Protocol Online logo
Top : New Forum Archives (2009-): : ELISA and Immunoassay

How to determine the optimal concentration of a primary antibody solution - (Jan/04/2012 )

I have looked around the ELISA and Immunoassay board a little bit, but I have not been able to find a post that addresses my problem. However, I am very new to all of this, and my ability to even recognize the answers to my questions is not very good. So if my question has been dealt with elsewhere, I apologize for creating a new topic.

I am currently trying to create an indirect ELISA protocol for measuring the competitive adsorption of human albumin, gamma globulin, and fibrinogen onto various polyethylene samples. The samples are rather large (20X15X3 mm) and, in the first step, require incubation with 5-10 ml of protein solution (Albumin - 4000 ppm; gamma globulin - 1000 ppm; and fibrinogen - 300 ppm). In the second step, the samples (after washing) will be incubated with goat anti-human antibodies for albumin, gamma globulin, or fibrinogen. However, I do not have very much experience with ELISA (I have performed ELISA kit experiments before, but nothing more), and I am unsure in what concentrations I should prepare the primary antibody solutions. In references I have seen people quote concentrations ranging fro 1/400 to 1/5000. However, if I want the antibodies to be in excess of the attached protein, wouldn't I need to use a concentration greater than 1/400? Especially if my initial protein solution has a concentration of 4000 ppm (using albumin as an example)? Any help or guidance that you can provide would be very much appreciated. Thank you.

-Water Bear Minimum-

You are basically performing the step used to block microtiter plates after adsorption of antibody or protein conjugate. You have to insure that the surfaces are completely coated or your antibody and conjugate would also adsorb. I would place the samples in small test tubes, coat overnight at RT, and also have samples uncoated, samples post coated and blocked with something such as seablock (fish). Thus, test uncoated, coated, and also coated and blocked for evaluation.
Wash several times pbs/tween or pbs triton x and transfer into clean test tubes after each assay step.
Then react with ab. You will have to titer your conjugate and ab solutions. Take one condition ie albumin and run serial 1:10 dilutions of your antibody to determine the concentration you need.


If you are looking at competitive absorption, then you might consider labelling each of your proteins with I-125 or biotin or fluroscein or some other tracer and look at competition of your labelled molecule with your competition agent (s).

If you are set on detecting with antibodies, I suspect that if we base the mass calculations of required detection reagent on affinity, then we'll end up in a depletion situation (so all of it gets bound to a low proportion of available albumin, globulin, or fibrinogen molecules) simply because we are expecting large concentrations of bound material. If you can do some ballpark calculations to determine the high end concentration of bound protein, then you can do some rough calculations to determine how much detection agent to add to maintain an excess.

When you say 4000 PPM, I think we mean 4 g of albumin for each litre of solution, or 40 mg of albumin in the 10 mL volume that you use as the first step of the assay, and if we say that the surface can bind up to 1% of this, we need sufficient detection reagent to saturate that 1% or 0.4 mg of the albumin. So to saturate this with the detection reagent, we would need at least 1 mg of detection reagent (based on 60 kDa for albumin and 150 kDa for IgG). So you might have to use the detection reagent without dilution or even concentrate it or do the assay in litre sized vessels.

If these assumptions are anywhere near ballpark accurate, we are beyond the concentration range of feasibility for immunoassays. If you are able to reduce the polymer size down to 1 mm x 1 mm x 1 mm, then we might be in the realms of feasibilty.

If the expected degree of binding is vastly less than 1% then the approach in the previous post should be adopted.

-Ben Lomond-

Thank you both very much for your advice!

@Ben Lomond, We don't have the facilities for I-125 labeling, and I am concerned about fluroscent labeling possibly altering the activity of the proteins. Biotin labeling is intriguing, but I don't have any experience with it. Also, from published results, it looks as though the degree of binding will be roughly 1 to 2 orders of magnitude less than 1%. So I think ELISA is feasible.

@PAO_ahac, I think I will take your advice and test different concentrations of primary antibody in the method that you suggested.

-Water Bear Minimum-

Hello again. I am sorry, but I have a follow-up/related question.

Obviously, protein adsorption is as much a function of the protein as it is the material. However, is there a standard material used for handling and storing protein samples? Polypropylene, glass, etc.? Any help would be very much appreciated.

-Water Bear Minimum-

Purchased antibodies are sent in glass and polypropylene containers. Avoid polystyrene.

good luck. FYI seablock is made by East Coast Biologicals