Do you know which is the protein you are trying to quantify?

If you run your serum and run a SDS Gel, can you identify the protein by Coomasie Blue or Silver staining?

I will get back to you during the day.

Ricardo Espinola

Yes. Its S100A14 (Very small protein about 12kD). Its quit hard to distinguish that protein by Coomasie blau staining.

I am more concerned about non-specific re-activity of second Ab. I test 3 different goat anti-mouse IgG labelled with per oxidase (from different companies), and all are reacting with human sera. And reaction is really high. 

have you try to dilute further the concentrations of the second Ab?

Have you considered to do an affinity chromatography to isolate the protein?

Second Ab recommendation by company is 1:2000 and I am using 1:4000.

No. I never tried to isolate the protein. But I have purified recombination proteins (Commercial).

have you ruled out that the 70 kD protein you see in the blot is not Albumin?

the 15 kD band is the protein you are talking about. Did you perform a sequencing of that fragment ?

No I haven't sequenced the protein next to 15 kD. ... and I sure that its not proteins what I am looking for. Because I can see this faint band only with secondary Abs (That is against Mouse IgG). It means that 15kD band is result of something from second Ab.   

Secondary Ab is binding unspecifically to several proteins in H.serum. That is what I could conclude from ELISA and W.blot experiments. Most surprisingly, I tested 3 secondary Abs from different companies, and one among them is Human sera adsorbed. 

are these antibodies polyclonal or monoclonals?

This info is not available... Please see the information paper about all three products.

https://www.kpl.com/catalog/productdetail.cfm?Catalog_ID=17&Category_ID=408&Product_ID=787

http://southernbiotech.com/techbul/1190.pdf

http://tools.lifetechnologies.com/content/sfs/manuals/616420_Rev1208.pdf

I think that you have a problem with cross reaction. Goat anti-Mouse IgG labelled with HRP, probably has a strong cross reaction with human immunoglobulin because recognize a common and similar epitope present in mouse and human species (these two types of immunoglobulins (mouse and human) species are in proximity and evolutionary conserved common epitopes, are from two mammalian species).

I suggest you that you could buy an secondary antibody without cross reactions with human immunoglobulins.

Your observations are very perplexing, but I suspect artifacts are being created by diluting the human serum in alkaline conditions -- there is no reason to do this.  But overall, your ELISA design has little chance for providing accurate quantitative information because it is dependent on the amount of S100A14 bound to plastic, which is affected by irrelevant proteins competing for plastic binding sites, not necessarily the amount in serum.  The only feasible way of carrying out this protocol is to capture your protein using a pure antibody bound to plastic.  By this approach you will need a second antibody to another epitope of S100A14 followed by a third anti-secondary (or a enzyme-labelled secondary).  You would then create a standard curve with the recombinant 15KDa protein. If you don't have these antibodies, you might be able to set up a semi-quantitative assay in a Western blot format, although even this approach would be problematic because there are probably numerous 15KDa proteins in human serum, such as histones, that could compete for binding sites on the transblot.

Please correct me @Robert L Rubin, are you talking about pro-Zone / hook effect.

Firstly I was doing ELISA by coating Human sera with PBS and faced same problem. 

for the moment, I am concerned about reaction of Secondary Abs with human (non-Specific binding) and the reaction is very high. I tested 3 secondary Abs from different companies (all are Human Sera adsorbed). And all reacting with Human sera, You can see the W.blot figure (attached with question), that they are binding clearly with 3 different human proteins. 

Aaqib: I’m not questioning your observations on the reaction of anti-mouse Ig with components of human serum. I thought it might be due in part to an artifact of your stated use of high pH (denaturing) solvent, although it now seems more likely that it is merely cross-reaction of the commercial anti-mouse antibody reagents (which, may have been adsorbed only with human IgG, not whole human serum) between human and mouse serum proteins. I doubt that the manufacturers of these antibodies ever imagined they would be used in the way you are employing them, so would not have tested for such reactions. Perhaps replacing your secondary Ab with peroxidase conjugated Protein G would minimize background with spurious serum proteins, although it would react with human heavy chain, so this would probably not work in your system either. Ultimately, you cannot create a quantitative assay unless you capture your protein of interest with a specific antibody as I previously described.

@Robert L Robin,

If I coat human sera in dilution, 1:5, 1:10, 1:50, 1:100, 1:200, 1:250, 1:500 and 1:1000.

I can see one very strange thing that at lowest dilution (1:5) the read is very low (about nothing), but its increase very fast till 1:250, after that dilution makes no effect. I mean 1:250 and 1:1000 has same readouts!!!!!!!!!!!

Dilution is directly proportion to OD readouts.

No effect of dilution after after 1:250.

You couid treat the human serum so as to remove many of the interfering materials.  Look at the purification methods to see what steps you might take.  Also I would get the first antibody conjugated to a label that captures enzymes that recognize the label.

Aaqib: There are two opposing phenomena taking place as you dilute the sample: lowering the concentration of antigens in solution and reducing the amount of non-antigenic material that can compete with antigens for binding to the solid phase. At low dilution the plastic is saturated with antigenically-inert material, preventing antigenically-active material from binding to the solid phase; as the former is diluted out, some antigenic material has capacity to bind. You have several macromolecules that are antigenically active; in addition to their different concentrations, they each have characteristic affinity for plastic and for the primary antibody, so their behavior in this system can be unique. In addition there may be protein-protein interactions that mediate binding to the plastic. Furthermore, 1:1000 serum dilution is still >50 ug protein/ml, well in excess of a plastic-saturating concentration. Signal will eventfully disappear if you go to higher dilution.

