How to interpret ELISA results - Antibody titre from serum (Sep/14/2010 )

Hi!

I'm setting up an ELISA to look at antibody titres in mouse serum and I've never done this before (nor has anyone in my lab as we're a plant lab).

This is the background:
I have tobacco plants expressing a bacterial antigen. I want to see if that antigen can elicit an immune response. So I have five groups of mice as follows:

1. Mice gavaged with soluble fraction of tobacco extract (containing 100ug of the antigen) and CTB as an adjuvant
2. Mice gavaged with soluble fraction of wild-type tobacco extract and CTB (this is my negative control)
3. As group 1, with a subsequent sc boost of purified antigen (with alum)
4. Mice injected with purified antigen (with alum)
5. Mice injected with purified antigen (with alum) (with a few weeks' gap).

I have collected serum and BAL from these mice, but for now I just want to see the levels of antigen-specific IgG1 and IgG2a in the serum. (Later I will check the IgA in the BAL).


I know roughly how to do an ELISA (I've used it to determine how much of my antigen I have in my tobacco extracts), but I'm not sure on a few things.

1. Do I still need to leave a blank on the plate with just PBS? Or do I count my negative (i.e. wild-type sample) as my blank?
2. Do people find it better to leave the serum on the plate for 2 hours at rt or overnight at 4C?
3. Is PBST with 5% milk powder an adequate block?
4. When I get my results, how do I interpret them? How do I go from a page of absorbance values to a graph?

Sorry if I haven't been very clear, this is a bit confusing to me.

Thanks for any advice you can give me.

1. It costs you nothing to include blank with PBS and you get information how high really is the background from the negative control.
2. When I was working with rabbit serum or ammonium sulphate purified antibodies and plant proteins (MW ~40-70kDa) I tried both and haven't noticed any difference so I used one that was much more convenient for me.
3. Yes.
4. The most simple way is to assume that positive = 100%, negative = 0%, and use linear regression. Although when you make a dilution series they quite often follow sigmoidal curve and that requires more specialized software to manipulate.

K.B. on Tue Sep 14 13:34:23 2010 said:

1. It costs you nothing to include blank with PBS and you get information how high really is the background from the negative control.
2. When I was working with rabbit serum or ammonium sulphate purified antibodies and plant proteins (MW ~40-70kDa) I tried both and haven't noticed any difference so I used one that was much more convenient for me.
3. Yes.
4. The most simple way is to assume that positive = 100%, negative = 0%, and use linear regression. Although when you make a dilution series they quite often follow sigmoidal curve and that requires more specialized software to manipulate.

Table Curve (from Systat.com) is an inexpensive product that will suite your requirements. For Data analysis you can use Med Calc for a limited number of times without purchase.