The only issue with selecting transfected cells via FACS is that you will select all cells expressing the reporter gene, regardless of whether the construct has stably integrated into the cell's genome, or is merely being transcribed off of a transient (unintegrated) plasmid. Maintaining the cells under selection for an extended period of time (this varies) weeds out the cells that fail to integrate, and subsequently lose, the plasmid.

The only issue with selecting transfected cells via FACS is that you will select all cells expressing the reporter gene, regardless of whether the construct has stably integrated into the cell's genome, or is merely being transcribed off of a transient (unintegrated) plasmid. Maintaining the cells under selection for an extended period of time (this varies) weeds out the cells that fail to integrate, and subsequently lose, the plasmid.

Forest is completely right. You have to combine both approaches - selection via neomycin (G418) or puromycin with subsequent flow sorting. If you sort them, try to isolate at least two different pupulations - bright and dim cells. As I often observed in my own experiments - the brightly fluorescent cells are not always the best choice, since GFP at higher levels may change biological properties of your cells of interest.

Good luck and best regards,

Darius

Theoretically you could also do FACSorting after a longer period of time (e.g. 1 - 2 weeks after transfection), when the cells that did not incorporate the plasmid into the genome degraded the transiently expressed plasmid. However, there are several drawbacks: 1) the percentage of stable transfectants is very low - so you have to sort very rare events (and need a perfect sorting strategy); 2) your sorted cells will be of mixed clonal origin.

Therefore it is better to use the selection strategy: usually First you linearize the plasmid (to increase the probability of integration into the genome; the cutting site should be between the expression cassettes - i.e. in front of the promoter of the gene of interest or behind the Poly-A site. Then you transfect the linearized plasmid with your method of choice. 1 d after transfection you start with the selection (usually G418 at a concentration known to kill your target cells). The killing takes several days (because it affects only dividing cells). Approx. 2 weeks after the transfection you can see colonies of cells that survived the selection (which originate from individual transfected cells) - and you can check, whether they are fluorescent. The fluorescent colonies can be picked (e.g. by careful scraping and sucking with a pipette tip under the microscope). Then you transfer the colony in a new well (with a drop of trypsin solution to separate the cells for 5 min - followed by addition of medium) - this results in a transfectant clone originating from a single cell (with only this specific genomic integration). You can pick several clones and test them for expression etc.

Good luck and kind regards, Hannes

The MCF10A cells are very sticky ans it is difficult to harvest them as single cell. I always have difficulty to sort these cell. It's will be much better to to try as recommended above.

Hi Jack

You've got some excellent advice above. I've made stable lines many times and it (generally) works well (although now I do it all using lenti). The only input I can add is to make sure you do a kill curve using a range of doses of whatever selection drug you choose - this will give you your selection dose and the time-frame necessary to kill all the non-integrated clones. Some people, once they have selected their clones, then move this dose down to around 1/10 of the selection dose to a 'maintenance' dose to increase the growth rate of the cells but to still maintain a small amount of selective pressure on the cells. I think whether to keep selection, move to maintenance or remove drug alltogether is still a matter that you need to debate yourself dependent on your own experimental conditions.

Also, depending on your final application, you may not necessarily want to pick individual clones. My rationale here is that all the clones that grow out will be puro or G418 resistant, however this has no bearing on their ability to express your F-actin-GFP, and gives you no information of levels of GFP expression (which may be very important depending on your imaging setup). Pre-linearisation of your plasmid backbone will increase the chance that your GFP expression is preserved during integration, however ultimately you will have to screen your resultant cells for GFP expression by some means (either qualitatively by microscopy when you pick single clones, or quantitatively by FACS for either individual clones or a pool of clones). Years back I used to pick single clones, but my experience is that single clones often show considerally poorer fitness, greater clone-clone variation and aren't as representative of the parental polyclonal population of MCF10A's that you started with. The approach I moved to was to select with my drug (4d with puro) then, once I had killed off most of the non-integrants, flow sort based on GFP expression for a level of expression that I required (generally I chose a broad range of GFP to give me as polyclonal a population as possible, so for example I might gate cells with a fluorescence intensity of >10 fold over background). This has worked very well for me and allows you to retain the heterogeneity within the parent population (of which there is plenty in MCF10A's.. both genetic and epigenetic).

Hopefully this helps.

Good luck

Al

I definitely second Alasdair's advice. The kill curve is needed to optimize your protocol & the selection of polyclonal populations can frequently save you a lot of time and effort ahead.

It is possible to select clones by FACS sorting. I did this several times. To get monoclonal cultures, just sort into 96 well plates, one cell per well. In my experience, linerization is not necessary, but increases the amount of stable clones. For my cells, this does not last longer than selection via G418/antibiotics. There is one advance for sorting: you will select for cells expressing your gene of interest, not for cells expressing the resistance.

Thanks everyone. I need to do a lot more reading. I have a physics background not a biology background ;)