Drug delivery of siRNA therapeutics: potentials and limits of nanosystems

Comparison of siRNA with other RNAi therapeutic classes 

Two of the major advantages of siRNA over small molecule drugs are that sequences can be rapidly designed for highly specific inhibition of the target of interest. Also, the synthesis of siRNAs does not require a cellular expression system, complex protein purification, or re-folding schemes, and is relatively uncomplicated.9

There are four major types of anti-mRNA strategies: first there are single-stranded antisense oligonucleotides (ODNs), wherein a synthetic, small, single-stranded ODN inhibits the translation of a specific gene by hybridizing to the corresponding mRNA through Watson-Crick binding.10 The second anti-mRNA strategy involves ribozymes, which are catalytically active RNAs that cleave single-stranded regions in RNA through trans-esterification or hydrolysis reactions,11, 12 and third there are siRNAs. Bertrand et al compared the knockdown effects of antisense ODNs and siRNA in cell culture and in vivo, and concluded that siRNA is more efficient. Compared with antisense ODNs, siRNAs inhibit the synthesis of target proteins with high specificity and at a much lower dosage. In one head-to-head comparison, siRNA knocked down gene expression about 100- to 1000-fold more efficiently than antisense ODNs.13 Vickers et al performed another comparative study in human cells. The potency, maximal effectiveness, duration of action, and sequence specificity of optimized RNase H–dependent ODNs and siRNA ODN duplexes were evaluated and found to be comparable. Effects examination of 80 siRNA ODN duplexes designed to bind to RNA from four distinct human genes revealed that, in general, activity correlated with the activity to RNase H–dependent ODNs designed to the same site, although some exceptions were noted. The one major difference between the two strategies is that RNase H–dependent ODNs were determined to be active when directed against targets in the pre-mRNA, whereas siRNAs were not.14 Some differences and similarities of ODNs, siRNA, and ribozymes are shown in Table 1.

Table 1. Differences and similarities between antisense oligonucleotides, siRNA, and ribozymes

		Oligonucleotides (ODNs)	siRNA	Ribozymes
	Formation	Short (∼18–25 bp) nucleic acids; single-stranded	Short (∼19–25 bp) nucleic acids; double-stranded	Three-dimensional catalytically active RNA, classically composed of three helices
	mRNA cleavage by	Hybridization to corresponding mRNA	Unwinding the duplex, enabling the recognition of mRNAs, incorporated into RNA-induced silencing complex	Catalyze either their own cleavage or the cleavage of other RNAs by trans-esterification or hydrolytic reactions
	Stability in serum	Less resistant than siRNAs13	Several minutes to about an hour15, 16, 17, 18	10 seconds to a few minutes19, 20
	Delivery to mammalian cells	Due to the negative charge, ODNs pass through cell membrane to some degree	Due to high negative charge, siRNA cannot pass freely through cell membrane	Due to negative charge, ribozymes pass through cell membrane to some degree

Finally, there are so called microRNAs —endogenous, short, double-stranded, and noncoding RNA molecules, that have been identified in a variety of organisms and certain viruses. This group of “new” molecules is transcribed mainly from the introns and/or exons or intergenic regions and plays important regulatory roles in development and gene expression. Mature microRNAs are typically 20–24 nucleotides in length and regulate target mRNAs post-transcriptionally by interactions with partially mismatched sequences in the 3′-untranslated regions of these messengers. These interactions result in the suppression of translation or degradation of target mRNAs.21

Advantages and drawbacks 

RNA interference has already been observed in most organisms, from plants to vertebrates. Furthermore, it has had an immense impact on biomedical research and will lead to novel medical applications in the future. RNAi could provide an exciting new therapeutic modality for treating infections, cancer, neurodegenerative diseases, antiviral diseases (e.g., viral hepatitis and human immunodeficiency virus 1, HIV-1), Huntington's disease22, 23 hematological diseases,24 pain research and therapy25, 26 dominantly inherited genetic disorders,27 and many other illnesses. However, until now there have been a few hurdles to overcome. Major ones include the poor pharmacokinetic property of siRNA and major biological restrictions, such as off-target effects and interferon response.28 In particular, siRNAs longer than 30 nucleotides, in specialized highly sensitive cell lines and at high concentrations, lead to the activation of the immune system.29, 30, 31 Also, a low transfection efficiency, poor tissue penetration, and nonspecific immune stimulation by in vivo–administered siRNAs have delayed their therapeutic application. Still, the lack of an efficient delivery system to target and deliver the siRNA to the desired cells is of particular limitation for the full therapeutic potential of this approach. Many groups are searching for an optimal delivery tool that can be systemically administered, safely and repeatedly, and will deliver the siRNA specifically and efficiently to the target tissue.

