Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

Inaccurate processes on cellular level are the origin of various severe diseases. Depending on the nature of the genetic disorder, two main therapeutic strategies can be pursued. One possibility to adjust the life machinery is the replacement of missing or defective genes by introduction of correspondingly encoding DNA into cells.

Alternatively, inaccurate or overexpressed genes with pathological consequences can be deactivated. This latter strategy has an interesting history. More than 30 years ago, the first steps were made towards knocking down genes by the antisense approach using single-stranded oligonucleotides (Zamecnik and Stephenson 1978), but only 13 years ago, the RNA interference pathway was discovered (Fire et al. 1998). Silencing of target genes may take place by small-interfering RNA (siRNA)-mediated mRNA degradation (Hannon 2002; Meister and Tuschl 2004; Zamore and Haley 2005).

Two main possibilities of utilizing the RNAi pathway for therapeutic purposes exist. One is based on the discovery that extracellularly applied artificial double-stranded siRNAs, 20–24 nucleotides long, as soon as taken up into the cell, can enter the RNAi pathway through cytosolic RISC incorporation (Elbashir et al. 2001). The other approach is based on plasmid-based intranuclear expression of short-hairpin RNA (shRNA) (Brummelkamp et al. 2002), which is processed by the cell after entering the cytosol into the corresponding siRNA. This method leads to stable gene silencing, while artificial siRNA attains only transient knockdown, as it is diluted and lost during cell division.

Another naturally occurring class of double-stranded oligonucleotides, namely, microRNAs (miRNAs), utilize RNA interference for regulation of many targets at the same time (Bartel 2004). Deregulation of microRNA expression in tumors (overexpression of oncomiRs, loss of expression of suppressor miRs) suggests their role as either therapeutic targets or therapeutic agents (Lu et al. 2005; Gregory and Shiekhattar 2005).

However, before acting on intracellular level, oligonucleotides have to reach the target tissue and enter the cells, which is a challenging task. Systemic application involves several risks for “naked” oligonucleotides, such as enzymatic degradation during blood circulation and rapid renal clearance. Another challenge is reaching the target tissue and passing the barrier cell membrane before reaching the site of action. So as to overcome these challenges, accurate protection of the nucleotide cargo by chemical modification and/or packaging into viral or nonviral nanocarriers is mandatory and has been widely studied during the last decades.

This review presents the latest developments of oligonucleotide encapsulation and packaging methods and their potential for therapeutic use.

2 Nucleic Acid Encapsulation and Packaging Methods

Reaching the target tissue and entering cells without loss of activity is, as already mentioned, an insuperable challenge for systemically applied naked oligonucleotides. The size of such particles is a very important issue. Large particles may get stuck in tight blood vessels and lead to tissue hypoxia or even death. Capture and inactivation by the reticuloendothelial system is another undesired outcome. If nanoparticles are too small, like naked oligonucleotides, they may be subjected to rapid renal clearance without having the possibility of systemic circulation sufficient to reach the target tissue.

As the natural ideal nucleic acid carriers, namely, viruses, have a size between 20 and 300 nm, the same range is valid for particles formed with nonviral vectors and the nucleic acid. Carriers for plasmid DNA-based gene delivery have been evaluated for decades now, but unfortunately the requirements for the delivery of oligonucleotides or siRNA are not exactly the same (Scholz and Wagner 2012; Kwok and Hart 2011). Still, since the discovery of RNA interference in C. elegans (Fire et al. 1998) and in mammalian cells (Elbashir et al. 2001), research in this field has tremendously advanced towards therapeutic application in humans.

Already in the 1990s, antisense oligonucleotides were developed for clinical application, and siRNA for local treatment of CMV retinitis was successfully tested in clinical trials (Azad et al. 1993; Crooke 1998; Grillone and Lanz 2001). However, short in vivo half-life and transient transfection hamper the use of naked oligonucleotides. Therefore, packaging methods and delivery strategies have been developed in order to overcome these drawbacks.

3 Viral Vectors for Oligonucleotide Delivery

Viruses are natural carrier systems for the delivery of any kind of nucleic acid optimized in host interactions over very long time. They are programmed by nature to reach the target tissue, enter cells, and set free the nucleic acid for their own replication. Therefore, it seems rightly reasonable that these highly specific natural shuttles, due to their ability to adapt to changing environmental conditions, found application as nucleic acid vectors with very promising therapeutic potential. Different viral vector systems have been evaluated during the past years for oligonucleotide delivery. Replication-defective adeno-associated viruses (AAV) packaging their genome into pure protein capsids, and also lentiviruses (LV), a subclass of retroviruses, containing the nucleic acid/protein capsid within a lipid envelope, showed high efficiency combined with the best safety profiles (Couto and High 2010; Sliva and Schnierle 2010). These vectors are able of transducing both dividing and nondividing cells, thus allowing high gene transfer rates.

