Do you have negative sera samples? I mean pre-immune sera or sera from animals not immunized with the target antigen. If so, are them negative as expected?

In principle, you dont have to include these non-coated wells, although I usually do  something similar (not with uncoated wells, but with wells coated with a non-relevant protein). But the fact is that you included this control and the results indicate a high serum non-specific background in your assay. This certainly creates serious doubts about the response you are measuring. Then, it is very important to see the results with negative  sera.

If you are measuring natural reponses (not in immunized or infected animals) it is difficult to define negative sera, but you can still do it by adding a huge excess of soluble antigen to normal serum. This should block the specific response and make the samples negative. Is that your case or you are working only with immune sera? In the second case, it would be very easy to have non-immune negative sera as controls. 

Probably you need to improve blocking in your assay. I always use skim milk (4%) to block before adding the samples and also to dilute samples and conjugate, unless there is a theoretical reason to think that milk can interfere in the assay. BSA is another option, although I prefer milk in general. If a better blocking make your samples negative perhaps this is because the responses you are measuring are just non-specific background.

Tween alone is not blocking. You need inert protein too for this, usually BSA in ELISA or something else as detailed above nicely. I think sec ab conjugate just sticks to the empty wells where you do not use any protein before, and give a strong signal. Practically, this type of sample/well is not used in ELISA, i think, mightbe i am wrong. I never use it.

Gertrudis Rojas: Thanks for your response. Actually, we are measuring natural responses, that means that the serum we use, except for the positive control, are extracted from no-immunized rats. We will try the solutions you proposed: add antigen to our serum to block the specific response and coat the wells with a different antigen which should not trigger a response. About the blocking, we have tried with or without BSA blocking, and different incubation time. In all the contexts, the results were the same: high results in unocated wells.

Attila Lehotzky: Thanks for your response. As you mentioned, we also thought that the high response in these "unocated" wells was due to the sticking of the secondary antibody, so we did another ELISA skipping the coating and rat serum steps, we added secondary antibody in the budder and then added the substrate, and any response was observed... thus we confirmed that the high signal was no the result of the second antibody sticking.

Thamsanqa Emmanuel Chiliza : Thanks for your response. So you think the high signal is normal in this case? What we observed could be the general immunoglubulinemic response ? As I mentioned before, when we block with BSA, we had the same pattern of results: high signals in uncoated wells. 

As the problem certainly comes from the rat serum properties; we also tried to centrifugate them before adding them to the wells. the results were the same.

AFter reading all your responses, where you said that these wells are normally not included, my question is : should I take the results of the uncoated wells into consideration, i mean do they really put in doubt the results of the "antigen-coated"wells? 

if you use "normal" serum to detect some natural response, i think your dilutions are quite low... rat serum contains about 50 to 70 mg IgG, you should well lower the dilution values, and in your case it is always better to investigate dilution curves to detect (immune) response.  Might be i am wrong. What is the maximal dilution of your positive control serum where it still gives signal?

I am worriing a little bit, whether your response is antigen specific. Untill you do not got negative response as wrote by Dr Rojas against an inert protein, there could be some unspecific issue. Again, mightbe i am wrong, or just did not understand something well.

Attila Lehotzky:

So If I understand i should use more diluted serum ? What could be a good dilution factor for you? The dilution of the positive control I use is 1:2000, but we have to take into consideration that the positive control is serum extracted from immunized rats which is not a natural response, obvisouly the quantity of specific antibodies is higher than the other serum we used to measure natural response. 

We will try other coatings, such as BSA, milk and gelatin coatings, so if we observe high signals, it will be possible to conclude about the specificity of our response. Do you consider these coatings appropriated or should we use different ones? 

 Also, we did a western to check the specificity of the serum to the antigen and we got bands, however it was really dirty.But it was a promising result, we just need to decrease those negative controls or get rid of them if we could... 

The problem when measuring natural responses is that you normally have to use low sera dilutions and sera have a huge amount of antibodies that can produce non-specific signals against almost everything. So you need to be extremely careful with your assay conditions. If you are able to remove any indication of non-specific background and still detect a response, this can be considereded a true natural response. I would use as the control the same ELISA, but coated with an antigen towards which you dont expect to have a natural response as a control (better than uncoated wells). And also, I would block with skim milk (not BSA), but I would also include milk in the dilution buffer for sera and conjugate. Casein (the milk protein) is one of the best blocking agents you can use. If there is still a specific response and this response can be abrogated by pre-incubating sera with your antigen, the chances that you are detecting a real natural antibody response are very high. Otherwise, using low dilutions and no blocking agent, the chances that you are just detecting background are too high. Even though uncoated wells are not normally used, I think  you should pay attention to this background problem you found.   