Research and development 

So far, the vast majority of in vivo studies using RNAi technology have used high doses of nonmodified naked siRNAs, usually delivered through a hydrodynamic injection.32, 33, 34, 35, 36, 37, 38, 39 Direct delivery of naked siRNA results in very small gene-inhibiting effects both in vitro and in vivo, as a result of their poor intracellular uptake and the rapid enzymatic degradation in serum conditions (Figure 1). The half-life reported for unmodified siRNAs in serum ranges from several minutes to about an hour.15, 16, 17, 18 Chemically modified siRNAs, such as sugar modifications (e.g, 2′-O-methyl and 2′-deoxy-2′-fluoro (OMe/F)40) or backbone modifications (e.g., phosphorothioate linkages41), are often more stable against nucleases, enhance nuclease stability, and prolong siRNA half-life in serum, while permitting its function. Backbone modifications and bioconjugation with lipids (e.g., fusogenic liposomes42) and peptides (e.g., cell-penetrating peptides (CPPs)43), in particular, are known to improve the stability and cellular uptake of siRNAs (Figure 1). CPPs—short peptides of less than 30 amino acids—are able to penetrate cell membranes and translocate different cargoes into the cells. Current CPP-mediated siRNA delivery approaches can be divided in two main categories: noncovalent CPP-siRNA in complex by virtue of electrostatic interactions, and CPPs covalently attached to siRNA duplexes through disulfide bond formations. Noncovalent electrostatic complex formation is a technically simple approach utilizing the cationic charge of CPPs to condense siRNAs into aggregates or nanoparticles with an overall positive charge. This process, of inducing complex formation between CPPs and nucleic acids, has been previously used in gene therapy applications.44 Covalent linkage of CPPs to siRNAs to create small, monomeric CPP-siRNA molecules remains the “holy grail” of CPP-mediated siRNA delivery, bypassing multiple in vivo complications shown for cationic polymer condensation and liposome encapsulation methodologies used for nucleic acid delivery.44 The mechanism behind the cellular translocation of CPP-siRNA molecules is not exactly known, but it seems to be receptor- and energy-independent, although in certain cases translocation can be partially mediated by endocytosis.

View Large Image Download to PowerPoint

Figure 1. Schematic illustration of different uptake pathways of siRNA into the cell. (A) Spontaneous siRNA uptake. (B) Cell-penetrating peptides (CPPs). (C) Electroporation. (D) Nanoparticles, nanocapsules, liposomes. (E) Fusogenic liposomes. (F) Microinjection.

One main problem in RNAi technology is to internalize siRNA specifically into the cell. To achieve cell-specific delivery of siRNA, several targeting ligands, such as cell membrane receptors and antibodies, have been successfully explored that ensure targeted delivery of siRNA.45, 46 Santel et al investigated the biodistribution characteristics of naked siRNAs using Cy3 dye and reported that naked siRNAs are found primarily in the kidney 20 minutes after injection, with no detectable signals in other organs (pancreas, heart, lung, liver, and spleen).47 In addition, naked siRNAs-Cy3 accumulate in the pole and lumen of the proximal tubules and in the urine 5 minutes after intravenous injection. These observations suggest that naked siRNAs are not targeted toward any cell type of the tissues investigated (kidney, pancreas, heart, lung, liver, and spleen) in vivo, and this is most likely due to instant renal excretion.47

Most carrier systems are taken up by endocytosis and will probably reach the lysosomes, where siRNA will be degraded by lysosomal enzymes (Figure 1). Nonetheless, there are some possible methods of avoiding lysosomal degradation. In vitro, chemically synthesized siRNA is most effectively introduced into cells using electroporation,48, 49, 50, 51, 52 nuclear or cytoplasmic microinjection, or commercially available lipid reagents such as Oligofectamine1, 53 and Lipofectamine54 (Invitrogen, San Diego, California); Metafectene (Cambio Ltd., Cambridge, United Kingdom)52, 55 and siPORT Amine (Applied Biosystems, Foster City, California)56, 57 (Figure 1). Because of the invasiveness of the electricity and cytotoxicity of some lipid reagents, these transfection methods are not commonly used for in vivo delivery. Still, some studies have been published on electrically mediated nucleic acid delivery in vivo, reporting the successful use of DNA and siRNA for gene silencing in rodent model systems.58, 59, 60

