From your description anove, it sounds like you are getting non-specific binding which could occur at multiple steps within your assay. I personally prefer antigen capture ELISA. That said, herein is the procedure I use for sandwich ELISA methods:

Sandwich ELISA Protocol:

1. Before the assay, both antibody preparations should be purified and one must be labeled.

2. For most applications, a polyvinylchloride (PVC) microtiter plate is best; however, consult manufacturer guidelines to determine the most appropriate type of plate for protein binding.

3. Bind the unlabeled antibody to the bottom of each well by adding approximately 50 μL of antibody solution to each well (20 μg/ml in PBS). PVC will bind approximately 100 ng/well (300 ng/cm2). The amount of antibody used will depend on the individual assay, but if maximal binding is required, use at least 1 μg/well. This is well above the capacity of the well, but the binding will occur more rapidly, and the binding solution can be saved and used again.

4. Incubate the plate overnight at 4°C to allow complete binding.

5. Wash the wells twice with PBS. A 500 ml squirt bottle is convenient. The antibody solution washes can be removed by flicking the plate over a suitable container.

6. The remaining sites for protein binding on the microtiter plate must be saturated by incubating with blocking buffer. Fill the wells to the top with 3% BSA/PBS with 0.02% sodium azide. Incubate for 2 h to overnight in a humid atmosphere at room temperature.

Note: Sodium azide is an inhibitor or horseradish peroxidase. Do not include sodium azide in buffers or wash solutions if an HRP-labeled antibody will be used for detection.

7. Wash wells twice with PBS.

8. Add 50 μL of the antigen solution to the wells (the antigen solution should be titrated). All dilutions should be done in the blocking buffer (3% BSA/PBS). Incubate for at least 2 h at room temperature in a humid atmosphere.

9. Wash the plate four times with PBS.

10.Add the labeled second antibody. The amount to be added can be determined in preliminary experiments. For accurate quantitation, the second antibody should be used in excess. All dilutions should be done in the blocking buffer.

11.Incubate for 2 h or more at room temperature in a humid atmosphere.

12.Wash with several changes of PBS.

13.Add substrate as indicated by manufacturer. After suggested incubation time has elapsed, optical densities at target wavelengths can be measured on an ELISA plate reader.

Note: Some enzyme substrates are considered hazardous, due to potential carcinogenicity. Handle with care and refer to Material Safety Data Sheets for proper handling precautions.

I include, for example, a "0 pg", low ~ 35 pg, medium 375 pg, and high 825 pg control which are different values than those on my standard curve. My "0 pg or buffer control gives absorbency values of < 0.02 and my top standard gives absorbency values of ~ 1.8 under assay conditions. All standards, controls and samples should be diluted in blocking buffer which becomes your assay buffer. Typically an ELISA assay gives most accurate results over one log of values For example 10 pg to 100 pg. You may be stretching the limits of accuracy by trying to achieve a two log assay. you will likely need to test each step for non specific binding of HRP-strepatvidin. For example, blocked well + HRP-strepatvidin, Blocked well +NGAL + HRP-strepatvidin etc. The HRP-strepatvidin may also require greater dilution. Under assay conditions titration (serial 5 fold )_of HRP-strepatvidin with each dilution evaluated against your standard curve including "0 pg / ml" should demonstrate the achievable spread within your assay. If this fails to produce results you may need to perform a box titration of each of the active components to achieve the sensitivity and specificity required.

Dear Frank,

thanks a lot for your explanations. I was doing the similar procedure step by step but the incubation time and duration and number of washing steps were not the same as you suggested. Instead of PBS with 3% BSA I am using 1% BSA and block just for 1h. Could this affect and cause nonspecific interactions? I want to ask You something else. You mention that I should titrate antibodies and antigen, so can You explain me how you perform this? Titration is usually doen with flow cytometer but I don´t have this instrument in the lab.Is there any other option to perform titration assay in order to minimize the consumption of antibodies and antigen?Thanks a lot.

Best regards.

