8 replies to this topic

[Final year phd at wits end]

At great length I generated a conditional knockout construct containing

5' arm of homology > LoxP > Targeted exon > FRT > Neomycin > FRT > LoxP > 3' arm of homology

At great length I screened 180 colonies to find only 2 positive clones. These were electroporated and used to generate chimeras. The chimeras were confirmed and mated to B6 to confirm germline transmission. Floxed hets were identified and then mated to beta-actin driven cre to generate a systemic knockout.

So I should have mice that are:
cre positive, neomycin negative, deleted allele
cre positive, neomycin negative, wild type allele
cre negative, neomycin positive, floxed allele
cre negative, neomycin negative, wild type allele

BUT I have no mice that are cre positive, neomycin negative, with the deleted allele; instead I have mice that are cre positive, neomycin positive, wild type allele.

i.e. the presence of cre should result in recombination at LoxP sites and the neomycin lost along with the targeted exons.

I checked the LoxP sites of the construct before it was injected. I'm in the process of sequencing the LoxP sites in the current progeny.

Could it be the cre recombinase? I was under the impression that you only need one copy of cre for recombination to occur.

Has anyone ever come across this/have any ideas what may be happening??? Any suggestions would be gratefully received!

Thanks
Any ideas?

Edited by Final year phd at wits end, 09 September 2010 - 12:50 PM.

[Rsm]

Are you sure about correct targeting of ES cells? Have you checked both external and internal probe?

How about your beta-actin Cre? It is transgenic, I assume, maybe you have silencing of the locus? Have you done some expression analysis of Cre, like RT-PCR, western (from blood/spleen is rather easy)?

rsm

[Final year phd at wits end]

[quote name='Rsm' timestamp='1284102511' post='86444']
Are you sure about correct targeting of ES cells? Have you checked both external and internal probe?

How about your beta-actin Cre? It is transgenic, I assume, maybe you have silencing of the locus? Have you done some expression analysis of Cre, like RT-PCR, western (from blood/spleen is rather easy)?

rsm


I am fairly certain about the correct targetting. I used a 5' Southern probe. And I was unable to get (5) 3' Probes to work so I confirmed this by long-range PCR and RT-PCR across the break point of the 3' arm of homology, that would only amplify the recombinants and not wild type.

The beta-actin cre is a transgenic, whilst I've checked for the presence of cre by PCR I haven't checked it's expression - good idea!

Thanks!

[Final year phd at wits end]

I'm sorry, this is a stupid question, but the beta-actin cre line I'm using has been used before by others within the department so we know the transgene isn't integrated into a transcriptionally silenced region - how/why would this change?

[dpo]

are you sure the first loxP site is integrated? If the fragment containing the targeted exon is quite big, recombination can also occur in this region, resulting in Neo resistant colonies, but without the first loxP... It depends on the restriction enzyme you used in the Southern, will this result in a different band when the loxP is present?

Edited by dpo, 10 September 2010 - 12:27 AM.

[Rsm]

Final year phd at wits end, on 10 September 2010 - 12:08 AM, said:

I'm sorry, this is a stupid question, but the beta-actin cre line I'm using has been used before by others within the department so we know the transgene isn't integrated into a transcriptionally silenced region - how/why would this change?

That depends on the nature of the transgene. If you have multiple integration sites (which is usually the case), and cross with wildtype mouse, you may end up with Cre present in a single inactive locus.
Your long-range PCR shows you correct integration into your locus, but does not give you an idea on how many copies are integrated into the whole genome (like a transgenic effect). But that shouldn't matter much, Cre should still delete all LoxP sites.

rsm

[Final year phd at wits end]

dpo, on 10 September 2010 - 12:26 AM, said:

are you sure the first loxP site is integrated? If the fragment containing the targeted exon is quite big, recombination can also occur in this region, resulting in Neo resistant colonies, but without the first loxP... It depends on the restriction enzyme you used in the Southern, will this result in a different band when the loxP is present?

I see what you're saying! I am fairly sure!
The targeted exon is 771bp, the 5' arm of homology is ~3Kb. Upon homologous recombination an extra restriction site is introduced which reduces the size of the band detected by Southern by ~2Kb.
Several standard PCRs have generated products of the correct size when amplifying from the 5' arm to the neomycin cassette, which would suggest that there aren't any gross abnormal recombination events. But I am currently in the process of sequencing this region to check the sequence of the loxP sites to ensure nothing weird has happened to them and in doing this it will also confirm (I hope) the present of the 5' LoxP.

[Final year phd at wits end]

Rsm, on 10 September 2010 - 02:19 AM, said:

Final year phd at wits end, on 10 September 2010 - 12:08 AM, said:

I'm sorry, this is a stupid question, but the beta-actin cre line I'm using has been used before by others within the department so we know the transgene isn't integrated into a transcriptionally silenced region - how/why would this change?

That depends on the nature of the transgene. If you have multiple integration sites (which is usually the case), and cross with wildtype mouse, you may end up with Cre present in a single inactive locus.
Your long-range PCR shows you correct integration into your locus, but does not give you an idea on how many copies are integrated into the whole genome (like a transgenic effect). But that shouldn't matter much, Cre should still delete all LoxP sites.

rsm

Ah, I see! Thanks!
In theory (ahem!) there should only be one integrated copy! As I forgot to mention, my construct also contained a thymidine kinase cassette to allow for the negative selection of clones containing randomly integrated events from targeted clones - Of course with genetics, nothing is simple and anything is possible; the construct could have lost the TK cassette and randomly integrated wherever it wished... sigh...

[jasminelj8890]

Hi,
I have been having the similar problem with my conditional KO mice. I have the same construct design as yours, and I am using ER-cre sistem. I checked the expression level of my target gene with RT-PCR a week after tamoxifen inducible deletion, but no deletion at all. Could you kindly share some experience on this? What happened with your mice? Have you eventually figured out why it cannot been deleted? Thanks a lot in advance.

Jas