I think this may be associated with the plate used, what plate are you using (Gibco, TPP) ? The plastic used for the well plates differ and i observed TPP gives a more uniform reading.

You may do a background subtraction by subtracting the reading of the empty well from your blank and samples as well.

Hope this helps.

Hi Carmen

What wavelength are you using? What I want to know is do you do background correction for the plate itself?

For example, I am supposed to read my plates at 450nm, but to remove optical imperfections in the plate, I subtract the absorbance from 540nm (or 570nm) fromthat of 450nm. That gives very uniform readings.

Also, I am curious. Is your detection antibody directly conjugated to the HRP or is it biotinylated?

The detection antibody is directly conjugated to the HRP . I used a wavelength at 450nm

Yeah, I think you should try deducting the reading at 570nm from the reading at 450 nm. It should remove any inconsistencies that are due to the plate.

Dear Carmen

Why do you use background control in a Sandwich ELISA like that? I think it is not necessary. On the other hand, I agree with Tasfia, you can use background correction if necessary.

Dear Carmen

I suggest that you need to look at what is happening on the plate. The suggestions above are all very good but the system you use involves the same plate with different well contents hence it is unlikely that you are finding differences in the plastic. Similarly measuring the plate at different wavelengths will not change the fact that you obtain different results when the wells have different contents. Clearly the contents of the wells are the key to your problem. If the blank - capture antibody alone - is producing a higher absorbance result than the combination of capture antibody plus second antibody conjugated to HRP and the substrate, then the problem would appear to be that the capture antibody absorbs at the wavelength being used. However all the wells have the capture antibody and it is the addition of the substrate and second antibody which appears to lower the absorbance. Consider how this is possible. I assume you are using the same buffer system. I also assume that the washing steps are identical and hence the absence of the analyte should result in the removal of the antibody conjugate from the wells when they are washed prior to substrate addition, thus the enzyme-antibody conjugate should have no influence on this system. The final steps in most ELISA systems are the addition of the substrate to the washed plate, incubation to allow colour to develop and then addition of the stop agent such as sulphuric acid. Therefore the difference in absorbance must be the result of the presence of the substrate which in the final analysis is the only difference that there should be between the 2 systems being measured. You need to review your methodology to ensure this is correct. You should then also check that all the wells are being washed to the same extent and that the wells all go through the same steps. In your system the processes should end with 2 types of well - those with substrate and stop solution and those with only stop solution. Your results suggest that stop solution is absorbing light whilst stop solution plus substrate is not.

I suggest that you methodology needs revision to ensure the steps are all correct as from the arguments above the results you have can only occur if you have not correctly washed the wells between reaction steps.

Yes, I agree with Christopher. I was going to ask how many times have you repeated the experiment? Do you see the same result every time?

Also Christopher, I have seen in the absorbances at 540 or 570 nm vary between wells even in the same plate. Just from my experience I think it is always a good idea to deduct the absorbances due to inconsistencies of the plate.

I think you are asking the question in opposite way, you shld ask how to decrease the value of your background not to increase your blank control, cause the less the value of the background, the better.

Firstly, in agreement with Oscar I am suggesting not to use background control as it is not necessary when you are using the same buffer system and 570nm correction.

second, you already got several important suggestions I am not repeating, try those and check the time period you are maintaining the reaction after substrate addition, there shldnt be any visible color in the background if you are getting it reduce the time period. Lastly, I want to know what are you using to stop the reaction, any commercial stop solution or H2SO4? 

I assume that you are using TMB as the substrate? What is the OD reading of just the substrate?  What is the reading of just the substrate with stop?  These backgrounds seem high considering the TMB substrate alone should provide no backgrounds.  I would also try a less sensitive kinetic TMB as this can eliminate low positives. If you want to try some, I can send you some.

The values of the Blank in ELISA can be increased or decreased by

(1) Selecting the plate (polystyrene, polyethylene, etc.

(2) Selecting different commercial sources of the plates.

(3) using protein (hu Albumin; BSA/Caesin/)for coating buffer.

(4) adjusting the concentration of the protein and detergent solutions.

Finally, see a paper to prove the above issues.

For further questions you may reach [email protected]

J Immunol Methods. 1994 Mar 10;169(2):257-72.

Factors affecting the fine specificity and sensitivity of serum antiganglioside antibodies in ELISA.

Ravindranath MH1, Ravindranath RM, Morton DL, Graves MC.

I agree with most of the above comments. The lower the backgrounds the better as a significant increase in test samples will be a much lower reading.

What is the variation in your background reading across all the wells of a plate.  In some assays there can be an edge effect were outside wells differ from the centre wells but I have seen less of this recently.

In some assays, standing the plate full of wash buffer, say for 1 minute, between each wash can be beneficial.  This is most useful for the conjugate wash but some assays benefit at all steps.  I had one assay which had 1 minute stands for each of the 6 washes, another had 3 minute stands alternate fillings for the 6 washes.  This reduced background levels and variability.

A lot of good suggestions have already been made. You should consider the relaibility of the blank wells and whether one well or perhaps two wells are necessary. I don't see the value in the background control as well. If your assay is consistent and reliable enough you should not need this. An effective blocking step and consistent wash steps should help with keeping blank values consistently low. Wavelength correction for 450nm at 570nm should iron out plate effects seen.

Don't care that much. The reference of your cuve starts at antigen concentration = 0. All other blank, negative control are shift your calibration curve only a little bit up or down. 

You some unspecific reaction. The can reduced as described from Mepur and/or by adding the double amount of NaCl into your PBS to a final concentration of 0.3 mol/L. 

Dear Carmen:

Both your blank and background are excellent. Correct you experimental value against background value. (Exp OD - Background OK).  Then test with serial dilutions of the capture antibody.  This would be a logical strategy under the conditions described.

Alternatively, Check the substrate purity and concentration before you do the experiment. The concentration of the substrate can be lowered but i do not see that in your case.

Good Luck!

Thank you very much for the reply. It was a great help.