Hi Sabine,
kifunensine would work, but it is expensive, so I would say the method of choice really is to use the GnTI- cell line (CRL-3022) you described. We are doing piggyBac in it regularly for the same reasons. The trick with this cell line in general, at least in our hands, is that you need to compensate for its lack of complex negatively charged glycans upon transfection, i.e., you need to get positively charged lipoplexes or polyplexes of DNA+your transfection reagent of choice “closer” to the membrane to get them in and have the cells transfected at all. Our solution is so-called “high-density transfection” protocol, described in Bláha et al., Protein Expr Purif 109:7-13 (2015); doi: 10.1016/j.pep.2015.01.006; PMID: 25623399 – I will send you the paper by email. This helps to get this cell line properly transfected. For piggyBac, I am not sure what other parameters my students optimized, but I believe they optimized concentrations of selection antibiotics which might be lower for this cell line than for the regular HEK cells, also sometimes we need to re-optimize this for some recalcitrant constructs – so basically the way to go is to split the transfected cells to several wells and select simultaneously at different concentrations of ATB and find where they will not die completely, but also not just grow happily ever after – but you know all this. Just try our high-density transfection and then try to select at three different ATB concentrations, you should get some stable pool. If not, I can link you to my PhD students for the current version of the “best protocol” they use 😉
Cheers,
Ondrej

O.K., I will try to get the details about the selection, then.

Dear Sabine,

I would try to answer some points and describe what works in our hands, in Ondrej’s lab:

– High density transfection – While working with HEK293S GnTI-, we are transfecting cells in ExCell medium at density of 20 E6/ml (typically for PB – 30 E6 cells in 1,5 ml on 12well plate, more detailed – first I resuspend 30 E6 cells in 1,2 ml of fresh ExCell, I add DNA diluted in 300 ul of PBS and then transfection reagent directly to the cells – we are using lPEI in 1:5 weight ratio, this should be incubated on the stirrer in incubator for 4 hours and then diluted to 1 E6/ml with fresh ExCell in small culture flask). This is just “down-scaling” what we use for transient transfections.

– Selection – For HEK293S GnTI- we are adding selection antibiotics 48 post transfection (when the cell density is 3-4 E6/ml out of initial 1 E6/ml), if you add antibiotics just after transfection, the pool usually dies. As for concentrations – puromycine at 5ug/ml (could be lower), G418 at 50 ug/ml – which is quite low, but sufficient to kill untransfected cells and gently enough to HEK293S GnTI-. It takes around 14 day to re-establish the cell growth.

– Be careful that HEK293T was originally transfected, so they are also resistant to G418 (but if you are not cultivating them routinely under selection pressure, it will take them some time to grow nicely again), but PB has a solution for HEK293T as well as for HEK293E – you can use different helper plasmid – PB-RB (with blasticidine resistance, we can provide it) instead of PB-RN.

We are routinely using HEK293S GnTI- (with G418 and puromycine) as well as HEK293T (with blasticidine and puromycine, both at 5 ug/ml) for PB pools. If I compare these two cell lines, I have to say that HEK293S is very sensitive, so you have to lower antibiotics concentrations, you have to wait with selection and it takes 2-3 times longer to have growing pool.

I hope this might help, of course I will try to answer further questions.

Best,

Barbora