Polymeric nanoparticle-mediated siRNA delivery into primary human bronchial airway cells

Abstract

Small interfering RNAs (siRNA) are powerful tools to target cellular protein expression, making them promising candidates for therapeutic applications. siRNA-based approaches to target detrimental mechanisms in airway diseases such as asthma and COPD are highly appealing. However, such delivery systems must be non-toxic, protect siRNAs from degradation and enable intracellular uptake that accesses cytosolic RNA machinery. The present study examines the mechanisms by which guanidinium-functionalized poly(oxanorbornene)imide polymer (PONI-Guan) nanoparticles can effectively and safely deliver siRNA in human bronchial epithelial (BEC) and airway smooth muscle cells (ASM). PONI-Guan polymers were engineered to self-assemble with siRNA through electrostatic interactions. Primary BEC and ASM cells were preincubated with methyl-β-cyclodextrin, dynasore, dansylcadaverine chlorpromazine, latrunculin B or cytochalasin D followed by incubation with nanoparticles at a single concentration but different guanidinium/phosphate ratios (G/P). BEC and ASM treated with methyl-β-cyclodextrin and dynasore demonstrated significant decrease in nanoparticle uptake. Additionally, BEC showed decreased uptake with latrunculin B. Minimal BEC toxicity was observed with 20, 30 and 40 G/P ratios; ASM showed some toxicity with 40 G/P. Transepithelial electrical resistance readings were stable with 20 and 30 G/P while 40 G/P showed significant but transient changes, with barrier integrity restored in ~6h. 30 G/P did not induce markers of necrosis, apoptosis or inflammation in BEC or ASM. Transfection of BDNF siRNA in ASM and Arginase 1 and 2 siRNA in BEC showed significant decrease in corresponding mRNA and protein expression. Overall, these data indicate that PONI-Guan polymers can deliver siRNA safely and effectively in bronchial cells, primarily through caveolar or macropinocytosis uptake, offering a promising tool for future siRNA based therapies.

Graphical Abstract:

Uptake of PONI-Guan nanoparticles in human airway cells is modulated by different mechanisms. In human bronchial epithelial cells, uptake of PONI-Guan nanoparticles is primarily through caveolar-mediated endocytosis, or through macropinocytosis, while displaying minimal cytotoxicity; in contrast, uptake of PONI-Guan nanoparticles in human airway smooth muscle cells is predominantly through caveolar endocytosis while also displaying minimal cytotoxicity except at higher doses. Figure created in https://BioRender.com.

New and Noteworthy:

siRNA delivery to target cellular processes is an appealing area in lung diseases such as asthma. To realize this potential requires efficient, targeted cytosolic delivery without cellular toxicity or siRNA degradation. We demonstrate the efficacy of PONI-Guan nanoparticles in accessing airway epithelial and smooth muscle cells with specific siRNA targeting of mRNAs of interest, while maintaining barrier integrity and avoiding cellular toxicity.

INTRODUCTION

Chronic lung diseases such as asthma or COPD are characterized by structural and functional changes including thickened airway epithelium and excess mucus production creating a challenging barrier that can limit delivery of therapeutics to intended cell types such as the epithelium itself and the underlying layer of airway smooth muscle (ASM) (1–3). Numerous therapies for asthma or COPD such as corticosteroids and beta adrenoceptor agonists are routinely administered through inhalation, but often require high doses to be effective, and their long-term use leads to steroid resistance (4), systemic effects, and sometimes even worsening of disease. Furthermore, access to targeted cell types is not always possible, especially given the need for mucus penetration. Thus, there is a recognized need for drug delivery systems that can overcome physical and cellular barriers to target specific cell types and safely deliver therapies without producing toxicity. In this regard, nanoparticles (NPs) have appeal as they are highly tunable to control size and toxicity as well as cellular access. A variety of NP delivery platforms have been developed, including polymers (5, 6), lipid nanoparticles (LNP) (7, 8), and inorganic NPs (9, 10). Currently there are several FDA-approved NP delivery systems employing LNP and siRNA-GalNAc conjugates, although they have been used to target liver disease (11). Different types of NPs potentially useful to the treatment of airway diseases have been explored including polymeric NPs, solid lipid NPs, extracellular vesicles, and metal NPs (12). Thus, the potential exists to use NP delivery systems in airway disease.

siRNA therapeutics represent a promising approach to treat lung diseases such as asthma, COPD, lung cancer and respiratory infections (13). Delivery of siRNAs to the lung is commonly achieved through intratracheal, orotracheal or intranasal routes (14, 15) although other routes, such as transdermal and intravenous administrations have been reported (16). For asthma, administration of both non-formulated (17) and encapsulated (18) siRNAs have been reported, with one siRNA treatment entering phase I clinical trials but since discontinued in phase II (19), and two newer trials for RNAi-based treatments targeting RAGE and MUC5AC have been attempted. Thus, the potential to deliver siRNAs in lung diseases is appealing if using efficient delivery mechanisms.

One key challenge in delivering siRNA therapeutics is that the chemical design of vectors dictates their ability to cross biological barriers and deliver efficiently to the cytosol. Most of the current siRNA delivery systems are taken up by endocytotic processes (20) with the majority remaining in endolysosomal compartments (21) resulting in the need for higher siRNA doses which increase the risk of off-target effects (22, 23). These limitations necessitate a vector design that can more effectively deliver siRNAs in a cell- and target-specific fashion with minimal off-target or toxic effects.