For persistent intracellular siRNA activity, a DNA-based method for sustained intracellular siRNA expression was developed (Brummelkamp et al. 2002; Paddison et al. 2002). Plasmid DNA cassettes encoding for the corresponding shRNA are transcribed in the nucleus and further processed in the cytosol, leading to sustained gene knockdown. Thus, in a further step, such shRNA-encoding cassettes were encapsulated in viral vectors for targeted delivery. Beverly Davidson used adeno-associated viruses (AAV) for therapeutic use by shRNA expression (Xia et al. 2004). This strategy was further pursued and evaluated in mouse models, aiming at treatment of various diseases such as prion disease or Huntington’s disease, viral infections like HBV, HCV, and HIV (Couto and High 2010). AAV were also used for the delivery of shRNA for eradication of prostate cancer xenografts (Sun et al. 2010).

Furthermore, Anderson et al. (2007) combined three anti-HIV RNAs in one lentiviral vector. In a xenograft mouse model, by the simultaneous delivery of shRNA, TAR decoy, and anti-CCR5 ribozyme into human CD34+ cells, T cells with maintained functionality but HIV resistance were generated. This threefold combination vector is used at the moment in a clinical trial for AIDS/lymphoma patients (DiGiusto et al. 2010).

Viral vectors can also be used for substituting downregulated suppressor miRNAs in tumors. Kota et al. (2009) have shown that systemic administration of an AAV vector expressing the suppressed miRNA in a mouse liver cancer model dramatically reduced tumor progression.

Although the results achieved with viral vectors in RNAi are very promising, this application for therapeutic purposes is still accompanied by a few challenges. shRNA expressed by viral vectors, for instance, can induce immune responses such as interferon (IFN) induction (Kenworthy et al. 2009; Fish and Kruithof 2004) or toxicity can occur due to oversaturation of the RNAi pathway (Grimm et al. 2006). Another significant drawback is the safety issues related to viral vectors, such as their potential of immune response induction or the risk of insertion mutagenesis. Furthermore, low loading capacity and difficult large-scale production may limit therapeutic application (Tomanin and Scarpa 2004; Devroe and Silver 2004).

4 Synthetic Carriers for Oligonucleotide Delivery

The research work dedicated to development and optimization of nonviral nucleic acid formulations is motivated out of the hope for a virus-like high transfection efficiency with minimal or no side effects. Viral features, though, can be considered and imitated, as viruses represent natural examples of most effective carrier system. Several classes of synthetic carriers have evolved out of this aim. Like in viruses, the formulation of siRNA or other oligonucleotides may follow significantly different rules depending on the chemical (lipid-free or lipid-based) composition of the carrier.

In the following sections, different classes of synthetic formulations used for oligonucleotide encapsulation for therapeutic purpose will be described.

5 Natural Cationic Polymers

Natural cationic polymers such as chitosan, atelocollagen, or protamine have the advantage of being biodegradable and nontoxic. Packaging of the oligonucleotide occurs via electrostatic interactions between the positively charged polymer and the negatively charged backbone of the cargo, thus forming nano-sized polyplexes, usually bearing a positive surface charge. Furthermore, encapsulation of siRNA has already been shown to enable biocompatible and efficient gene silencing.

Atelocollagen-siRNA polyplexes, for instance, were used for tumor growth inhibition in a germ cell tumor xenograft mouse model (Minakuchi et al. 2004). Also, vascular endothelial growth factor inhibition achieved with such polyplexes suppressed tumor angiogenesis and growth (Takei et al. 2004).

The natural, arginine-rich cationic polypeptide protamine has the ability for efficient condensation of nucleic acids. Complexes consisting of a cell-penetrating peptide derived from natural arginine-rich protamine and siRNA suppress tumor growth in a mouse hepatocarcinoma model (Choi et al. 2010) without induction of immunostimulatory effects.

Nanoparticles formed with the natural cationic polysaccharide chitosan, for instance, were shown to mediate effective gene silencing after either intravenous (Pille et al. 2006) or intraperitoneal (Howard et al. 2009) administration while at the same time displaying favorable safety profiles.

