Still same KD but different % activity

Hello. I performed this experiment on a mAB vs a target. Here are the experimental conditions. a) capture method with Protein A immobilized on surface, b) Mw of target: 17300. c) 2x dilutions (nM): 0.195 0.391 0.781 1.563 3.125 6.250 12.500 25 50, using 1x HBS P+ with 0.01% BSA. d) used 1:1 binding. I tried to process the sensorgram several ways (ie. deleted top points, set RI=0), to get a better fit. Whatever I do it looks like I am still getting the same KD, around 0.2nM. But I do see differences in the % activity. Some higher than 100%, one at 84% (Figure 4). I did see a dip one of the sensorgrams (Figure 2) and applied the RI=0. Fit was worse and still no change in KD. Which figure would be the correct one? From what I recall from some of my readings, to get a good KD, Rmax should theoretically be around 50-100 RU's. The 84% activity makes more sense. I can't imagine an Ab would be more active than 100%. Thanks.

Hi,
How do you calculate the % activity? The numbers are hard to read in the figures.

When I understand correctly you have a protein G chip and capture an antibody (150 kDa), then your target (17.3 kDa) is bound.
The amount of bound antibody determines the theoretical Rmax. But keep in mind that the capture of antibody to the protein G does not say anything about the integrity of the antibody --> that it will bind to your target.
Looking at the four sensorgrams the high analyte concentrations have a minor bulk effect. You could try a fit with RI and fill in the starting values of the fittings you show. Possible the fittings will be a bit closer.
However for sensorgram 4 I think the association time is too short. Even the highest concentration has not much curvature in the line. Curvature is necessary to have a proper fit.
Overall, I think your fittings are good and when several fittings all report the same kinetic constants I would not worry too much (about activity).
In SPR we try to have low response levels to avoid mass-transfer and other possible detrimental side effect of high response level surfaces ( www.sprpages.nl/best-results/artifacts ). But there are no absolute 'the best' response levels.
As a bonus tip you should consider the Short @ Long strategy to shorten the experimental time ( www.sprpages.nl/experiments/strategy ).

Kind regards
Arnoud

Hi Arnound,
I used the formula (figure 5) to calculate for % activity. I also did a screenshot of the tiny table underneath the sensorgram for easy viewing. (figure 6). The nos. on the left refer to the different figures (1 was underneath Figure 1, 2 was underneath Figure 2, etc...). I also went ahead and did another experiment (extended association time from 120 secs to 360 sec. With conc. from 0.195nM to 6.25nM. (Figure 7). Figure 8 is the same actually with the RI set to zero. Still getting pretty much the same KD, % activity.
Thanks.

HI again. Sorry, looks like the attachment tables are still small. I just listed here what I got for figure 7 and figure 8. Figure 7 had: ka = 1.30E6, kd = 1.90E-4, KD = 1.46E-10, Rmax = 85, RL = 319. % activity calculated - 115%. Figure 8 had ka = 1.15E6, kd = 1.00E-4E-4, KD = 8.69E-10, Rmax = 85, RL = 319. % activity calculated - 115%.