Tatjana

I notice that your labels are in a different order than I typically used in my sandwich ELISA assays. My secondary antibody is avidin or streptaviden conjugated and my HRP is biotin labeled. The reason for this is that the avidin or streptavidin is an amplification step capable of binding four biotin- peroxidase. Thus the signal increases up to four fold. Your system will also work but I suspect you will get less amplification of signal. Under your assay conditions the antigen is titrated. That is your standard curve. I would perform two additional titrations. The first titration is the secondary antibody (anti NGAL monoclonal antibody biotynilated) and the second titration is the streptavidin peroxidase. I presume that the first and second monoclonal antibody are to two independent epitopes on NGAL or that the epitope is repeating on the molecule. If not your assay will suffer from epitope binding competition. First coat several plater with your primary antibody (50 microliter at 20 micrograms / ml of primary antibody / ml in protein free coating buffer) for 2 hours at RT. Wash and block with your blocking buffer. Row one "0" antigen NGAL protein, Row 2: 20 - 50 pg NGAL / ml, Row 3: 100 - 250 pg NGAL / ml Row 4: 500 - 1000 pg NGAL / ml (50 ul each well). The range I give is arbitrary and whatever you prepare as a standard within this range is appropriate. The point is to have 4 values which span the range of the assay. Secondary antibody titration: Typically you would start with a 1:5000 dilution of secondary antibody (anti NGAL monoclonal antibody biotynilated) and prepare serial two fold dilutions in blocking buffer. Add 50 microliter of each dilution to each of the four antigen concentrations in duplicate. Incubate one hour, wash x 5 by filling each well with blocking buffer and flicking the plate to remove wash solution. Blot plate against blotting paper to remove excess fluid from wells and add your streptaviden-biotin reagent at the concentration typically employed. React 1 hour, repeat washing steps, blot dry, add substrate, develop and stop reaction as usual. You are determining the secondary antibody (anti NGAL monoclonal antibody biotynilated) dilution which produces absorbency values which range from < 0.02 (0 antigen) to ~ 1.80 highest NGAL concentration used within measurable absorbency range.

With the determined dilution of secondary antibody and keeping this concentration and all other assay perimeters constant repeat this titration method for the HRP-strepatvidin starting with the concentration you typically employ. Serially diluting HRP-streptaviden reagent to determine minimal concentration required to produce the desired curve shape within the assay range you require.

Keep me posted on your results.

I think you should take 2 types of control. one is plate you don't coat the well with primary antibody and second is you coat the well with primary antibody but don't put antigen (you need to add only diluent). Please increase the concentration of blocking agent BSA 1% to 5%. I think by these two controls you will be able to find out what is reacting with your substrate. Please make sure to wash the plate 5 times after each step even after coating the plate with primary antibody before blocking.

You should optimize all the reagents in your assay. To start you will need to determine optimal concentrations of your detector system. This is probably why you are getting signal everywhere. Coat multiple wells with your capture antibody. I usually start at 1ug/ml. You may adjust this later. Then to half of your wells add just buffer, to the other half buffer plus an amount of target antigen that you feel you should see in your samples. For a decent antigen capture assay 500 ng/ml is probably a good place to start. Then add progressively less detector and HRP. Titrate the detector down the plate and titrate the HRP across the plate. You are looking for the best signal to noise ratio between the wells with target antigen and the wells without. This is a place to start optimizing all other aspects of the EIA in a similar manner (coating buffer, coating conc., blocking buffer, blocking conc. wash buffer, sample buffer, sample conc., detector Ab, HRP conc. etc.). If you have not done so already, I would check to make sure the washer is working as expected. If so, you may also want to add a few extra wash steps to the procedure.

Good luck

Dear all,

I try few stuffs that you suggest me. So I coat the plate (polystyrene with high binding capacity) with 50μl 5 μg/mL (I didn't have enough material to use more as Frank suggest me)primary antibody in PBS pH=7.4 . Incubation was performed during the night in the fridge. Then I block the plate with 3% BSA in PBS for 2 h. Then I titrate 50 μL antigen (I put 4 different concentration of antigen on the plate from 1000 to 125 pg/mL,dilution were made directly on the plate, serial dilution). Antigen was incubated for 2h. Then I titrate secondary antibody (make serial dilution from 1 to 0.25 μg/mL and incubate for 1h. Then I add HRP and TMB and stop solution. Volume of all reagents was 50 μL and after incubation I did washing 3 times with PBS 0.5 % Tween. I also put few different controls on the plate. One one without primary antibody,

one without antigen, one without secondary antibody. After ELISA again colour was obtained everywhere even in controls without some of the components. So I was thinking probably as most of you told me there is unspecific binding on the plate. Probably blocking buffer is still not good enough and still there are free places on the plate for binding of HRP??????Do you think that washing buffer has to much detergent inside and this detergent destroy interaction between plate and BSA, and BSA is released from the plate????? Then I did one pilot experiment to see what is the difference of different blocking agents. As Parvez suggest me I try with fat free skimmed milk (I bought one in the shop) or you suggest some special one???? I block microtiter plate with 200 μL PBS (+3% BSA), with PBS (+5% BSA) and with PBS (5% milk). Plate was blocked for 4 h at 37 C. Then I washed 3 time with PBS Tween. Then I add 100 μL HRP and incubate for 1h at 37 C. Again I washed 3 time and add 100μL TMB and stop solution. At the end there was colour in all of the wells maybe the smallest (about 0.3 and more which is really high) absorbance was in the wells blocked with PBS 5% BSA.. I am wondering should I addmore extra washing step? Why this blocking agents didn't work fine????? Should I change a plate to some which will not have high capacity ? Do you have idea what else can I try in oreder not to have unspecific binding of everything? Thanks a lot in advance.

