I have been having problems immobilising vancomycin using a standard amine coupling protocol, as outlined in this paper: www.sciencedirect.com/science/article/pii/S0003269704008541
Briefly, unclogging and desorb were performed before changing the flow buffer to HBS-EP and priming. CM3 chip was docked, primed, and normalisation with glycerol performed before further priming. Flow buffer was run until constant baseline before checking the detector output. Flow buffer was then injected, followed by injection of analyte in order to assess any non-specific binding, and then washing with NaOH (50 µM).
Amine coupling was performed by injection of NHS (0.1 M) / EDC (0.4 M) (35 µL) at 5 µL min-1 before injection of vancomycin (35 µL, 1mM in 10 mM sodium acetate, pH 6.0) and deactivation with ethanolamine hydrochloride (35 µL, 1 M, pH 8.5). A reference surface was prepared in the same manner, omitting the vancomycin injection.
I have attached the sensorgram for this experiment, as well as one which is zoomed in in order to see the changes in baseline. There are a couple of problems which stand out to me which I do not know how to fix:
The EDC/NHS injection increases, as far as I am aware the line should be flat and an increase of ~200 RU in the baseline is all that should be seen. I do not know why this is happening or what it indicates.
Whilst the baseline does increase by ~100 RU following the vancomycin injection, deactivation / washing seems to completely remove whatever was bound / adsorbed by the surface.
If anyone could shed any light on this it would be greatly appreciated! Apologies for the wall of text.
Thanks in advance,
Your 'wall of text' is much appreciated because it shows that you are on the right way. One thing I noticed is that you used a CM3 chip. This type has a lower capacity that a CM5 due to the shorter matrix. This can explain a lower activation level. It does not explain the increase while acitvating. I don't have an explanantion for this, since you washed the system.
The only thing I can imagine is that either the EDC or NHS was not working and therefore the activation was not 'good'.
You could do an pre-concentration experiment to asses the best immobilization pH and vancomycin concentration. In general I choose a pre-concentration speed (slope) of 20 - 50 RU.s-1.
I have also tried the same immobilisation procedure with CM5 and had the same problems, I believe the increase from EDC/NHS alone in this case was ~2000 RU which seemed abnormally high, although I would have to go back and check this. The reason I am now opting for CM3 is due to the final application as I aim to inject molecularly imprinted polymers to test their binding to vancomycin, and due to their size (~200 nm with a MW in the MDa range) I thought CM3 would be better - however neither are useful if I cannot even immobilise vancomycin in the first place!
I have tried pre-concentration experiments previously, I will go back and check the sensorgrams to see what exactly the speed was and maybe repeat this, I do not believe the speed was as high as 20-50 RU.s-1 and so this could suggest another problem that needs to be addressed. I've also never played around with the vancomycin concentration so that is something else to try.
The increase during EDC/NHS seems to me like some contaminant could be getting into the system? I was planning on thoroughly cleaning the machine using one of the superdesorb procedures as mentioned under the "clean machine" section of the site, would you have any recommendation on which would be better to use out of the protocols suggested by Biacore or Myszka? I'll follow this up with sanitize, just to be sure this isn't a problem with the machine itself.
I'll also try another source of EDC/NHS for future experiments, although I'm slightly reluctant to assign the blame to the EDC/NHS as the increase during injection seems to only happen randomly, even though the EDC/NHS I use is from the same stock of frozen aliquots - I would expect if there was a problem with the EDC/NHS I would see this everytime. The increase was also observed with freshly made EDC/NHS as well, as I thought the frozen aliquots could have been the problem.
Normally i do only the desorb once a week and sanitize once a month unless there are signs that there is a problem. I never had to use the protocol of Myska but keep in mind that you probably need some time to equilibrate the system again. Also, we do the sanitize at the end of the day so we can wash the system overnight.
Is there a way to confirm the integrity of the vancomycin? Maybe that can give a clue?
A colleague of mine has successfully performed immobilisation of biotin / streptavidin today using a different stock of EDC after seeing the same problem as me previously, so hopefully the problem simply comes down to a bad batch of EDC! I'll verify this myself and post the results here in case anyone else has the same problem in the future.
As for the pre-concentration, I dug out one of my previous experiments which I have attached the sensorgram of, showing injection of vancomycin in acetate buffer from pH 5.5 - 4.0 in 0.5 intervals. Pre-concentration is not something I have much experience with, however the first three injections do not look how I would expect, I was under the impression that these injections should be near-enough straight lines (like in the case of the 4th injection) rather than curves?
I never did pre-concentration experiments with small molecules. Surface saturation will probably at lower response levels than with proteins. I agree that the first three injections look weird compared with the last one. But with the size of vancomycin in mind the first injections could be real. To test, you can try to inject lower vancomycin concentrations during the pre-concentration to see if the pre-concentration is slower but still will reach a saturation plateau.