Also, high-efficiency transfection of siRNA shows major difficulties. To achieve maximum effectiveness of exogenously introduced siRNA, transfection optimization experiments are required. Parameters, which are a precondition for the efficiency, include culture conditions in vitro, choice and amount of transfection agent, exposure time of transfection agent to cells, and siRNA quantity and quality. Transfection efficiency can be increased significantly by the association of targeting ligands (active targeting), such as antibodies,61, 62, 63 peptides,63, 64 or proteins65, 66, 67 into complexes, which exploit the diversity of receptors present at the surface of each cell type. Therefore, drug targeting will be a promising strategy to achieve systemic delivery of nucleic acids in the future. However, the transfection efficiency does not necessarily correlate with cellular uptake, because high transfection efficiency is influenced by ferrying the siRNA into the nucleus and unpacking the RNA from its vector.66

Another major issue in the application of siRNA is that nucleic acid drugs are highly charged and do not cross cell membranes by free diffusion. Therefore, the delivery of RNAi therapeutics must use a targeting technology that permits the RNAi therapeutic to cross biological membrane barriers. For the above-mentioned reasons, there is a need for suitable carriers that will provide a stable complex and protection to ensure targeting and delivery of siRNA as well as be able to cross the gap between cell culture and animal models to allow for an efficient siRNA delivery in vivo. This review summarizes the established nanoscaled delivery systems and the recent progress made in the drug delivery of siRNA.

Potential therapeutic applications of siRNA 

The field encompassing the therapeutic applications of siRNA is versatile and includes siRNAs in central nervous system therapeutics, as antiviral and anticancer agents, in inflammation or cardiovascular therapeutics, and pain research as well as many other therapeutic applications of siRNA. Figure 2 presents some indications addressed by gene therapy in current clinical trials.

View Large Image Download to PowerPoint

Figure 2. Indications addressed by gene therapy clinical trials (modified from Ref. [68]).

Potential nanocarriers for siRNA application 

Carrier-mediated delivery has several advantages over the delivery of individual nucleic acid molecules. In essence, mediators of RNAi can be differentiated into viral and nonviral vectors (a summary of common delivery vehicles for siRNA is schematically depicted in Table 2), and both systems have their pros and cons.

