Steven Zimmerman, PhD , University of Illinois (Urbana-Champaign)
We made significant progress on several fronts. First, we developed a general route to alkyne core HPGs that involved polymerizing glycidol with a propargyl alcohol initiator. The crude polymer was purified by methanol/acetone fractionation which afforded some molecular weight fractionation and a narrowing of the MW range (better PDI). Each polymer contained a single alkyne group for subsequent azide alkyne cycloaddition. These polymers were further allylated to form polyallyl HPGs to allow for ring-closing metathesis-mediated cross-linking. In addition, an alkyne modified polyethylene glycol (PEG) polymer, was also synthesized from propargyl bromide and Mn = 2000 PEG monomethylether and this served as a control compound.
Many commercially available, reactive dyes like FITC contain an amine reactive group. We developed a general synthesis of PG-dye conjugates by converting an amine reactive group to an azide into a click partner for the alkyne cored HPG. To illustrate this approach with FITC, a triazide amine linker was synthesized following a procedure published by Diaz and coupled with FITC to yield a fluorescein unit carrying three azide groups. The triazide was clicked to an allylated HPG polymer (Mn = 2000) with sodium ascorbate and copper sulfate using a variation of the conditions described by Kim and coworkers and this compound could be made water-soluble by treatment with osmium tetraoxide. The allylated HPG was also intramolecularly cross-linked using Hoveyda-Grubbs catalyst prior to osmium catalyzed dihydroxylation to produce the cross-linked dye. A third water soluble fluorescein derivative was synthesized which incorporated PEG (Mn = 2000) chains in place of the HPG.
The fluorescent properties of these three macromolecules were compared and it was found that the HPG shell did increase the photo-stability of the fluorescein core and also protected against quenching by known quenchers. However, the RCM treated dendrimer was negligibly different than the uncross-linked compound indicating that the “shrink-wrapping” in this case did not increase the photo-stability. The origin of this effect is not known.
A similar synthesis and study was carried out with the perylenediimide fluorophore. PDIs represent a highly fluorescent and stable class of dyes. However, they are highly insoluble, especially in aqueous solutions where the tendency to aggregate results in reduced fluorescence. We started with the bay substituted tetrachloro PDIs as described by Müllen in his synthesis of tetrasulfated PDIs and clicked on the HPGs as described above. The same comparison was made with and without RCM and to a PEG control compound.
Overall, we developed simple synthetic routes to dendritic dye conjugates that impart enhanced photophysical properties to the dye molecules. Because of the use of click chemistry to attach the polymers to the dyes there was no sign of intermolecular cross-linking during the dye polymer conjugation. In the case of fluorescein the dendronized dyes showed improvements in photostability over the free dye and a comparable fluorescein PEG conjugate. Conjugation of the hyperbranched polymer to PDI produced water soluble dyes with a quantum yield four times greater than a similar PEG conjugated dye. These polymers could be cross-linked to reduce their size, however the cross-linking did not provide additional photostability and caused a decrease in quantum yield, an effect that we are looking into currently.