Previous work by our group has shown that guanidinium-functionalized poly(oxanorbornene)imide (PONI-Guan) polymers and siRNA targeting tumor necrosis factor α (TNF-α) readily self-assemble and, when administered intravenously in an LPS mouse model of inflammation, result in significant reduction in serum levels of TNF-α with significant distribution of the PONI-Guan/siRNA complexes in the lung, demonstrating high target delivery and protein suppression efficacies (24). Furthermore, prior studies demonstrate that PONI-Guan NPs interact with the cell membranes of macrophages resulting in membrane-fusion like delivery to the cytosol with relatively low cytotoxicity (25, 26). Taken together, these studies suggest that PONI-Guan NPs may offer a potential platform for delivering siRNA to cells in the lung relevant to asthma such as bronchial epithelial cells (BECs) and ASM. The goal of the present study was to determine if PONI-Guan NPs could effectively deliver siRNA in human primary BEC and ASM, establish the mechanisms of uptake into these cells, and evaluate potential cytotoxic effects, overall assessing their potential as a delivery tool for therapies for lung diseases.

Towards demonstrating the relevance of the PONI-Guan NP based approach in blunting airway mechanisms relevant to lung disease, we tested two proteins that we and others have previously shown to be important in airway biology and asthma. Arginases (Args) which hydrolyze L-arginine into urea and L-ornithine, the latter being the first step in collagen synthesis, are expressed by BECs (27–32) but their roles in asthma are still under investigation. Lung homogenates and airways from asthmatics show more Arg1, while Arg1 inhibition blunts BEC remodeling and ASM hyperplasia (30, 33, 34). Thus, Args are a potential therapeutic target for asthma. Separately, previous work has established the importance of the neurotrophin brain-derived neurotrophic factor (BDNF) in modulating calcium/contractility and airway remodeling through autocrine/paracrine effects on ASM and its regulation by cytokines such as TNF-α (35–42). Circulating and lung BDNF are increased in asthma (43–45). Thus, BDNF within ASM represents another target to alleviate asthma pathophysiology.

MATERIALS AND METHODS

Human BECs and ASM

All protocols involving human lung tissue and primary human airway cells were approved by the Mayo Clinic Institutional Review Board (IRB #08–002518). Bronchial epithelial cells (BECs) and ASM cells were prepared as described previously (46, 47). Briefly, surgical lung specimens from patients undergoing thoracic surgery at Mayo Clinic (typically resections up to lobectomy for focal, noninfectious disease) were obtained following patient consent, and normal areas of third- to sixth-generation bronchi were identified and dissected for further use. BECs were isolated by scraping the lumen of the bronchi using a nasal brush, and the brush placed in PBS with agitation to free the cells, followed by pelleting cells by centrifugation at 300 xg for 5 minutes. The resulting pellet was resuspended in Bronchial Epithelial Growth Medium (BEGM; # CC-3170; Lonza Bioscience, Walkersville, MD) and plated for further use. BEC phenotype was confirmed by immunofluorescence staining for E-cadherin and cells with less than two passages were used for experiments. Human ASM cells were isolated and cultured using previously established procedures, cultured (<passage 5) in phenol red-free Dulbecco’s modified Eagle’s medium/F12 (Life Technologies, Rockville, MD) supplemented with 10% FBS and 1% penicillin-streptomycin (Life Technologies, Carlsbad, CA) and grown in 5% CO₂ humidified incubators. ASM phenotype was routinely evaluated by immunofluorescent staining for α-smooth muscle actin and smooth muscle myosin.

Compounds and treatments

Cy3-Negative Control siRNA was purchased from MilliporeSigma (#SIC003, Burlington, MA). Methyl-β-cyclodextrin (#C4555), Chlorpromazine hydrochloride (#C8138), Dansylcadaverine (#D4008), and Cytochalasin D (#C8273) were purchased from Sigma-Aldrich (St. Louis, MO). Dynasore (#2897) and Latrunculin B (#3974) were purchased from Tocris-BioTechne (Minneapolis, MN).

Preparation of PONI-Guan nanoparticles

The homopolymer PONI-Guan (Mw 60 kDa) was synthesized using a previously established protocol (25). Briefly, a guanidinium-functionalized monomer and Grubb’s catalyst (3^rd generation) were separately dissolved in dichloromethane. The solutions were degassed by exposing them to multiple freeze-thaw cycles before letting them warm to room temperature. Thereafter, the degassed monomer solution was mixed with the catalyst solutions and stirred for 30 min under an inert atmosphere. Next, the reaction was quenched by adding ethyl vinyl ether and stirring the solution for an additional 30 min. The polymer was obtained from the solution by precipitating it in hexane or anhydrous diethyl ether, followed by filtration, washing, and drying under reduced pressure, yielding a faintly yellow colored solid. Subsequently, the polymer was dissolved in dichloromethane and trifluoroacetic acid (1:1) and stirred for 4 h to remove the Boc functionalities. Thereafter, excess trifluoroacetic acid was removed via azeotropic distillation with methanol. The resulting polymer was redissolved in minimal amounts of dichloromethane and precipitated in anhydrous diethyl ether, followed by filtration, washing, and evaporation of the remaining solvent under reduced pressure. Polymers were then dissolved in water and transferred to Biotech CE dialysis tubing membranes with a MWCO of 10000 g/ mol and dialyzed against Milli-Q water for 3 days.

To characterize the nanoparticle preparations, PONI-Guan polymer was mixed with siRNA by pipetting and incubating for 2, 6, and 24 h. A Malvern Panalytical Advanced Zetasizer Pro was used to measure the size and polydispersity indices (PDIs)) in Dulbecco’s Phosphate-Buffered Saline (DPBS). To assess the stability of the nanoparticles, we measured the diameter at three different time points (2h, 6h, and 24h) and observed consistent results (Figure 1A). Low dispersity without significant aggregation was verified by the PDI values across different time points and G/P ratios (Figure 1B). A transmission electron microscopy (TEM) image was obtained using a Lacey Carbon Film-Cu grid (300 mesh, 50 μ, CC; Electron Microscopy Science) and captured with FEI Tecnai 12 TEM. The TEM data verified spherical morphology of the particles (Figure 1D, 1D). Longer term stability (up to 4 weeks) has been previously demonstrated (48).

