*3.2. Biological E*ffi*cacies of siRNA Complexes Based on PPI-G4-Y- or Various Tyrosine-Modified PEIs*

The beneficial effects of tyrosine-modification also became evident when analyzing knockdown efficacies. The use of PPI-G4 dendrimers for siRNA transfection into stably luciferase expressing PC3-EGFP/Luc reporter cells yielded no reduction of luciferase at all, independent of complex stoichiometries up to mass ratio 5 (Figure 2A, left).

**Figure 2.** (**A**) Knockdown efficacies of PPI-G4 dendrimer-based siRNA complexes with (right) or without (left) tyrosine modification, as determined by luciferase knockdown in PC3-EGFP/Luc reporter cells. Bars show results upon transfection with complexes containing negative control siRNA (black) or luciferase-specific siRNA (grey), respectively, with black bars normalized to 100%. UT: untreated. (**B**) Knockdown efficacies upon pre-incubation of the complexes in the presence of FCS at various concentrations. (**C**) Knockdown efficacies of tyrosine-modified PPI- or PEI-based siRNA complexes in HT29-EGFP/Luc (upper panel) or MV4-11-EGFP/Luc cells (lower panel). (**D**) Dose-dependent knockdown efficacies of various tyrosine-modified PPI- or PEI-based siRNA complexes in HT29-EGFP/Luc cells.

In stark contrast, > 90% reduction of luciferase activity was observed upon incubation of cells with PPI-G4-Y/siRNA complexes. In agreement with the above results, ratio 2.5 proved to be sufficient for reaching maximum knockdown efficacy (Figure 2A, right).

When analyzing luciferase activity values normalized to untreated control cells, rather than those normalized to the respective PPI-G4-Y/negative control siRNA complexes as in Figure 2A, it was also seen that PPI-G4-Y/siRNA complexes prepared at ratios 1.25 or 2.5 exerted very few non-specific effects, with luciferase activity upon PPI-G4-Y/siCtrl transfection remaining comparable with untreated (Figure S2A, right). In contrast, at least in this cell line, non-specific reduction of luciferase activity was observed at ratios 3.75 or higher. Knockdown efficacies were not impaired by high protein content, i.e., no decrease in siRNA activity was seen in the presence of FCS at a concentration as high as 50% (Figure 2B). In fact, higher protein concentrations rather protected knockdown activity upon storage at various temperatures for three days (Figure S3). Gene targeting on the protein level, as determined by reduced luciferase activity, was also paralleled by decreased mRNA levels (Figure S2B), further substantiating the specificity of the siRNA-mediated knockdown. Results were comparable to previously described tyrosine-modified PEIs, e.g., to P5Y/siRNA complexes.

Knockdown efficacies and the definition of optimal complexes, however, were also found to be dependent on the cell line. In HT29-EGFP/Luc colon carcinoma cells, PPI-G4-Y/siRNA complexes showed better knockdown activity than their counterparts based on tyrosine-modified linear 5 kDa or 10 kDa PEI (LP5Y/siRNA or LP10Y/siRNA complexes; Figure 2C). In particular the LP10Y/siRNA complexes which had previously been found to efficiently mediate gene knockdown showed only moderate ~ 50% reduction of their target gene (Figure 2C, upper panel). In contrast, inMV4-11-EGFP/Luc B-myelomonocytic leukemia cells the same complexes were particularly efficient while little knockdown was achieved when using PPI-G4-Y (Figure 2C, lower panel). Maximum achievable knockdown efficacies, however, also proved to be dose-dependent. The lesser activity of LP10Y/siRNA complexes in HT29-EGFP/Luc cells observed above was not seen any more when doubling complex amounts: while at 15 and 30 pmol/well major differences between both polymers were seen, 60 pmol led to a very profound > 90% reduction of luciferase also in the case of LP10Y-based complexes (Figure 2D, right). This was comparable with LP5Y/siRNA complexes, whereas their counterparts based on branched 5 or 10 kDa PEI (P5Y, P10Y) showed somewhat lesser efficacy in HT29 cells (Figure 2D, left and center). In other cell lines, knockdown efficacies were rather comparable and were also seen when targeting other genes. In PC3-EGFP/Luc cells, the transfection of siRNAs specific for EGFP led to profound reduction of EGFP fluorescence, as determined by flow cytometry (Figure 3A, Figure S2C).

**Figure 3.** (**A**) Knockdown efficacies of various tyrosine-modified PPI- or PEI-based siRNA complexes targeting EGFP in PC3-EGFP/Luc cells (left). EGFP levels were determined by flow cytometry (see right panel and Figure S2C for original data). (**B**) Reduction of GAPDH mRNA levels in PC3 cells.

In the case of all polymers, the direct comparison between untreated and negative control siRNA transfected cells revealed no differences, indicating the absence of non-specific effects. Finally, when switching to an endogenous gene rather than stably, but ectopically expressed reporter genes, similar results were obtained. As shown for glyceraldehyde 3-phosphate dehydrogenase (GAPDH), PPI-G4-Y/siRNA complexes led to an ~ 90% reduction of the target gene (Figure 3B).

Taken together, this identifies complexes based on PPI-G4-Y, as alternative to tyrosine-modified linear or branched PEIs, as very efficient for siRNA-mediated knockdown in various cell lines, with some differences seen dependent on the cell line and the complex amounts used for transfection.