Cyclodextrin-based oligocations have also proven their capability of acting as biocompatible and efficient nucleic acid carriers. Targeted siRNA-cyclodextrin nanoparticles, with transferrin as targeting ligand and polyethyleneglycol (PEG) for shielding, have been used in mouse models for treatment of metastatic Ewing’s sarcoma (Hu-Lieskovan et al. 2005). Such nanoparticles were also intravenously administered to patients with solid tumors in a clinical trial (Davis et al. 2010), achieving specific gene inhibition.

6 Synthetic Cationic Polymers

A synthetic polymer-based backbone offers the possibility for a wide spectrum of modifications with regard to optimization for a given cargo. One of the leading synthetic polycation for pDNA delivery is the “gold standard” polyethylenimine (PEI) (Fig. 9.1c). PEI has outstanding features like excellent interaction with nucleic acids, pronounced endosomal buffering capacity, which is important for the endosomal escape, and high accessibility for modification (Boussif et al. 1995; Akinc et al. 2005). Therefore, it seemed reasonable that PEI, either linear or branched, might be useful for oligonucleotide and siRNA delivery. In order to avoid toxicity after systemic administration (Moghimi et al. 2005), various modifications of PEI were investigated, such as PEGylation (modification with polyethylene glycol; PEG) (Tsai et al. 2011; Beyerle et al. 2011), introduction of targeting moieties or lytic domains (Kwon et al. 2008), deacylation (Thomas et al. 2005), alkylation (Fortune et al. 2011), or succinoylation (Zintchenko et al. 2008). PEI modified with stearic acid was used as cationic counterpart for the formation of siRNA polyplexes for induction of tumor apoptosis in a melanoma cell line (Alshamsan et al. 2010).

Fig. 9.1
figure 00091

Examples of carriers used for oligonucleotide encapsulation. (a) Adenovirus. (b) Liposome consisting of a bilayer membrane with the lipid tails clustered together. (c) Linear PEI. (d) DOTAP. (e) Exemplified structure of sequence-defined polymers consisting of a cross-linking domain, a succinoyl tetraethylene pentamine unit and a bis (fatty acid) tail

Full size image

Schaffert et al. (2011b) combined different domains in one carrier, each for fulfilling another task, for delivery of the apoptosis, and for immune response inducing double-stranded RNA polyinosinic:polycytidylic acid, poly(I:C). Branched 25 kDa PEI was modified with PEG as a shielding domain, the tumor cell targeting moiety epidermal growth factor (EGF), and melittin for enhancing endosomal escape. The group could show growth retardation of human A431 xenograft tumors in mice.

The polypeptide polylysine (PLL) was also used as a cationic polymer for siRNA delivery. PLL on its own does not mediate efficient transfection and additionally has been shown by Symonds et al. to induce apoptosis in three different human cell lines (Symonds et al. 2005). Therefore, it was modified by Meyer et al. with different moieties, each responsible for facilitating a step in the delivery pathway, all together resulting in a very efficient delivery construct (Meyer et al. 2008, 2009).

7 Precise and Sequence-Defined Synthetic Carriers

Most of the polymers mentioned above and also others developed up to now have a significant drawback, which is polydispersity of the carrier. This negative feature hampers precise conclusions on structure–activity relationships, as the efficiency or also toxic effects cannot be traced back to one single compound. Also, large-scale GMP production requires identical reproducible batches; polydispersity impedes achieving this aim.

Development of dendritic polymers was a step towards solving this drawback. Different synthetic strategies enable the development of branched three-dimensional molecules with a well-defined structure and a narrow polydispersity range. The options for modification and improvement are very broad and have been widely investigated for gene delivery. Dendritic gene carriers such as polyamidoamine (PAMAM), polypropylenimine (PPI), or dendritic polylysine (PLL) with different surface and internal modifications have been investigated for delivery of antisense nucleotides, siRNA, and DNAzymes. A more detailed overview is given in Ravina et al. (2010).

In contrast to polymers, peptides designed for oligonucleotide and siRNA delivery (Andaloussi et al. 2011; Ezzat et al. 2011; Leng and Mixson 2005) benefit from precise sequence-defined assembly by solid-phase-supported synthesis (SPS). Hartmann and Börner took first advantage of this methodology for the generation of precise, sequence-defined poly(amidoamines) (Hartmann et al. 2006). These compounds were shown to efficiently condense pDNA (Hartmann et al. 2008).