Table 2. Some examples of delivery vehicles for siRNA applications

	siRNA Delivery system	Diameter (nm)	Type	Targeted mRNA/tissues/diseases	Refs.
	Cationic lipids
	DOTAP (1,2 Dioleoyl-3-trimethylammonium-propane)	Not shown	Nonviral cationic liposomes	Various cell types (mouse in vivo)	[69]
	DOTAP/DOPE (1,2-dioleoyl-sn-glycero-3-phosphatidylethanolamine)	Not shown	Nonviral cationic liposomes	Targeting the luciferase mRNA in mouse L-cells	[70]
	DOTAP/DOPC (1,2-dioleoyl-sn-glycero-3-phosphatidylcholine)	Not shown	Nonviral cationic liposomes	Targeting the luciferase mRNA in mouse L-cells	[70]
	MVL5 (multivalent lipids)/DOPC	Not shown	Nonviral cationic liposomes	Targeting the luciferase mRNA in mouse L-cells	[70]
	DOPE/CDAN (N-cholesteryloxycarbonyl-3,7-diazanonane-1,9-diamine)	300–500	Nonviral liposomes	HeLa and IGROV-1 cells	[71]
	Polymers
	P(MDS-co-CES) poly {(N-methyldietheneamineseacate)-co-[(cholesteryloxocarbonylamidoethyl) methylbis(ethylene)ammoniumbromide] seacate}	175	Nonviral cationic core-shell nanoparticles	MDA-MD-231 human breast cancer cells	[72]
	Poly(ethylene glycol)-block-poly(aspartic acid) (PEG-PAA) with calcium phosphate	100–300	Nonviral nanoparticles	HeLa cells	[73]
	MPEG/PCL (Methoxy(polyethyleneglycol)/poly(ε-caprolactone))	150	Nonviral di-block co-polymeric nanoparticles	Not shown	[97]
	Lactosylated PEG	Not shown	Nonviral PIC micelles	Not shown	[97]
	PEG-poly(methacrylicacid)	Not shown	Nonviral PIC micelles	Not shown	[97]
	PEG – DPT (PEG-poly (3-[(3-aminopropyl)amino]propylaspartamide)	Not shown	Nonviral PIC micelles	Not shown	[97]
	Chitosan	200–500	Nonviral cationic polysaccharide nanoparticles	H1299 human lung carcinoma cells CHO K1 and HEK 293 cell lines	74, 75
	Polyisobutylcyanoacrylate	325	Nonviral noncationic aqueous-core nanocapsules	Ewing sarcoma (metastatic bone cancer)	[76]
	Hyaluronic acid	200–500	Nonviral nanogel	HCT116 human colon carcinoma cell lines	[77]
	Polyethylenglycol–polypropylenesulfide-peptide (PEG–PPS-peptide)	171–601	Nonviral ABC tri-block co-polymer	HeLa cells	[78]
	Atelocollagen	100–300	Nonviral nanoparticles	Human nonseminomatous germ cell tumor	79, 80, 81
	Polyethyleneimine (PEI)	Not shown	Nonviral cationic polymer nanoparticles	U87 glioblastoma cells	[82]
	Peptide/Protein-based
	Acetyl-GALFLGFLGAAGSTMGAWSQPKKKRKV cysteamide	200	Nonviral noncovalent nanoparticles	Various cell types	[83]
	Ternary proticles (HSA-Protamine-ODN)	202	Nonviral self-assembled nanoparticles	Murine fibroblast	[84]
	Cell-Penetrating Peptides
	Penetratin	Not shown	Nonviral	Mouse lung in vivo	[43]
	Tat-(48-60)
	Viral Vectors
	Adenovirus	60–90	Viral	HeLa, U251, MCF-7 cells	[85]
	Adeno-associated virus	Not shown	Viral	HEK 293 cells	[75]
	Retrovirus	Not shown	Viral	HEL cells	[86]
	Lentivirus	Not shown	Viral	HCC cells	[53]

The application of nonviral siRNA delivery systems is limited by packaging efficiency, colloidal stability, target internalization, and endosomal escape; moreover, it lacks the efficiency of gene transfer compared with viral vectors.

Viruses have developed a range of mechanisms to permit their genomes to penetrate into the target cell's cytoplasm. These mechanisms can potentially be very helpful in solving the problem of siRNA delivery. Further advantages of viral carriers are high transduction efficiency for different quiescent and dividing cell types, and high levels of short-term expression to provide therapeutic benefits. But there is the possibility of immune and toxic reactions in addition to the potential for viral recombination.

In general, the delivery vehicles can affect the resulting biodistribution through passive and/or active targeting. Passive targeting occurs as a result of the intrinsic physicochemical properties of the carrier. Important aspects that control the biodistribution and transfection are the charge and size of the delivery vehicle.87

Physicochemical requirements 

The charge of the delivery vehicle significantly impacts its interaction with components of the bloodstream; highly charged particles can lead to complement activation, whereas near-neutral particles show reduced phagocytic uptake. Specifically, cationic polymers such as polylysine and polyethyleneimine (PEI) have been shown to activate the complement system, whereas increased polycation length and surface charge density leads to higher complement activation and/or cytotoxic effects due to electrostatic interactions with negatively charged cell membranes.88 Rapid binding of charged molecules by complement proteins or other opsonins can lead to immune stimulation and rapid clearance of the delivery vehicles from the bloodstream. The size of the delivery vehicle is also of concern for systemic delivery. Similar to plasmid DNA and ODNs, siRNA are probably taken up by cells through endocytosis, and if siRNA is not effectively delivered to the cytoplasmic compartment, it will not induce RNAi.