Even if you had a purely mono-specific primary antibody, reacting only with S100A14, the assay protocol you want to use for quantitative determinations will not work because of competition by other proteins for binding to the plastic. You must eliminate this confounding variable by immobilizing this protein independently of other materials, and this is generally done by capturing the protein with a solid phase antibody.

Prof. R. L. Rubin 

I did an ELISA by coating plate with antibodies, but the signal was very low(near zero).

But I know where could be problem, for example I used 0.319µg/ml i.e. a high dilution.

and other I have to adjust serum dilution. I used 1:5, 1:10 and 1:50. All gave me same readouts.

Anyhow, I will coat plate with low Antibody dilution and test at least 6 serum dilutions. 

Thank you so much!! I will write back after this experiment.

I think a good option could be to label with biotin your primary antibody. It is very easy, there are some good kits from GE Healthcare. THe you can use as secondary reagent streptavidin labelled with PO. This gives cleaner blots and increases the sensitivity of the detection.

Aaqib: If your primary (capture) antibody is pure IgG (no carrier protein) and either affinity purified on the antigen or is a monoclonal antibody, 1 ug/ml should be more than enough. In general 2-5 ug/ml will be a saturating protein concentration for most polystyrene microtiter plates. However, if this is not a high quality Ab, your assay signal will remain low. If your Ab prep contains non-Ig carrier protein, you will then either have to physically isolate the IgG fraction or you could pre-coat the plate with anti-mouse IgG which will then provide the capture surface for the mouse antibody. The next problem is that you will need another anti-S100A14 antibody, preferably labeled as Luisa Villar suggested, that reacts with a different epitope on the protein than the S100A14-capturing Ab reacts with, because it is unlikely that both the capturing and detecting Abies can bind simultaneously to the same molecule. For the purpose of developing an ELISA for a macromolecule, some companies sell such paired antibodies of different fine specificities.

This false positive reaction is because of non-specific binding of human (? a kind of evolutionary) proteins (present in very minute quantities in pico/nanogram levels) with proteins of rodents, ruminants, canine and vertebrate class of animals. This is a known phenomenon, and we had observed this about 25years ago while developing antigen detection for diagnosis of infectious diseases of central nervous system (CNS) (Re: Serodiagn Infect Dis, 1990; 4; 441-450). This was overcome by incorporating normal serum of that animal species that secondary antibody was raised in. Hence I would suggest you to develop avidin-biotin based elisa in order to make it more specific and sensitive by capturing your proteins of interest by coating antibodies raised against protein of interest and detecting with mouse antibodies (against of your proteins of interest) which is biotinylated. In other words you are employing mouse (poly clonal) antibodies (raised against your protein of interest) and same antibody ( aliquot required to be biotinylated) used for detection. This reaction is finally detected by Avidin-peroxidase and subsequently color development with substrate addition. Most important step is blocking non-specific reaction between human and mouse protein. For this normal mouse serum could be included while incubating with biotinylated Abs. In every assay,  antibody, antigen control, blank, a known positive control and a known negative control should be included. In your current system you are using human, mouse and goat, naturally you 'll have all kinds of signals. For your curiosity you may check reaction between normal human serum with other animal sera on blots and compare with blocking. good luck.

Fiest, there a couple of conservative protein which can be recognized by anti-mouse-IgG-HRP (as described above by Muralidhar). Second, most of the commercial antibodies are specific for the proteins described on their label but are not purified via negative affinity chromatography on potential cross reactive species /proteins.

How to get this solved: 

1. Try to get an anti-mouse-IgG which didn't cross react. Ask the manufacturer this question: Most of them will / can not answer. One who can dos this is Seramun: http://www.seramun.com/products/antibodies/labelled.html

2. Do the negative affinity chromatography by yourself. For this purpose you can use either CNBr-Sepharose or latex particles. The latex particles should be magnetic. this makes washing etc. much more easier. Latex will bind like polystyrene microtiterplates. You need only higher concentrations due to the large surface. About 40 - 50 µg/mL (in your case: human serum) in a 0.1 mol/L carboate buffer pH 9.6. The normal serum total protein is at 70 g/L = 70 mg/mL so you can dilute the serum about 1:1500. Wash the latex 5 times with PBS 0.1% Tween 20. Put your anti-Mouse-IgG on this latex for about 30 - 90 min. The supernatant is free of cross reactive antibodies. 

Bound antibodies can be removed by 10 min incubation in 0.01 mol/L Gyline/HCl-buffer pH 2.0. Wash with this buffer 3 times plus 5 times with PBS Tween. Than you can reuse this latex for the next experiment.

Sure you have a problem of cross-reaction. The goat polyclonal antibody cross-react with human serum immunoglobulins adsorbed in the microtiter plate, remember that there are many common epitopes between mammalian immunoglobulins. Therefore I recommend you use a secondary antibody made in chicken or change your screening by this: monoclonal antibody bound directly to the plate and adding the biotinylated protein with streptavidin to reveal later.

I hope that you eliminate the background..