questions with NTA chip

Hi,

I am new to SPR.
1) I seem to have a very fast assoc/dissociation situation (NTA chip, Ligand 108KD, analyte 500Da), which makes me wonder if it is a real specific binding. The Kd obtained from the kinetics (kd and ka are out of the limit indicated by the program, so I believe it is not reliable) is different from the steady-state fit. Is that correct?
2) I tried to fit the sensorgram manually and found that the association is better fitted with two exponential functions. Does that me the initial phase is contributed by bulk refraction? But the amplitude is proportional to the concentration of the analyte, I am confused.
3) I also noticed that with 350mM EDTA, the sensor can not be fully regenerated. The capture of the ligand is getting lower, so I used a single-cycle mode. Should I include 50mM NaOH or Imidazole in my routine regeneration step to clean the chip?

Thanks.

Jane

Hi,
1) first compare the reference and specific surface. The reference can be the NTA-surface alone but a non-related HIS-tagged protein is better. Is the bulk effect different enough to conclude that there the analyte binding to your ligand? I assume that the sensorgram you show is referenced subtracted. This kind of curves cannot be fitted with a normal 1:1 fit or other model. You have to use a steady state fit as in the left figure you show. The problem I see is that the ligand is not saturating which is a sign of non-specific binding as opposed to real kinetics ( www.sprpages.nl/troubleshooting/low-affinity ).
2) see 1)
3) two or three short pulses are better than one longer. You can add imidazole to the buffer or use it in a separate regeneration injection ( www.sprpages.nl/sensor-chips-intro/biacore-sensor-chips/nta ).

kind regards
Arnoud

Thank you very much! Arnoud.
I tried a shorter contact time as you suggested. The kinetics is better, but I still did not see the saturation with high conc. of the analyte. Also, the Kd values from the kinetics and affinity are quite different. Any suggestion?
Thank you.

Jane

Hi Jane,
I have some doubts about the kinetic fitting. When in the fitted parameters the RI-parameter is much larger than the Rmax, there is probably no interaction or the interaction is so weak that a normal kinetic fitting is unreliable. In addition, from the steady state plot it appear that the highest concentration is below the K_D, which means that here also the K_D can not properley be determined.
I would suggest first to find out if there is an other means (not SPR) to confirm interaction between your molecules. You may want to look up FRET, which is very sensitive. Search for HTRF technology.
Arnoud

Thanks, Arnoud.
I got the result from the FRET experiment. I just want to make sure it's a specific binding and how high the affinity is.

Jane

Hi Arnoud,
I played with the analyte preparation method and got this sensorgram. It seems normal to me though I still see high RI at high analyte concentration, which makes hard to fit with 1:1 model (Should I use other model?).
I am planning to reduce the ligand capture in next run. Any other suggestion to improve it? Thanks.
I really appreciate your help.

Jane