I have a lentivirus that has GFP and my gene of interest under a tetracycline-dependent promoter and a ubiquitin c promoter driving the rtTA and puromycin with an IRES before the puro gene. My problem is that I cannot select cells with puro. I've done a kill curve and 1000ng/ul is the lowest concentration that kills all non-transduced cells. However, I use the same conditions on my transduced cells and they also die. I have sequenced the puro gene and there are no mutations. I have also tried 6 different concentrations (from 750ng/ul up to 1500ng/ul) in parallel and with non-transduced cells alongside, and there is no difference in cell death at any concentration. I'm not entirely sure on the MOI, but less than 25% of cells were GFP when I added dox to the medium. Therefore, in my last experiment I used a smaller number of cells and more lentivirus so that all cells are transduced. I turned on expression of half of the cells after plating and there are still non-transduced cells.

When I turn on GFP expression with dox, according to the original reference, there is increased read through of the puro gene, which should increase the amount of puro in the cell. However, the whole point of my construct is to be able to turn on expression whenever I like and look at the immediate effects. I'm not sure if there will be permanent effects to the cell if I turn on expression for selection, then let it go off until I'm ready to activate it again. If the gene induces apoptosis, for example, perhaps all the susceptible cells will die on the first pulse of expression and there will be no effect on the second pulse.

Please, can anybody offer some suggestions? I also have an undergrad working for me doing this and it is killing both of us doing this - we are dead in the water at the moment and both of us are getting very frustrated! One thing I thought about is trying to pick individual colonies and then split them when they are confluent and use one plate to look for 100% gfp expression. However, I think it might be better to have a polyclonal population. Do you think this is the case?

I'm not sure i get the whole picture, but I will try anyway.

As i understand, your vector has an inducible promoter (GFP + Your POI) and another constitutive promoter controlling the rtTA-IRES-puro. But as you said, adding tetracyclin should increase the amount of the puro gene. Therefore, more cells should survive when tetracyclin is added.

I've been told by several persons that IRES sequences work, but not as much as it should. Maybe your cells don't have enough resistance and get killed.

I've worked with lentiviruses with puromycin resistance, and it worked like a charm. So i recommend you try with even more viruses to get a polyclonal population. Individual clones should not be used with such viruses since they can enter the genome anywhere.. Good luck!

Thanks for your help Madrius. Do you think that if I get a 100% transduced polyclonal population, without puro selection some cells will lose the integration/GFP expression at a later point when I actually want to turn on gfp expression? Because normally puro will select for cells that maintain expression and a non-silenced integration but without the selection it could be silenced? Does this make sense?