I read that transgene insertions can cause off-target mutations that result in a phenotype being overly ascribed to the transgene, e.g. the increase of lifespan being ascribed to Sir2 overexpression (Burnett et al, 2011, https://www.ncbi.nlm.nih.gov/pubmed/21938067). In that particular study, they outcrossed the transgenic strain several times with wild-type to remove these off-target effects, a mutation in Dyf2 being named in particular.

I have a poor background in genetics, so I don't really understand how transgene insertions can cause off-target mutations, nor how outcrossing can isolate the intended mutation.

  • $\begingroup$ Could you give a link to the Pubmed page of the paper you mention? $\endgroup$ – terdon Feb 21 at 20:17
  • $\begingroup$ Depends on how the transgene was inserted. Transposon and retrovirus mediated methods cause random insertions in the genome and can disrupt other functional parts of the genome. If you provide the link to the paper then it would be easier to figure out what the exact reason was. $\endgroup$ – WYSIWYG Feb 21 at 22:21
  • 1
    $\begingroup$ Edited to include link $\endgroup$ – BatWannaBe Feb 21 at 23:39

The C.elegans used in the study you linked was originally made by Tissenbaum & Guarente using the gamma irradiation protocol.

In short, the protocol provides the injection of the desired DNA construct in the worm's gonad followed by gamma irradiation of the worms.

Upon radiation exposition, the genomic DNA of the worm will break at random sites and will undergo repairing by the appropriate cellular enzymes, during the latter stage the exogenous DNA can be integrated (ligated) into the genome. However, the genomic DNA can also acquire random mutations due to the radiation. This is how transgene insertions can cause off-target mutations in this case.

With other organisms or cells, the protocols are different. To engineer mammalian cells, for example, radiations are never used but still, you have to generate breaks in the genome to edit it. Depending on the method the "off-target rate" of a given experiment may vary, but it is never zero (not yet).


When you cross a mutant organism with a wildtype one (backcrossing), in each offspring half of the alleles originate from the mutant parent and half from the wildtype. Another way to look at this is to say that approximately half the offspring will contain any given mutation, assuming it is heterozygous in the mutant organism, and doesn't cause a fitness differential.

Because of this, on average, each time you backcross to the wildtype you lose half the mutations. The probability of a mutation remaining after several rounds of backcrossing reduces very quickly (0.5 after one generation, 0.25 after 2, 0.125 after 3, 0.0625 after 4, 0.03125 after 5 etc).

At each round of backcrossing you test offspring and select the ones that have the intended mutation and use these for the next round, so while the probability of keeping the intended mutation stays high (~1.0) the probability of keeping unintended mutations decreases very rapidly. This all assumes that the offtarget mutations are on different chromosomes, or far enough away from the intended mutation (unlinked) that they segregate separately and are not carried forward into the backcross generations by 'linkage drag'.

  • $\begingroup$ Edited to replace 'plant' with 'organism', and mention fitness differential $\endgroup$ – Jonathan Moore Feb 23 at 20:53

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.