 Thanks Gertrudis Rojas for all the help. Would you have any protocol for the pre-incubation sera with the antigen? 

thanks a lot

Could be just adding the antigen (as concentrated as possible) to  already diluted sera including the blocking agent and incubate the mixture at your assay temperature (I guess you use RT in ELISA) for 2h. Then, the samples will be ready to be added to the coated and blocked ELISA plate and continue the assay. The same sera samples must be treated in the same conditions, but without adding the antigen (adding only the same volume of the antigen buffer) to assess maximal reactivity.

It is suggested to first determine optimal conditions for ELISA by checker board method under defined conditions before venturing into test conditions. Lowest antigen concentration with highest dilution of primary antibody giving OD of 1.00 with blank as 0.00 should considered as optimal. For every assay on must include antigen control, antibody control and blank in addition to known positive and negative control, to know any non-specific binding. Blocking is a must with skim milk or BSA and dilution of test samples and conjugate can be made in blocking buffer.

A lot of good things already was written above. I strongly suggest to use dilution curves to detect signal of individual serum... On western blot, i would use decreasing amount of antigen in a few steps, that is the routine to check specificity and sensitivity for a given serum. Depending of your exposure type and route (if it is a planned experiment), not only IgG, but other Igs could be important in humoral response, and even the response can be effected by the route (ie daily exyposure is weaken the response rather after a certain time, etc.). And a 2,000 dilution is on the limit considering an immunized serum's usability (but it can be originated from a weak antigen). etc, etc.

Yes, (serum) proteins stick to plastic wells. That is how you are able to coat wells  with antigen in the first place! To do an ELISA correctly, you coat, wash, and block . The negative control is to coat with an irrelevant antigen. Then you add dilutions of either your antiserum or a normal serum ( control) . You then wash and develop with several  different dilutions if the the conjugated secondary to make sure you are getting a signal but not using too much. Generally 1/1000-1/5000 works.  ( read what the kit directions suggest). Wash and add substrate. 

The  specific curves should rise and plateau. You compare the curves with the relevant vs irrelevant antigen and with the antiserum vs normal serum. 

ElISA methods are described in the literature. Youn need to read up on them. 

I agree with previous comments about blocking - I'd add a suggestion that if your BSA blocking buffer did not previously work, you could try increasing the concentration of the protein in the buffer. If you expect a lot of interference in the assay, adjusting the tween/protein levels might help with this. I'd consider adding protein (e.g. bsa or casein) to the buffer that your antibody and HRP are diluted in also. Good luck!

thanks to all for your help, we will try teh different solutions you mentioned and when we get results, we will communicate them here!

Do you think that coating with gelatin or bsa could be a negative control?

It depends on your experimental animal. if you still in issue, I suggest to examine naive (pre-experimental) animals sera against different blocking proteins, and choose that one (protein) which against you get the weakest response in general. (It should be a routine during experimental work, i think.) Again, use dilution curves. Problem could be arised from the routine chow of the animals. You give a very nice question, anyway, thanks for it.

Hello, 

It is perfectly clear that you have missed a major step called BLOCKING. Use BSA or Skimmed milk powder (3 to 5 %) dissolved in PBST. If your working volumes of antigen and antibodies are 100 uL them Block the plate with minimum 200 uL. You will definitely vanish the problems of background.

Tween is a surfactant, it can not be used for blocking, Use BSA or high qulity Skimmed milk powder. It will work out fine

Hi 

I have one more question. I am planing one more elisa and i would like to include the negative control one of you proposed : i will coat the well with a inert protein, i think it will be BSA. The question is: the concentration of bsa coating should be the same that my antigen coating right, however  bsa and my antigen of interest have different  molecular weight, should i take this into account?

Thank you

Trying again.... My answer disappeared.

Just use the same ug/ml of your control coating protein. You need not match molarity. Then wash and block. 

Tween 20 is a surfactant, it will wash up the adhering proteins from the well. It is not advised to use Tween as BLOCKER. Any non reactive Albumin proteins will serve the purpose to the point of question.

could you also suggest me... how can i standardize my indirect elisa? i am geting absorbance from each well including the blank. any Suggestion? 

Use a blocking step after coating with buffer containing irrelevant proteins (could be PBS/ 4% skim milk). BSA could also be used. Dilute the primary antibodies and the conjugate in the same solution.

Try different dilutions of both the primary antibodies and the conjugate. The best dilutions will be those producing a higher signal to noise ratio (comparing a positive sample and the blank or negative control).

I dont know what is exactly your problem and the assay you are using, but these general advices about blocking and titrating the reagents to reduce background could be useful. 

For every assay you need a positive and a negative control. The positive control contains the antibodies in quest, the negative control doesn't but is otherwise as similar to the positive control as is possible. The specific signal is the difference between the positive and the negative control. You do not mention using a negative control, i.e., a normal rat serum. You will never be successful with this assay if you do not adhere to the fundamentals of testing!