Hi, I am working with Hela cells and my treatment condition is always UV light. One of the assays that I work all the time is ELISA on UV photoproducts, more specifically cyclobutane pyrimidine dimers. The issue is that i cant get to make the assay robust. It seems that it was always unstable in the lab and the way it has been coped with was to basically bombard the bench with a million of experiments and hope that some will work out fine with all the controls behaving as expected and then basically consider your other samples. I am afraid this is not a possibility anymore, I am running out of time and I just need to make sense of this situation as very often than not I just spend weeks on getting my samples and then all this work goes to trash. Anyways, I am pretty confident that the cells are fine, that the irradiation step is correct, and the ELISA per se is fine. I always run 3 technical replicates per condition and they are always satisfactorily similar. I additionally trust my pipetting and the fact that the technical replicates are fine prompts to think that whatever the problem is, it is in the tubes/samples at some point in the process. The hypothesis I am now considering is that maybe UV irradiated DNA is not stable in the sense that, freezing and thawing is able to break the DNA with a articular affinity to the sites of UV photoproducts. this would explain the situation really beautifully, as I can easily see how I am loosing my epitopes in the tube when they break in a somewhat "stochastic" way. It additionally aligns perfectly with the notion that the higher the DNA yield I obtain, these samples behave better, as there is this rule of thumb rule that the higher the concentration of a chemical the more stable it is. I am of course testing this possibility empirically but I would like to know if anyone noticed this or if this is an actual notion in the field. I have tried to find the answers online but cant get any such studies and I am just so baffled that if this is the case how it is possible that no one ever told me before. Additionally, since I am thinking this way could it be that the vortexing is also doing the same thing? this would be a massive pain in the ass, as I handle many samples/experiment and i need to be super precise with the amount of DNA i load, and avoiding vortexing would make the whole thing far more time consuming and additionally could introduce variability due to the loss of precission in my nanodrop quantifications and when I prepare my dilutions from the samples.

Thanks a lot