Many studies on targeted drug delivery have demonstrated the importance of particle size for cellular uptake. In general, particles with an average diameter between 200 and 500 nm could be introduced into target cells via endocytosis and subsequently permeate the nuclear membranes through the nuclear pores. Therefore, nanoscaled systems in the range from 100 to approximately 500 nm are required. The appropriate particle size depends on the field of application; to target tumors, for example, the carrier should be as small as possible displaying a diameter lower than 100 nm. Fang et al have investigated the influence of methoxypolyethyleneglycol (MePEG) on the molecular weight and particle size of stealth nanoparticles and on their in vivo tumor targeting properties.89 Three sizes (80, 170, and 240 nm) of poly(methoxypolyethyleneglycol cyanoacrylate-co-n-hexadecyl cyanoacrylate) (PEG-PHDCA) nanoparticles loading recombinant human tumor necrosis factor- (rHuTNF-α) were prepared at different MePEG molecular weights (MW = 2000, 5000, and 10,000) using a double emulsion method. The results of the study showed PEG-PHDCA nanoparticles with higher MePEG molecular weight and smaller particle size, which could achieve higher in vivo tumor targeting efficiency.89

Different pathways for drug release and cellular uptake of nanoparticles are schematically described in Figure 3.

View Large Image Download to PowerPoint

Figure 3. Schematic illustration of different uptake mechanisms for drug-loaded nanoparticles into the cell. (A) Extracellular drug release. (B) Drug release with targeting ligands. (C) Exposure of a positive surface charge. (D) Sterically stabilized nanoparticle drug release.

Depending on the formation construction and basic material there are different possibilities for packaging drugs into a nanoparticles matrix (Figure 4). Rejman et al have investigated the effect of particle size on the pathway of entry and intracellular fate in nonphagocytic B16 cells.90 Latex particles of up to 200 nm were internalized exclusively by clathrin-mediated endocytosis, whereas larger particles entered the cells via a caveolae-dependent pathway. Mouse melanoma B16 cells took up particles of up to 500 nm in size, but there was no uptake of particles 1 μm in size. Relative to internalized 50-nm particles, the uptake of the 100-nm beads was diminished three- to four-fold, whereas the internalization of 200- and 500-nm beads was reduced 8–10 times. Furthermore, internalization of smaller particles (50, 100, 200 nm) started immediately at 37°C, and 50% of the cell-associated beads were internalized and distributed throughout the cell, whereas 500-nm beads were internalized over a period of 30 minutes, and significant accumulation within the cells could only be detected after 2–3 hours.90

View Large Image Download to PowerPoint

Figure 4. Various types of drug-loaded nanoparticles. (A) Solid colloidal nanoparticle (NP), drug embedded. (B) Solid NP, drug adhered. (C) Nanocapsules. (D) Solid colloidal NP, drug embedded with a cell-surface ligand.

Cellular uptake routes 

There are a number of different endocytic pathways to internalize substances by eukaryotic cells. These pathways include phagocytosis, macropinocytosis, clathrin-mediated endocytosis, non-clathrin-mediated endocytosis, and caveolin-mediated uptake. Cationic-lipid-DNA complexes, for example, are internalized via clathrin-mediated endocytosis.91 Each of those pathways delivers siRNA to specific cellular compartments. For an efficient RNAi activity, nucleic acid tools have to reach their cellular targets after gaining entry into the cell. This ability is based on their internalization pathway and it is highly dependent on the delivery system used.

Nonviral in vitro and in vivo delivery of siRNA for several disorders 

Nonviral vectors seem to be promising tools for gene delivery, because they are relatively safe and can be modified through the incorporation of ligands for targeting specific cell types. However, the levels of gene expression mediated by these vectors are low as compared with viral vectors.

Nanosized particles such as liposomes, polymeric micelles, lipoplexes, and polyplexes are often studied as drug carrier systems for nucleic acid delivery. Commonly used delivery vehicles are summarized in Table 2.

There are three major vectors that are convenient for nonviral siRNA delivery: polymer-based, lipid-based, and peptide- or protein-based gene delivery systems.

