br Results and discussion br Characterization of
2. Results and discussion
2.1. Characterization of R646/siRNA nanoparticles
Bioreducible PBAEs were synthesized via polymerization of monomer bis(2-hydroxyethyl) disulfide (BR6) with monomer 4-amino-1-butanol (S4), followed by polymer end-capping with 2-((3-amino-propyl)amino)ethanol (E6), to yield polymer BR6-S4-E6 (R646), based on the bioreducible linear BR6-S4 polymer that we found promising for oligonucleotide delivery . R646 (Fig. 1A) spontaneously self-as-sembled into nanoparticles when mixed with siRNA via electrostatic interactions. Nanoparticle tracking analysis (NTA) showed that the mean and mode hydrodynamic diameter of the nanoparticles in PBS buﬀer was 137 ± 6 nm and 115 ± 4 nm, respectively (mean ± standard deviation of n = 3 particle batches), indicating that the par-ticles are within the ∼100 nm size range conducive to cellular uptake via endosomal engulfment . This measurement was verified by transmission ML162 microscopy (TEM) (Fig. 1B), with the dry Biomaterials 209 (2019) 79–87
diameter of the nanoparticles being slightly smaller than the hydro-dynamic diameter measured by NTA. Without siRNA being present, the homogenous spherical polyplex nanoparticles did not form (Fig. 1B). The zeta potential of the R646 siRNA nanoparticles was positive, measured at 18 ± 1 mV (mean ± standard deviation of n = 3 particle batches) (Fig. 1C). This is expected due to the cationic nature of R646, as is typical of PBAE-based nanoparticles, and may aid in facilitating particle uptake by cells. Particle size distribution of the self-assembled R646 siRNA NPs was found to be approximately monodisperse (Fig. 1D).
siRNA is hypothesized to be released upon degradation of the PBAE. Because R646 contains both ester and disulfide bonds, its degradation can be caused by either hydrolysis in aqueous conditions or reduction in reducing environments. A gel retardation assay was used to assess the rate of siRNA release from particles after incubation in an aqueous, physiological salt solution with or without the reducing agent glu-tathione (GSH). siRNA still bound to the PBAE would be unable to move through an agarose gel, whereas a distinct band in the gel indicates siRNA release from the polymer. As can be seen in Fig. 1E, R646/siRNA particles incubated at physiological temperature in artificial cere-brospinal fluid (aCSF) did not begin to release siRNA until 4 h, and release was not complete until > 8 h. On the other hand, R646/siRNA incubated in PBS with 5 mM GSH showed some siRNA release almost immediately, and release appeared to be complete after approximately 30 min (Fig. 1F).
Because all of the siRNA is complexed with the R646 polymer upon nanoparticle formation (Fig. 1E and F), by measuring the concentration of nanoparticles in the suspension, it is possible to calculate the average number of siRNA molecules per particle using NTA . As 720 nM RNA was encapsulated by the particles and the average concentration of particles was 2.9 ± 0.2 × 1011 particles/mL, each particle con-tained approximately 1520 ± 90 siRNA molecules.
2.2. siRNA delivery to human GBM and healthy brain cells
To verify that R646 could deliver siRNA into cells and enable gene knockdown, we formed nanoparticles with R646 and AllStars Human Cell Death Control siRNA (siDeath), a blend of siRNA oligos that kill human cells upon successful intracellular delivery. We used a scram-bled-sequence control (scRNA) as our negative control RNA. Cells were treated with either R646/siDeath or R646/scRNA nanoparticles, and transfection eﬃcacy was measured as the diﬀerence in cell death be-tween the two groups. Primary human GBM cells obtained in-traoperatively from four separate patient samples served as our GBM cell model, and healthy neural progenitor cells (NPCs), obtained from three separate primary human tissue samples served as our non-cancer cell model [35–38]. As shown in Fig. 2A and B, GBM cells treated with R646/siDeath showed significantly lower viability after five days than GBM cells treated with the R646/scRNA control. Despite heterogeneity among the GBM tissue samples, the response we observed was con-sistent, with 72 ± 9% of GBM cells killed by siRNA-induced gene knockdown.
Importantly, there was significantly (p < 0.01) less siRNA-induced cell death seen in the transfected NPCs (Fig. 2A, C). Interestingly, this diﬀerence between GBM and healthy cells does not appear to be due to diﬀerences in the overall nanoparticle uptake eﬃciency. Fig. 3 shows that 94 ± 2% of GBM cells and 94 ± 1% of NPCs internalized parti-cles (Fig. 3A). There was also no statistically significant diﬀerence in the amount of particles taken up per cell, measured as the mean fluorescence intensity per cell (120 ± 20 vs. 170 ± 60 normalized fluorescence units in GBM cells and NPCs, respectively) (Fig. 3B). While this does not account for potentially diﬀerent routes of particle uptake between the cell types , it indicates that total nanoparticle uptake by cells is not the limiting factor to successful transfection in this system.