The high background could be due to non-specific binding of the 2 antibodies. it also good to note that performing a 2 step ELISA can be challenging. Since you did not add the antigen, but later added the antibody, the antibody especially when purified,  can also bind the ELISA plate, bind the secondary antibody, hence give the background. Having said that, high background could be due a number of reasons like the type of ELISA plate, the reagents used etc. Try to use a different batch of reagents, trying one reagent at a time to see where the problem is.  Your negative control should be coated too, with a negative antigen.

You could speculate that not coating wells could leave more "space" for unspecific binding of antibodies to the well. I actually don´t know if omitting coating-buffer (without antigen) could effect physical properties of the well - compared to adding coating buffer without antigen. Don´t know which of these you did?

Anyway, your negative control should be situation 2 as you describe (dilution buffer without analyte). 

It is true that non-coating wells are not a usual negative control. But you included them and the result is telling you that your sample antibodies can bind in a non-specific fashion. That could be due to two main reasons

1. The antibodies are too concentrated. Try to dilute them more. Any molecule, if its concentration is huge, can in principle bind to anything.

2. You are not properly blocking non-specific interactions of the sample. You should dilute the antibodies in the same blocking buffer you are using to prevent non-specific interactions. And if BSA doesnt work, I would suggest to switch to 4% milk as blocking buffer, and aslo to dilute primary and secondary antibodies. 

I dont use normally uncoated wells as control, but I always use wells coated with an unrelated antigen. In many cases (for instance when using phage-displayed antibodies) this control tells you that the interactions you are detecting are just non-specific background, at least for the lowest dilutions. This is not an unusual finding, You can also see many previous discussions about that in researchgate. I think it is exactly the same with uncoated wells.

Removing this background by diluting more the samples or changing the blocking solution is crucial to obtain reliable results. But if you dont use this kind of control, you just ignore the problem, which (again) is quite common.

The problem you are encountering may be higher cocentrations of primary and secondary antibodies you might be using in your assays. You may over come by titreting and checking them one by one simply using blocked plates without antigen coated to look for any non-specific binding persist. If you find non-specific binding you prepare different dilutions of first primary antibody the dilution of secondary antibody to maitain what you were using and perform the assay. If you encounter the same problem then it suggest that the dilution of second antibody being used in the assay is concentrated. You can further dilute second antibody and performe the assay in blocked plates to know non specific binding only. After optimization of primary and secondary antibody dilutions perform the assay.

For further clarification pl. get back to me. 

With best wishes

Yes I also agree with Caroline  that your negative control should be coated with coating buffer without antigen. You may try following:

1. find out the exact dilution of antibody

2. block non-specific interactions with 3% BSA or with 5% skimmed milk for 2 hours.

Hi, your description of the different wells is conceptual, but the actual complete details of the steps would be needed to make a judgement on the possible origin of the problem. Several colleagues have already pointed out that a proper dilution pre-screening of all antibodies should help, and those you can select by preparing complete curves for the 'step' and choosing the dilutions that provide the best signal to noise ratios with a minimum of saturation of the system. I hope this helps!

To your credit you have uncovered a serious problem with the assay – you cannot get meaningful results in ELISA if there is substantial background binding.  And what you call “actual negative control” provides a false sense of confidence in the assay. But as Dr. Lamonte has pointed out, you do not provide enough detail to solve this problem, i.e., incubation durations, antibody diluents, washing solutions and volumes.  My guess is that most of the non-specific binding is due to poor blocking.  BSA is not a particularly good blocking agent, and while non-fat milk is, as Drs. Rojas and Nagar suggest, there can be some materials in this crude solution that can lead to idiosyncratic results with some antibodies.  My suggestion is to use 0.1% gelatin as the sole blocking solution and to add gelatin, bovine gamma globulin and 0.05% tween-20 (along with BSA) to the antibody diluents.

Thank you all for your suggestions.

I tried 2% BSA, 3% BSA and 5% BSA. I tried two hours and over night blocking time. All the situations given the similar result. Assay was working fine with all other controls , test samples and spike samples as expected.

Only the control without coating is showing high back ground. I know that uncoated ELISA control never be a right control for ELISA.

I will try using

1. Unrelated antigen for coating

2. Using various dilutions of antibody.

Will update you the results.

Thanks

Jaya

Jaya:  While you could use an “unrelated antigen” as a negative control as suggested, your problem is the plate is not properly blocked to prevent binding of primary and/or secondary antibody.  By luck you might find a protein that could prevent this non-specific binding, but what would be the rationale for this search?  A gelatin solution at 0.1% is an overall excellent blocking agent (post-coating agent) if prepared properly as follows:

Gelatin Stock Solution (1%):  Dissolve 10 g of gelatin (1-2124 Type B, J. T. Baker Chemical Co./Avantor Performance Materials) in 1 liter of water by heating (not boiling).  Store frozen in 100 ml aliquots.

Gelatin Postcoating Solution (0.1% gelatin in PBS with 0.01% thimerosal as antibiotic if desired):  Mix 100 ml of gelatin stock solution (preheated to 50 to 60°C) with 880 ml of water and 20 ml of stock (0.5 M) phosphate buffer, pH 7.2.  Dissolve 8.2 g of NaC1 and 0.1 g of thimerosal in this solution.  Alternatively, diluting 100 ml 1% gelatin in 900 ml PBS would be OK.

Dear Jaya

Can you tell me about your results update? What did you find out after trying all the suggestions? I am facing the same. Need your help.

Thank you.

Hi Jaya

I am also facing the same issue with my assay. Can you please share how you finally solved the issue?

Thanks

I think the problem is not how you block the wells before you adding the first antibody, 1-5%BSA should all work well. the problem is on the first antibody, which might be not diluted enough, as my experience it should at least be diluted by your blocking buffer more than 20 times, otherwise increase the BSA and tween concentration when you dilute the antibody. hope this can help you.