br Fig Physicochemical characterization of
Fig. 1. Physicochemical characterization of R646/siRNA nanoparticles. siRNA-containing nanoparticles (NPs) based on (A) polymer structure R646 are of (B) Spherical morphology, (C) Small ∼100 nm size, and have positive zeta potential. (D) Particle size distribution of the self-assembled R646 siRNA NPs is approximately monodisperse. Gel retention assay results of siRNA release from NPs over time show (E) Persistence of the NPs over longer times in aCSF, mimicking the extracellular space, compared to (F) Faster release of siRNA from the NPs in conditions mimicking the cytosol with 5 mM GSH. Scale bars = 200 nm.
2.3. Knockdown of therapeutically relevant genes
Moving forward, we completed all in vitro experiments with patient-derived primary human GBM cell line 612, as we found this cell line to have the highest transfection eﬃcacy in our previous experiment (Fig. 2). Additionally, GBM line 612 has been shown to be highly mi-gratory, a trait which has been shown to negatively correlate with positive patient outcomes [21,27,40,41]. Cells were transfected with siRNA oligos targeting Robo1, YAP1, NKCC1, survivin, and EGFR. Our first goal was to determine the timeline of gene knockdown. We sepa-rately targeted Robo1, YAP1, and NKCC1, using scRNA as a negative control, and harvested DiD perchlorate at days 3–7 post-transfection for analysis via Western blotting (Supplementary Fig. S1). For all targeted genes, we saw a decrease in protein expression vs. scRNA over all days tested, and for all further Western blotting analyses, cells were harvested at day 3 post-transfection. For all 5 gene targets, we then needed to determine which siRNA oligo would lead to the strongest decrease in protein ex-pression. Each siRNA oligo targets only a region of a gene's mRNA transcript, and mutations and splice variants can lead to oligos being ineﬀective in some cell types. Using scRNA as a negative control and a blend of three siRNA oligo variants as a positive control, we were able to determine which particular oligo was most eﬀective, and then to use only this one for all future experiments (Supplementary Fig. S2, Supplementary Table S1).
We completed a dose-response test for siRNAs individually, tar-geting either Robo1, YAP1, or NKCC1 and analyzed the reduction in protein expression via Western blotting (Supplementary Fig. S3). In this experiment, the total siRNA dose was maintained at 120 nM, but the percentage containing targeting siRNA was varied, using scRNA to fill
Fig. 3. Nanoparticle uptake is not statistically diﬀerent between primary human GBM cells and NPCs. (A) The percent of cells positive for fluorophore-labeled siRNA is not statistically diﬀerent between GBM and NPC samples. (B) The geometric mean fluorescence of fluorophore-labeled siRNA within each cell, a measure of nanoparticles per cell, is not statistically diﬀerent between the GBM and NPC samples.
the remaining dose. For all gene targets, we found protein expression reduction when 10% of the dose contained targeting siRNA to be equal to the reduction when 100% of the dose was targeting siRNA. We also found some reduction in protein expression when 5% and even 1% of the dose was targeting siRNA, versus 0% (scRNA only). This indicated that 90% of our deliverable siRNA was in biological excess for gene knockdown, and led us to test the hypothesis that we could knockdown more than one gene simultaneously.
For functional analysis of GBM behavior both in vitro and in vivo, and to demonstrate robustness of the nanoparticles, we performed
Fig. 2. Bioreducible PBAE nanoparticles selectively transfect primary human GBM cells versus primary human neural progenitor cells (NPCs) (A) siRNA-mediated death 5 days following transfection using a death positive control siRNA shows significantly more cell killing in the (B) Four GBM cell samples tested versus (C) Three NPC cell samples tested with low non-specific toxicity due to the polymer. Scale bars = 100 μm.
Fig. 4. (A) Primary human GBM cells GBM1A transfected with a single for-mulation of bioreducible R646 nanoparticles containing a mixture of five siRNA sequences showed simultaneous knockdown of all five targeted genes. (B) Cells transfected with 5x doses of each individual siRNA sequence showed reduced viability compared to control cells as did the siRNA mixture combination condition. (C) Reducing the functional siRNA dose while keeping the total siRNA amount the same by mixing in control scRNA results in an even more pronounced eﬀect of the combination therapy. (D) Cells transfected with cer-tain single siRNAs showed reduced migration capacity compared to control cells. Cells transfected with the mixture condition of all five siRNA sequences in combination showed optimal cell killing with a strong reduction in migration capacity. For B-D, a one-way ANOVA with post-hoc Dunnett's test was used to determine statistically significant diﬀerences from the scRNA control (*p < 0.05; **p < 0.01) or the combination treatment (#p < 0.05, ##p < 0.01). A Bonferroni correction was used for the separate comparisons.