Figure 1. Characterization of PONI-Guan nanoparticles.

A) Hydrodynamic diameters (by number) was assessed at 2h, 6h, and 24h using dynamic light scattering, showing consistent maintenance of size. B) Polydispersity indices were measured at 2h, 6h, and 24h demonstrating maintained dispersion without significant aggregation. C,D) Representative transmission electron microscopy (TEM) images of guanidinium/phosphate ratio (G/P) 30 PONI-Guan nanoparticles.

For siRNA studies, 2 μL Cy3-Negative siRNA was mixed by pipetting with 4 μL of 75 μM PONI-Guan NPs (1:2 v/v) and then incubated for 10 min followed by dilution into 1 mL medium for treating cells, with an effective 25 nM siRNA concentration and 300 nM NP concentration. Pilot studies were performed using nanoparticles with different G/P ratios but at a fixed concentration to determine efficacy of siRNA delivery to BECs and ASM vs. cytotoxicity. We determined that 30 G/P was most effective in delivering siRNA to BEC and ASM without cytotoxicity at the concentration used, and this G/P was used for subsequent uptake experiments.

Testing adverse effects of PONI-Guan nanoparticles

Human BECs or ASM seeded onto 8 well chambered coverslips were treated with 20 G/P, 30 G/P, or 40 G/P PONI-Guan NPs. After 2h, 6h, and 24 h, cell viability was determined using a Live/Dead assay kit (#L3224; Invitrogen-ThermoFisher Scientific; Waltham, MA) according to manufacturer’s protocol. Briefly, a working solution of 2μM calcein AM and 4μM ethidium homodimer-1 was prepared in Hank’s Balanced Salt Solution (HBSS), the cell medium aspirated, and replaced with working solution. Cells were incubated in the dark for 30 min at room temperature, washed 3x in HBSS and fluorescence visualized using a Nikon Elements Ti microscope with GFP (live) and Cy3 (dead) filters. Cell viability was determined as the percentage of calcein AM positive cells compared to total number of cells with at least 300 cells counted/group using Nikon Elements software. Positive controls for dead cells were measured by incubating with 0.1% saponin for 10 min.

In additional experiments, markers of necrosis (HMGB1, LDHA), apoptosis (Caspase 3, Caspase 9, cytochrome C), induction of inflammation (TNFα, IL-1β and IL-6) and cellular proliferation (PCNA, Ki67) were tested in BECs or ASM exposed to 30 G/P PONI-Guan NPs for 24h as indication of adverse effects.

Testing PONI-Guan nanoparticle effect on epithelial integrity

BECs were seeded into 6.5mm Transwell plates with polyester membrane and 0.4μm pores (#3470; Corning Inc-Life Sciences, Oneonta, NY) at 50,000 cells/well. Cells were grown for 21 days to confluence and media from upper chamber aspirated to create an air-liquid interface (ALI) for an additional 7 days to allow for cell differentiation and formation of tight junctions. Cells were then treated with 20 G/P, 30 G/P, or 40 G/P PONI-Guan NPs (fixed 300 nM concentration) suspended in BEGM and transepithelial electrical resistance (TEER) measured after 2h, 6h, and 24h using an EVOM2 volt/ohmmeter and TEER chopstick electrodes (#STX2; World Precision Instruments, Sarasota, FL).

Nanoparticle uptake in BECs and ASM

Cells were seeded onto 8-well chambered coverslips (Ibidi, Grafelfing, DEU) at 5,000–10,000 cells/well and allowed to grow for 24h. To determine caveolar/lipid raft mechanisms of NP uptake, cells were pre-exposed to 5 mM methyl-β-cyclodextrin (MβCD; lipid raft/caveolae inhibitor) for 1h, followed by addition of PONI-Guan/Cy3-Negative siRNA complex in the continued presence of MβCD for 24h. Additionally, 80μM dynasore (dynamin inhibitor), clathrin-mediated endocytosis inhibitors dansylcadaverine (200μM) and chlorpromazine (30μM), and macropinocytosis inhibitors latrunculin B (20μM) and cytochalasin D (10μM) were tested by preincubation for 1 h prior to addition of PONI-Guan siRNA. After 24 h, cells were visualized using a Nikon Eclipse Ti imaging system, LED fluorescence light source and 16-bit high-sensitivity CCD camera using filters for Cy3 detection. Regions of interest were identified in 10–20 cells/treatment group and uptake was quantified by determining relative fluorescence after background subtraction using Nikon Elements software.

