Inconsistent ELISA results - (Jun/18/2009 )

Hi all, I am getting inconsistent results for repeating sandwich ELISA runs with the same sample. Two plates were run in parallel for twice (at different day) with the same sample, but all 4 results were different. The standard curve is fitting well with an R^2 of greater than 0.95. The OD reading is obviously higher at column 3, 4 and 5, 6 as compared to the other columns (7 - 12; column 1,2 are standard). What are the possible reasons? How to troubleshoot this?

Thanks

In order to answer your question, it would help to know the reagents that you are using and the outline of the protocol. Inconsistent ELISA can occur because of many reasons. What kind of value differences are you getting? What is your background reading?

you may try to

1. Calibrate the ELISA plate reader

2. use a very good multichannel pipette

I hope this may help.

kausikdatta on Jun 21 2009, 09:14 AM said:

Thanks for answering my question.

I am coating the plate with anti-mouse IgG from goat in carbonate/bicarbonate buffer, incubate for 2 h at 37 degree C. Wash 4x with NaCl/Tween 20 and block with BSA/Tween 20 overnight at 4 degree C. Standard mouse IgG and samples from cell culture supernatant were loaded after washing process as above. After incubating at 37 degree C for an hour and half, repeat the washing process. The capture Ab is anti-mouse IgG HRP from goat, diluted in PBS at 1:1000 and incubate at 32 degree C for 1 1/2 h. Wash plate as before, then load ABTS solution and incubate for 45 min at 35 degree C. Read at 405 nm.

Example of values after calculation: sample A = 0.472, 0.454, 0.362, 0.649 ug/ml; sample B = 0.749, 0.299, 1.165, 0.324 ug/ml. Is the backgroud reading here means OD reading of empty plate? If yes, it is about 0.040. If you mean blank or negative control (diluent without sample), the OD reading ranging from 0.37 - 0.54.

Minnie Mouse on Jun 22 2009, 11:45 AM said:

Thanks, answering the question already help.

I have just calibrated the multichannel pipette, but it may not be a good one. Yes, calibrating the plate reader might be one of the step that I should try.

Thank you very much.

Check your washing step, it's the most likely cause for the bad results if your pipetting is OK.
Let the wash solution soak for 30 seconds between the washes.

Gerard on Jun 22 2009, 06:44 PM said:

Yes, I had already added 30s soaking step into my washing and even turn the plate after 2 washes each time.

Can someone comment on the negative control? To me it appears the signal is very high for just buffer alone (no binding). If the negative control is a true blank it should read close to the 0 standard of the assay.

I think residual conjugate is left in the wells. Make sure you blot...bang out the plate on papper towels between each wash (manual).

Your %CV is very high for your samples...are you getting high CV for duplicates of your dose response curve as well? YOu can easily check the pipettes with an analytical balance.

good luck

sgt4boston on Jun 24 2009, 03:56 AM said:

My negative control is the culture medium, not buffer. I have tested with PBS, basal medium and culture medium as the diluent. They gave different range of standard curve, so I chose the one close to my sample, which is culture medium.

I did tap the plate on paper towel after the washing step, as dry as I can ....

Currently, %CV of my standard curve is very big, ranging from 8 - 35. The standard is normally place at column 1 and 2. %CV of column 3 and 4 is lower, at the range of 5 - 15. The rest of the columns are well below 10.

The pipettes were either new or just came back from calibration.

Ok...lets confirm you do wash several times between additions of sample, conjugate etc?

You blot out the wash between each and every wash step.

Your pipettes are calibrated. (including your multichannel pipette if you use one)?

Your %CV is very high for your standards OD values...and also for your samples.

Your sample values will be much closer if the CV of your curve is tighter....should be less than 10%.

Let us know if your samples fall within the linear range of the curve. And what the analytical range of the curve is (including the ODs for the 0 and highest point)?

You are running multiple assays...are you running frozen aliquotes of samples...are samples changing over time?

I noticed surfactant in the blocking solution...could the tween be interfering with bsa blocking the plastic?

Are you sealing your plates during those incubations? Are you mixing the plate before the incubation of reagents?