Plasmid vs Lentivirus as a transfection vector - Opinions? (Jan/12/2008 )
I'm looking into creating a stably transfected variant of a few human tumor cell line that I'll eventually use for xenograft experiments. I'd like to insert a tag for in vivo imaging (IVIS). I'm thinking GFP or luciferase would be suitable, but I'm not sure what is a better vector. I have a pEGFP plasmid left over from a earlier student in the lab that I could use with lipofectamine, but someone suggested a lentiviral GFP would be better. My specific questions are:
Would I expect a difference in transfection stability, assuming G418 selection?
After selecting G418-resistant clones, can I remove the G418 before xenograft?
Would one or the other have a greater trasfection efficiency?
Any opinion would be appreciated! Cheers-JAH
lentivirus are good for cells that are hard to trasnfect using lipids. the choice of reporter system depends on which imagine system do you have. Luc is better and more quatitative.
For your specific questions:
1) stability is more or less gene specific. ie. is the gene you want to express grwoth promoting or inhibitory. Cells will drop the inhibitory gene over time if you dont use a regulated expression system. Both routes can generate stable cell lines.
2) Yes, you can remove selection pressure for a few generations.
3) Its cell type dependent. Lenti have more chance to work but making virus is more exprensive and takes longer to get the virus.
Would I expect a difference in transfection stability, assuming G418 selection?
i don't strickly remember the mechanism of virus insertion in genomic dna but i still belive that the DNA is circular or should be opened for linear insertion. In viruses, the opening point is quite strict, and does not interferes either with your selection gene or the gene of interest. So it's better than plasmids in this way, as the opening point wich is not compatible with the selectio marker may occur in the gene of interest.
After selecting G418-resistant clones, can I remove the G418 before xenograft?
yes you can remove it but for few generations as mentioned previously.
Would one or the other have a greater trasfection efficiency?
it depends on the cell type. for hela 293 and such cells, transfection by plasmids gets a high efficiency. On primary cell lines, transfection has a poor efficiency and lentiviruses have better delivery efficiency in general. so retroviruses are better choice (if you can afford costs)
My preference would be to go with plasmid transfection than lentiviral infection as virus preps are more time consuming and an additional step. Also to choose between GFP and luciferase, it depends on the experiments you wish to do. if you wish to use them for invivo, then having gfp will help in easy readout. But something quantititaive, then luc might be better.
In terms of stability, both methods should be kind of similar.
One can get high transfection efficiency with both systems, ofcourse depending on cell types. but something like 293 or hela, it should be fine.
Thanks again for all of your comments. I think for my purposes, I don't necessarily need quantitative data, so I think GFP will be sufficient. Basically, I'm orthotopically inoculating human rhabdomyosarcoma cells into the thigh muscle of nude mice and would like to be able to verify tumor engraftment after injection and localize metastases, should they occur; I'd also like to perform FFPE histology with the affected tissues, so I think GFP will facilitate this better than a luciferase tag. I've not tried to transfect anything into these cells so I don't know yet if thay are as genehunter-1 put it "hard to trasnfect using lipids". Reading these responses, I'm inclined to try the plasmid vector first; As such a few more questions came to mind....
"Yes, you can remove selection pressure for a few generations." (genehunter-1, regarding G418 selection)
"yes you can remove it but for few generations as mentioned previously." (fred_33, regarding G-418 selection)
1) I can't very well put G418 in to mice though....should I just wean the selected cells off of the G418 prior to inoculation, and just hope they don't kick out the plasmid?
Quoth fred_33:"i still belive that the DNA is circular or should be opened for linear insertion.....So it's better than plasmids in this way, as the opening point which is not compatible with the selection marker may occur in the gene of interest."
1) So, should I compare a linearized plasmid in parallel with a circular/supercoiled one?
2) Assuming the plasmid is linearized by RE digestion, is there a particular end (blunt/sticky/overhang) that I should strive for?
3) Where should the digestion site be...obviously not through the GFP or selection gene or their promoter/regulatory regions so as not to alter the ORF?
Quoth genehunter-1 "is the gene you want to express grwoth promoting or inhibitory. Cells will drop the inhibitory gene over time if you dont use a regulated expression system. "
- I don't believe the plasmid I have is explicitly responsive to growth regulatory process...I believe the promoter was cloned from a 'housekeeping' gene, though I'm trying to figure out what this regulatory element is. Anyway, the cells I'm going to try are show very aggressive growth, doubling in ~18h, and do not appear to be contact inhibited at all. But I'm not sure if the plasmid harbors something that might change this behavior, but will be sure to compare them vs. non-transfected cells before in-vivo use.