Schaffert et al. (2011a) further developed this strategy by the generation of precise cationic oligo(aminoethane)-based building blocks suitable for SPS, thus enabling fast and parallel synthesis of structural variations. The oligoamino acid building blocks were designed to contain the 1,2-diaminoethane motif (Wagner 2012) which is part of a series of very effective polycationic carriers such as PEI or related degradable carriers (Miyata et al. 2012). In a subsequent publication, an excerpt was given over selected compounds out of a library of more than 300 structures with differing topologies and containing activity-enhancing moieties such as cysteines or fatty acids (Fig. 9.1e). These precise oligomers did not only show efficient transfection for pDNA, reaching levels equivalent to PEI, but also knockdown of the target gene down to approximately 10 % with very good maintenance of cell viability and low in vivo cytotoxicity (Schaffert et al. 2011c). Structure–activity relationships of these carriers with siRNA were further evaluated in vitro and in vivo (Frohlich et al. 2012). Apart from linear topologies, also branched structures such as four-arms and the influence of various different building blocks were tested (Salcher et al. 2012). Also, novel constructs generated by SPS containing of PEG moieties and targeting ligands such as folic acid could efficiently complex and deliver siRNA and knock down the target genes (Dohmen et al. 2012; Martin et al. 2012).

8 Cationic Lipids and Liposomes

Synthetic lipids containing a polar head group and a lipid tail have the ability to form so-called liposomes (Fig. 9.1b). Hydrophilic structures like oligonucleotides can be encapsulated in the aqueous core of such liposomes. Encapsulation may take advantage of inclusion of cationic lipids in the bilayer. Electrostatic interactions between the cationic lipids and the negatively charged nucleic acid strongly enhance the encapsulation efficiency. Alternatively, cationic liposomes may also form complexes (lipoplexes) with nucleic acid without encapsulation into the interior of lipid bilayer membranes. In both types of the formed particles, the lipidic structures can provide oligonucleotide cargo protection after systemic delivery. Moreover, a crucial advantage is the interaction of the cationic lipid with anionic lipids in the cellular and endosomal membrane, thus facilitating cell entry and endosomal escape as basis for cargo release into the cytosol.

The cationic lipid DOTMA (N-[1-(2,3-Dioleyloxy)propyl]-N,N,N-trimethylammoniumchlorid chloride) was the first synthetic lipid used for gene delivery to mammalian cells and was presented in 1989 (Felgner and Ringold 1989), followed by the delivery of antisense oligonucleotides and ribozymes (Akhtar et al. 2000; Hughes et al. 2001). Since then, these promising vectors were further improved and evaluated for oligonucleotide delivery. The cationic lipid DOTAP (Fig. 9.1d), for instance, was used for siRNA encapsulation and subsequent melanoma inhibition (Tran et al. 2008).

Cationic lipids were also commercialized, such as Lipofectamine 2000 (Invitrogen) and i-FECT (Neuromics), and widely used for siRNA encapsulation and transfection (Yokota et al. 2007; Kumar et al. 2006). Additionally, Lipofectamine was used for inhibiting HIV-I expression by transfection of encapsulated locked nucleic acid (LNA)-modified antisense oligonucleotides and DNAzymes (Jakobsen et al. 2007).

Particle stability is also a crucial issue concerning systemic delivery. Hence, stable nucleic acid lipid particles (SNALPs) got more and more into focus. SNALPs are PEG-lipid-coated liposomes consisting of different cationic and fusogenic lipids. The coating provides particle stability during formulation and also hampers undesired interactions with blood components.

In vivo studies conducted by Morrissey et al. (2005) showed inhibition of HBV replication by the administration of anti-HBV siRNA-SNALP complexes. Furthermore, systemic delivery of siRNA-SNALP formulation targeting apolipoprotein B (ApoB) in nonhuman primates led to efficient protein reduction lasting for 11 days after injection (Zimmermann et al. 2006; Frank-Kamenetsky et al. 2008). The SNALP formulation contained the lipids 3-N-[(methoxypoly(ethylene glycol))carbamoyl]-1,2-dimyristyloxy-propylamine (PEG-C-DMA), 1,2-dilinoleyloxy-N,N-dimethyl-3-aminopropane (DLinDMA), and 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC) and cholesterol at specified ratios.

Highly effective guinea pig postexposure protection from the lethal Zaire Ebola virus was shown by treatment with SNALP-encapsulated siRNA (Geisbert et al. 2006). In a subsequent study, a combination of SNALP-embedded siRNAs targeting the expression of three viral proteins was used for preventing macaques from lethal hemorrhagic fever induced by the Ebola virus (Geisbert et al. 2010). Due to their high efficiency and positive safety profile, SNALPs are currently being evaluated in clinical trials (Burnett and Rossi 2012).

Semple et al. used the ionizable cationic lipid 1,2-dilinoleyloxy-3-dimethylaminopropane (DLinDMA) as a base substance for the design of a library of 1,200 lipid nanoparticle compounds. These were screened for siRNA delivery, and the best performer was used for SNALP formulation and characterization. Gene silencing was tested in rodents and nonhuman primates with a successful outcome at low-dose application (Semple et al. 2010).

PEGylation of cationic lipoplexes (Sonoke et al. 2008) leads to increasing plasma concentration of the lipoplex and also to a higher accumulation in tumor. However, PEGylation also hampers cellular uptake and endosomal escape. Hatakeyama and colleagues therefore introduced between PEG and their multifunctional envelope-type nano-device (MEND) an enzymatically cleavable bond which should release the PEG moiety and thus facilitate cellular uptake. The molecule was further modified with the pH-sensitive fusogenic peptide GALA for enhanced endosomal escape, and encapsulated siRNA showed efficient knockdown after intratumoral delivery compared to unmodified MEND (Hatakeyama et al. 2009).

Toxicity, though, is an issue with cationic lipid particles, as has been reported both in vitro and in vivo (Ma et al. 2005; Lv et al. 2006; Akhtar and Benter 2007). The use of neutral lipids for encapsulation and delivery of siRNA should help to avoid toxicity associated with positively charged polymers. Peer et al. (2008) developed liposomes (neutral) coated with hyaluronan for stabilization and covalently attached monoclonal antibody on the surface. Condensed CyD1-siRNA was loaded into these particles, and efficient gene silencing in an intestinal inflammation murine model after IV application could be shown. In another study, siRNA targeting oncoprotein EphA2 encapsulated into neutral liposome 1,2-dioleoyl-sn-glycero-3-phosphatidylcholine (DOPC) (Landen et al. 2005) led to inhibited tumor growth in mice bearing ovarian tumors.

In order to circumvent undesired side effects caused by synthetic lipid structures, natural lipid analogs such as cardiolipin were evaluated for siRNA delivery (Chien et al. 2005; Pal et al. 2005).

A large library of lipid-like materials, so-called lipidoids, was developed and tested for the ability to transfect siRNA in vitro (Akinc et al. 2008), thus finding a leading structure for in vivo formulations. Further studies with this leading structure encapsulating siRNA showed efficient and sustained gene silencing in lung and liver targets in mice, rats, and cynomolgus monkeys (Akinc et al. 2008; Frank-Kamenetsky et al. 2008; Epiphanio et al. 2008).

Saw and colleagues loaded oligodeoxynucleotides complexed with cell-penetrating peptide (CPP) into CPP surface-modified liposomes (anionic) for successfully sensitizing glioblastoma cells to the effects of chemotherapeutic agent paclitaxel (Saw et al. 2010).

Another class of lipid-based vectors are solid lipid nanoparticles (SLN) consisting of natural or synthesized lipids, resulting in biocompatible and biodegradable lipid systems (zur Muhlen et al. 1998) which were primarily used for gene delivery (del Pozo-Rodriguez et al. 2007; Pan et al. 2009; Choi et al. 2008). SLN were also used for encapsulation of anti-microRNA oligonucleotides (AMO) in order to decrease human lung cancer cell functions related to microRNA-21 (Shi et al. 2012).

9 Conclusion and Prospects

Oligonucleotides possess a great therapeutic potential, as their gene sequence specificity can be exploited for finding the best strategy for disease treatment at its genetic origin. Delivery of oligonucleotides for therapeutic purposes has made vast progress during the last decades. Different strategies have been evaluated, all pursuing one aim—treating humans suffering of severe disorders. Nevertheless, appropriated shuttles have to be found for target cell transfection. Suitable encapsulation protects the cargo during blood circulation, mediates cellular uptake, and, depending of the payload nature, can lead to a sustained effect on the target gene.

Various encapsulation methods offer several options, all still bearing advantages but also drawbacks, though. Viruses, for instance, are the perfect nucleic acid carriers designed by nature. Thus, their outstanding properties concerning cargo protection and delivery were widely exploited for therapeutic purposes. Nevertheless, despite the highly promising results achieved with viral vectors, safety issues still hamper establishment as therapeutic agents. Therefore, nonviral oligonucleotide encapsulation methods and delivery strategies are getting more and more into focus. Either synthetic or natural polymers, cationic or neutral lipids have been evaluated for their therapeutic potential. The type of delivery system also has to be well chosen, though, as cationic lipids and polymers may have an impact on siRNA-caused immune stimulation and undesired off-target effects (Akhtar and Benter 2007; Hollins et al. 2007; Omidi et al. 2003, 2005). Even so, other features such as a positive safety profile, low production costs, and high loading capacity have led such carrier systems towards clinical trials, thus marking the first steps towards curing genetically based disorders by tackling the problem at its origin.