Viral in vitro and in vivo delivery of siRNA for several disorders 

For many disease models, the most desirable cell types to use, such as the immune system or primary cells, cannot be efficiently transfected by lipid delivery reagents. Thus, viral delivery of RNAi cassettes-containing vectors is a powerful alternative to lipid transfection. Viral delivery is applicable for any mammalian cell type, including hard-to-transfect, primary, and even nondividing cell types. Some advantages of viral vectors are their ability to infect both dividing and nondividing cells (e.g., adenovirus), the stability of recombinant vectors, the large insert capacity, and the potential to be produced at high titers.85 Viruses are naturally evolved nanoscaled vehicles that efficiently transfer their genes into host cells. This characteristic makes them desirable for engineering virus vector systems for the delivery of therapeutic genes. The viral vectors that have recently been used in the laboratory and clinical practice are based on RNA and DNA viruses processing very different genomic structures and a range of hosts. Particular viruses have been selected as gene delivery vehicles because of their capacities to carry foreign genes and their ability to efficiently deliver these genes, thus correlating with efficient gene expression. These are the major reasons that viral vectors derived from retroviruses, adenovirus, adeno-associated virus, herpesvirus, and poxvirus are used in more than 60% of clinical gene therapy trials worldwide.105 Vectors that are currently used in gene therapy clinical trials are listed in Figure 5.

View Large Image Download to PowerPoint

Figure 5. Vectors that are currently used in gene therapy clinical trials (modified from Ref. [68]).

Adenoviruses are a family of DNA viruses that can infect both dividing and nondividing cells; they are nonenveloped viruses of 60–90 nm in diameter, containing a linear double-stranded DNA genome. There are more than 40 serotypes for the adenovirus, but the most commonly used recombinant vectors are generated from serotypes 2 or 5 of subgroup C. The life cycle does not normally involve integration into the host genome; instead, they replicate as episomal elements in the nucleus of the host cell, and consequently there is no risk of insertional mutagenesis. The initial interaction of adenovirus with cells occurs through binding of the distal knob domain of the fiber to a host cell's surface molecule, coxsackie and adenovirus receptor, which mediates internalization of adenovirus.85 The wild-type adenovirus genome is approximately 35 kilobases (kb) long, of which up to 30 kb can be replaced with foreign DNA.106 There are four early transcriptional units (E1, E2, E3, E4), which have regulatory functions, and a late transcript, which codes for structural proteins.

Replication-deficient adenovirus vectors can be generated, for example, by replacing the E1 gene, a gene essential for viral replication, with the gene of interest and an enhancer promoter element, which are than replicated in cells that express the product of the E1 gene in very high concentrations (>1011 adenovirus particles per milliliter).107 As a result of the lack of essential genes for replication, cells infected with recombinant adenovirus can express the therapeutic gene without replicating the vector itself.108

Uchida et al have developed an adenoviral vector harboring a tandem-type siRNA expression unit, in which sense and antisense strands composing the siRNA duplex were separately transcribed by two human U6 promoters. Targeting survivin, an antiapoptotic molecule widely overexpressed in malignancies but not detected in terminally differentiated adult tissues, this type of adenoviral vector (Adv-siSurv) successfully exerted a gene knockdown effect and induced apoptosis in HeLa, U251, and MCF-7 cells. These cancer cells, once infected with Adv-siSurv, displayed remarkably attenuated growth potential, both in vitro and in vivo. Moreover, intratumoral injection of Adv-siSurv significantly suppressed tumor growth in a xenograft model using U251 glioma cells.109 Another study demonstrated an adenoviral vector capable of expressing siRNA molecules targeting p53 or VprBP/KIAA0800, a cellular protein that interacts with the HIV auxiliary protein viral protein r (Vpr). In both cases, specific reduction in the target protein level was observed after an adenoviral infection and correlated with a specific reduction in the mRNA level.110

Adeno-associated viruses are nonpathogenic, single-stranded DNA human parvoviruses, which can infect variable cell types, and are dependent on a helper virus, usually adenovirus or herpes simplex virus, for proliferation. Similar to the adenovirus, an adeno-associated virus is capable of infecting both dividing and nondividing cells, yet they can accommodate only 3.5–4.0 kb of foreign DNA, which is the reason for excluding larger genes.108 Gorbatyuk et al have demonstrated that the expression of rhodopsin can be downregulated in vivo by adeno-associated virus–delivered siRNA for the allele-independent treatment of autosomal dominant retinitis pigmentosa. Thereby, HEK 293 cells were co-transfected with a plasmid carrying mouse RHO complementary DNA driven by the cytomegalovirus promoter and a chemically synthesized siRNA duplex of 21 nucleotides. Reduction of RHO mRNA was confirmed by reverse transcriptase–polymerase chain reaction.75

