So after bacteria have been transformed to perhaps grow up a plasmid of interest, why pick only a single bacterial colony from a selective plate for further expansion?

I understand that this is to ensure that you are only working with a single genetic makeup because each separate colony is derived from only a single bacterium. What I can't rationalize is that if I am trying to expand and isolate a plasmid of interest, all colony expansions on a selective antibiotic plate should contain my plasmid. So is it really necessary to take only a single colony, because all colonies I pick should be useful to me.

  • 2
    $\begingroup$ If you want to do multiple preps, then by all means pick multiple colonies, but keep them separate. The point of picking a single colony is that it's (theoretically) a single clone, so you're sure that your starting material isn't contaminated with another sequence. $\endgroup$
    – MattDMo
    Jun 9, 2015 at 23:14
  • $\begingroup$ Right, when I make a plasmid, I typically grab 24 colonies because my centrifuge can hold 24 tubes. Keep them all separate and labeled. There have been times when only 1 of the 24 had what I wanted. $\endgroup$
    – user137
    Jun 10, 2015 at 2:47
  • $\begingroup$ Why risk a heterogeneous culture? Pragmatic workflow is always best. $\endgroup$
    – AliceD
    Jun 10, 2015 at 2:55

2 Answers 2


This is a matter of pragmatism in the culture process.

Taking 100 colonies instead of 1 increases the inoculation volume by a factor of 100, which then saves you perhaps 2 hours of bacterial growth time before your culture reaches the OD you want.

However, mutations and loss of plasmid in culture, while unlikely, are possible, especially if the bacteria were not cloning strains with their recombinases knocked out. In such a case, you would risk the downstream experiments being contaminated with mutants, be that protein expression, retroviral plasmid production, or something else.

Worse still, the mixture of bacteria containing mutants is less likely to be detected, since had you picked a single mutant colony, there would be absolutely no result, but a contaminated culture would produce a low but still positive result. Therefore, you may end up wasting many days or weeks troubleshooting your downstream experiments.

Therefore, picking single colonies is simply a matter of "better safe than sorry".


While all colonies on a selective media SHOULD contain a resistance gene, that's a long way from them containing the plasmid you want. A typical cloning experiment will insert a gene of interest into a empty vector backbone to make an intact plasmid. The backbone usually contains some antibiotic resistance gene. If your gene of interest is not correctly inserted, but the empty vector religates on itself, it will make a viable colony without your gene.

Additionally, I have seen mutations happen even when an insert is correctly inserted, and if you're doing blunt end ligation, there is always a risk of getting your insert in backwards or ligating two vectors together.

  • $\begingroup$ And that all make sense but let's say for instance that we are growing up a plasmid from a stock solution which has already been sequenced and the resistance gene and the gene of interest are both on a single plasmid. Now is it relatively safe to just pick any number of colonies and assume that they would all have the gene of interest? $\endgroup$ Jun 9, 2015 at 23:11
  • 1
    $\begingroup$ You can easily test this. Resuspend a few scrapings of your frozen stock in a small volume of media and plate serial dilutions on both plates with drug and plates without drug. Let the same culture grow in a shaker for 90 min (with drug) and do the same two sets of serial dilutions. If the frozen stock is pure the counts will be the same plus/minus drug. Likewise after growing in liquid. If you grow the liquid long enough the drug will get used up and you may start to see cells that lack a plasmid start to accumulate. $\endgroup$
    – mdperry
    Jun 10, 2015 at 1:37
  • 1
    $\begingroup$ @TheNightman You will get your plasmid but it is not relatively safe. If the downstream analysis/experiments are complex then you want to make sure you start with the best material you can. You might extract plasmids with mutations if pooling colonies together and this could be very problematic for subsequent experiments. Sometimes very little details make an experiment succeed or fail. $\endgroup$ Jun 10, 2015 at 3:47

You must log in to answer this question.

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