Gelatin and lipidoid integrate to create gelasomes to enhance siRNA delivery with low toxicity

Abstract

RNAi therapeutics possess the potential to cure many uncurable human diseases. For instance, RNAi therapeutics using liposomes showed remarkable survival benefits in patients with liver diseases. However, the extension of liposomes to deliver RNA to cure other ailments has largely been unsuccessful. Therefore, researchers are focusing on designing and testing different combinations of materials for versatile RNA delivery applications. Yet, an efficient and safe RNA delivery platform has not been identified. In this work, we have developed a new class of RNA-delivery vehicle called “Gelasomes,” using an incongruous combination of gelatin and lipidoid to exploit each material's unique properties while overcoming their inherent limitations. The low in vivo toxicity of Gelasomes is attributed to the exterior gelatin layers that shield the exposure of cationic lipidoid-siRNA clusters and yet present a biocompatible surface. Indeed, toxicity studies in mice indicate that repeated administration of Gelasomes (up to 48 mg/kg BW) is well-tolerated with no notable changes in body weight, hematology, or serum chemistry. Interestingly, the gelatin outer layer efficiently protects siRNA from serum degradation (48 h), preserving its functionality beyond two months of storage. Notably, Gelasomes possess dual siRNA conjugation modes, i.e., electrostatic binding with lipidoid core and covalent attachment to gelatin surface. The bivalency coupled with lipidoids' high transfection efficiency rendered Gelasomes with remarkably high gene silencing efficiency (>90 %) at very low treatment doses in vitro (40 μg/mL). In vivo studies further confirmed the high gene silencing ability of Gelasomes in non-small cell lung tumor mouse models. This new platform is tunable on all fronts: size, degree of surface coating, and biomolecule functionalization. Truncating the lipidoid C14-tail to a C8-tail yielded Gelasomes of reduced size. As lipidoids with different carbon lengths are synthesizable, we can develop a library of Gelasomes with different sizes. The surface coating with less gelatin resulted in high transfection efficiency at low doses of Gelasomes. The structure of Gelasomes offers chemical handles to couple target-specific molecules like antibodies to tune their properties for efficient biological application.

Graphical abstract

Highlights

• •Designed a novel class of bioactive materials called “Gelasomes.”.
• •We demonstrated that gelasomes can store labile payloads such as siRNA and release within the cells.
• •Gelasomes release siRNA within cells and downregulate the target protein by >90 %.
• •We demonstrated the non-toxic nature of this material and used tumor mice models to show the target protein downregulation.

1. Introduction

RNA interference (RNAi) is a promising technology for cancer treatment [[1], [2], [3], [4]]. This technology uses sequence specific small interfering RNA (siRNA) molecules to silence oncogenes that are responsible for the growth, proliferation, treatment resistance, and metastasis of cancer [1,2]. Delivery of siRNAs to the disease site can be achieved by either encapsulating or covalently attaching them with nanoparticles of lipids, polymers, or inorganic material [1,[5], [6], [7], [8]]. Researchers have been using these materials to design delivery systems that require a low dose to achieve maximum gene silencing efficacy [9]. Despite extensive research, only moderate success has been achieved to date, and only a few candidates are in clinical use. The scarcity can be attributed to the limitations of the delivery system in terms of siRNA loading, stability, toxicity, cytoplasmic release, and transfection efficiency [5,10]. Taken together, the best combination of nanoparticle-siRNA construct is still elusive.

Gelatin is a popular choice for drug delivery applications due to its biocompatibility, biodegradability, non-immunogenicity, and abundant pendant functional moieties for bioconjugation [[11], [12], [13]]. It has been rated as a “generally regarded as safe (GRAS)" material by the U.S. FDA, making it a desirable option for clinical use [11,14]. In addition to its non-toxic nature, the physical properties of gelatin make it more attractive and have been formulated into various drug delivery systems such as nanoparticles, hydrogels, or films [12,15]. Among these, gelatin nanoparticles (GNPs) have gained prominence as drug delivery agents, as they can be chemically modified to improve their bioactivity and stability [16,17]. Despite the advantages, there are limited studies on using GNPs for RNA delivery. Gelatin is a polyampholyte polymer carrying both positive and negative charges in its backbone [18]. Thus, unmodified GNPs lack adequate positive charge to condense and stabilize the large negatively charged RNA molecules. Similarly, encapsulation of siRNA in the core of GNPs is not highly favorable [19]. As demonstrated with liposomes, efficient encapsulation of siRNA within nanoparticles requires a hydrophilic core for cohesive interactions and a hydrophobic shell to prevent the rapid escape of the drug [19]. GNPs do not offer a definite hydrophilic core or hydrophobic shell resulting in poor encapsulation efficiency [20]. Thus, the physico-chemical features of GNPs make it difficult to load (or encapsulate) genetic material electrostatically or physically. Our previous work focused on covalently attaching RNA to the GNPs surface only made slight advances and had limited success [21,22].

To overcome this issue, researchers modified gelatin with cationic molecules (spermine, 1,2-Ethanediamine, cholamine, or polyethyleneimine) and synthesized highly positively charged GNPs to achieve efficient electrostatic conjugation with siRNA [[23], [24], [25], [26], [27], [28], [29]]. Though this approach improved the siRNA loading of GNPs, it created a new problem, toxicity. Systemic administration of cationic species is known to cause multifaceted toxicity both in vitro and in vivo rendering them less useful for further applications [8,[30], [31], [32], [33], [34], [35], [36]]. Thus, cationizing GNPs is not a constructive strategy for safe delivery application.

A more promising approach is to utilize a cationic molecule to efficiently condense siRNA and encapsulate the electrostatic complex within a gelatin nanoparticle. This method could potentially enhance loading efficiency and reduce toxic effects on cells. Among the cationic species screened in the recent past for RNAi, lipidoids have emerged as a promising class of material with remarkable potential in RNA delivery [9,37,38]. Lipidoids are composed of cationic amino heads and alkyl tails with excellent RNA condensing ability [9,[37], [38], [39]]. Lipidoid-based nanoparticles were shown to facilitate efficient delivery of siRNA to diverse cell types and organs to effect simultaneous and persistent silencing of multiple genes at very low doses [9,37]. Thus, utilizing lipidoids to efficiently condense siRNA and shielding the cationic adduct within a biocompatible gelatin matrix would result in a novel nanocarrier with high RNA delivery efficiency and low-toxicity. However, stable interactions between gelatin and lipidoid are unlikely due to their opposing physico-chemical properties. Lipidoid is highly cationic and insoluble in water; gelatin is polyampholyte, amphiphilic, and alters its structural confirmation and molecular interactions based on solvent composition, pH, temperature, and co-materials. Hence, the synthetic strategy for developing a stable gelatin-lipidoid composite is not straightforward and has not been previously explored. Previous research efforts to integrate gelatin with lipids are very few where gelatin is used as a stabilizer or bioactive coat for solid lipid nanoparticles, and active processes such as emulsification or spray drying were required to stabilize the interactions between the two materials [[40], [41], [42]]. In these studies, the lipids are not cationic and structurally different and hence interact differently with gelatin than cationic lipidoids.

In this work, we report a novel synthetic strategy to produce gelatin-lipidoid nanocomposites or ‘Gelasomes’ with high siRNA transfection efficiency and no systemic toxicity. We describe the optimization of several synthesis parameters such as solvent composition, mass ratio of gelatin/lipidoid/siRNA mixture, pH, temperature, and crosslinking agent to produce stable and delivery efficient Gelasomes. Gelasomes comprise a cationic lipidoid-siRNA core that is uniformly shielded within a layer of gelatin matrix. The pendant functional groups on the gelatin surface were further activated to conjugate PEG, targeting antibody and additional siRNA. The salient features of Gelasomes are as follows: (1) Bivalency: siRNA is attached in two different binding modes, i.e., electrostatic binding with lipidoid core and covalent conjugation to gelatin surface. Thus, Gelasomes offer versatility to engineer the same or different siRNAs in two binding modes and enhance the siRNA loading and gene silencing efficacy; (2) Biocompatibility and stability: Gelasomes are negatively charged and behave like gelatin nanoparticles with regards to interaction with the physiological environment as the lipidoid-siRNA complex is encapsulated within. The protective gelatin layer renders additional stability to siRNA. Further, the presence of PEG improves the pharmacokinetics. (3) Size tunability: the length of alkyl tails of lipidoid can be changed to tune the particle size of Gelasomes. (4) High gene silencing efficiency: Bivalency coupled with lipidoid-mediated efficient siRNA loading and release yields high gene silencing ability at very low doses. The paper presents the synthesis optimization, characterization, in vitro and in vivo toxicity and delivery efficiency of Gelasomes.

