

In college I did research on those sonically (sp?) induced bubbles--if you induce them with a sufficiently rhythmic sound wave, they can actually be made to give off pulses of light. Which no one really understands. I quite like hearing that they have applications elsewhere, even if unpredictable and frustrating.

Good luck with your shattering!

[holmes-iv.livejournal.com]

Wasn't that light emission effect at the center of one of the more recent Cold Fusion false starts?

[meepodeekin.livejournal.com]

Could be. I haven't been keeping up since college, but the amount of energy density focussing going on could easily relate to a cold fusion hope.