triple knockout

7 posts / 0 new
Last post
Immuno-competent
Immuno-competent's picture
triple knockout

I want to know the best strategy to create tripple knockout or double knockout with transgene expression.  What is the frequency of getting a triple knockout embryo through interbreeding? 

Ivan Delgado
Ivan Delgado's picture
 

 
In my experience most double and triple knockouts are generated by crossing established lines to each other. So if there is anyway you can do this then I suggest you take that route. If the genes you plan on knocking out are all novel (i.e. there are no lines already established), then it may be advantageous to simply generate each line separately but concurrently, that way you minimize the time you will need to generate the knockouts. 
Alternatively it is not uncommon to simply inject two constructs and then double select. I personally have not done this or seen it work successful, but I've heard that it work (not very efficiently, but it can save you a ton of time if it does). 
The frequency of getting a triple KO through interbreeding depends on many factors. Most of the time you will come across fertility issues once you start piling on the KO genes. I do not think you can get a fair frequency that is not significantly biased by the genes you are working on. 

Omai
Omai's picture
If the genes have no effect

If the genes have no effect on viability (or fertility as Ivan points out), and are not found on sex chromosome, you can expect proper Mendelian ratios. When I was crossing animals, I found that putting together Punnet squares was exceptionally useful.
Here is a great exercise to get used to using Punnet Squares.
If you have other questions with your own specific genes, please post them and we will do our best to help with your crosses.
 
Triple knockouts can be very time consuming and "mouse expensive" (large number of cages) to produce. Its best to make a detailed breeding plan at the very beginning to cut down on mistakes.
 
Omai

Ivan Delgado
Ivan Delgado's picture
 

 
I agree with Omai that a detailed breeding plan can be critical when venturing in an experiment such as this, especially if one or more of your genes is not a typical one that behaves in normal mendelian fashion. You can save yourself a ton of time (and money) by taking the time to plan the most efficient way of doing this before you even start the injections and/or crosse. Another suggestion is to share your plan with someone else. Invariably if you can find someone that has done this kind of work before they will likely be able to look at your breeding plan and suggest a shortcut or better yet identify a mistake.
As a side: if you do not already keep track of your breeding colonies using software (or an Excel spreadsheet) you may want to start doing it. It is another way of making the process more efficient.
Good luck

Immuno-competent
Immuno-competent's picture
Thank you for your

Thank you for your suggestions.  I thought I might get good suggestions, but never thought that I might get a reply so soon.
Well, if you are willing to help, I am willing to provide information.
Here is the generic situation:
A classical ko of a gene of interest (goi) on chromosome 7 leaves the KO mouse with mild autoimmunity and enhanced cancer immunity.  The gene has impacts on both T cells and dendritic cells, thus I want to propose to create a loxP-flanked conditional allele for the gene (first ko mouse of the tripple combination).  No problem with fertility or viability.
Want to delete goi in T cells and ascertain its ability to control spontaneous occuring tumor, such as ApcMin/+ mouse model. 
So, I want to use CD2-driven-cre deleter transgenic line to get rid of T cells (second of the triple cross).
Want to then cross the CD2-cre/goi floxed mice with ApcMin/+ mouse (has point mutation that truncates  APC gene prematurely (third cross). Mice die in 3-4 months
O.K.  The above scenario is probably not too difficult for those in the know to crack.  What would be the frequency of obtaining such cross?
 
Now, here comes the harder one.
It appears that the deletion of my goi impacts only a sub-population of T cells called Treg, which is differentiated sub-set of T cells and is defined by the exclusive expression of a transcription factor called FoxP3, which is located on Y chromosome.
I have a deleter mouse strain that have IRES-Cre casette knocked into the 3' untranslated region of FoxP3 allele.  Thus, cre is expressed only in FoxP3+ cells through CAP-independent translation.
I want to cross goi flox mice with FoxP3 cre deleter and then cross it with ApcMin/+ mouse.  I don't mind using male only if it increases the frequency of triple transgenic.
I will be very grateful for any help you can provide.
Thank you.

Immuno-competent
Immuno-competent's picture
Forgot to ask one more

Forgot to ask one more question:
I assume that it is better to cross goi flox mouse with cre deleter knockin, creating bigenic mouse.  Then, create another bigenic line with ApcMin/+ mouse X credeleter knockin.  Since Cre deleter knockin is present in both bigenic mice, the frequency of getting trigenic mice is better than if one were to serially breed bigenic with the third gene knockout, right?

Omai
Omai's picture
 Hey Immuno-competent,

 Hey Immuno-competent,
Your triple cross looks pretty straight forward. I've done a similar triple knockout in heparan sulfate biosynthetic genes including a lck-cre, 1 gene that was floxed, and one gene that's knocked out.  I'll call it the CD2-cre or the FoxPcre (the crosses should be thesame with either cre, but I think they need to be done separately), f/f (floxed/floxed goi) and ApcMin/+ (Does this mean the Apcmin acts as a dominant mutation so it carries in the heterozygote?) for the separate alleles. I'll use 1/4 to represent 1 in 4 animals of the noted genotype.
I agree that I would first take the strategy of putting the cre onto the floxed line. 
 
Cre is a dominant allele and can carry through as a het at 50% so your first cross will yield 1/2 cre+, f/wt animals. When the cre+ f/wt animal is crossed back onto a f/f animal, it will yield a 1/2 cre+ and 1/4 f/f to yield a 1/8 cre+ f/f animals.
Now you have your Cre+ f/f goi animal.
Cross this animal to your ApcMin/+ animal. (It sounds like you only need a het for this gene?)
This will yield a 1/2 Cre+, all f/wt, 1/4 ApcMin/+. So 1/8 will be cre+, f/wt, ApcMin/+. Cross this animal back to your  f/f goi animal. This will yield a 1/2 cre+, 1/4 f/f, 1/4 ApcMin/+  for a 1/32 chance of a cre+, f/f, ApcMin/+. This is your experimental animal. You can take other animals from this last cross (alternate genotypes) and use them to get a higher chance of getting an experimental animal.
I hope this helps and wasn't too confusing. 
Again, I don't think the breeding changes for either cre. I believe you have to do the whole breeding scheme for each cre.
Let me know if you have any questions.
Here is a link to my paper where I did a similar cross.
 
Omai