Knockdown with shRNA not stable (Why?) - (Apr/02/2014 )

I have troubles with knock downs of different target genes in different cell types. They all behave the same.

Protocol: Spinning protocol for lentiviral treatment with polybrene and puromycin treatment after 2 days.

Used cell types: Jurkat, CEM and adherent carcinoma cells

Used targets: 3 different targets (some of the shRNAs are validated by Sigma)

The problem is as follows: After the lentiviral transduction (on day 2), I harvest some cells for FACS or western blot analysis. I always see a knock down of at least 50 % of my target genes/proteins compared to wild type and control (scramble shRNA) cells. I now begin with the puromycin treatment (Depending on the cell type with up to 5 µg/mL puromycin). Untreated wt cells die, lentiviral cells stay alive, like they should behave.

The problem is now, when I test them again on day 7 after lentiviral transduction in FACS or western blot, the knock down, compared to wt and control cells, is gone again. I cannot see any differences! The cells are resistant to the puromycin treatment, since they don't die, but wt cells do.

As I said: Some of the shRNAs are validated and I have seen data in other publications, indicating that they work fine. Some of the constructs also already worked in our lab for other cell types. Maybe I am doing somethign wrong? But where can be a problem, since the cells are resistant?

Maybe you can help me?

I don't know much about lentiviral stuff - so the first of these may be somewhat naive:

Are the puro resistance and shRNA on the same construct?  If they are - see below.  If not - the cells could easily be losing the shRNA and keeping the puro as the shRNA is not actually under selection.

If you are targeting essential gene pathways the cell may find a way of inactivating the shRNA but keeping the puro resistance.   I don't know if this would work for transduction, but it seems to for other types of DNA insertion.

Yes, the resistance gene and the target shRNA are encoded on the same plasmid (pLKO.1)

We were thinking the same, like processing the RNA by the host cells. But still, some of the constructs are validated, published and used in our lab in other cell lines to generate knock downs. And why should this processing only occur after several days and not immediately? These genes should (as far as we know) not be essential for survival of the cells.

Can you check that the shRNA is actually being produced (qPCR perhaps?)?