We have been developing an assay for sometime. However now we are about to go into validation we are questioning if it is good enough. 10% of the assays have random wells with high inter-well CV%. We use triplicate wells. This can be as high as 50%. Or just above our limit of 15%. Normal inter-well cv% are really low ( 6 months in small single use aliquots stored at - 80 where low Cvs are through the whole plate.
The assay uses Greiner medium bind.
The antigen is morphine conjugated to BSA coated in carbonate buffer 37°C for one hour at 1 ug/mL
. Block is 0.1% BSA.
Washes three times with statmatic, 300 uL PBS 0.5% Tween20 between each step
primary is a polyclonal rabbit anti morphine 1 hour shaking Room temp
secondary anti-rabbit HRP 1 hour shaking room temp
TMB/Acid
I am about to go through all the numbers to decide whether it is worth proceeding or back to change some aspect of the assay. Of course is business issue of delay to project vs risk of continuing.
What could it be?. Last time it happened the operator reported seeing a little high intensity dot of blue in the TMB incubation in a well that later was twice the OD of it's sister wells. However I cannot say that this was observed in the previous occasions.
It does seem to happen in little spurts, so for three days it is observed, then nothing for two weeks.
Tough to troubleshoot from afar. But here are some notes:
I used Costar 9018 high-binding plates for all my ELISAs.
I assume you did an antigen binding profile to determine saturation, but a lot of my antigens bound to plates at much higher concentration (10 ug/ml, sometimes more) and I was using different high-binding plates.
Antigen binding 1 hr 37C in carbonate buffer should be fine, but I bound antigen overnight at 4C in PBS (100 uL).
Blocked with 1% BSA in PBS (150 uL) 1hr 37C. Employed Sigma A-7906 BSA; found that different BSA sources can differ (solution 0.8 micron filtered for 4C storage).
Washed 3x 300 uL PBS, 0.01% Tween 20 between additions.
Later dilution buffers varied according to protocol, but concentration-saturation curves were obtained beforehand.
Otherwise your procedures appear workable and standard. I had the most trouble with well to well variance due to plates. Good luck.
Random wells with high signals are often caused by particles, either in the reagents used or in the samples. Try to filtrate or centrifuge (high speed!) all solutions carefully. Some reagents tend to form particles during storage. In addition, if the manufacturing of the plates is not performed in a cleanroom, the particles might be introduced by air.
We decided to re-examine the assay. And so far so good. We switched form one hour coat to an overnight coat, as we thought the coating step was most likely to cause this issue. I will update later as to whether it did fix the problem, hopefully this might help someone else. In fact we are going to default back to overnight coats for all future assays. 37 C for one hour is convenient, can do an assay on a Monday, but this is not the first assay we have had strange random things happen within it, which we never saw when we coated overnight.
- Badges
- Science topic
- Similar topics
- Engineering
- Materials Engineering
- Coating