I am using an ELISA kit that states that the supernatant must be diluted 1/40 with dilutent buffer. I have previously tried a diluted and undiluted sample in the same kit and yielded a negative result with the diluted and a positive result with the undiluted sample. What is the best method to use going forward?
Probably when you use undiluted supernatant, the chances of having a high non-specific background are high and this is the reason why they recommend to use the dilution buffer, which should contain blocking components. In that case, what you are detecting in the undiluted sample is non-specific (your analyte is not present) and you should continue using the recommended protocol. To confirm that possibility, I would add an undiluted supernatant similar to the sample but known to be analyte-free, if possible. If there background signal, you would know for sure that undiluted samples cannot be used.
Alternatively, the concentration of your analyte in the sample is lower than expected and it cannot be detected after dilution. If you have some of the analyte, you could add it to analyte-free supernatant to make an artifical sample of known concentration. Using this, you could determine what is the range of concentrations of the analyte in the supernatant that can be detected using different dilutions.
You probably have a low analyte concentration, below the detection limit. I would prepare a series of dilutions with your sample to try to bring the concentration into the detection range of the ELISA. If you are successful, you will obviously have to remember account for the dilution factor when calculating your final results
One reason the kit instructions say to dilute out samples that much is to minimise matrix components such as other proteins, surfactants, etc. which may interfere with quantification giving inaccurate results. Gertrudis' suggestion to also test an undiluted negative control sample will help if the matrix is causing inaccuracy on the high side (high background), but not if it's causing inaccuracy on the low side (interfering with binding of your antigen of interest).
One way around this, if 1:40 dilution brings your sample below the limit of detection, is to dilute the kit standards in the same matrix as your samples. For instance if your samples are from tissue culture, dilute the kit standards in the tissue culture medium. At my previous company we've had to try this on vaccine samples for impurity assays where the impurity, by definition, ought to be at extremely low concentrations, so we diluted the kit standards in the same buffer formulation as the vaccine.
I agree with all recommendations above. How is your standard curve? If the standard curve is OK and your 1:40 (negative) as recommended and undiluted as well negative, your ELISA is below the level for detection if your analyte is present in the supernatant. Is your supernatant positive on Western-Blotting?
In the past, trying to measure IL-4 in plasma with an expensive ELISA from R&D the data was consistently negative. An alternative method for us and I do not know if possible for your analyte is to use the Rolling-Circle Amplification method (multiplex assay) and the data was then positive although indeed below the detection of the commercially available ELISA.
- Badges
- Science method