PONI-Guan NP-mediated protein suppression

Towards demonstrating relevance of the PONI-Guan NP based approach in airway diseases, we tested efficacy of suppression of arginases (Args) in BECs (27–32) and brain-derived neurotrophic factor (BDNF) in ASM (35–42). ASM (200k cells/well) in 6-well plates were transfected using 30 G/P NP and 25nM Neg control siRNA (#D-001810-10-20, Horizon Discovery Biosciences, Waterbeach, UK) or 25 nM BDNF siRNA (#L-017626-00-0020, Horizon Discovery Biosciences) for 24h. Cells were incubated in serum free culture medium for 24 h then treated with or without TNF-α (#210-TA-10, R& D Systems) for 24h and cell pellets collected. TNF-α was used as a surrogate for airway inflammation and to stimulate BDNF production thus testing the potential for NPs to blunt protein production in the context of disease. BEC were transfected similarly using ARG1 (#L-009922-00-0005, Horizon Discovery Biosciences) or ARG2 (#L-009454-01-0005, Horizon Discovery Biosciences) siRNA for 24 h. Total RNA was extracted using RNeasy micro kit (#74004, Qiagen, Germantown, MD). Complementary DNA (cDNA) was synthesized using Transcriptor First Strand cDNA Synthesis kit (#04896866001,Roche, Indianapolis, IN), and amplified using a Veriti thermocycler (Applied Biosystems Corp, Waltham, MA). Standard qPCR techniques using a Roche Lightcycler 96 were used. mRNA expression in ASM was determined using BDNF primers (#QT00235368, Qiagen), NTRK2-FL forward primer is 5’ ACTACTACAGGGTCGG 3’, reverse primer is 5’ CCCTAGCCTAGAATGTCC 3’ (IDT Coralvill, IA), and S16 forward primer is 5’ CAATGGTCTCATCAAGGTGAACGG 3’, reverse primer is 5’ CTGACGGATAGCATAAATCTGGGC3’ (IDT). ARG1 mRNA expression is BEC was determined using forward primer 5’ TTGGAGCTCCTTTCTCAAAGGGACAG 3’ and reverse primer 5’ GCTTGCTTTTCCACAGACCTTGGAT 3’; ARG2 expression using forward primer 5’ ACAGGGTTGCTATCAGCACTGGATCT 3’ and reverse primer 5’ TGGTGAACTGGGAGTAGGAAGTGGT 3’. For both ASM and BEC mRNA expression was determined for the following using primers from Qiagen: HMGB1 (QT01002190), LDHA (QT00001687), CASP3 (QT00023947), CASP9 (QT00036267), CYCS (QT01672342), TNF-α (QT00029162), IL1β (QT00021385), IL6 (QT00083720), KI67 (QT00014203), and PCNA1 (QT00024633). Data are presented as fold change in comparison to an endogenous control (S16) and relative to untreated control.

JESS Analysis

BEC and ASM cells were cultured and harvested post-transfection with negative control and targeted siRNAs as described above. A bicinchoninic acid (BCA) Protein Assay Kit (ProteinSimple, San Jose CA) was used to measure protein concentration. Protein expression was measured using the JESS capillary electrophoresis system (Protein Simple, San Jose, CA) according to the manufacturer’s protocol and described previously (49). 3μl of each sample were loaded onto the 12–230 kDa cartridge plate. The target proteins were immune-probed with anti-Arginase 1 (Abcam-ab96183, Cambridge, MA), anti-Arginase 2 (Abcam-ab228700, Cambridge, MA), or anti-TrkB (Abcam-ab33655, Cambridge, MA) diluted in antibody buffer at a 1:25 ratio followed by horseradish peroxidase (HRP)-conjugated secondary antibodies. 10μl of primary and secondary were loaded for each sample. Capillaries and the plate were loaded into the JESS device. Protein separation, blocking, antibody incubation, and signal detection were conducted automatically via standard manufacturer protocol. The vendor-supplied Compass software was used for data analysis. Electropherogram data were digitally displayed and normalized to total protein.

BDNF secretion

BDNF secretion in cell culture medium was determined via ELISA. Briefly, ASM (200k cells/well) seeded in 6 well plates were transfected and treated with TNF-α as described above. Supernatants (1mL) were collected and concentrated to approximately 125μL using Pierce-3K MWCO Centrifugal filters (#88512, ThermoScientific). BDNF was quantified using a Quantikine ELISA kit (#DBNT00,R&D Systems, Minneapolis, MN) according to manufacturer’s protocol.

Statistical Analysis

Data analysis and figure preparation were performed using GraphPad Prism 10.0 software (GraphPad Software, San Diego, CA, USA). Statistical tests were unpaired Student t-test for two groups (NP exposure vs. vehicle, effect of uptake modulators) and 2-way ANOVA for multiple groups (siRNA effect and NP exposure) with Bonferroni correction. Outliers were determined by Grubb’s outlier test. “n” values represent samples from individual patients per treatment/exposure group. Data are represented as mean ± SEM and p < 0.05 used for statistical significance.

RESULTS

Effect of PONI-Guan NPs on BEC viability and epithelial integrity

Key to the use of NPs as delivery systems will be to demonstrate minimal cytotoxicity, and in the case of BECs, show minimal disruption to the membrane barrier. Here, PONI-Guan NPs of different G/P ratios but at a fixed concentration (300 nM) were incubated with BECs over 24 h with cell viability determined using Live/Dead assay at 2h, 6h, and 24h timepoints. Calcein/AM was used to assess live cells, indicating intracellular esterase activity to cleave the Calcein/AM ester groups producing green fluorescence, and red fluorescence ethidium homodimer-1 used to assess dead cells due to loss of membrane integrity. No significant changes in the percentage of live cells were noted at 2h, 6h, or 24h with 20 G/P, 30 G/P or 40 G/P PONI-Guan NP exposure compared to untreated controls (Figure 2A and B). PCR analysis of BECs exposed to 30 G/P PONI-Guan NPs for 24h showed no change in any of the markers for necrosis (HMGB1, LDHA), apoptosis (Caspase 3, Caspase 9, cytochrome C), inflammation (TNFα, IL-1β and IL-6) or proliferation (PCNA, Ki67) (Figure 2C).

Figure 2. PONI-Guan nanoparticles, epithelial cytotoxicity, and barrier disruption.