2. Materials and methods

2.1. Materials

Gelatin (Bloom type B; Cat. No. G9391-500G), EDC (CAS No. 25952-53-8), Sulfo-NHS (CAS No. 106627-54-7), Sulfo-SMCC (CAS No. 92921-24-9), and glutaraldehyde (CAS No. 111-30-8) were purchased from Sigma-Aldrich (USA). G0C14 was purchased from Ruixi Biotech, China. Sterile ultraclean distilled water (Cat No. 10-977-015), sodium hydroxide (NaOH; CAS No. 1310-73-2), hydrochloric acid (HCl; CAS No. 7647-01-0), 200-proof ethanol (CAS No. 64-17-5), acetone (CAS No. 67-64-1), sodium chloride (NaCl; CAS No. 7647-14-5), sucrose (CAS No. 57-50-1), paraformaldehyde (PFA; Cat. No. 50-980-486), RPMI-media (Cat. No. A1049101) and fetal bovine serum (FBS; Cat. No. 26-140-079) were purchased from Thermo Fisher Scientific (USA). HS-PEG-COOH (Cat. No. CM-PEG–SH–2000) was purchased from Laysan Bio (USA). siRNA for AXL (5′ S–S.G.G.A.A.C·U.G.C.A.U.G.C·U.G.A.A.U.G.A.U·U 3’) was purchased from Horizon Discovery, USA. Cetuximab antibody was purchased from Syd Labs, USA.

2.2. Instrumentation

Automated liquid pipetting was performed on a single and multi channel INTEGRA pipette. pH measurements were performed using a Mettler Toledo pH-meter. Controlled addition of solvents was achieved using a KDS-200 Syringe pump. Sonication was performed on a Branson Bransonic® M Mechanical Ultrasonic Bath 3800. Hydrodynamic size and zeta potential were measured (by dynamic light scattering and laser doppler velocimetry techniques) on a Zetasizer Nano-ZS (Malvern Panalytical Ltd). Absorbance spectroscopy was performed on a BioTek Synergy H1 and Cytation-3 multi-mode microplate reader respectively. HR-TEM imaging was performed on a JEOL JEM-1400 high contrast TEM equipped with a 300 kV field emission gun (FEG) high resolution TEM/STEM. Protein expression was quantified using the traditional and Jess automated western blotting system (ProteinSimple).

2.3. Synthesis of Gelasomes

An aqueous solution of gelatin (10 mg/mL) was prepared by dissolving 50 mg of Type B Gelatin (Bloom strength 225) in 5 mL of ultra-pure water at 37 °C, 750 rpm. In another vial, a lipidoid-siRNA complex was formed by smoothly mixing 500 μL of G0C14 (2.5 mg/mL in acetone) with 20 μL of siAXL (200 μM) and incubating at room temperature for 1 min. To the gelatin solution, the lipidoid-siRNA complex was added dropwise at 37 °C, 750 rpm. The pH of the mixture was adjusted to 7.0 using 0.2 N NaOH. To this mixture, around 8 mL of warm ethanol was added dropwise at 37 °C, 750 rpm to trigger the formation of nanoparticles. The formation of nanoparticles was indicated by the appearance of turbidity. After 15 min, the nanoparticles were crosslinked by dropwise addition of 4 ml of the ethanol-water mixture (60 % ethanol) containing 48 μL of glutaraldehyde (25 % aqueous solution). The heat was turned off at this point. After an overnight reaction at 25 °C, 750 rpm, the excess glutaraldehyde was quenched by adding 5 μL of 1 M glycine to the reaction mixture. The nanoparticles were washed with water by 5 cycles of centrifugation (15,000 g for 8 min) and redispersion (35 mL of water). The washed pellet was dispersed in 250 μL of water to obtain Gelasomes. The above is the optimized procedure for synthesizing Gelasomes (1 d) where the siRNA to gelatin ratio is 1:1000. Most of the procedure remains the same for synthesizing 1a, 1 b, and 1c, except that the initial amounts of gelatin taken for preparing 10 mg/ml solution are 20 mg, 30 mg, and 40 mg respectively. The volumes of 25 % glutaraldehyde used for crosslinking 1a, 1 b, and 1c are 19.2 μL, 28.8 μL, and 38.4 μL respectively. Similarly, volumes of 1 M glycine used for 1a, 1 b, and 1c are 2 μL, 3  μL, and 4 μL respectively.

2.4. Surface conjugation of antibody and PEG to Gelasomes

Gelasomes were conjugated to cetuximab antibody and NH₂-PEG-COOH (2 kDa) by activation with EDC and sulfo-NHS mixture. Briefly, 8 mg of Gelasomes was dispersed in 500 μL of MES buffer (0.1 M, pH 4.5). To this dispersion, 200 μL of MES buffer containing 1.52 mg of EDC and 1.76 mg of sulfo-NHS was added. The reaction mixture was incubated at 25 °C, 800 rpm for 90 min followed by centrifugation at 15,000g for 6 min to obtain a pellet of activated Gelasomes. The pellet was redispersed in 500 μL of 1X PBS and mixed with 1.2 mg of cetuximab antibody and 4 mg of NH₂-PEG-COOH and the pH was adjusted to 7.2. The reaction mixture was incubated at 25 °C, 800 rpm for 16 h followed by centrifugation at 15,000g for 8 min. The pellet obtained was washed once with 1.5 mL of water and dispersed in 200 μL of water.

2.5. Surface conjugation of siRNA to Gelasomes

The Gelasomes-Ab-PEG were activated using sulfo-SMCC to conjugate additional siRNA on the surface of NP. Briefly, 8 mg of Gelasomes-Ab-PEG was dispersed in 0.5 mL of 1 XPBS, mixed with 160 μL of an aqueous solution containing 0.8 mg of sulfo-SMCC, and the pH was adjusted to 7.2. The reaction mixture was incubated at 4 °C, 800 rpm for 2 h, followed by centrifugation at 15,000 g for 8 min. The activated pellet was washed once with 1.5 mL of water, redispersed in 0.5 mL of 1X PBS, mixed with 20 μL of thiolated siAXL (200 μM), and the pH was adjusted to 7.0. The reaction mixture was incubated at 4 °C, 800 rpm for 16 h followed by centrifugation at 15,000 g for 8 min. The pellet obtained was washed once with 1.5 mL of water and dispersed in 200 μL of water to obtain dual siRNA Gelasomes.

2.6. Synthesis of G (siRNA)-Ab-PEG NP

G (siRNA)-Ab-PEG NP serve as control for Gelasomes and are synthesized using a similar protocol used for Gelasomes except that siRNA is encapsulated without the help of lipidoid. Briefly, 50 mg of Type B Gelatin (Bloom strength 225) was dissolved in 5 mL of ultra-pure water at 37 °C, 750 rpm. To this solution, 20 μL of 200 μM siAXL was added and the pH of the mixture was adjusted to 7.0. To this mixture, around 9.5 ml of warm ethanol was added dropwise at 37 °C, 750 rpm to trigger the formation of nanoparticles. The nanoparticles were crosslinked by dropwise addition of 4 ml of the ethanol-water mixture (65 % ethanol) containing 48 μL of glutaraldehyde (25 % aqueous solution). After an overnight reaction at 25 °C, 750 rpm, the excess glutaraldehyde was quenched by adding 5 μL of 1 M glycine to the reaction mixture. The nanoparticles were washed with water by 5 cycles of centrifugation (15,000 g for 8 min) and redispersion (35 mL of water). The protocol for conjugating antibody and PEG to these nanoparticles is the same as the one used for Gelasomes.

2.7. Synthesis of G-Ab-PEG-siRNA NP

G-Ab-PEG-siRNA NP serves as another control for Gelasomes. The first step involves the synthesis of empty gelatin nanoparticles with no siRNA encapsulated within. The procedure for this step remains the same as the one used for synthesizing siRNA encapsulated gelatin nanoparticles in G (siRNA)-Ab-PEG except that siRNA is not added to the gelatin solution. In the next step, the antibody and PEG are surface conjugated to the empty gelatin nanoparticles using the same procedure adopted for Gelasomes. Finally, siRNA is surface conjugated to the nanoparticles using the same procedure followed for Gelasomes.

2.8. Measurement of hydrodynamic size and zeta potential

Dynamic light scattering (for measurement of hydrodynamic size) and zetapotential measurements were performed with a Malvern Zetasizer Nano ZS (Malvern Instruments Ltd, USA) equipped with a 633 nm He–Ne laser operating at a backscattering angle (NIBS) of 173°. The Zetasizer Software version 7.13 was used to collect and analyze the data. The nanoparticle dispersion of 0.01 mg/mL was sonicated for 1 min before transferring to the measurement cuvette. Protein was chosen as the material to set the refractive index as 1.45, water was selected as the solvent and the temperature was equilibrated to 25 °C.

For hydrodynamic size, 1.0 mL of sample was measured in low-volume semi-micro disposable sizing cuvettes (Fisher Scientific, USA) with a path length of 10 mm. Triplicate measurements were made at a position of 4.65 mm from the cuvette wall with an automatic attenuator, with water as the dispersant and no equilibration time. The size distribution, the Z-average diameter (Z-ave), and the polydispersity index (PDI) were obtained from the autocorrelation function using the “general purpose mode” for all nanoparticle samples. For zeta potential, the laser Doppler microelectrophoresis technique using a non-invasive backscatter technology was used. 0.7 mL of sample was measured in folded capillary cells using the Debye-Huckel approximation.

2.9. High-resolution transmission electron microscopy (TEM) analysis

A dilute suspension of nanoparticle sample (0.01–0.05 mg/mL) was coated on a 200-mesh carbon-coated Cu-film grid. The TEM images of the samples were obtained using a JEOL JEM-1400 Transmission Electron Microscope, (Jeol Inc, USA) at 120 KV. Normal beam alignment was performed prior to image acquisition. The particle size was estimated using ImageJ software.

