Are you working with normal or cancer-associated fibroblast (CAF). Antibodies to several markers which distinguish CAF from its normal counterpart.

I am comparing two non cancerous fibroblast. One with a point mutation in a gene. I am trying to see if it effects the levels of cmyc in the cell. I will also access by western blot

Hi Lauren,

   This is a long-winded answer because I REALLY don't want to grade exams.

   Immunostaining typically is not great for quantifying subtle changes in protein levels, however, it can be done.  If you are looking for changes in the sub-cellular distribution of your proteins of interest, immunolabeling is an excellent approach.  If your goal is just to detect differences in protein levels across treatment conditions, your best bet might be to do some careful western blotting using a loading control that you know remains constant in your cells.

    If all you want to do is to compare relative levels of your proteins of interest between culture conditions (for example, untreated control vs drug-treated cell cultures), then you may be able to do it by immunostaining, as long as the protein level changes aren't too subtle.

   Because immunolabeling is dependent on the local concentration of the antigen, it may be possible to titrate your primary antibody concentration to a point where you can detect differences in labeling intensity between your treatment conditions. To do this I would recommend running a dilution series with your primary antibody on a set of control and experimentally-treated cultures in parallel. You might try staining the cultures using a series of different primary antibody dilutions at regular intervals. For example, if your primary antibody normally works at 1:200 dilution, you might try staining a series of control cultures and experimental cultures in parallel using your primary antibody at 1:20, 1:50, 1:100, 1:200, 1:500, 1:1000, 1:2000, 1:500, and 1:10,000 to determine the threshold at which your immunolabeling becomes undetectable in your control vs. experimentally-treated cultures. (All secondary antibody conditions have to be held constant across all the cultures). A difference in this threshold between control and experimental cultures would indicate a change in protein level. If the dilution for the detection threshold for the experimental cultures is higher than for the controls (for example, if the detection threshold is at 1:1000 for controls and at 1: 5000 for experimentals), this would indicate that the amount of the protein of interest in the experimental cultures is increased compared to controls because it can be detected using much less antibody.  If the dilution for the detection threshold for the experimental cultures is lower than for the controls (for example, if the detection threshhold is at 1:1000 for controls and at 1: 20 for experimentals), this would indicate that the amount of the protein of interest in the experimental cultures is decreased compared to controls, because it requires much more primary antibody in order to detect it.

    As it turns out psychophysical detection by an individual observer tends to be quite consistent (this means that if you look at all the cultures, you will tend to be internally consistent with what you are calling the detection threshold). Somebody else in your lab might put that threshold at a different level.  If you use the same observer's observations over several trials, you can reach a statistically legitimate conclusion.  All that said, your reviewers will probably demand that you measure labeling intensity more objectively using intensity measurements taken from a series of micrographs captured using constant acquisition settings, so you might as well take your images that way from the start.

    If you get a massive redistribution of your protein within your cells between the control and experimental conditions, all bets are off. Its impossible to compare the labeling intensity very well under those conditions to determine overall concentrations, although you might be able to say something about local concentrations at specific structures (for example if you were examining the levels of some protein specifically at focal adhesions, the detection threshold might still be useful).

     I would recommend a method like immunofluorescence where the amplification of your signal should be linear, rather than enzyme-based visualization (such as ABC), in which the amplification step can be very non-linear and exagerate differences.

   If you need to do something more quantitative with these measurements, you might take a look at the following paper, which did a beautiful job of quantifying amino acid neurotransmitter levels across cells in the retina:

Patterns of glutamate immunoreactivity in the goldfish retina.

Marc RE, Liu WL, Kalloniatis M, Raiguel SF, van Haesendonck E.

J Neurosci. 1990 Dec;10(12):4006-34.

Good Luck,

Dave Sherry