When transforming a K. pastoris strain, is there a way to predict how many copies of the gene will be integrated following the homologous recombination? More specifically is it something that a molecular biologist is able to control in a simple way, or, is it usually left to the chance (and determined afterward using gene sequencing)?

Alternatively : if I want to transform an electrocompetent wild type strain such as Pichia X33 using a plasmid vector, is the multiple integration an iterative process I control or is it random?)

enter image description here

Edit: Thank you all for the answers. Finally if I understand well the process is pretty random but you can control it to some extent. @gaspanic: I keep the idea of using fluorescent reporters and FACS for that ;)

  • 1
    $\begingroup$ Do you mean if you target the same locus in a diploid/polyploid species or do you mean off-target integration? $\endgroup$ – Cell Aug 13 '20 at 2:22
  • $\begingroup$ Yes if I am targetting a same locus in haploÏd or diploÏd specie. $\endgroup$ – Strack Aug 13 '20 at 9:57
  • $\begingroup$ I only worked with with S. Cerevisiae and as long as you use 50 nt flanking regions it should only go to your region of interest unless your background has defective DNA repair mechanisms etc. In my lab we typically don't do whole genome sequencing just to verify a transformation. With conventional transformation techniques you only target 1 site in a diploid because the efficiency is so low, but with the cas9 enzymes you could integrate in both loci of a diploid. $\endgroup$ – Cell Aug 15 '20 at 15:25
  • $\begingroup$ Also your schematic won't work because the background will already have zeocin resistance. $\endgroup$ – Cell Aug 15 '20 at 15:29
  • $\begingroup$ Out of curiosity, is there a specific reason for not verifying systematically the transformation of final clone by sequencing? Apart from cost and time reasons? It is not the first time I see that. I guess yes, in the future CRIPR would be more used. (You are right about my schematic, it was only for illustration purposes) $\endgroup$ – Strack Aug 17 '20 at 13:24

To predict the number integrations is probably not possible, but you can favor multiple integrations, if that's what you are after. At least in Saccharomyces cerevisiae integration of a first plasmid appears to stimulate the insertion of additional plasmids at the same site (1). It is therefore possible to favor multiple integrations (2) by increasing the DNA-to-cell ratio in the transformation, thereby increasing the chance that any competent cell receives more than one copy of the plasmid, which could then integrate in tandem. The easiest way is to simply use more DNA, but depending on your cell competence you might need to compensate this increase by using less cells for your transformation (or simply plate less cells on each plate) as you will probably also increase the overall colony yield.

One way to actively select for multiple integrations is to use a competitive inhibitor for the selection marker. Again, in Saccharomyces cerevisiae transformations with HIS3 selection can be combined with 3-AT (3). This allows you to inhibit HIS3 activity just enough to only permit the growth of cells that have integrated a desired number of copies of the plasmid. Of course, this method is far from fool-proof, as you might also select for mutants that have increased HIS3 expression for other reasons. Edit: I actually found one paper doing this in K. pastoris (4).

Finally, if your implementation allows it you can facilitate the screening for multiple integrations by including a fluorescent protein-expressing cassette in your plasmid. That way you can easily screen for different numbers of integrations by fluorescence microscopy or FACS.


  1. Orr-Weaver et al. (1983): https://mcb.asm.org/content/3/4/747
  2. Plessis and Dujon (1993): https://doi.org/10.1016/0378-1119(93)90172-Y
  3. Wikipedia (retrieved 15 Aug 2020): https://en.wikipedia.org/wiki/3-Amino-1,2,4-triazole
  4. Menéndez et al. (2018): https://doi.org/10.2144/00295rr05
  • $\begingroup$ Welcome to Biology.SE! How would using less cells compensate for having more DNA — did you mean more cells? This looks like a good answer, but answers are much more likely to receive a favorable response if you edit to include supporting references (primary literature is best). Without that support, your answer is indistinguishable from opinion. This is a good example of how to format references. ——— You may also want to take the tour and then consult the help pages for additional advice on How to Answer effectively on this site. Thanks! 😊 $\endgroup$ – tyersome Aug 14 '20 at 18:49
  • $\begingroup$ Thanks for your comments! I clarified the text a bit and added links to relevant papers. If you still think I should clarify it better, please let me know! $\endgroup$ – gaspanic Aug 15 '20 at 12:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.