I am developing an LFIA to detect a certain type of bacterium and using in-house produced antibodies for the test line as well as AuNP-mAb conjugates. The sandwich antibody pairs that we've chosen do not cross-react with each other in ELISA assay. Does anyone have any idea or comment on possible causes of false positive results? We observe both test and control lines when we run only buffer on the test system.
Thanks in advance,
Gamze Kuku, PhD
In many cases it is necessary that your buffer contains Tween 20 (typically 0.1 %) and/or BSA (typically 1 %). PBS is not sufficient to avoid nonspecific binding.
As Michael G. Weller states, adding a surfactant such as Tween-20 at a concentration of 0.01 - 0.1% along with BSA at concentrations of around 0.5 - 2.0 mg.mL-1 is common.
It is also normal to add salt, such as NaCl up to 0.5 M. Some phosphate buffered salines (such as P4417) contain a formulation to match those of the human body and therefore also include KCl. Sigma also supply P3563 which contains 0.05% Tween-20.
Dear Michael G. Weller and Jules Lloyd Hammond , thank you for your suggestions. By meaning "buffer" do you mean the sample application buffer or any of the buffers that I use in pre-treatment procedures or the buffer that contain the AuNP conjugate? I have tried Tween-20 at various concentrations in conjugate, and sample pads as well as in membrane to pre-treat these components. I then moved to IGEPAL CA-630 instead because it resulted in better background clearance and better conjugate pad release. After all my optimizations, I now use 5% IGEPAL CA-630 in the sample pad. I don't include any other detergent in other components for it decreases the stability of my conjugate.
I keep AuNP conjugate in 1% BSA and 5% sucrose in borate buffer prior to application to conjugate pad. The thing is, when I exclude the pads from the system -in dipstick format- I still see both control and test lines.
Iris Miraballes coming to your question, I apply both at 1 mg/ml in PBS containing 1% methanol.
I mean the sample application buffer. It should contain 1% of BSA.
OK. I will try and hopefully the problem will be solved.
Thank you for your kind suggestion.
Dear Michael G. Weller , I tried with BSA, with, 0.5%, 1%, and 2% the signal intesnsity was both reduced on the test and control lines but to same extent. So, practically it did not work. I also tried PVA but the situtation was the same. I am having the same problem in my work on LFIA of another bacteria type, too. It seems I have a fundamental problem but I really never faced such an issue before. I appreciate all the help and suggestions. Thanks in advance.
From my point of view, you do not have a problem with non-specific binding, but some of your antibodies do not work as expected. You cannot resolve such problems by blocking. Are you using polyclonal antibodies in a step?
Both of the antibodies against bacteria are in-house-developed monoclonal antibodies (IgG1 and IgG3) produced in mice and we have not faced any problems with their ELISA assays, no cross-reaction etc.
Perhaps you should specify more clearly, what you mean with false positives.
Is there no concentration dependency in the signal? Did you optimize the antibody and conjugate concentrations? What kind of gold nanoparticles did you use? Did you check the activity of the conjugate and the test line, e.g. with the respective ELISA reagents?
There is no concentration dependency in the signal, I am just giving blank buffer to the system and it is the same as when I give the bacteria sample.
~40 nm citrate reduced AuNPs were not conjugated electrostatically but they were conjugated covalently to the antibodies and yes we optimized their ratio.
I just have not checked the AuNP conjugate for its activity in ELISA, which I added to my worklist now. Will be back with the results. Thank you.
- Badges
- Science method