A) To assess potential cytotoxicity of NP, bronchial epithelial cells (BECs) were incubated with increasing G/P ratios of 300 nM PONI-Guan NPs over 2, 6 and 24h, then cells stained with 2μM green-fluorescent Calcein/AM to indicate intracellular esterase activity (live) and red-fluorescent ethidium homodimer-1 (4μM) to indicate loss of plasma membrane integrity (Dead). Inset shows BEC treated with 0.1% saponin for 10 min as positive control for ethidium homodimer-1 staining. B) Representative fluorescent images of Live-Dead assay were quantified by cell number. No significant changes in the number of live cells with the increasing concentrations of PONI-Guan NPs were observed over 24h, suggesting little cytotoxicity. C) BEC were treated with 30 G/P PONI-Guan NPs for 24h and mRNA for markers of necrosis, apoptosis, inflammation and proliferation assessed. No significant changes are noted in comparison to untreated controls. D) To assess the effect of PONI-Guan NPs on epithelial barrier integrity, the time course of changes in transepithelial electrical resistance (TEER) was evaluated in air-liquid interface cultures on transwell membrane inserts (baseline TEER ~300 ohms, indicating formation of tight junctions). BECs were incubated with NPs of different G/P ratios, and TEER recorded at 2h, 6h and 24h. Initial TEER readings with 20 G/P and 30 G/P at 2h and 6h indicated no or slight drop in resistance compared to untreated controls but full recovery within 24h. Cells treated with 40 G/P showed a decrease in TEER at 2h and 6h, but also fully recovered within 24h. Data are represented as mean ± SEM; n=4–5 BEC lines per group; *P < 0.05; **P < 0.01.

To determine the effects of PONI-Guan NPs on BEC membrane integrity, TEER measurements were utilized. When TEER values measured ~300 ohms, indicating formation of tight junctions, BECs were incubated with PONI-Guan NPs of different G/P ratios (fixed concentration) over 24h. TEER values at 2h and 6h showed no significant changes with 20 G/P and or 30 G/P PONI-Guan NPs. However, BECs treated with 40 G/P PONI-Guan NPs showed a significant but transient drop in TEER values at 2h and 6h compared to untreated controls, that fully recovered within 24h (Figure 2D).

Effect of PONI-Guan NPs on ASM viability

Conditions for treating ASM to determine cytotoxicity of PONI-Guan NPs were the same as for BECs. ASM were incubated with 300 nM of 20 G/P, 30 G/P, or 40 G/P PONI-Guan NPs over 2h, 6h, and 24h, then stained using the Live/Dead assay to determine cell viability. No significant changes in numbers of live cells with 20 G/P ratio PONI-Guan NPs were observed at 2h, 6h, or 24h compared to controls. No significant change in viability was noted at 2h or 6h with 30 G/P compared to controls; however, a significant difference at 24h was observed though the number of cells stained with Calcein/AM remained at 80%. In contrast, significant cytotoxicity was noted with 40 G/P PONI-Guan NP at 2h, 6h and 24h (Figure 3A, 3B). PCR analysis of ASM exposed to 30 G/P PONI-Guan NPs for 24h showed no significant change in markers for necrosis (HMGB1, LDHA), apoptosis (Caspase 3, Caspase 9, cytochrome C), inflammation (TNFα, IL-1β and IL-6) or proliferation (PCNA, Ki67) (Figure 3C).

Figure 3. PONI-Guan nanoparticles and airway smooth muscle (ASM) cytotoxicity.

A) ASM were incubated 20 G/P, 30 G/P, or 40 G/P PONI-Guan NPs over 2h, 6h, and 24h, then cells stained using the Live/Dead assay to determine cytotoxicity. B) Cell viability was quantified by number of Calcein-AM (green) positive stained cells compared to total cell number. No significant changes in the number of live cells with the 20 G/P Cy3-tagged PONI-Guan NPs were observed over 24h suggesting little cytotoxicity. Some cytotoxicity was observed with 30 G/P after 24h; however, cell viability remains above 80% live cells. Significant changes in live cell numbers were noted with 40 G/P at 2h, 6h and 24h. Images are representative of typical experiment. Inset shows hASM treated with 0.1% saponin for 10 min as positive control for ethidium homodimer-1 staining. C) hASM were treated with 30 G/P PONI-Guan NPs for 24h and mRNA for markers of necrosis, apoptosis, inflammation and proliferation assessed. No significant changes are noted in comparison to untreated controls. N=4–5 ASM cell lines per group. **P < 0.01; ***P < 0.001; ****P < 0.0001.

Mechanisms of PONI-Guan nanoparticle uptake in BECs and ASM

Previous work has established the importance of caveolae, a component of lipid rafts, in BEC and ASM cell function (50, 51), including a role in endocytosis (52, 53). To investigate the role of lipid rafts/caveolae, BEC and ASM were pre-incubated with 5 mM methyl-beta-cyclodextrin (MβCD) for 1h, followed by addition of Cy3-Negative siRNA (25nM)/NP complex (30 G/P ratio) in the continued presence of MβCD for 24 h. Quantification of Cy3-Negative siRNA fluorescence showed NP uptake was significantly decreased in BECs (Figure 4) and ASM (Figure 5), −77%±11 and −71%±10, respectively, in comparison to untreated controls.

Figure 4. Mechanisms of PONI-Guan nanoparticle uptake in BECs.

A) To inhibit lipid rafts, BECs were exposed to 5 mM methyl-beta-cyclodextrin (MβCD) for 1h, followed by addition of Cy3-Negative siRNA (25nM)/NP complex (30 G/P ratio) in the continued presence of MβCD for 24h. Dynamin inhibition in BECs was investigated by incubating with 80 μM dynasore for 1h prior to addition of NPs. To disrupt clathrin-mediated endocytosis, 200 μM dansylcadaverine (DC) or 30 μM chlorpromazine were used, also for 1h. Macropinocytosis was prevented with inhibitors latranculin B (20 μM) and cytochalasin D (10 μM). Following 24h incubation the cells were visualized using fluorescence microscopy for Cy3. B) MβCD and Dynasore significantly inhibited Cy3-Negative siRNA/NP complex uptake after 24h in comparison to untreated control BECs. Likewise, incubation with latrunculin B also resulted in significant reduction in Cy3-tagged nanovectors vs. control cells. Data are represented as mean ± SEM; n=4–6 BEC lines per group; *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 5. Mechanisms of PONI-Guan nanoparticle uptake in human ASM.

