Diblock Copolymer Hydrophobicity Facilitates Efficient Gene Silencing and Cytocompatible Nanoparticle-Mediated siRNA Delivery to Musculoskeletal Cell Types

Abstract

pH-responsive diblock copolymers provide tailorable nanoparticle (NP) architecture and chemistry critical for siRNA delivery. Here, diblock polymers varying in first (corona) and second (core) block molecular weight (M_n), corona/core ratio, and core hydrophobicity (%BMA) were synthesized to determine their effect on siRNA delivery in murine tenocytes (mTenocyte) and murine and human mesenchymal stem cells (mMSC and hMSCs, respectively). NP-mediated siRNA uptake, gene silencing, and cytocompatibility were quantified. Uptake is positively correlated with first block M_n in mTenocytes and hMSCs (p ≤ 0.0005). All NP resulted in significant gene silencing that was positively correlated with %BMA (p < 0.05) in all cell types. Cytocompatibility was reduced in mTenocytes compared to MSCs (p < 0.0001). %BMA was positively correlated with cytocompatibility in MSCs (p < 0.05), suggesting stable NP are more cytocompatible. Overall, this study shows that NP-siRNA cytocompatibility is cell type dependent, and hydrophobicity (%BMA) is the critical diblock copolymer property for efficient gene silencing in musculoskeletal cell types.

Figure 1

NP facilitate significant and differential siRNA uptake across multiple cell types and species. Murine tenocytes (A), murine MSCs (mMSC) (B), and human MSCs (hMSC) (C) were treated with 30 nM fluorescent siRNA complexed to NP at a charge ratio of 4:1. Median fluorescence intensity (MFI) was detected via flow cytometry and normalized to negative controls within each cell type. *p < 0.05 compared to negative control (–), ^#p < 0.01 compared to positive (+) control. Negative controls (–) are comprised of untreated cells and positive controls (+) are cells treated with 30 nM siRNA using Lipofectamine2000.

Figure 2

Normalized MFI data were fit using a multiparameter linear regression to determine if siRNA uptake is dependent on NP properties. (A) p-values less than 0.05 (bold) indicate that siRNA uptake is dependent on NP parameters, and the R² values illustrate what percentage of the behavior can be adequately described by the predictive model. (B) Effect sizes (E) from significant regression models show the extent that a single property correlates with MFI. p < 0.05 indicates the polymer property significantly influences MFI. Significant effects are in bold.

Figure 3

NP-mediated delivery of siRNA results in robust gene silencing across multiple cell types and species. Murine tenocytes (A), murine MSCs (mMSC) (B), and human MSCs (hMSC) (C) were treated with 30 nM siRNA targeting glyceraldehyde 3-phosphate dehydrogenase (GAPDH) complexed to NP at a charge ratio of 4:1. Gene expression was detected using RT-PCR and normalized to negative controls in each cell type. GAPDH expression was normalized to B-Actin expression for murine cell types and peptidylprolyl isomerase B (PPIB) expression for hMSCs. *p < 0.0001 compared to negative control, ^#p < 0.001 compared to positive controls (+) control. Negative controls (–) are comprised of untreated cells and positive controls (+) are cells treated with 30 nM anti-GAPDH siRNA using Lipofectamine2000.

Figure 4

Relative GAPDH expression data was fit using a multiparameter linear regression to determine if gene-silencing capability is dependent on NP properties. (A) p values less than 0.05 (bold) indicate that gene silencing is dependent on NP parameters, and the R² value describes what percentage of the behavior can be adequately described by the predictive model. (B) Effect sizes (E) from significant regression models show the extent that a single property correlates with MFI. p < 0.05 indicates the polymer property significantly influences GAPDH expression. Significant effects are bolded.

Figure 5

pH-dependent hemolysis ability of NPs. *p < 0.05 compared to pH 7.4 for a given NP using two-way ANOVA with Dunnett’s correction for multiple comparisons. ^#p < 0.0001 comparing hemolysis across indicated NPs using two-way ANOVA with Tukey’s post hoc test for multiple comparisons. n = 8 from two independent experiments using blood from two separate donors. Error bars represent standard deviation.

Figure 6

NP-siRNA cytocompatibility via assessment of metabolic activity. Murine tenocytes (mTenocyte) (A), murine MSCs (mMSC) (B), and human MSCs (hMSC) (C) were treated with 30 nM nontargeting negative control siRNA complexed to NP at a charge ratio of 4:1. Metabolic activity was quantified as a measure of cytocompatibility and normalized to negative controls. *p < 0.05 compared to negative (–) control, ^#p < 0.05 compared to positive (+) control. Negative controls (–) are comprised of untreated cells and positive controls (+) are cells treated with 30 nM nontargeting negative control siRNA using Lipofectamine2000.

Figure 7

Relative metabolic activity in treated cells was fit using a multiparameter linear regression to determine if cell viability is dependent on NP properties. (A) p values less than 0.05 (bold) indicate that cell viability is dependent on NP parameters, and the R² value describes what percentage of the behavior can be adequately described by the predictive model. (B) Effect sizes (E) from significant regression models show the extent that a single property correlates with cell viability. p < 0.05 indicates the polymer property significantly influences cell viability. Significant effects are bolded.

Scheme 1. Diblock Copolymers Form Cationic Nanoparticles (NP) via Self-Assembly That Can Complex with Negatively Charged siRNA^a

^a(A) Diblock copolymer structure showing cationic first block composed of dimethylaminoethyl methacrylate (DMAEMA), and a pH-responsive hydrophobic tercopolymer second block composed of DMAEMA, propylacrylic acid (PAA), and butyl methacrylate (BMA) that drives self-assembly and allows endosomal escape. R and Z are functional end groups. (B) Self-assembled cationic NP electrostatically complex with negatively charged siRNA.