4
$\begingroup$

We are regularly doing bacterial transformation and subculture from plate to liquid media to extract DNA. This usually goes very well and is straightforward, but occasionally, the colonies that grew well on a plate, will not grow in liquid culture (with the same antibiotics that were in the plate). This has happened now to three persons on multiple plasmids.

For example, I was trying to ligate an ampicillin resistance gene (with its promoter) inside pENTR4 plasmid (already with Kanamycin resistance). My plate and LB both contained the two antibiotics, but the colonies only manage to get slightly turbid at a 16h timepoint.

Phage contamination is very unlikely, and the effect cannot be due to a toxicity of the inserted fragment, as both insert and vector have been grown in these same bacteria multiple times without problem.

Help?

EDIT1 - We use Carbenicillin in place of ampicillin in the plate, but not in the liquid media.

EDIT2 - After managing to grow some colonies in liquid, I managed to get a small amount of the plasmid (profile was good; concentration = 70ng/ul). The weird thing is that when I retransformed this DNA in Stbl3 bacteria, I had absolutely no trouble. Do you think the bacteria stock could be responsible?

$\endgroup$
12
  • 2
    $\begingroup$ Did you change your batch of antibiotics, or media recently? $\endgroup$
    – Rover Eye
    Commented May 10, 2015 at 14:27
  • 2
    $\begingroup$ Simplest explanation is something is culture tubes or media prep bottles that inhibits growth (detergent residue?). But lots of people have seen what you describe if the colony-bearing plates have been stored too long at 4 C. One way to test (besides comparing media +/- drugs) is to "patch" out colonies on a new gridded plate--you may do this routinely, lots of people do. So either after finding cfu patch candidates to a master grid, or, when you innoculate the tubes for mini-preps, also patch out the cfu on a fresh plate. If the patched grid plate grows but the culture doesnt, voila $\endgroup$
    – mdperry
    Commented May 10, 2015 at 14:45
  • $\begingroup$ In very rare cases patching shows healthy tiny lawns with pin-prick holes from phage contamination. Also, if it is phage contamination from a lytic phage, in my experience the broth will clear and accumulate a small stringy clot of lysed cell debris, that floats. What if there was a non-lytic phage like M13? Sounds like you are carefully considering the "obvious" reasons. There definitely ARE toxic inserts that are virtually impossible to propagate in regular E. coli, but then you would only find transformants without inserts (typically), but that is fortunately a pretty rare occurrence $\endgroup$
    – mdperry
    Commented May 10, 2015 at 14:49
  • $\begingroup$ The antibiotics are stocks stored at -20C, and they have been used simultaneously by a colleague of mine that had the same problem on some of his vectors (2/8). Since we have used the same type of tubes successfully, I doubt it could be the reason. $\endgroup$
    – WilliamL
    Commented May 10, 2015 at 15:09
  • $\begingroup$ did you try positive control transformation? Of some test vector, like pEGFP-type $\endgroup$ Commented May 10, 2015 at 15:16

1 Answer 1

2
$\begingroup$

I think there is more than one possibility for this.

  1. Plates bearing colonies are stored for too long in 4C(as @mdperry suggested). This causes the colonies to lose the plasmid over-time as the antibiotic in the plate gets degraded over-time. There is a study on this and I've shared the link. (http://homepages.bw.edu/~mbumbuli/biotech/colin/index.html).
  2. The insert is toxic to cells and promoter is not strong enough. There is a leaky expression causing cells to die.
  3. Check the competent cells and their storage condition as it would greatly affect the efficiency of transformation. Do a mock transformation with plain plasmid without insert and spread them on an antibiotic plate.
$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .