Recent advances in the field of nanotechnology offer an unprecedented opportunity to enhance the power of siRNA mediated gene therapy by providing both an efficient delivery system as well as targeted specificity. Small interfering RNA (siRNA)-mediated knockdown of gene specific messenger RNA (mRNA) levels is a great therapeutic strategy 12345678, in which, double-stranded RNAs are cleaved by the cellular nuclease Dicer into short 21–22mer fragments referred to as siRNA, which enter a ribonuclease protein complex termed the RNA-induced silencing complex. This complex mediates a specific degradation of the corresponding mRNA. Nanoparticles can form stable complexes with siRNAs (nanoplex: a complex between a nanoparticle and an siRNA) that can overcome all the impediments associated with siRNA in the free form. siRNA technology is a novel method to achieve complete and persistent knockdown of gene expression. Stable and sustained gene silencing is a key determinant as to whether or not RNAi-based therapeutics will have clinical relevance and gene delivery encompasses extracellular transport of the nanoplex to target cells, its intracellular RNA trafficking, and processing12345678.

Gold nanoparticles are particularly attractive for therapeutic applications due to their biocompatibility and ease of complex formation with biomolecules. Gold-nanorods (GNR) have far reaching potential for the study of intracellular processes at the single-molecule level, using high-resolution cellular imaging, long-term observation of cell trafficking in vivo and gene silencing 910111213. The hydrodynamic size of these GNR-nanoplexes under physiological condition is <100 nm, making them ideal as intracellular delivery agents. Gold nanoparticles have been used for more than a decade as a gene carrier for plasmid DNA and oligonucleotides, but recently our group was the first to report the use GNR for the delivery of siRNA against DARPP-32 in vitro 1415161718192021. DARPP-32 plays a significant role in the activation of the dopaminergic signaling pathway which is related to addictive behaviors. DARPP-32 was initially discovered as a major target for dopamine-activated adenylyl cyclase and protein kinase A (PKA) in the striatum and is recognized to be critical to the pathogenesis of drug addiction. Drugs of abuse act on the dopaminergic system in the brain and perturb DARPP-32 function. Successful suppression of DARPP-32 gene expression may efficiently facilitate addiction therapy by inhibiting the drug abusing behavior. To silence DARPP-32 gene expression in dopaminergic neuronal cells, we used cationically charged PEGylated gold nanorods electrostatically coupled with negatively charged siRNA to form stable nanoplexes. We evaluated these nanoparticles as nonviral gene carriers by investigating their specificity and efficiency in achieving DARPP-32 gene silencing in primary dopaminergic neuronal cells, which are typically difficult to transfect. Gold nanoparticles (GNP) are an attractive drug delivery vector due to the ease with which biomolecules such as siRNA can be attached to the gold surface using thiol chemistry. This process can also allow attachment of multiple targeting or functional groups to the nanoparticle surface to produce a multifunctional nanoparticle 22232425.

We and others have developed colloidal gold nanostructures as siRNA nanocarriers to improve cellular penetration, endosomal release, carrier unpacking and intracellular transport 141516171818202122. However, if we want to extend our in vitro studies to in vivo or pre-clinical evaluations, issues such as stability, specificity and long term safety are still major concerns. For example, the major problems of naked siRNA delivery in vivo are rapid enzymatic degradation and substantial liver and renal clearance 19. The intrinsic cytotoxicity and high reactivity of the GNR in the physiological milieu have limited their applications in vivo 23. Our group has demonstrated that PEGylated gold nanorod formulations can be a promising nanocarrier in vivo because of their low toxicity and ability for prolonged systemic circulation up to 40 hours which allows nanoparticles to reach their targeted site 2223. It appears, that the surface coating of nanoparticles is the key factor that determines it stability, cytotoxicity and efficacy and therefore their utility for in vivo studies 2425. Therefore, the aim of this study was to develop a stable, nontoxic and cationic GNR based nanoparticle gene delivery system using four different types of modifications of the GNR surface coatings, which include, amino terminated polyethylene glycol (GNR-PEG), poly (diallyldimethyl ammoniumchloride) (GNR-PDDAC), polyethyleneimine (GNR-PEI) and Chitosan (GNR-CIT). These four separate cationic GNR with the above four different surface coatings were used as vehicles for DARPP-32 siRNA delivery in dopaminergic neuronal (DAN) cells. We did a systematic characterization of all four GNR formulations, investigated their ability to electrostatically bind to the DARPP-32 siRNA to form a nanoplex and evaluated their stability, cytotoxicity and gene silencing efficiency in an in-vitro dopaminergic neuronal cell culture system.

PEI and PDDAC coating were performed using a layer-by-layer regime. A thin layer of anionic polymer poly(styrenesulfate) (PSS) was deposited on the CTAB layer to facilitate the further attachment of biocompatible cationic polymers (GNR-PSS). Many reports suggested that the synthetic parameters such as the concentration and the ionic strength of anionic polymer in layer-by-layer (LBL) preparations should be precisely controlled in order to obtain well-defined nanoparticles with high yield 282930313233. In our case, fast injection of excess PEI or PDDAC into concentrated GNR-PSS on vigorous stir plate were employed to give well-dispersed, stabilized and high positively charged nanoparticles.

Another formulation took advantage of the unique covalent bond of thiol-Au(0) by replacing CTAB with amine-poly(ethylene glycoside)-thiol. GNR-PEG was accomplished by ligand exchange process. PEGylated GNR was obtained after 24 hours of stirring the mixture in a glass container. GNR-CIT hybrid nanoparticles were prepared by crosslinking chitosan using glutaraldehyde. Low molecular weight chitosan in 1% acetic acid was used to form a neat protection layer to the GNR and thus reduce the toxicity originated from CTAB. The use of EDTA served as a buffer layer between two cationic moieties, CTAB surface and chitosan chain. The GNR-EDTA-chitosan mixture was stirred for 4 hours to homogeneously disperse the solution. All formulations were purified from excess chemicals by centrifugation and redispersion in HPLC-grade water. Figure 1 illustrates the different strategies of coating GNRs with cationic materials and are defined as GNR-PEI, GNR-PDDAC, GNR-PEG and GNR-CIT. Morphological images and uniformly polymer coating of all formulations were shown in Figure 2. The length and width of GNRs are estimated to be 50 nm and 15 nm, respectively. Note that the core-shell structure of GNR-CIT was clearly shown under high resolution TEM (Figure 2c).

After the surface modifications, red shifts in the longitudinal surface plasmon resonant peak, which can be accounted for the localize refractive index changes on the gold nanorod surfaces, were observed (Figure 3a and 3b). Zeta-potential of all coating formulations was recorded to assure that GNRs would electrostatically complex with negatively charged siRNA. It is also a good indicator to confirm the successful ligand exchange or polymer deposition. In Figure 3c, the highly positive charge of the bilayer CTAB was clearly reversed to negative after PSS deposition. The reversals of Zeta-potential were observed again after the growth of PDDAC or PEI layer. Ligand exchange from CTAB to NH₂-PEG-SH and the formation of core-shell structure of GNR-CIT were also indicated from the reduction in Zeta-potential.

The Dynamic light scattering (DLS) technique is used as a valuable parameter to confer stability. The unstable and abrupt change of the hydrodynamic diameter of nanoparticles in biological medium greatly affects their function and behavior in vitro and in vivo. We studied the stability of GNRs with different surface modifications for 100 hours in different biologically relevant media such as water, phosphate buffered saline (PBS) and Dulbecco's Modified Eagle Medium (DMEM). As shown in Figure 3(d-f), all formulations showed stable dispersions in water. The nanoparticles are normally stabilized in a medium by their surface charges. When the nanoparticles are subjected to the high ionic strength biological fluids, self-assembled aggregation may occur due to the neutralization of surface charge by the ionic species. Short term stability of GNR-PDDAC, GNR-CIT and GNR-PEI were observed in PBS (Figure 3e). Their hydrodynamic diameters gradually increased after 24 hours. The presence of high content of amino acids and glucose in DMEM decreased the stability of the nanoparticles. Aggregation of GNR-CIT and GNR-PEI were found after 1 hour incubation in DMEM. GNR-PDDAC also aggregated after 24 hours of incubation in DMEM as shown in Figure 2f. It appears that the stability of the nanoparticle is highly correlated with its surface charge. GNR-PDDAC, with the highest Zeta–potential, exhibits more stability in media which can be attributed to a longer neutralization time by the ion containing liquid. Our results confirm reported literature that demonstrates a drop of Zeta–potential is observed in high ionic strength medium 32.

Our study revealed that GNR-PEG is extremely stable although it carries the least positive charge. As shown in Figure 3(d-f), the hydrodynamic diameter of GNR-PEG was maintained in all medium for 100 hours. The extra stability was conferred by the surface passivation with the PEG layer which has been known to prevent protein opsonization. It is worth mentioning that the aggregation and serum absorption of nanoparticles can disrupt its function and cause adverse effects such as degradation of passivation surfaces, the exposure of the unprotected core to the biological samples and cytotoxicity.

Cytotoxicity of the various GNR formulation were tested using the MTS assay. Various concentrations of the different GNR formulations were incubated with DAN cells in DMEM with 10% fetal bovine serum (FBS). No cytotoxicity to DAN cells was observed after 24 or 72 hours of incubation even at the highest concentration of 80 µg/ml (Figure 4a and Figure 4b), indicating that aggregation of nanoparticles has no adverse effect DAN cells.

In vitro gene silencing studies were performed in DAN cells. DARPP-32 gene expression, was suppressed using the different GNR formulations complexed to DARPP-32 siRNA. All GNR formulations were incubated with the same concentration of DARPP-32 siRNA to form nanoplexes (GNR-PDDAC-siRNA, GNR-PEI-siRNA, GNR-PEG-siRNA, GNR-CIT-siRNA, Lipofectamine-siRNA) by electrostatic binding. The commercial available agent, Lipofectamine, was used as the positive control. DAN cells were treated with the nanoplexes and incubated for 48 hours. The efficiency of gene knockdowns were evaluated by using quantitative real-time (Q)-PCR technique 48 hours post-transfection. Figure 5 demonstrates that DAN cells treated with the various GNR siRNA nanoplexes, showed significant decrease in a DARPP32 gene expression. The percentage suppression in DARPP-32 gene expression was 38%, 40 %, 36%, and 39% for GNR-PDDAC-siRNA (p<0.01), GNR-PEI-siRNA (p<0.01), GNR-PEG-siRNA (p<0.01), and GNR-CTI-siRNA (p<0.01) respectively as compared to their respective controls that were not complexed to DARPP-32 siRNA. All the GNR nanoformulations showed gene silencing efficiencies comparable to the commercially available Lipofectamine reagent (42% gene suppression).

Due to the strong plasmonic scattering of GNR, dark field microscopy was used to study the cellular uptake of the GNR-siRNA nanoplexes. Figure 6 shows representative dark field images of DAN cells without and with the treatment of GNR-PEG-siRNA. Intense light scattering from GNRs is clearly seen in Figure 6b when comparing with the untreated DAN cells image as shown in Figure 6a. This strong light scattering is associated with cellular uptake of GNR-PEG-siRNA nanoplexes. The cellular uptake of GNR-PEG-siRNA was also confirmed using confocal microscopy. 6-carboxyfluorescein, fluorophore (FAM)-labeled siRNA (siRNA^FAM) was used to provide the fluorescence signal. As shown in Figure 6d, DAN cells treated with GNR-PEG-siRNA^FAM nanoplexes showed stronger fluorescence signal than DAN cells treated with Lipo-siRNA^FAM (Figure 6c).

Our data suggest that the gene silencing efficiency is similar for all GNR nanoformulations we tested and is comparable to the commercially available transfection agent lipofectamine. The GNR nanoformulations however, have an increased shelf life and stability as compared to lipofectamine. The accumulation of GNR inside cells has both positive and negative outcomes for biomedical applications. The positive outcome is that they can serve as multifunctional imaging and therapeutic agents for specific targeted cells. On the other hand, the negative outcome is the non-specific uptake of GNR by cells which cause an accumulation of GNR within the body and the physiological implications of which are unknown. The surface coating of GNR by PEG chains and other formulations as described in this report, is one an effective way to ensure their clearance from systemic circulation.

In summary, we have reported different coatings of GNRs for gene therapy. All coatings were confirmed by characterization techniques. Our prepared gene delivery nanoplatforms manifest a high biocompatibility and all of them are stable in biological medium. The suppression of DARPP-32 gene with all the four GNR –siRNA nanoformulations (GNR-PDDAC-siRNA, GNR-PEI-siRNA, GNR-PEG-siRNA, GNR-CIT-siRNA,) have comparable transfection efficiency with the commercially available transfection agent, Lipofectamine. In addition to efficient gene targeting, our nanoplatforms are capable of utilizing multimodal imaging. Further development and optimization of this technology could help stabilize other plasmonic nanostructures and could enable sensitive detection and specific gene targeting both in vitro and in vivo.