Hi,
we are trying to use a fluormeter to quantify GFP expression in transfected mammalian cells. The cells are mainly attached, with a few detached. Does anyone have any experience with this? In particular, I am wondering if it is important to lyse the cells to release the GFP into solution. If so, any suggestions on lysis buffers would be welcome.
atb, Liam
It is important to keep in mind that GFP is very pH sensitive. The level of fluorescence is depending on the pH of the intracellular compartment your (chimeric?) protein is expressed in, and may change during the time after expression depending on where in the production line it is if it is targeted to a low-pH compartment, and depending on the physiological regulation status of the cell. There are other fluorescent proteins that can be expressed that are more pH-stable. Furthermore: Dying cells autofluoresce in all channels, and with higher intensities than most expressed constructs. So you need to control for that.
Why don't you use flow cytometer to quantify the transfected cells? By doing this, you can have information regarding the percentage of transfected cells and GFP expression of each individual cell (by looking at the fluorescence intensity of each cell, that is an indirect measure of how many GFP molecules you have per cell). Of course, you have to take in account the problem that Tanja told you. GFP can be read in every flow cytometer using the FL1 channel (The same used to read FITC). I did this kind of analysis in human transfected T cells, cell lines and dendritic cells, and it worked pretty well.
Hi Mariana, thanks for the suggestion. We are going to do that; however, I am not sure it will be accurate.
We want a measure of total GFP production in a well, not GFP per certain number of surviving cells.
Some of the conditions we use result in toxic effects on cells, and in this case, I think FC will over estimate production (because the lysed cells will be invisible).
There are several Kits available to quantify GFP-expression with a 96-well plate reader. Lysis is definetly necessary to get a mean value of GFP, even if there are several platereaders which can measure a greater area of each well. Lysisbuffer should contain protease inhibitors. Regarding detergents it is more important to concentrate on a solution with low autofluorescence than on the stability of GFP since GFP is very stable. A good buffering system should be included as well like Tanja Kögel said.
Yes, I agree with Benjamin Gottschalk. Protease inhibitors and strictly low temperatures during the procedure are very important because proteases are released from confined organelles by cell lysis. Therefore, also here, you need to control the time you analyze after lysis. However, since GFP is very stable, maybe not the most strict regime on protease inhibitors is necessary. With a less strict regime you need to keep in mind that the GFP might survive the proteases, but they might chew away the remainder of the chimeric protein. So you might want to consider controlling the molecular weight of the protein by a westen blot.
If quantification becomes difficult because of autofluorescense of dying cells, and you want to use some money on the procedure, you could consider a purification step with antibodies. Through a column, for example.
But if you only need a rough estimation you might want to go for the flow cytometry.
Autofluorescence in the GFP/green spectrum is high in some cell types, for example HeLa. Autofluorescense is lower in the red color spectrum, so using a red color might provide a better signal/noise ratio.
If you need only relative quantitation, microplate fluorometer meaurements in live cells is useful, since fluorescence can be a surrogate indicator of functional expression. You can use non- transfected cells as a blank to account for non-specific background.
Another possibility is to fix and stain for GFP using mAbs labeled in another channel, say TRITC or Cy5. You can use a stable GFP expressing cell line to calibrate the sognal intensity correlation between GFP signals and 'stained' signals. In a lot of situations, though, fixed GFP fluor is lower yield than live, so you may want to compare live vs fixed data.
Plus (in addition to Chinten James Lin's post), membranes need to be permeabilized to reach the protein by the antibody if the protein is not expressed on the surface. However, permeablizing (and fixation) may loose cytoplasmic pools, and different detergents/exposure times will permeabilize different membranes (cell, nuclear, endomembrane systems) to different degrees.
Does anyone knows hoW to do all of this data analysis using the FCS Express 5 Software? I have trouble creating the appropiate gate and having all the data for GFP.
Hi Ana - the team at De Novo Software would be happy to help you with this type of data analysis in FCS Express 5. Please contact us directly at [email protected] to schedule a help session with our team of experts.
- Badges
- Science method