A) Human ASM cells were incubated with 5 mM MβCD for 1h to disrupt lipid rafts/caveolae, followed by addition of Cy3-Negative siRNA/GP complex (30 G/P ratio) in the continued presence of MβCD for 24h. ASM were exposed to inhibitors of clathrin-mediated endocytosis and/or macropinocytosis for 1h including 80 μM dynasore, 200 μM dansylcadaverine (DC), or 30 μM chlorpromazine. Macropinocytosis was prevented by inhibition of F-actin depolymerization using 20 μM latranculin B or 10 μM cytochalasin for 1h. B.) Uptake of Cy3-NP was analyzed with fluorescent microscopy. Preincubation with MβCD and dyansore resulted in significant reduction of NP uptake when compared to controls. Data are represented as mean ± SEM; n=4–6 ASM lines per group; *P < 0.05; **P < 0.01.

Dynamin is a GTPase protein that modulates clathrin-mediated endocytosis and caveolar internalization. BECs and ASM were preincubated with dynasore (80μM), a dynamin inhibitor, for 1h prior to addition of NPs. After 24 h incubation, NP uptake was significantly decreased in BEC (−63%±7; Figure 4) and ASM (−85%±4; Figure 5). To further investigate clathrin-mediated endocytosis, 200 μM dansylcadaverine (DC) or 30 μM chlorpromazine was used, also for 1h, prior to NP exposure; however, neither drug significantly effected NP uptake in either BEC (Figure 4) or ASM (Figure 5).

Macropinocytosis is an actin-dependent process that is important for cells for uptake of large molecules. Here, macropinocytosis was probed using F-actin inhibitors latrunculin B (20 μM) or cytochalasin D (10 μM). In BECs, following 24h incubation, latrunculin B significantly inhibited NP uptake (−53%±10; Figure 4). Cytochalasin D did not statistically decrease NP uptake. In contrast, in ASM, neither latrunculin B nor cytochalasin D substantially prevented NP uptake compared to controls (Figure 5).

PONI-Guan NPs and siRNA transfection in BECs and ASM

In BECs transfected with ARG 1 or ARG 2 siRNA, mRNA and protein levels showed significant knockdown (ΔΔCT<0.5 for mRNA) of both arginase isotypes (Figure 6A, 6B). In ASM, BDNF siRNA transfection using PONI-Guan NPs resulted in substantial knockdown of BDNF mRNA levels in the presence of TNF-α (−43%). In comparison, mRNA levels for BDNF receptor, neurotrophic tyrosine receptor kinase 2 (NTRK2), were unchanged in cells transfected with BDNF siRNA indicating specificity of the knockdown and any downstream functional effects of the siRNA will be specific to BDNF and not due to changes in NTRK2 (Figure 6C). Protein analysis of ASM lysates showed decreased NTRK2 with siRNA transfection, demonstrating specificity of the knockdown (Figure 6D). To determine if transfection of BDNF siRNA using PONI-Guan NP in ASM could have functional cellular effects, culture media was collected from transfected cells in the presence or absence of TNF-α, the media concentrated, and secreted BDNF levels quantified using ELISA. Cells treated with TNF-α showed significant increases in BDNF secretion in comparison to negative siRNA transfected controls(+247%). Significantly less BDNF (−56%) was present in culture media of cells transfected with BDNF siRNA compared to cell treated with TNF-α alone (Figure 6E). To determine transfection efficiency, BEC and hASM cells were transfected with Cy3-Negative siRNA using 30G/P PONI-Guan NPs for 24h and transfection efficiency evaluated by comparing the number of cells with Cy3 fluorescent staining to the total number of cells present (DAPI positive). Both BEC (84%) and hASM (82%) show substantial levels of Cy3-Negative siRNA uptake indicating a high degree of transfection efficiency (Figure 6F).

Figure 6. siRNA transfection using PONI-Guan nanoparticles.

A) BEC were transfected with 25nM Arg1 or Arg2 siRNA using 30 G/P NP for 24h and mRNA expression determined respectively. Both Arg1 and Arg2 mRNA levels were significantly knocked down compared to negative siRNA transfected controls (ΔΔCT<0.5). B) BEC were transfected with 30G/P NP using 25nM negative control or Arg1/Arg2 siRNA for 24h and protein expression determined via Jess system. Both Arg1 and Arg2 are significantly knocked down in comparison to negative control siRNA. Protein expression is normalized to total protein signal (TPx10⁵). C) ASM were incubated with 30 G/P NP and 25nM BDNF siRNA for 24h and mRNA expression determined for BDNF and NTRK2. BDNF mRNA expression was significantly increased in the presence of TNF-α, effects substantially inhibited by BDNF siRNA, but showed no decrease in TNF-α stimulated increase of NTRK2. D) hASM were transfected with 30G/P NP as previously stated using 25nM negative control or NTRK2 siRNA for 24h and protein expression determined. NTRK2 protein exptression is substantially knocked down in comparison to negative controls. Protein expression is normalized to total protein signal. E) ASM were transfected with BDNF siRNA as above, +/− TNF-α for 24h, and BDNF secretion measured in supernatants via ELISA. TNF-α stimulated a significant increase in BDNF secretion compared to Negative siRNA controls with significant inhibition observed in BDNF siRNA transfected cells. F) Transfection efficiency in BEC and hASM was evaluated by determining the number of positive Cy3-Negative siRNA cells compared to total number of cells (DAPI). Both BEC (84%) and hASM (82%) show substantial levels of Cy3-Negative siRNA uptake indicating a high degree of transfection efficiency. N=4–8 cell lines per group. *P < 0.05; **P < 0.01.

DISCUSSION

Current strategies for controlling airway diseases such as asthma typically involve inhalation of a drug such as beta agonists to stimulate bronchodilation or corticosteroids to inhibit inflammation (54–59): approaches that often require high doses and/or chronic exposure to ensure sufficient access to airway cells but resulting in undesired side effects. An exciting recent advance has been the development of NPs and siRNAs to target specific cell types involved in diseases and to knockdown specific proteins (60) where localized NP delivery to the airways via inhalation/nebulization is obviously appealing. This necessitates NPs to translocate beyond the initial epithelial barrier to underlying ASM and fibroblasts, deliver siRNAs effectively while having minimal deleterious effects. The primary purpose of the current study was to demonstrate the use of PONI-Guan NPs for siRNA delivery in human airway cells, exploring their mechanisms of uptake and effective siRNA delivery but the lack of toxicity or other adverse effects, thus providing proof of principle for their further development as therapeutic delivery agents.

One of the key requirements to provide effective delivery of siRNA into cells and produce the highest level of gene silencing is for the delivery system to efficiently provide access to the cytosol (61). PONI-Guan polymers offer some important advantages in this regard. First, the structure of the poly(oxanorbornene) provides a pre-organized semi-rigid platform for multivalent self-assembly. Second, the positively charged guanidinium moieties that enable self-assembly facilitate membrane fusion-like delivery when appended to preorganized macromolecular structures. Previous studies have validated the effectiveness of PONI-Guan nanoparticles in delivering siRNA in macrophages using fluorophore-tagged negative siRNA (24). Diffuse staining throughout the macrophages was noted, suggesting that uptake into the cell was directly to the cytosol versus a more punctate appearance, which would indicate involvement of endocytosis (25). In the present study, similar experimental conditions were utilized along with various inhibitors of endocytosis to determine the mechanisms of PONI-Guan nanoparticle uptake in primary BEC and ASM cells. Under control conditions significant uptake of the Cy3-tagged control siRNA was observed with some punctate appearance noted in both BEC and ASM. In the presence of MβCD, a lipid raft inhibitor, and dynasore, an inhibitor of caveolar internalization, there was significant reduction in the amount of observed uptake with a diffuse, punctate appearance suggesting that caveolar endocytosis may be one of the primary mechanisms by which NPs are translocated intracellularly in BEC and ASM. Both BEC and ASM are abundant in caveolae (51–53, 62, 63) and involved with modulating numerous mechanisms important to normal cell function. Thus, a role for caveolae in uptake of PONI-Guan NPs would be logical. Likewise, in BECs, exposure to latrunculin B, an inhibitor of F-actin polymerization, also resulted in significant reduction of PONI-Guan NPs. Macropinocytosis is the process by which cells take up large fluid amounts or large molecules intracellularly with macropinosomes moving along the actin filament, thus the reduction in NP uptake by latrunculin B would suggest the involvement of macropinocytosis. Taken together these data would seem to suggest that multiple mechanisms are involved in PONI-Guan NP uptake in airway cells. While direct cytosolic delivery through membrane fusion may be present, endocytotic processes are also clearly involved. However, endocytosis followed by rapid endosomal degradation for siRNA delivery cannot be ruled out and future studies will need to be conducted to determine gene silencing efficiency.

A major aspect of any future therapeutic application of NPs in airways is ensuring a lack of toxicity and, in the case of BEC, limiting barrier dysfunction. The significance of our studies lies in the strategic design of PONI-Guan NPs to validate their potential for applications in targeting asthmatic airways. Recognizing the difficult problem of NP toxicity has limited its use in airway diseases (64). The idea that altering NP properties can be a controllable parameter with which to enhance access and beneficial effects in the asthmatic airway while minimizing toxicity make the development of the PONI-Guan polymers even more significant. While there is some previous work on NP interactions at the alveolar level (65, 66), much of it is conflicting. In this study BEC and ASM were exposed to PONI-Guan NPs over a period of 24h and their cytotoxic and other effects determined by a battery of tests: Live/Dead assay, mRNA analysis of necrosis, apoptosis, inflammation, and cell proliferation, and TEER measurements (for BECs). In BECs, increasing G/P ratios of PONI-Guan NPs (20 G/P-40G/P) at 2h, 6h, and 24h had no significant effect on cell viability with only 40 G/P at 6h and 24h demonstrating a live cell percentage of below 95% (albeit not that substantially lower). Likewise, under the same conditions measurement of TEER showed that membrane barrier integrity was consistently maintained with 20 G/P and 30 G/P while 40 G/P PONI-Guan NPs show a significant decline in TEER at 2h and 6h; however, within 24 h TEER returns to control levels suggesting that membrane integrity can been restored with these NPs. Importantly, 30 G/P PONI-Guan NPs showed no discernible effect on multiple markers for necrosis, apoptosis, inflammation, or proliferation, suggesting that these NPs are overall non-toxic to BECs. This is significant given the critical barrier role of BECs in the airway, and thus the use of PONI-Guan NPs for siRNA delivery is thus facilitated by our findings.

Like BECs, in ASM PONI-Guan NPs had little effect on altering cell viability over 24h with 20 G/P and 30 G/P ratios. There was some statistically significant toxicity with 30 G/P after 24h; however, cell viability remained above 80%. In contrast, the 40 G/P NPs demonstrated significant toxicity even at early time frames suggesting that ASM may be more sensitive to deleterious effects of NPs but only at higher concentrations. Nonetheless, our data indicates that PONI-Guan NPs are relatively non-toxic at low to moderate concentrations and may be a useful tool for airway cell targeting. Indeed, the lack of adverse effect of 30 G/P PONI-Guan NPs on markers of necrosis, apoptosis, inflammation, or proliferation again highlights the advantages these NPs bring as delivery vehicles in airways.

Our results highlight the many advantages of the PONI-Guan NP technology for use in airways. However, their further development and application require demonstration of their superior (or at least comparable) efficacy and safety compared to other NP technologies. We have previously shown that the ~60 kDa MW PONI-Guan homopolymer provides the highest efficacy for cytosolic delivery (67). Our approach also provided high control over molecular weights and limited batch-to-batch variations as previously demonstrated (24). We have also verified NP-siRNA complex sizes and zeta potentials using dynamic light scattering and TEM finding spherical morphology, an average diameter of 170 nm, and average charge increasing with G/P ratio (24). Importantly we have found siRNA encapsulation efficiency to be ≥97% at most G/P ratios, far exceeding LNP based efficacies. This efficiency is attractive, particularly when payload is delivered to the cytosol with protection of siRNA against nuclease degradation as previously found (24). These capabilities highlight control of the tunability of NP properties in the context of siRNA delivery. Taken together with our current findings of lack of cytotoxicity, induction of inflammation or cell proliferation, these many factors of NP design, properties and cellular effects indicate distinct advantages of the PONI-Guan NP technology compared to other polymer based drug delivery systems.

Towards establishing potential therapeutic relevance of PONI-Guan NP based siRNA delivery into airways (e.g. via an inhalational route in vivo), we explored the application of these NPs in targeting arginases in BECs. The contribution of arginases has recently been explored in adult asthma (27–34) with lung homogenates and airways from asthmatics showing increased levels of arginase I compared with healthy adults (68, 69). In animal models, Arg1 is increased, and its inhibition blunts epithelial remodeling and ASM hyperplasia.(28). Thus, arginase I is implicated as an enzyme highly relevant to the airways. The importance of arginases, especially in airway remodeling, lies two-fold. First, increased arginase activity diverts L-arginine from the NOS/NO pathway resulting in impaired bronchodilation (27–32). Secondly, further downstream via ornithine, arginases can increase polyamines, which can regulate proliferation, and increase production of collagen, a key component of extracellular matrix, thus the targeting of arginases offers potential for alleviating airway epithelium mediated remodeling (30, 33, 34). Using PONI-Guan NPs complexed with ARG1 or ARG2 siRNA, our data demonstrates the presence of both arginase isotypes in BEC which is significantly knocked down through siRNA transfection as measured by mRNA expression compared to controls. Furthermore, preliminary experiments indicate that TNF−α induced increases in NO production in BEC, via DAF-2 fluorescence measurement, can be reversed by both Arg1 and Arg2 siRNA transfection using PONI-Guan NPs (unpublished observations). These data highlight the possibility of using PONI-Guan NP/siRNA complexes as a means of alleviating epithelial driven remodeling in the airway.

We further explored the potential therapeutic relevance of PONI-Guan NP based siRNA delivery via targeting of BDNF in ASM. BDNF and its receptors, especially the high-affinity receptor tryosine kinase NTRK2, have long been recognized in the nervous system but have more recently been found to be highly expressed in the lung with elevated levels patients with asthma and allergic rhinitis (36, 37, 70–72). We showed that ASM is a substantial source of BDNF in the airways and can produce autocrine/paracrine effects resulting in increased potentiation of airway contractility and remodeling (36, 37, 39, 40). Thus, BDNF can sustain the effects of inflammation making it a potential attractive target for siRNA therapeutics by inhibiting its production and functional effects by limiting remodeling. In the present study, PONI-Guan NP/BDNF siRNA complexes were used to target BDNF in ASM in the presence of the pro-inflammatory cytokine, TNF-α. Results demonstrate significant increases in BDNF and NTRK2 mRNA expression with TNF-α exposure; increases on BDNF are substantially inhibited by siRNA with no change on NTRK2 expression showing the effects of the PONI-Guan NP/siRNA is specific and targeted. Likewise, use of PONI-Guan NP/BDNF siRNA shows significant decrease of BDNF secretion, measured via ELISA, from ASM in the presence of TNF-α. Taken together these results suggest that the PONI-Guan NPs could be an effective tool for targeting BDNF driven remodeling in ASM.

In summary, the use of NPs and siRNA-based treatments which currently display some limited success can be facilitated using PONI-Guan NPs as a platform to enhance delivery of siRNA with little cytotoxicity. Here, we show that PONI-Guan/siRNA polyplexes are readily taken up by primary human BEC and ASM cells via multiple mechanisms, and that the PONI-Guan system displays low cytotoxicity at different G/P ratio and a reasonable concentration. Furthermore, additional studies have explored the effectiveness of PONI-Guan/siRNA complexes in specific gene silencing in two different cell types relevant to airway function and asthma pathophysiology. In this regard, while our demonstration of the safety and efficacy of PONI-Guan NPs in siRNA delivery to two key airway cell types is encouraging, we recognize the importance of other cell types in asthma and remodeling. In particular, it will be important to demonstrate efficient targeted delivery to resident immune cells in the airway such as dendritic cells, macrophages or neutrophils with suppression of key inflammatory mediators, allowing for a multipronged approach to blunting various features of the asthmatic airway. Furthermore, we demonstrated lack of inflammatory effect of PONI-Guan NPs in BECs and ASM but there is need to demonstrate lack of such effects at different NP doses, particularly in inflammatory cells. Regardless, our study sets the stage to further explore the use of PONI-Guan/siRNA complexes as effective delivery mechanisms in the airway in vivo for modulating airway hypercontractility and remodeling.