*3.1. Generation and Analysis of Tyrosine-Modified PPI-G4 and (L)PEIs, and Their Corresponding siRNA Complexes*

Tyrosine-modified ("Y") fourth generation ("G4") polypropylenimine ("PPI") dendrimers (PPI-G4-Y; Figure 1A) were generated according to the synthesis scheme shown in Figure S1A. Based on our previous studies for the tyrosine-modified branched PEIs, all 32 primary amines of the PPI dendrimer were subjected to tyrosine modification. More specifically, the primary amines of the PPI dendrimer were coupled with tyrosine by using PyBOP and DMSO as solvent. The Boc deprotection with TFA and extensive purification by dialysis finally yielded the modified PPI dendrimer as white, fluffy powders. 1H-NMR analysis confirmed the structure of PPI-G4-Y (Figure S1B) and the average number of tyrosine per dendrimer was calculated to be 30. Tyrosine-modified branched and linear PEIs (Figure 1A, lower panels) were prepared as described previously [50,54].

**Figure 1.** (**A**) Structures of tyrosine-modified PPI-G4 ("PPI-G4-Y"; upper panel) as well as tyrosine-grafted branched or linear polyethylenimines (PEIs) (lower panels; numbers indicate molecular weights of PEI). (**B**) Analysis of PPI-G4-Y-mediated siRNA complexation efficacy, dependent on polymer/siRNA mass ratios. (**C**) Determination of PPI-G4-Y/siRNA complex stability by heparin displacement assay. (**D**) Size and zeta potential of PPI-G4-Y/siRNA complexes at different mass ratios.

The analysis of complexation efficacies at various polymer/siRNA stoichiometries by agarose gel electrophoresis revealed complete siRNA complexation already at a PPI-G4-Y/siRNA mass ratio of 2.5, as indicated by the absence of the free siRNA band (Figure 1B). Thus, mass ratio 2.5 was used in the further experiments, unless indicated otherwise. The finding of a particularly high complexation efficacy of the tyrosine-modified PPI dendrimer was comparable to previous results on linear or branched low molecular weight PEIs upon their tyrosine-modification, with similarly low mass ratios being sufficient even for the complexation of small RNA molecules like siRNAs. Likewise, PPI-G4-Y/siRNA complexes showed very high stability against heparin displacement, with no siRNA release even at heparin concentrations as high as 60 IU/0.2 μg siRNA (Figure 1C). While similar increases in complex stability had been obtained previously upon tyrosine-modification of PEIs, the complete absence of siRNA release upon heparin treatment as seen here for PPI-G4-Y/siRNA complexes was reminiscent to LP10Y/siRNA complexes [54]. Considering, however, that LP10Y/siRNAas well as PPI-G4-Y/siRNA complexes display biological activity which must be based on the release of intact siRNA molecules, this demonstrates that heparin displacement seems to somewhat over-estimate complex stabilities and thus only poorly reflect the situation in biological media. Still, it should be noted that PPI-G4-Y/siRNA complexes, comparable to LP10Y/siRNA complexes, show particularly high stabilities even when prepared at very low polymer/siRNA ratios.

Zetasizer measurements revealed rather large PPI-G4-Y/siRNA complex sizes with diameters of almost 600 nm already at mass ratio 2.5. They increased even further when using more polymer for complexation and were then in the range of ~ 800–900 nm (Figure 1D). Thus, despite essentially complete siRNA complexation already at mass ratio 2.5 (see above), the addition of more PPI-G4-Y dendrimer still contributed to the complex formation. This was also associated with alterations in the zeta potential. More specifically, while PPI-G4-Y/siRNA complexes were slightly negative at mass ratio 2.5 and, to a lesser extent, at mass ratio 3.75, they turned even slightly positive at the higher mass ratio 5 (Figure 1D). In conclusion, comparably large complexes with very little surface charge were obtained when using PPI-G4-Y.