Retroviruses are a class of enveloped viruses containing a single-stranded RNA molecule, which converts into double-stranded DNA in the infected cell. By integrating into the host's genome, it becomes expressed in the form of proteins. Some retroviruses contain proto-oncogenes, which can possibly mutate, thus leading to cancer. During the production of vectors, these components are removed. Retroviruses can also metamorphose cells by integrating near to a cellular proto-oncogene and driving inappropriate expression by disrupting a tumor suppresser gene. This event, termed insertional mutagenesis, though extremely rare, could still occur when retroviruses are used as vectors. The available carrying capacity for retroviral vectors is approximately 7.5 kb, which is still too small for some genes even if complementary DNA is used.108 A critical limitation of retroviral vectors is their inability to infect nondividing cells.111 Ling et al showed that embryonic development–associated gene (EDAG)/siRNA can silence the expression of EDAG in HEL cells. Downregulation of EDAG expression by retrovirus-mediated siRNA inhibited cell proliferation and tumor growth. Knocking down EDAG expression by siRNA is also associated with decreased expression of the antiangiogenic factor interleukin 8, suggesting that EDAG stimulates tumor growth, at least in part, by regulating angiogenesis.86

Lentiviruses also belong to the retrovirus family, but in contrast to retroviruses they are able to infect both proliferating and nonproliferating cells.112 The best-known lentiviral vectors are derived from HIV. These vectors can be produced at concentrations of >109 virus particles per millilitre.113 Data have shown that when lentiviral vectors are injected into rodent eyes the expression of the transgene persisted for at least 12 weeks with no apparent decrease.114 When injected into rodent brains the lentivirus vector system was able to efficiently and stably infect quiescent cells in the primary injection site with transgene expression for over 6 months.115 Chen et al have investigated a lentivirus-mediated RNAi to knock down EZH2 expression in human hepatoma cells so as to study the function of EZH2 in tumorigenesis and evaluate treatment efficacy. Lentivirus-mediated RNAi effectively reduced EZH2 expression. Suppression of EZH2 in hepatocellular carcinoma cells significantly reduced their growth rate in vitro and markedly diminished their tumorigenicity in vivo.116

What makes for an ideal nanovector? 

For therapeutic applications, siRNA technology represents a promising tool by offering new types of drugs that are relatively easy to design with very high target selectivity, inhibiting a specific gene expression. They are expected to have low toxicity because they are metabolized to natural nucleotide components by the cells' endogenous systems. Nevertheless, there are many obstacles for delivering siRNAs to the cells in vitro and in vivo, such as degradation by enzymes in the blood, interaction with blood components, and nonspecific uptake by the cells, which is associated with the biodistribution in the body. In addition, the immune responses against siRNAs must be taken into account when considering the application of siRNAs for in vivo therapy. To achieve the knockdown by siRNAs in vivo, many siRNAs delivery systems based on physical and pharmaceutical approaches have been proposed.

So what makes for an ideal nanovector? Important for biological function, siRNA requires protection from enzymatic degradation and cellular uptake without lysosomal compartmentalization; finally, it must encounter its target mRNA to ensure RNAi.117, 118 Furthermore, convenience and reproducibility of drug production, the ability to target the desired cell type, and a lack of immune response are desirable. The large majority of current nonviral methodologies have relied on nanoparticles or insoluble-complex formation to protect the siRNA from the RNase-rich in vivo environment as well as help siRNA cross cellular membranes. Unfortunately, nanoparticle delivery systems have been shown to have limitations in vivo due to insufficient biodistribution, low transfection efficiency, rapid plasma clearance, and cellular toxicity.44 Moreover multiple non-viral-based delivery methods have been used in vivo for delivering siRNA, including hydrodynamic injection, cationic liposome encapsulation, formation of cationic polymer complexes, and antibody-specific targeting delivery systems.

Most of the current gene therapy approaches make use of viral vectors. Due to the high transduction efficiency for different quiescent and dividing cell types, viral delivery systems require a powerful technique to deliver DNA to cells. High-titer concentrations (>108 viral particles per milliliter) allow many cells to be infected; however, problems such as the danger of viral toxicity and relatively strong host responses resulting from the activation of the human immune system are to be solved.

There is still no perfect nanoscaled delivery system, which achieves all of the requirements. Each of the current methods of gene delivery, whether viral or nonviral, have some limitations, and maybe there will never be a generalized delivery system for all applications; instead, the choice of vector would depend on its use. All of these desired properties exist in various delivery systems, so an ideal vector may have properties from both types of system—viral and nonviral.