Chemistry Reference
In-Depth Information
Figure 13.18 Comparison of calculated and experimental results: (*) experimental results
from Figure 13.17 and (—) calculated results for a 16
˚
volume shell.
Calculated line-broadening effects for end groups occupying at least a 16
˚
volume
shell show good agreement with the experimental values. Addition of further back-
folding volume into the calculation does not change the predicted line-broadening
effects significantly (Walter et al. 2005).
Experiments with sequential additions of NCS-TEMPO plus other isothiocyanates
indicate that end groups are randomly distributed in relation to one another using this
methodology. Even mannose-functionalized dendrimers appear to have a random
end-group distribution (see below). As with any random population, it is possible
that some carbohydrate-functionalized dendrimers will have higher activity in
binding studies than other populations.
To ensure that we are measuring the results of bulk properties rather than amplify-
ing results from a small, idealized subpopulation in our protein-binding studies,
we performed affinity chromatography experiments as shown in Figure 13.19. We
found that line-broadening effects before and after chromatography with resin-
bound ConA were unchanged, and we were unable to isolate any subpopulations
of lower and higher affinity dendrimers. From these experiments, we concluded
that the randomly carbohydrate-functionalized dendrimers did not have subpopu-
lations where the carbohydrate distribution on the dendrimer was either highly favor-
able or highly disfavorable for protein binding. Rather, the measured results for
protein binding represent the result for the complete population of carbohydrate-
functionalized dendrimers (Samuelson et al. 2004).
In our next series of experiments with spin-labeled dendrimers, we attempted to
alter the presentation of nitroxides on the dendrimer. Our synthetic strategy is
displayed in Figure 13.20. We invoked a boron protecting group strategy to hold
three terminal alcohols in proximity. After blocking the remaining alcohols using
ethyl chloroformate, we released the boron tether and added a 2,2,5,5-tetramethyl-
1-pyrrolidinyloxy (PROXYL) spin label. All compounds were characterized using
MALDI, and
11
B-NMR was used to characterize the intermediate borate esters. We
found that up to three borate esters per dendrimer could be formed. Additional
borate triesters could not be formed, presumably because of the highly strained
system this would create.
