This might be a naive question but it puzzles me sometimes
The reference channel often has no protein immobilized on it and its purpose is reference subtraction from the data from other channel where protein is immobilized. Thinking about what happens on the chip : Channel with protein has more surface density than reference channel which means the baseline RU levels will be higher for channel with protein on it. In case of a analyte which may or may not has some non specific binding , how does the difference in surface density is adjusted?
Should n't reference channel also have surface blocked with unrelated protein ( of similar MW) to improve the quality of data ?
All of your kind suggestions are much appreciated.
I am not sure, whether I got your question right, but I hope my answer will help you:
The purpose of the reference cell is to correct the signal from the reaction cell for the bulk refractive index. Thus the resulting reference cell subtracted signal ideally doesn't have any more bulk shift. Non-specific binding to the reference cell is also subtracted although in most cases refining your experiment is preferable.
It is not necessary to take the differences in baseline RU into account because you will most likely align your sensorgrams before data analysis, which means that you define a portion of the baseline before start of the injection as 0 RU (getting the relative response). (Also double referencing is standard procedure: by subtracting the averaged blank sensorgrams from the reference cell subtracted data you will eliminate any baseline drift and injection artifacts resulting from e.g. valve switching.)
All what you said makes sense. I think I did not frame my question directly.
Lets consider Ni-NTA chip system. The WT protein is functional protein fit for binding and mutant protein can not bind to analytes. WT protein go on Ch1 . for referencing which system is better
A reference channel which is not activated
or a reference channel which is activated and has a active site mutant protein ( has same size as WT) ?
If you have an active site mutant which can not bind to your analyte, that is the best reference. In addition you could use a non-protein reference.
I would use (if you have four channels)
1: non protein reference to look for non-specific surface binding
2: mutant protein to show that your mutant is not biding
3: wild type to show it is binding to the analyte.
In addition, don't forget blank runs for double referencing to compensate for ligand loading diffrences.
That said, one comment on the word immobilization. In case of a Ni-NTA interaction I would call it capturing because there is no covalent bond involved.