2.10. siRNA load of Gelasomes

The siRNA was loaded in Gelasomes in two steps: i) electrostatic complexation with G0C14 and encapsulation of the adduct within the gelatin matrix; and ii) covalent conjugation of siRNA to the gelatin surface. In both steps, the amount of siRNA loaded in the Gelasomes was calculated by subtracting the amount of siRNA that remained unconjugated from the amount of siRNA added to the reaction. siRNA-Cy5 was used for the reactions. The amount of siRNA-Cy5 that remained unconjugated was estimated by quantifying the siRNA-Cy5 present in the supernatant of the reaction mixture (after centrifugation of the reaction mixture) using the standard curve of Cy5. The standard curve of Cy5 was obtained using a plate reader with excitation and emission set at 649 nm and 666 nm respectively. The siRNA load per one mg of Gelasomes was calculated by dividing the total siRNA loaded in Gelasomes by the effective weight of Gelasomes.

2.11. Stability of siRNA

The stability of siRNA conjugated to nanoparticles in serum was assessed over a period of 2 days using the Polyacrylamide gel electrophoresis (PAGE) technique. siRNA-Cy5 was used for this experiment to enable fluorescence-based monitoring of siRNA. Briefly, the nanoparticle dispersion was mixed with RPMI 1640 media containing 10 % serum and incubated at 37 °C. At different time points, the mixture was subjected to gel electrophoresis. Stable siRNA-Cy5 that remains conjugated to nanoparticles cannot pass through the gel and would remain at the top of the well.

The stability of the gelatin coat surrounding the Gelasomes was examined by incubating 100 μL of 2 mg/mL Gelasomes in 100 μL of human plasma at 37 °C, 800 rpm. At regular intervals (up to 48 h), a small amount of Gelasomes was sampled and its zeta potential was analyzed. A stable coating of gelatin around the Gelasomes would result in a negative zeta potential.

2.12. In vitro cytotoxicity

A549 human adenocarcinoma epithelial cells (ATCC, USA) were grown in RPMI 1640 medium supplemented with 4.5 g/L d-glucose, 25 mM HEPES, 0.11 g/L sodium pyruvate, 1.5 g/L sodium bicarbonate, 2 mM l-glutamine, 10 % heat-inactivated fetal bovine serum (Atlanta Biologicals, USA) and 0.1 % v/v gentamycin. Cells were cultured in a humidified atmosphere of 95 % air and 5 % CO2 at 37 °C (Thermo Scientific, USA). For determining in vitro cytotoxicity of the nanoparticles, an MTT assay was performed on A549 cells. For the assay, 10,000 cells were seeded per well in a 96-well plate. After 12 h, the adhered cells were treated with different concentrations of nanoparticles diluted in media. After 24 h of treatment, 10 μL of MTT solution (Promega, USA) was added and the plate was incubated at 37 °C for 4 h. Crystals formed were dissolved in 100 μL solubilizing buffer and the plates were kept at 37 °C for 2 h. The absorbance of the solubilized formazan was measured using a Biotek Cytation 3 spectrophotometer at 570 nm. The viability of the cells transfected with various samples was then calculated by normalizing absorbance with untreated controls.

2.13. Western blot analysis

The gene silencing efficiency of the siRNA-loaded nanoparticles was evaluated by analyzing the reduction in the target protein levels in the treated cells by western blotting. Briefly, 1 × 10⁶ cells of A549 were incubated in a 6-well plate for overnight adherence at 37 °C in a 5 % CO₂ atmosphere. After overnight incubation, the media was removed, and the cells were treated with different concentrations of nanoparticles in serum-free media for a period of 60 h. As a positive control, siRNA transfection was performed using TransIT-X2 dynamic delivery system transfecting agent (Mirus Bio) as per the manufacturer's instructions. Whole-cell lysates were prepared using RIPA lysis buffer with MS-SAFE protease and phosphatase cocktail inhibitor (Sigma-Aldrich) and the protein amount was equalized using the Pierce Bicinchoninic acid (BCA) assay. Proteins were separated by 4–15 % PAGE (Bio-Rad) and were transferred onto nitrocellulose membranes (Bio-Rad). The membranes were incubated with primary antibody overnight at 4 °C, washed with 1X TBST buffer (5 cycles of 10 min each), incubated with secondary antibody at room temperature for 2 h, and washed with 1X TBST buffer (5 cycles of 10 min each). The membranes were developed with peroxidase-labeled anti-rabbit IgG (Cell Signaling Tech.) using enhanced chemiluminescence substrate (Thermo Fisher Scientific) and imaged on Bio-Rad Chemidoc imaging system. Protein expression levels were quantified by densitometry analysis. Actin was used to normalize the protein levels. Target protein expression of the in vivo tumor samples was analyzed using the Jess Automated Western Blot System (ProteinSimple) with the 12-230 kDa Fluorescence Separation Module and secondary anti-rabbit NIR module. Anti-human AXL Rabbit primary mAb was used at a dilution of 1:20 based on standardization experiments. Tumor samples were lysed using a rotor stator homogenizer in RIPA buffer with protease inhibitor cocktail at a ratio of 1:10 (10 µL for every mg of tissue). Protein concentrations were estimated using the Pierce BCA Protein Assay Kit. Diluted samples, protein ladders and total protein normalization reagent were loaded into the Jess cartridge according to the manufacturer's instructions. Protein signal intensity was quantified using the Jess software. Signal intensities were normalized to total protein levels for each sample.

2.14. Shelf-life study

Freshly prepared dispersion of Gelasomes was mixed with sucrose (10 % final concentration). Sucrose acts as a stabilizing agent for nanoparticles. The Gelasomes solution was lyophilized and stored at −20 °C. After 60 days, the lyophilized sample was homogenously redispersed in water. The gene silencing efficiency of the obtained Gelasome dispersion (60 days old) was evaluated and compared with that of freshly prepared Gelasomes to determine the long-term stability of stored Gelasomes.

2.15. In vivo animal studies

All animal studies were conducted following the University of Missouri's Animal Care and Use Committee (MU's ACUC) approved protocols as well as with federal guidelines. In vivo repeat dose toxicity studies of Gelasomes were performed in CF-1 female mice. Four-week-old female CF-1 mice were purchased from Envigo and housed in the animal research facility in a temperature- and humidity-controlled room with a 12 h light/12 h dark schedule for 1 week. On day 0, the mice were grouped into 4 groups with 6–8 animals per group: i) Untreated control (n = 6); ii) Gelasomes- 16 mg/kg BW dosage (n = 7); iii) Gelasomes- 32 mg/kg BW dosage (n = 8); iv) Gelasomes- 48 mg/kg BW dosage (n = 8). The Gelasomes treatment groups were intravenously administered with 0.2 mL (uniform aqueous suspension) of the respective dose two times in a week for three weeks (total of 6 injections). On every day of treatment, the animals in all four groups were monitored for changes in body weight and behavior. At the end of the study (day 21), the animals were sacrificed, and blood was collected for hematology analysis.

In vivo gene silencing efficiency of Gelasomes was evaluated using athymic nude mice. Five-weeks-old Athymic Nude-Foxn1^nu mice were purchased from Envigo and housed in the animal research facility in a temperature- and humidity-controlled room with a 12 h light/12 h dark schedule for 1 week. A549-Luc cells were trypsinized, suspended in 1:2 matrigel:PBS, and injected subcutaneously into the right flanks of 6-week-old Athymic Nude-Foxn1^nu mice (10 million cells per mouse). Three weeks after inoculation, the ten mice were randomly divided into two groups of 5 animals each: i) Untreated control; ii) Gelasomes treatment. The treatment group was intratumorally administered with a single dose of 50 μL of Gelasomes (3 μg of siRNA). Two days after the treatment, all the mice belonging to the two groups were sacrificed and the tumors were collected. The levels of AXL protein in the tumor tissues were measured using capillary western blot analysis.

2.16. Statistical analysis

Graphing and statistical analysis were performed using Prism software (GraphPad, San Diego, CA, USA). Two-tailed unpaired t-test was used to perform the statistical comparison between two groups. A p value less than 0.05 was considered statistically significant.

3. Results and discussion

3.1. Design of gelasomes

In this study, we synthesized a novel hybrid nanoparticle composed of gelatin and lipidoid to deliver siRNA safely and efficiently. The particle represents a structural advancement to the conventional lipid-based liposomes, which are well-known for drug delivery applications [[43], [44], [45]]. Liposomes suffer from instability and drug leakage due to the fluidity of the lipid layer [[43], [44], [45], [46]]. Further, cationic liposomes favorable for RNA delivery suffer from toxicity issue [33,47]. To overcome these problems, we integrated lipid with gelatin biopolymer to improve the stability and efficiency of the nanocarrier. We named the new class of material “Gelasomes” to reflect the composition of the nanocarrier. The schematic of Gelasomes is presented in Fig. 1A. The core of the Gelasomes is constituted by a cationic lipid-like material called lipidoid (G0C14) that forms a stable electrostatic adduct with siRNA. The lipidoid comprises a cationic head of generation 0 (G0) polyamidoamine dendrimer and hydrophobic tails of 14 carbon alkyl chains (C14). The rationale for choosing G0 is based on its minimal toxicity compared to higher generations while offering adequate positive charge to condense siRNA [48]. Importantly, lipidoids with C14 alkyl chains were shown to possess superior siRNA transfection efficiency [9]. Surrounding the lipidoid-siRNA core is a uniform layer of gelatin matrix that provides stability and prevents the direct exposure of cationic lipidoid. Additionally, the gelatin surface is engineerable through its abundant –COOH and –NH₂ moieties and functionalized with PEG, target antibody, and additional siRNA. Thus, Gelasomes are uniquely designed to be bivalent to facilitate loading of the same or different siRNA in two different conjugation modes, i.e., electrostatic encapsulation via lipidoid and covalent surface binding through gelatin. The bivalency of Gelasomes can be exploited to i) enhance the load of a single siRNA per particle, ii) load two different siRNAs and effect simultaneous downregulation of respective genes using a single particle, iii) direct the release of each siRNA by different stimuli.

Fig. 1.

A) Diagram representing the structure of Gelasomes- Gelasomes are hybrid nanocarriers comprising lipidoid-siRNA electrostatic adduct shielded by gelatin matrix. The surface gelatin layers are covalently conjugated with target antibody (Ab), NH₂-PEG-COOH (2 kDa), and additional siRNA. Thus, Gelasomes are bivalent by offering dual siRNA binding modes. B) Stepwise process for assembly of bivalent Gelasomes: -i) G0Cx (length of the carbon tail can be varied, x = 8 or 14) and siRNA are mixed in 25:1 mass ratio in acetone/water solvent (25:1, v/v) mixture to form an electrostatic adduct; ii) The adduct is added to gelatin solution, pH adjusted to 7 and temperature maintained at 37 °C. Ethanol is added to the mixture (ethanol:water-60:40, v/v) to initiate folding of gelatin fragments around G0Cx-siRNA adduct to form nanoparticulate Gelasomes (1). The amount of gelatin used to wrap the adduct can be varied to yield Gelasomes with decreasing amounts of gelatin coat (siRNA:gelatin mass ratio in the Gelasomes variants- 1a-1:400, 1b-1:600, 1c-1:800, 1d-1:1000). The Gelasomes are stabilized using glutaraldehyde crosslinker (gelatin:glutaraldehyde-1:0.45, w/w); iii) The –COOH groups on gelatin surface are activated using EDC-NHS chemistry to derive Gelasomes functionalized with PEG and Ab (2); iv) Finally, the NH₂ groups on gelatin surface are activated with SMCC and conjugated with thiolated siRNA to produce dual siRNA loaded Gelasomes (3). C) Summary chart of optimization of several reaction parameters and mass ratios of constituents for the synthesis of Gelasomes. The left side of the chart describes the particular step of the synthesis and the specific parameter (in bold) that is optimized. The right side mentions the variations in the specific parameter that are tested (in rectangular boxes). Yellow highlighted boxes represent the optimal condition that yielded the most stable and functionally efficient Gelasomes. D) Optimization of acetone/water ratio (v/v) in the solvent mixture to form a stable G0C14-siRNA complex. Picture showing that G0C14-siRNA complex precipitates when the ratio of acetone:water (v/v) in the solvent is 12.5:1, while remaining stable at 25:1. E) Optimization of mass ratio between G0C14 and siRNA to form Gelasomes with high gene downregulation (DR) efficacy. Western blot (WB) analysis of percentage DR of AXL protein in A549 cells after 60 h of treatment with Gelasomes synthesized using different mass ratios of G0C14 and siAXL. 25:1 and 50:1 showed similar and higher DR efficiency than 10:1. F) Optimization of methods for efficient encapsulation of G0C14-siRNA adduct within gelatin matrix. Two methods were tested: M1-the adduct is premixed with ethanol that is added to the gelatin solution to trigger Gelasomes formation; M2-the adduct is preincubated with gelatin solution for 10 min, then pure ethanol is added to the mixture to form Gelasomes. WB analysis showing that Gelasomes obtained from M2 showed superior DR efficiency of AXL protein in A549 cell lines treated for 60 h. G) Optimization of pH of gelatin and G0C14-siRNA mixture to support the formation of stable and efficient Gelasomes upon the addition of ethanol. WB analysis showing that bringing the pH close to neutral resulted in Gelasomes with superior DR efficiency. H) Optimization of the reaction temperature in M2- WB analysis showing that reaction temperature of 37 °C yielded Gelasomes with superior DR efficiency than that of 25 °C.

We have used type B gelatin for synthesizing Gelasomes as it acquires a net negative charge at the physiological pH due to its isoelectric point (pI) of 4.8–5.4. Thus, type B gelatin, in addition to shielding the cationic lipidoid, also presents a negative surface thereby providing biocompatibility to Gelasomes. On the other hand, type A gelatin (pI 8–9) acquires a positive charge at physiological pH and will present a positively charged surface even after shielding the lipidoid. Thus, type A gelatin was not used for synthesizing Gelasomes as it does not help address the systemic toxicity associated with the cationic species. In this study, we have used siRNA for AXL protein to demonstrate the gene silencing efficacy of Gelasomes. AXL is a receptor tyrosine kinase that was implicated as a cancer driver in several tumor types and its downregulation was found to offer potential benefits in cancer treatment [49]. Hence, we have demonstrated the potential of Gelasomes to carry AXL siRNA and downregulate AXL protein. However, the design of the Gelasomes can be extended to any other therapeutic RNA molecule.

3.2. Optimization of synthesis of gelasomes

As illustrated in Fig. 1B, we accomplished the synthesis of Gelasomes in four steps:

• 1.Formation of an electrostatic adduct between G0C14 and siRNA.
• 2.Encapsulation of the adduct in a gelatin matrix and stabilization of the composite.
• 3.Covalent conjugation of PEG and target Ab on the gelatin surface.
• 4.Covalent conjugation of additional siRNA on gelatin surface.

The stability and downregulation efficiency of Gelasomes depends on the successful execution of the first two steps of the synthesis involving complex interactions between G0C14, siRNA, and gelatin. G0C14 is a cationic moderately sized amphiphilic molecule not soluble in water, while siRNA is an anionic large-sized hydrophilic molecule. Gelatin is a polyampholyte, amphiphilic biopolymer composed of a mixture of high molecular weight peptides whose overall charge, inter- and intra-molecular interactions, and structural conformation depend on several factors such as pH, temperature, solvent, and ionic strength of the medium [16,50]. Thus, considering the three materials' unique and competing physicochemical attributes, precise control of several process parameters is necessary to produce a stable and functionally efficient G0C14-siRNA-gelatin composite. Here, we briefly discuss the stepwise optimization trials that lead to reproducible synthesis procedures for Gelasomes (Fig. 1C).

The first step is the slow mixing of an aqueous solution of siRNA with G0C14 solution in acetone to form the G0C14-siRNA adduct. The volume ratio of acetone and water in the G0C14-siRNA mixture is critical to enhancing the interaction between the two species and maintaining the stability of the adduct. Due to its several hydrophobic tails, G0C14 does not favor an aqueous environment and forms an immiscible precipitate in an acetone/water mixture of 12.5:1 (Fig. 1D). We found that increasing the proportion of acetone in the mix (25:1 v/v ratio of acetone/water) yielded an adduct with extended stability. Next, we varied the weight ratio between G0C14 and siRNA (10:1, 25:1, 50:1) to determine the minimum amount of lipidoid necessary to efficiently condense and transfect the siRNA, resulting in high gene silencing efficacy. We observed that the Gelasomes synthesized using a 10:1 mass ratio of G0C14 is less efficient than those derived from a higher lipidoid amount (25:1) (Fig. 1E). Further, increasing the amount of lipidoid (50:1) does not significantly increase the gene silencing efficacy. Thus, we chose 25:1 as the optimum G0C14 to siRNA mass ratio for synthesizing Gelasomes.

The next step involves packing the G0C14-siRNA clusters in a uniform gelatin matrix to yield siRNA encapsulated nano assembly of gelatin-lipidoid composite or Gelasomes. To achieve this, we employed a desolvation technique to trigger gelatin fragments to fold into nano aggregates in the presence of G0C14-siRNA clusters. We used ethanol as a desolvating agent to dehydrate gelatin and wind it around G0C14-siRNA clusters. We experimented with two different methods to encapsulate G0C14-siRNA within the gelatin matrix: M1) add G0C14-siRNA along with ethanol to aqueous gelatin solution; M2) pre-incubate G0C14-siRNA adduct with gelatin solution before adding ethanol. We observed that M2 produced nano assemblies with superior gene silencing efficiency and therefore was chosen to synthesize Gelasomes (Fig. 1F). The preincubation step in M2 allows for non-covalent interactions between the two species and draws more G0C14-siRNA clusters inside the gelatin matrix during the nano aggregation process. Additionally, co-incubation with gelatin allows for efficient diffusion of G0C14-siRNA clusters to the inner space as the nano aggregates are formed. Next, we optimized the volume of the desolvating agent (ethanol) to be added to an aqueous gelatin solution to produce Gelasomes. The addition of slightly excess amounts of ethanol (70:30 or 65:30 v/v of ethanol/water) resulted in the collapse of the Gelasomes into macroaggregates. Conversely, lower amounts of ethanol (55:45 v/v of ethanol/water) significantly reduced the yield of Gelasomes. We found that the ethanol-to-water ratio of 60:40 (v/v) produced stable Gelasomes with good yield.

The pH of the gelatin-G0C14-siRNA mixture before adding ethanol influenced the stability and efficiency of the Gelasomes. Acidic pH resulted in uncontrolled aggregation of Gelasomes. This can be attributed to the net charge of the gelatin nearing zero as the pH approaches its isoelectric point (4-8-5.4) [51,52]. The neutral charge reduces the electrostatic repulsion between the nanoparticles, causing aggregation. Basic pH resulted in poor downregulation efficiency of the derived Gelasomes (Fig. 1G). This result was somewhat unexpected as we anticipated that gelatin's negative charge would increase at basic pH, resulting in enhanced interaction with positively charged G0C14-siRNA adduct. Alternatively, basic pH can cause deprotonation of tertiary amines in the cationic head of G0C14, diminish its positive charge, and weaken the electrostatic interaction with siRNA. This poor interaction results in Gelasomes with suboptimal siRNA loading, reducing their efficiency in downregulating target proteins. The near-neutral pH of 6.85–7.0 has resulted in Gelasomes with optimum stability and gene silencing efficiency. On the same note, we found that the reaction temperature of 35–37 °C was optimum for synthesizing Gelasomes as lower temperatures resulted in quick gelation and aggregation of nanoparticles (Fig. 1H). Finally, we have tested two crosslinking agents for binding the gelatin fragments surrounding the Gelasomes and imparting rigidity and stability to the nanoparticle: i) glyoxal and ii) glutaraldehyde. We observed that adding glyoxal invariably resulted in aggregation and flocculation of Gelasomes. Conversely, glutaraldehyde with a weight ratio of 1:0.45 (gelatin; glutaraldehyde) yielded stable, rigid particles and hence was chosen to crosslink Gelasomes. Additionally, we have found that adding a diluted solution (0.35 % final concentration) of the glutaraldehyde is essential to control the cross-linking rate and prevent the formation of flocculates.

3.3. Versatility of gelasomes

Following successful optimization of the synthesis of Gelasomes, we evaluated if the structure of the Gelasomes has the flexibility to accommodate variations in the constituent materials and tune its properties for efficient biological application. We used the optimized synthetic parameters to produce a library of Gelasomes with different physico-chemical or biological properties by modifying the lipidoid or gelatin content. For example, Gelasomes with smaller sizes were obtained by substituting G0C14 with a lipidoid with a smaller carbon tail (G0C8) (Figs. S1 and S2). Also, the zeta potential of these particles is close to neutral.

Similarly, we were able to synthesize Gelasomes with varying amounts of gelatin coating to determine the minimum amount of gelatin fragments sufficient to form a protective shield around the G0C14-siRNA adduct. The presence of excess gelatin other than what is necessary to shield the charged adduct can either result in excessive layering of gelatin around Gelasomes or the generation of empty gelatin nanoparticles with no payload inside. Generating bare and half-filled nanoparticles would increase the dose of nanoparticles being administered to obtain the desired biological efficiency. Administering low-dose and getting the maximum benefit is desirable, and that is the overall goal of designing an effective delivery vehicle. Therefore, we synthesized four different Gelasomes using the same amount of G0C14-siRNA adduct and changing the weight ratio of siRNA to gelatin (1:400 (a), 1:600 (b), 1:800 (c), 1:1000 (d)) (Fig. 1B). We labeled the simple gelatin-coated cores as Gelasomes 1a-1d, while 2a-2d represent the respective Gelasomes with the functionalization of Ab and PEG. Lastly, the dual siRNA Gelasomes were labeled as 3a-3d. All the four Gelasomes were purified and characterized.

3.4. Characterization of gelasomes

3.4.1. Size

Gelasomes were characterized using conventional analytical techniques, which include TEM and DLS. The TEM images of Gelasomes and other control gelatin nanoparticles are shown in Fig. 2A. They are spherical and monodisperse, with a size range of 220–260 nm. Interestingly, the size of Gelasomes is relatively unaltered, whether or not the lipidoid is present within the void. This observation suggests that the gelatin nanoparticles create a “defined” void space, and the lipidoid or its adduct with siRNA fits well within the space. More representative TEM images of these particles are shown in Figs. S3–S6. Of note, the size and shape of the Gelasomes are not significantly affected by reducing the amount of gelatin used for forming the protective coat (Fig. S7). As typically observed for gelatin-based formulations, the hydrodynamic size of the Gelasomes and the control particles is larger than the actual particle diameter due to the well-known nature of the gelatin to swell in water (Fig. S8) [53].

Fig. 2.

A) Representative TEM images of four gelatin-based nanoparticles that are evaluated for siRNA delivery application: i) G (si)-Ab-PEG - siRNA is encapsulated in the gelatin nanoparticle without the help of a lipidoid (G0); ii) G (G0)-Ab-PEG-si - G0C14 is encapsulated in gelatin matrix and siRNA is covalently attached to the surface; iii) G (G0si)-Ab-PEG or Gelasomes (3a)– G0C14-siRNA adduct is encapsulated in gelatin matrix; iv) G (G0si)-Ab-PEG-si or dual siRNA Gelasomes (3 d) - G0C14-siRNA adduct is encapsulated in gelatin matrix and additional siRNA is covalently attached to the surface. All four nanoparticles are surface functionalized with Ab and PEG. The nanoparticles (i), (ii), and (iii) serve as controls for iv (bivalent Gelasomes). B) Change in the zeta potential during the stepwise assembly of Gelasomes. The sudden shift in the zeta potential of G0C14-siRNA adduct before and after the formation of Gelasomes (from highly positive to negative) suggests that the gelatin matrix efficiently shields the cationic adduct from surface exposure. C) Comparison of siRNA loads of dual siRNA-Gelasomes (3 d) with other nanoparticle controls mentioned above. Bivalency of Gelasomes helps in achieving higher siRNA loads per mg of the nanoparticle. D) Comparison of siRNA loads of dual siRNA-Gelasomes synthesized using increasing amounts of gelatin coating (siRNA:gelatin mass ratio in the Gelasomes variants- (3a)-1:400, (3 b)-1:600, (3c)-1:800, (3 d)-1:1000). The siRNA load of Gelasomes increases as the amount of gelatin coat decreases. E) Gel electrophoresis images showing serum stability (48 h) of siRNA loaded in Gelasomes (3 d) in two different conjugation modes, i.e., electrostatic encapsulation (left), and covalent surface conjugation (right). No significant bands representing unbound, or fragment siRNA were observed in either case suggesting the stability rendered by gelatin coating.

3.4.2. Zeta potential

The primary objective behind the design of Gelasomes is to shield the cationic lipidoid-siRNA adducts with a uniform matrix of gelatin so that efficient delivery of siRNA is achieved without toxic implications associated with cationic species. During the formation of a hybrid nanoparticle, gelatin can get homogenously integrated with a lipidoid without forming a protective coat around the lipidoid. In that case, the particles would remain cationic due to exposure to the lipidoid and become inefficient for delivery purposes. Zeta potential is a vital characterization that validates the defined formation of Gelasomes by measuring its surface charge. We have measured the change in the zeta potential at every stage during the synthesis of Gelasomes (Fig. 2B). As known, lipidoid has a high positive zeta potential (>60 mV). Upon complexation with siRNA, the zeta potential is reduced to +50 mV, indicating the electrostatic binding of siRNA to the lipidoid. Of note, mixing the electrostatic adduct with gelatin and initiating the desolvation process has resulted in nanoparticles with negative zeta potential. This radical shift in zeta potential from +50 mV to – 15 mV suggests that the synthesis has resulted in the desired formation of Gelasomes where the gelatin matrix completely shields the cationic charge of the lipidoid complex. The negative charge of Gelasomes is due to the surface gelatin fragments, which acquire an overall negative charge at any pH above their isoelectric point (4.8–5.4). Thus, Gelasomes remain negatively charged at physiological pH (7.4), serving as safe delivery vehicles.

Additionally, the zeta potential remained negative for all four Gelasomes derived from decreasing amounts of gelatin coat (siRNA to gelatin mass ratio- 1:400 (1a), 1:600 (1b), 1:800 (1c), 1:1000 (1d)) (Table S1). These results indicate that 1a with minimum amounts of gelatin strands is sufficient to shield the electrostatic adduct effectively. The implications of these findings are significant for the translational opportunities of Gelasomes. The negative charge on Gelasomes remained unaltered following functionalization with Ab and PEG. The dual-siRNA Gelasomes, wherein siRNA is also attached covalently to the surface, showed minimal change in the zeta potential.

3.4.3. siRNA and antibody loading

In this study, we have used Cy5 labeled siRNA to enable fluorescence-based quantification of siRNA. We estimated siRNA loaded in Gelasomes (3a-3d) by measuring the amount of unconjugated siRNA in the supernatant using the standard curve of siRNA-Cy5 (Fig. S9). siRNA load was the highest for the bivalent gelatin nanoparticles compared to controls (Fig. 2C). Notably, the amount of gelatin used for coating the lipidoid complex significantly impacts the siRNA load of the Gelasomes. The lesser the gelatin coating, the higher the siRNA loading per mg of the Gelasomes (Fig. 2D). For example, 3d has a siRNA load of ≈12.5 μg/mg, while 3a has a significantly higher siRNA load of ≈28 μg/mg.

The –COOH functionalities on the surface of Gelasomes can be activated using EDC and sulfo-NHS to conjugate with antibodies that can help in active targeting of the payload. To demonstrate the ability of Gelasomes to couple with antibodies, we have studied the conjugation of Gelasomes with two different antibodies, i.e., cetuximab and anti-beta-globin antibody. The conjugation efficiency and loading of cetuximab and anti-beta-globin antibodies were estimated by measuring the amount of unconjugated antibody in the supernatant of the reaction mixture using the standard curves of respective antibodies. The BCA assay was used to obtain the standard curves of the antibodies (Fig. S10). The conjugation efficiency of cetuximab and anti-beta-globin antibodies were found to be 68 % and 72 % respectively. Similarly, the loading of cetuximab and anti-beta-globin antibodies were found to be 98 μg and 104 μg per mg of Gelasomes respectively. Thus, Gelasomes possess the versatility to efficiently conjugate with any antibody based on the intended application.

3.4.4. Stability

The effectiveness of the nanocarrier depends on its ability to safeguard siRNA from degradation by serum nucleases. To evaluate the serum stability of siRNA associated with Gelasomes, we used two different nanoconstructs that represent the two binding modes of siRNA: i) Gelasomes where siRNA is electrostatically attached with G0C14 and encapsulated within gelatin matrix; ii) Gelasomes with siRNA covalently conjugated on the surface. We used siRNA-Cy5 for this experiment and evaluated stability using gel electrophoresis. In both cases, siRNA remained highly stable after incubating in the serum for 48 h (Fig. 2E). The first type, where the siRNA is electrostatically attached, showed minor disintegration after 24 h. In contrast, the covalent attachment of siRNA rendered very high stability, and no disintegration was observed for 48 h. Combining both would provide a remarkable ability to retain the siRNA for 48 h. It is evident that Gelasomes impart high stability to siRNA and protect it from serum nucleases.

Further, we investigated the stability of the gelatin coat surrounding the G0C14-siRNA complex in human plasma to understand if the gelatin can continue to shield the cationic G0C14 complex and preserve the biocompatibility of Gelasomes in vivo. As shown in Fig. 2b, the degradation of the gelatin coat around the Gelasomes will expose G0C14 causing the zeta potential to become positive. However, the zeta potential of the Gelasomes incubated in human plasma remained stable and negative for up to 48 h thereby suggesting the stability of the gelatin coat (Fig. S11). Thus, Gelasomes may not exhibit toxic effects in systemic circulation as G0C14 is stably shielded by the gelatin coat.

3.5. In vitro cytotoxicity of Gelasomes

Cationic nanoparticles can pack more molecules of siRNA due to electrostatic solid interactions and cause efficient target gene downregulation [45]. Hence, the most common strategy adopted to enhance gelatin nanoparticles' siRNA loading and gene regulation efficiency is to cationize them by modifications with positively charged molecules or polymers [[23], [24], [25], [26], [27], [28], [29]]. However, the surface charge on the nanoparticles influences their cellular adherence, endocytosis, and toxicity [45,54]. Literature suggests that the use of cationic nanoparticles for cellular applications is limited by their toxicity arising from disruption of plasma-membrane integrity, damage to mitochondria and lysosomes, and generation of many autophagosomes [45,54]. Therefore, shielding the positive charge of nanoparticles used for siRNA delivery applications is essential. To elucidate the difference in the toxicity between negatively charged Gelasomes and regular cationic gelatin nanoparticles, we have synthesized protamine functionalized gelatin nanoparticle (GNP-Prot), which represents the cationized gelatin nanoparticle family, and compared its toxicity with Gelasomes. In this study, we used a cancer cell line (A549) as the test system; the rationale for this choice is we will be using the same cell line in the next step to evaluate the target gene knockdown efficacy.

Protamine (Prot) is a natural cationic peptide widely explored for RNA transfection [55]. Hence, we chose protamine to modify the surface of gelatin nanoparticles and cationize them. The nanoparticles (GNP-Prot) were further pegylated and electrostatically conjugated with siRNA for AXL protein (Fig. S12). GNP-Prot-PEG are spherical and are of similar size as Gelasomes (Fig. S13) As expected, GNP-Prot-PEG particles efficiently transfected siRNA and downregulated the target gene (AXL) at 0.25 mg/mL treatment concentration (Fig. S14). However, the MTT assay showed that these particles are highly toxic to cells at significantly lower concentrations (IC50 <0.008 mg/mL) (Fig. 3A). Reducing the toxicity of these particles is desirable for therapeutic applications. Thus, we have prepared a library of GNP-Prot-PEG particles with lowering amounts of protamine and increasing amounts of PEG and evaluated their toxicity. The rationale for adding PEG is based on the well-established fact that PEGylation decreases the toxicity of particles. However, all GNP-Prot-PEG variants were toxic to cells (Fig. 3A). IC50 was found at very low treatment concentrations even for constructs with very low amounts of protamine and high amounts of PEG. Simultaneously, we evaluated whether cells could tolerate Gelasomes in the concentration at which they can downregulate the target protein. The Gelasomes were well-tolerated by cells, and very minimal cell death was observed (<20 %), which is usually observed due to shock or stress (Fig. 3A). IC50 was not reached for the Gelasomes even at the highest concentration (0.25 mg/ml), suggesting the relative non-toxicity of the construct. We have observed that the toxicity of the nanoparticles is directly related to their cationic nature. As shown in Fig. 3A, the IC50 of the GNP-Prot-PEG was very low for the highly positively charged variant and it gradually increased as the surface positive charge (or zeta potential) is reduced by lowering the amount of protamine conjugated to the nanoparticle. As the surface charge is negative in Gelasomes, the IC50 is considerably higher than GNP-Prot-PEG suggesting the non-toxic nature of the construct.

Fig. 3.

A) Comparison of cytotoxicity of Gelasomes (3 d) with cationic GNP-Protamine-PEG constructs using MTT assay. Several variants of GNP-Prot-PEG with decreasing amounts of protamine and increasing amounts of PEG were synthesized to evaluate improvement in cytotoxicity. Toxicity was observed in all GNP-Prot-PEG variants. IC50 was found at low treatment concentrations even for variants with very low amounts of protamine and high amounts of PEG. Gelasomes on the other hand are non-toxic as IC50 was not found even at high treatment concentrations. B) Plot of IC50 of the nanoparticles as a function of their zeta potential (ZP). X-axis, from left to right- GNP-Prot-PEG variants with different mass ratios of gelatin:protamine:PEG, amount of gelatin is kept constant while the mass proportions of protamine and PEG were gradually reduced and increased respectively. The last construct on the X-axis is Gelasomes with gelatin:G0C14:PEG mass ratio of 1:0.18:0.5. Right Y-axis- The positive ZP of the GNP-Prot-PEG constructs decreases as the amount of protamine is reduced. ZP of Gelasomes is negative due to the shielding of cationic G0C14 by gelatin. Left Y-axis- IC50 of GNP-Prot-PEG variant is very low due to high positive ZP and gradually increases with reduction in ZP. IC50 of Gelasomes is considerably higher due to its negative ZP. C) Comparison of cytotoxicity of empty GNPs and one with G0C14 (Gelasomes). Statistical values are reported as mean ± standard deviation (n = 3). Significant differences between the groups are indicated by *P < 0.05,**P < 0.01.

Importantly, reducing the amount of protamine by several folds did not significantly reduce the cytotoxicity of GNP-Prot-PEG. For example, even when GNP:Prot:PEG mass ratio is modified to 1:0.00625:2, where the amount of protamine is 160 times lower than gelatin, the toxicity remains (IC50 of 0.1 mg/mL). The toxicity can be attributed to the residual positive charge of the construct contributed by small amounts of surface protamine (Fig. 3C). In addition to the persistent toxicity, the resultant GNP-Prot-PEG with low positive zeta potential (+3 mV) may not be able to efficiently condense and stabilize siRNA. In contrast, the mass ratio of GNP:G0C14:PEG in Gelasomes is 1:0.18:0.5, where the amount of cationic molecules (G0C14) is only five times lower than gelatin. Nonetheless, Gelasomes were found to be non-toxic to cells as the gelatin layer effectively shields the positive charge of the lipidoid. Thus, Gelasomes behave similarly to gelatin nanoparticles as the lipidoid is not exposed to interact with the cell membrane. To further support this, we have shown that Gelasomes and empty gelatin nanoparticles without lipidoids had similar effects on cells (Fig. 3B). Compared to cationic GNP, Gelasomes are less toxic and ideal for the safe delivery of siRNA.

The ability of gelatin to significantly reduce the toxicity of cationic species was previously reported by Mimi et al. [25] The authors utilized polyethyleneimine (PEI) as the cationic carrier because of its ability to efficiently condense siRNA and support endosomal escape for payload release in the cytoplasm. However, the use of PEI is hindered by its high toxicity. To address this, authors have used GNPs as a substrate to immobilize PEI and limit its cellular interaction. The resulting GNP-PEI core-shell conjugate was found to be four times less toxic than PEI. It is important to note that this strategy has only reduced the high toxicity of PEI to moderate levels, but not lowered it to tolerable levels. The IC50 of GNP-PEI is still low (33 μg/mL) suggesting the non-compatibility with cells. This can be due to the high positive charge of GNP-PEI (>+30 mv) as the PEI molecules on the GNP core are exposed to the surface. Thus, the use of gelatin matrix as a protective coat to shield the cationic species could be an efficient strategy for reducing cytotoxicity as demonstrated in Gelasomes.

3.6. In vitro gene silencing efficacy of Gelasomes

Having demonstrated the non-toxic nature of Gelasomes, the next step is to evaluate their efficiency in downregulating the target protein in cells. We chose A549 cancer cells for our investigation; it is a lung cancer cell line and overexpresses a protein called “AXL,” which is responsible for acquired drug resistance in non-small cell lung cancer (NSCLC). In this study, our objective is to evaluate the efficacy of Gelasomes in silencing AXL expression.

As a first step, we synthesized two gelatin nanoparticles without lipidoid for comparison: (i) G (si)-Ab-PEG and (ii) G-Ab-PEG-si. In the first one, siRNA is encapsulated within the gelatin nanoparticle without the help of a lipidoid. In the second nanoparticle, siRNA is attached covalently to the surface of the nanoparticle. These two nanoparticles serve as appropriate controls of dual siRNA Gelasomes, where the siRNA is encapsulated with G0 as well as covalently attached to the surface. The comparison would help us determine the effect of lipiodoid as a differentiation factor for this new class of material. Further, to evaluate the advantage of bivalency of Gelasomes, we synthesized a third control, i.e., Gelasomes without the siRNA on the surface (2d) and compared its gene silencing efficacy with dual siRNA Gelaomses (3d).

We treated different concentrations of all four types of gelatin nanoparticles on A549 cells. Fig. 4A shows that the control gelatin nanoparticles achieved only 30–60 % downregulation of AXL protein in A549 cells. In contrast, Gelasomes showed >97 % AXL downregulating efficacy. Based on the data obtained, we can conclude the following: (i) G0C14 plays a vital role in achieving higher downregulation efficiencies in Gelasomes when compared with the gelatin nanoparticles with no lipidoid; This can be attributed to the ability of G0C14 to facilitate high siRNA loading, and efficient cytoplasmic release. (ii) Bivalency helps in higher gene silencing efficacy as Gelasomes with additional siRNA on the surface (3d) showed better downregulation efficiency than those without (2d); iii). Gelasomes showed target protein downregulation similar to or better than transfecting agents (positive control). To validate the efficacy of Gelasomes in other cancer cell types, we treated SKOV-3 (ovarian cancer), OVCAR-8 (ovarian cancer), and H1975 (NSCLC) with Gelasomes. We have found that the construct was able to completely downregulate the target protein (AXL) in all the cell types (Fig. S15).

Fig. 4.

A) Western blot (WB) analysis comparing the downregulation (DR) efficiency of dual siRNA Gelasomes (3 d) with its three controls: i) G (si)-Ab-PEG – siRNA is encapsulated in gelatin nanoparticles without G0C14; ii) G-Ab-PEG-si – siRNA is surface conjugated to gelatin nanoparticles; iii) single siRNA Gelasomes (2 d) – siRNA is complexed with G0C14 and the adduct is encapsulated within gelatin. All the four nanoparticles were functionalized with PEG and Ab. The siRNA used is specific to AXL. They were treated at three different concentrations in the A549 cell line for 60 h. The dual siRNA Gelasomes (3 d) showed the highest DR efficiency compared to controls at all concentrations. B) WB analysis comparing the DR efficiency of single siRNA Gelasomes synthesized using decreasing amounts of gelatin (siRNA to gelatin ratio in Gelasomes variants– 1:800 (2c), 1:600 (2 b), 1:400 (2a). All three types of Gelasomes were treated at two different concentrations in the A549 cell line for 60 h. Gelasomes synthesized from the lowest amount of gelatin (2a) showed high DR efficiency at very low doses. C) WB analysis comparing the DR efficiency of dual siRNA Gelasomes synthesized using decreasing amounts of gelatin coating (3 d, 3c, 3 b, 3a). All four variations of Gelasomes were treated at three different concentrations (10, 15, and 20 μg/mL) in the A549 cell line for 60 h. Gelasomes with the lowest amount of gelatin (3a) showed the highest DR efficiency at all doses. D) Comparison of cytotoxicity of dual siRNA Gelasomes variants (3a, 3 b, 3c, 3 d) at concentrations used for downregulation study. IC50 was not reached for the variants at all concentrations. E) Comparison of DR efficiency (calculated from WB) of the freshly prepared dual siRNA Gelasomes (3 d) with that of the lyophilized sample stored for 60 days.

An essential objective of our study is to achieve high downregulation efficiency at very low doses of Gelasomes. Thus, we evaluated the effectiveness of Gelasomes containing less gelatin (2a, 2b, 2c) in downregulating the target protein. We treated cells with Gelasomes 2a, 2b, 2c at different concentrations and found that 2a exhibited >98 % downregulation efficiency at a lower treatment concentration than 2b and 2c (Fig. S16). This promising result encouraged us to conduct another experiment wherein we treated 2a, 2b, and 2c at different concentrations. Remarkably, 2a was able to downregulate 95 % of target protein in 0.5 million cells at a very low concentration of 90 μg/mL (Fig. 4B).

Encouraged by these findings, we synthesized dual siAXL Gelasomes using decreasing amounts of gelatin (3a, 3b, 3c, 3d). We treated all four variations of Gelasomes at three different concentrations (10, 15, 20 μg/mL) in A549 cell lines (0.5 million). At all three concentrations, 3a showed the highest downregulation efficiency (Fig. 4C). At lower concentrations of 15, 10 μg/mL, 3b, 3c, and 3d were not able to downregulate the target gene, while 3a showed significantly higher downregulation efficiency. None of the constructs were toxic at these treatment concentrations (Fig. 4D). To establish the excellent downregulation efficiency of 3a, we tested its efficacy in other human cancer cell lines such as OVCAR-8 and MDA-MB-231. As shown in Fig. S17, Gelasomes (3a) completely downregulated the AXL protein in these cell lines even at a very low dose of 50 μg/mL. Interestingly, the Gelasomes (3a) synthesized using lipidoid with a smaller carbon tail (G0C8) was found to be less effective in downregulating the target protein than that of Gelasomes (3a) derived from G0C14 (Fig. S17). Understanding the cause and further optimization of the synthesis of Gelasomes-C8 will be the focus of our future studies.

In addition to the high siRNA loading efficiency exhibited by the Gelasomes, another factor that contributes to its excellent gene downregulation efficiency is the ability to efficiently release siRNA in the cytoplasm. After endocytosis, prompt release of the nanoparticle from the endo-lysosomal compartments is necessary to prevent degradation of siRNA. The cationic G0-C14, due to its proton buffering capacity causes enhanced Cl^- and water accumulation in the endosome resulting in osmotically induced swelling and rupture of the endosomal membrane [56,57]. Thus, G0C14 plays an important role in the endosomal escape of the siRNA complex and its release into the cytoplasm. The G0C14-siRNA complex experiences a higher pH in cytoplasm than in endosome resulting in reduced interaction between the two species and release of siRNA [56]. The released siRNA gets loaded onto the RNA-induced silencing complex (RISC) and binds to the target mRNA in a sequence-dependent manner, causing its degradation and downregulation of the corresponding protein [58].

To highlight the potential of Gelasomes, we have compared the unique characteristics of Gelasomes to other gelatin-based siRNA delivery vehicles reported in the literature. The design of the majority of the gelatin-based siRNA delivery vehicles is based on modifying the gelatin with cationic moieties to facilitate better siRNA loading and delivery efficiencies (Table S2). These modifications alter the properties of gelatin, potentially leading to changes in its interactions with living tissues. Importantly, the administration of cationic species is known to exhibit toxic physiological effects [33,47,54]. Despite the cationization, the in vitro downregulation efficiency of these gelatin nanoparticles is lower than that of Gelasomes (Table S2). The maximum downregulation of 70 % was reported by Mimi et al. and Rafael et al. on treatment with 30 μg and 10²–10³ μg of their cationized gelatin nanoparticles respectively [25,26]. In our study, Gelasomes were able to achieve >90 % downregulation efficiency at smaller doses and on larger cell populations (Table S2). Thus, Gelasomes were able to achieve high downregulation efficiencies without the need for toxic modifications such as cationization of particles. These findings indicate that using Gelasomes is a practical approach to siRNA delivery, and we believe it has the potential to revolutionize the field of drug delivery.

3.7. Robustness of gelasomes

One of the important characteristics of a delivery vehicle that determines its suitability for the translation stage is its ability to maintain stability over time. As we are still in the early stages of discovery, we investigated if Gelasomes can retain their integrity and RNA downregulation efficiency for at least two months. We have chosen to examine it over two months because preclinical animal studies are typically conducted for 30–60 days. While solution phase storage is an option, a more permanent and convenient option would be to lyophilize Gelasomes. If the lyophilized powder is stable, we can store it for an extended period without worrying about disintegration. Therefore, in this study, we lyophilized Gelasomes and investigated its downregulation efficiency in A549 cells after two months of storage. We observed that a 60-day-old lyophilized Gelasomes sample (after redispersion) maintained the functionality of siRNA as indicated by its 90 % downregulation efficiency (Fig. 4D). Lyophilization did not cause any change in the size and zeta potential of Gelasomes implying that the structure and morphology of the nanocarrier are well maintained (Fig. S18). As the sample is lyophilized, we anticipate that it can be stored for an extended period without any degradation or loss of knockdown efficacy.

3.8. In vivo repeat dose toxicity of gelasomes

To validate the positive in vitro results, we evaluated the in vivo toxicity of Gelasomes in CF-1 mice. In this study, we intravenously (IV) administered Gelasomes (3d) twice a week for three weeks, a standard protocol used for therapy studies (Fig. 5A). The choice for 3d is based on the objective to check if the mice could tolerate the Gelasomes with the highest amount of gelatin coat. We treated at three different concentrations (16, 32, and 48 mg/kg BW) to check if the Gelasomes were well tolerated by mice at higher doses. Throughout the 20-day study period, we monitored the mice for weight loss and any adverse reactions. After 20 days, we evaluated the treated mice's hematology and serum chemistry values and compared them to untreated mice.

Fig. 5.

In vivo evaluation of toxicity and downregulation efficiency of Gelasomes (3d). A) Design and timeline of multi-dose toxicity study of Gelasomes in CF-1 female mice. Mice were randomly divided into four groups: i) untreated (n = 6); ii) Gelasomes, 16 mg/kg (n = 7); Gelasomes, 32 mg/kg (n = 8); Gelasomes, 48 mg/kg (n = 8). Gelasomes were administered (IV) to treatment groups 2 days a week for 3 weeks. The body weight of all animals was measured on treatment days. On day 21, animals were sacrificed, and blood was collected for analysis. B) Change in body weight plot. Evaluation of C) hematological parameters, and d) serum chemistry parameters in blood samples collected from animals belonging to all four treatment groups. There is no significant change in the levels of any of the parameters in the treatment groups compared to the untreated controls indicating the non-toxic nature of Gelasomes. E) Study design to evaluate gene silencing efficacy of Gelasomes (3d) in athymic nude mice. Mice were randomly divided into two groups: i) untreated (n = 5); ii) Gelasomes (n = 5). A single dose of Gelasomes was intratumorally administered (siRNA injected −3.0 μg). After 48 h, mice were sacrificed, and tumor tissues were collected and analyzed. F) Evaluation of AXL levels (capillary Western Blot) in the tumors of treatment groups in comparison to the untreated group. Statistical values are reported as mean ± standard deviation (n = 5). Significant differences between the groups are indicated by *P < 0.05, and **P < 0.01.

All mice, irrespective of the treatment concentrations, gained weight throughout the study period (Fig. 5B). Interestingly, mice dosed with the highest concentrations of Gelasomes, 48 mg/kg, retained appetite and showed no weight loss. No adverse events were observed. As depicted in Fig. 5C and D, even at the highest treatment concentration of Gelasomes (48 mg/kg), there was no significant change in the hematology and serum biochemical values of mice in comparison to that of the control, indicating no signs of toxicity.

While cationic lipids are highly efficient in siRNA delivery, their inherent toxicity is the major barrier to clinical use [33,35,36,59]. Researchers have used several strategies to counteract the toxicity of cationic lipids. Abrams et al. and Tao et al. resorted to co-administration of dexamethasone, a glucocorticoid receptor agonist and Janus kinase 2 inhibitor to mitigate the elevated serum chemistry levels and inflammatory cytokines on treatment with siRNA-loaded lipid nanoparticles (LNPs) [60,61]. However, phase I clinical trials suggested that the levels remain elevated despite co-treatment with these drugs [62]. Sato et al. observed that the toxicity of LNPs is due to their accumulation in liver sinusoidal endothelial cells as suggested by elevated serum chemistry levels including alanine transaminase (ALT) and aspartate transaminase (AST) [63]. To offset these effects, LNPs were actively targeted to hepatocytes using N-acetyl-d-galactosamine as a targeting ligand [59,63]. By doing so, the use of the nanoparticle design is restricted to a particular cell type and cannot be easily extended to deliver siRNA to other organs. Notably, Gelasomes, even at significantly higher doses (48 mg/kg) did not cause any hepatic toxicity as suggested by unaltered levels of hematological parameters including ALT and AST. Thus, the presence of surface gelatin layers in Gelasomes efficiently shields the exposure of cationic lipidoid and renders biocompatibility to the construct.

Post the release of siRNA, lipidoids eventually degrade in the body through lipolysis by lipases present in the gastrointestinal tract, and various tissues [64]. The incorporation of biodegradable functionalities in the lipidoid structure further accelerates the rate of degradation and their elimination from the body. For example, Maier et al. showed that lipid-based nanoparticles were degraded in the cells after passing the endo-lysosomal compartments [65]. The degraded lipids were excreted within 24 h with around 50 % found in urine and feces. Thus, adding biodegradable functionality such as esters in the tails of G0C14 would further improve the biocompatibility of Gelasomes. These investigations are part of our future studies.

3.9. In vivo downregulation efficiency of gelasomes

Loco-regional delivery through intratumoral (IT) administration of drugs or delivery agents is a preferred strategy to circumvent the physiological barriers [66]. This approach is particularly beneficial for patients with localized spread of tumors to avoid toxic effects associated with systemic administration [67,68]. IT administration is specifically useful for liposomal formulations because they fail to efficiently overcome the RES and tumor barriers and homogenously accumulate in tumor tissues resulting in poor treatment efficacy and drug resistance [66]. Of note, IT administration was more favored for cationic liposome-mediated gene therapy to avoid non-specific toxicity and has resulted in successful therapeutic outcomes [69,70]. Even though the widespread practical application of intratumor administration is still unclear, certain cancers benefit from this approach. This study aims to examine whether Gelasomes are efficient in downregulating the target protein by IT administration. We used A549 mice xenografts and administered 3d (containing 3 μg of siAXL) directly into the tumor site. After 48 h of treatment, animals were sacrificed, and the tumor tissues were collected (Fig. 5E). Another set of mice was left untreated as controls and tumor tissues were collected for comparison. Western blot analysis of AXL protein showed that a single injection of 3d reduced the target protein by 44.5 ± 8 % (Fig. 5F). The results further confirm that Gelasomes are effective under in vivo conditions to down-regulate the target protein.

Overall, we have formulated a new class of RNA delivery vehicle called Gelasomes and demonstrated its excellent gene silencing efficacy and biocompatibility in vitro and in vivo. The Gelasomes in the study are synthesized using gelatin derived from bovine sources and may elicit an immune response in humans. The scope of the current study is to establish a safe and efficient siRNA delivery vehicle. Our future studies are aimed at investigating the translational potential of Gelasomes by studying the immunogenicity of the nanoparticles and optimizing the formulation using recombinant gelatin.

4. Conclusion

We have developed a novel nanomaterial called Gelasomes, which can deliver siRNA safely and effectively for RNAi therapy. We have optimized the reaction parameters and mass ratios of gelatin, lipidoid, and siRNA to produce Gelasomes consistently. Gelasomes have a biocompatible and negatively charged surface that can interact with cells without causing any harm. The gelatin matrix provides remarkable stability to siRNA against serum degradation and long-term storage. Gelasomes possess dual siRNA binding modes, i.e., covalent surface conjugation and electrostatic encapsulation, resulting in high siRNA loading and efficient gene silencing. We have improved Gelasomes by minimizing the gelatin layering around the lipidoid-siRNA complex, which increases siRNA loading and makes them more effective even at lower treatment doses. Even at high doses, Gelasomes do not have any toxic effects on mice when repeatedly administered. Furthermore, Gelasomes successfully downregulated the target gene (AXL) in a mouse xenograft lung cancer model. These nanomaterials have the potential to revolutionize the way RNA is delivered to cells, which could have significant implications for treating various diseases.

5. Ethics approval and consent to participate

This study involves animal studies, approved by MU's Animal Care and Use Committee (ACUC).

CRediT authorship contribution statement

Abilash Gangula: Writing – review & editing, Writing – original draft, Validation, Investigation, Conceptualization. Dhananjay Suresh: Writing – review & editing, Writing – original draft, Validation, Methodology, Investigation, Formal analysis. Agasthya Suresh Babu: Methodology, Investigation, Formal analysis. Zhaohui Li: Methodology, Investigation, Formal analysis. Anandhi Upendran: Writing – original draft, Methodology, Investigation, Formal analysis. Raghuraman Kannan: Writing – review & editing, Writing – original draft, Conceptualization.

Declaration of competing interest

The authors declare no conflict of interest.

Appendix A. Supplementary data

Additional information: Additional synthesis protocols, supporting Figs. S1–S13 and Tables S1–S2 are available online in “Supporting Information (SI)”.

The following is the Supplementary data to this article: