Sometimes CreER is leaky. And even without Tamoxifen you see Cre activity. I suggest you use a different CreER construct. You can try reducing the amount you transfect, might help a little bit, but even if very little Cre is present, you will still see GFP

I assumed that CreER could control recombinase activity more tightly. After reading your reply, I have found several papers dealing with leakiness of CreER.

I will try a different CreER construct following your advice.

Thanks

IF you are culturing you ES cells with phenol red in the media, that could be the culprit!!! Phenol red is a weak estrogen, and could be activating the inducible Cre!

Thank you Timothy for your critical advice!

I never really imagined something contained in ordinary DMEM is the problem. Now I now why the exactly same CreERT2 works really well in mice.

I would try some phenol red-free medium and see what happens.

Just for information.

I tried phenol red free medium and it turned out that it did not improve the situation in my case.

I guess reducing the dose is necessary for my CreERT2 construct.

You should try with another CRE, i suggest you to try the ERT2-CRE-ERT2.

dear Yutato, I am having similar issue with my ERT2-dependent construct. interestingly, the same construct worked fine in melanoma cell line but when it was transfected into a breast cancer line, 4-OHT free culture still resulted in cre activity. wondering if you have any update on this topic?

thank you so much!

anyone else who may help interpret what's going on would be also greatly appreciated.

Dear Jianzhong

I actually do not try this experiment anymore. So I am not sure if ERT2 Cre ERT2 is much less leaky as Roberta suggests.

I guess that lentivirus would be the most robust and cell type-independent way.

And what about using doxycycline inducible Cre?

Hi Julio, not sure about the cre, but I have used a dox-inducible shRNA construct that appeared to be leaky as well. good luck!

Hi guys,

This chat is very interesting.

My problem is related to yours. I have an ERT2 mouse line and I'm hesitating whether to inject or not in the brain an AAV-DIO-shRNA. I'm afraid that I'll have some Cre activity even without tamoxifen injection which means that I'll have RNA inhibition even in control conditions and then a big loss of specificity.

If anyone could help me I would kindly appreciate this

Albert

Hi guys, I am also interested in CreERT2 leak. Is there any possibility that the amount of CreERT2 expression matters? It would be great if there is any information about the relationship between CreERT2 amount and the leak.

Best,

Motoko

Hi Motoko,

Although I do not have a direct evidence, I think that the expression level does matter considering that CreERT2 often shows leakiness even under the optimal condition (i.e. mice with CreERT2 knocked-in in one allele).

I also want to add for everyone that I realized that my case is not a good experiment to assess the leakiness of CreERT2.

It is possible that the plasmids recombinated in the cytoplasm are transfarred into the nucleus in our system.

So if you are to use CreERT2 system in vitro, it is necessary to use systems that transfers the reporter directly into nuclei (I am not sure if the virus system does this) or the reporter should be recombined into the genome beforehand.

Hi Yutaro,

Thank you so much for your kind and prompt reply.

I will discuss with my colleagues about the possible experiments to support my hypothesis.

Best,

M

Hey, Motoko!

We are using mice with a ROSA26 ER-Cre (So "optimal " conditions, as Yutaro called it above) and, although not yet confirmed by PCR, we have observed effects strongly suspicious of leakiness. We excise ex vivo though, so that may be different. Provided, what we observe is true leakiness (which I am strongly suspecting) and remembering some localisation experiments with overexpressed proteins I would strongly advise against transient expression for any kind of titration of the TMX, particularly with a very strong promoter such as CAG. If you perform IFL analysis on transiently transfected cells you will typically see pretty much any localisation patterns with any protein, as the expression level varies depending on how many copies the particular cell has taken up. This makes aberrant localisation. Its probably preferable to select a stable clone beforehand as the titration in transiently transfected cells will produce too high recombinase activity. If leakiness can be totally avoided, I am not certain. Or try Mx-Cre. That may be better. If it was just to get a good dose of TXF, you may know what you wanted to know already, but be aware that leakiness may be a problem, particularly in cancer-induction (one cell is enough to make a cancer). So is its still useful, it depends. You may want to include another type of negative control.

Good luck! Hope I could help a bit.

Thomas :-)

Hi Motoko

I also had the same issue, more or less... i think that leakage effect is an intrinsic part of this system, since it is linked to the KD equilibrium constant that defines the association/dissociation between Cre-ER and HSP90. As an equilibrium, the bonding between HSP90 and ER domain entails the existence of a freely moving fraction of the protein, always present in the cell. This free fraction is capable of reaching the nucleus and operating the recombination, regardless of tamoxifen addition and concentration. Since recombination is an irreversible event, the entity of leakage follows the statistics of radioactive decay and grows monotonically with time, KD, and the relative concentrations of Cre-ER and HSP90 in the host cell. That's my guess.

Ciaaooo