Tatjana

For an avidin biotin sandwich reaction your conjugate is quite high ( 1 ug/ml). This is only a 1:1000 dilution of conjugate. Try titrating your conjugate against one row performed with out antigen ("0" standard") and one row with antigen ("1 ng/ml") and look for spread. Make all your dilutions in blocking buffer (3% BSA + 0.1% Tween. Tween at 0.5% may produce micelles with some proteins. Other than coating with primary antibody make all dilutions in blocking buffer. It looks like your conjugate is way to high resulting in nonspecific binding. Are you reading absorbency or viewing color development with your eye? Makes a big difference as the eye can't see what the micro titer reader will read as differences between wells. Hope this works for you

Hi,

PBST is the standard wash buffer for ELISA, containing 0.05% Tween20.

I recommend you to wash the wells at least five times in each step.

Volume of wash buffer is also important to reduce the non-specific bind.

Though I don't know how much volume you use for washing step, 200-300 microl/well is recommended even for 50 microl reaction ELISA.

Hi Tatjana,

I'm not working on ELISA test, but I'm using a similar test to detect proteases and I had the same problem than you in my first test. I had absorbances between 1,0 and 2,0 in non coated wells. My problem was the High Sensitivity-HRP I was using. I bought a normal one and the problem was solved. What kind of HRP and dilutions are you using?

Dear Solimar,

thanks a lot for this information. This could be possible reason for high background. I check the product specification and it is a high affinity HRP-streptavidin conjugate. I will try to dilute HRP more and to test again if it is not working then I should buy new one. Now I am using HRP-streptavidin 1.25 mg/mL from Invitrogen and I am diluting it 1000X but I should dilute it more at least 2500x.

Thanks a lot for this information I hope it will help me.

Dear Tatjana,

I am using simple HRP and diluting it to 10,000X but you are using HRP-streptavidin I think you should dilute it further. and please make sure to wash the plate with 300ul/well wash buffer i.e. PBST containing 0.05% Tween-20. Please make all your dilution in PBS+BSA+0.05% Tween -20. BSA concentration will be less than you used for your blocking buffer (like if you are using 5% BSA in blocking buffer then use 2.5% BSA in Diluent).

I think this will work for you.

Dear all,

thanks a lot for your advices. I try and finally a got signals on ELISA plate which looks normal (I hope so).

- Microtiter plate (Greiner Microlon 600 high affinity) was coated with 50 µL 2 µg/mL of primary antibody overnight at 4 C.

-Then plate was washed 5x with 300µL PBS +0.02% Tween.

- Blocking was done with 200µL PBS+5%BSA for 4h at 37 C on shaking platform 200rpm.

-Plate was washed 3x (I was afraid that I will wash all BSA).

-Then I add 50 µL antigen (diluted in PBS +3% BSA +0.01 % Tween ) serially diluted 1:2 starting from 1000 pg/mL and I also put 0 pg/mL (just buffer for dilution). Inucubation 2h at 37 C on shaking platform.

-Washing 6X with 300µL PBS +0.02% Tween.

-Secondary antibody-biotin labeled was diluted (diluted in PBS +3% BSA +0.01 % Tween ) on the plate serial 1:2 starting from 2 µg/ml to 0.25µg/mL and 50 µl incubated for 2h at 37 C on shaking platform.

-Washing 6X with 300µL PBS +0.02% Tween.

-HRP-streptavidin (0.25 µg/mL-2500x dilution) (diluted in PBS +3% BSA +0.01 % Tween ) 50 µL was added and incubated for 1h at 37 C on shaking platform.

-Washing 6X with 300µL PBS +0.02% Tween.

-50 µL TMB was added and wait 10 min for color development

-50 µL stop solution.

Absorbance was measured at 450 nm and 650 nm.All the measurement was done in triplicate and absorbance values at 650 nm were subtracted from 450 nm.

Calibration curve was linaer (but also I try and I could make logistic curve)

After measurement absorbance for the highest concentration of Ag-1000 pg/mL was ~0.4 (in wells with 2 µg/mL of secondary antibody) ~0.35 in wells with 1 µg/mL , ~ 0.3 in wells with 0.25 µg/mL . In the wells with 0 pg/mL absorbance was the lowest about 0.02 but this negative control is increasing in the wells with more diluted secondary antibody. How can I choose which concentration of components is optimal concentration for me?????? Should I take into account just absorbance signals and dilute as much as possible in order not to lose sensitivity and also I should take into account background (which is increasing by diluting secondary antibody)?????? Please if someone can answer I will really appreciate since all of you help me a lot.

Thanks a lot.

Your ELISA should include both positive and negative controls. Positive control wells (i.e., those that will show color) should include a standard curve response to Ag, controls testing for primary Ab binding to wells, matrix interference, specificity of the secondary Ab for Ag and HRP, and other conditions of the assay (e.g., enzyme activity). Negative controls (i.e., no color) should demonstrate the lack of response when one or more components of the test are left out (e.g., you do not want a positive color response to nothing in a well).