Rolling Circle Replication for Engineering Drug Delivery Carriers

Keywords:

• amplification
• DNA
• drug delivery
• nanotechnology
• RNA
• rolling circle replication
• transcription

The technique to generate long concatenated nucleic acids capable of constructing nano-, micro- and macro-scopic structures is a powerful platform for designing drug delivery carriers.

Rolling circle replication (RCR), including DNA and RNA replication, is a process found in some bacteriophages or viroids for replicating the DNA or RNA genomes. In vitro versions of this natural phenomenon are the rolling circle amplification (RCA) [1] and rolling circle transcription (RCT) [2] that enabled cost-efficient synthesis of concatenated DNA and RNA. Φ29 DNA polymerase, the most popular DNA polymerase for RCA, is capable to replicating the full-length genome of Φ29 bacteriophage [3]. The superb RCA performance of Φ29 DNA polymerase is attributed to its high processivity and strand displacement ability, making it possible to generate long consecutive DNA products even in the presence of topologically complicated DNA templates. In addition, Φ29 DNA polymerase reacts in an isothermal condition that obviates the need for a thermal cycler for the reaction. Other isothermal polymerases that can be employed in RCA include Bst DNA polymerase [4], Vent exo-DNA polymerase [1] and so on. For a typical RCA, the three major components are the DNA polymerase, a single stranded DNA (ssDNA) template and a ssDNA primer. The fact that the product is an amplification of the ssDNA template leads to extensive study of RCA for signal amplification in biodetection [1]. Recently, the polymeric property of RCA products has attracted a lot of attention due to the development of DNA nanotechnology [5]. The programmability of the DNA template makes RCA a highly versatile platform to generate DNA particles or gels for biomedical applications. Functional DNA sequences, such as aptamers and DNAzymes, can be incorporated into the RCA products for applications including targeted drug delivery or bioimaging.

Similar to RCA, RCT generates periodic RNA products from a ssDNA template. A typical RCT reaction requires only two major components, a DNA-dependent RNA polymerase (generally T7 RNA polymerase) and a ssDNA template containing the binding site of the RNA polymerase [2]. The high programmability of the DNA template makes RCT easy to manipulate. The flexible base-pairing rules of RNA, such as noncanonical base pairing, make RNA nanostructures more versatile and thermally stable than their DNA counterparts, giving RNA protein-like diversity in functions [6]. Functional RNA molecules such as aptamers, ribozyme, miRNA and siRNA have greatly expanded the toolbox for designing RNA carriers for drug delivery.

In this editorial, we highlight recent advances in using RCR techniques for engineering DNA- and RNA-based carriers for drug delivery. The carriers are categorized into three groups based on their structures. Future opportunities and challenges are also discussed.

RCR product self-assembled micro-nanoparticles

The products of RCR are linear single-stranded DNA or RNA chains with extremely high molecular weight, behaving more like generic materials than genetic materials when synthesized in vitro. Inter- and intra-molecular hybridizations tend to fold these linear chains into particles with the size affected by factors such as the selection [4,7] and concentration [8] of the polymerases, the elongation time [9] and so on. The magnesium pyrophosphate generated during the synthesis process, as a side product of the polymerized nucleotides, was recently reported to be necessary for the structural integrity of the nucleic acid particles [7,10]. In the pioneering work demonstrated by Hammond and coworkers, RNA microparticles composed of concatenated siRNA (more than half a million repeats per particle) were developed for siRNA delivery [2]. Intracellular machineries processed the cleavable RNA particles into mature siRNA and silenced target genes. In our recent work, RCA amplification was employed to prepare DNA nanoparticles (~150 nm) for delivering anticancer drug doxorubicin (DOX) [4]. GC-pair rich sequences were programmed into the DNA nanoparticle due to the preference of DOX to intercalate into GC-rich regions. A high DOX loading capacity (~70%) was achieved by this strategy. To further control the release of DOX from the DNA nanoparticle, an acid degradable polymeric nanogel was employed to encapsulate DNase I into positively charged single-protein nanocapsules. Electrostatic interactions between the DNase I nanocapsules and the negatively charged DNA nanoparticle assembled them into homogenous nanoassemblies with size around 180 nm. An acid triggered DOX release profile was observed. To further enhance the anticancer efficacy of this system, a targeting ligand (folic acid) was hybridized onto the surface of DNA nanoparticle via a linker oligo. It was observed that rapid nucleus localization of DOX occurs within the first 0.5 h after incubating the cells with the particles. In addition to our GC-rich RCA nanoparticle, other groups also demonstrated the incorporation of functional DNA motifs into the RCA particles for drug delivery. For example, DNA aptamers that recognize specific receptors on cancer cell surface were incorporated into RCA nanoparticles for targeted delivery of DOX [9]; antisense nucleotides were encoded into DNA microparticles, which were subsequently condensed with polyelectrolytes into nanoparticles, for gene therapy against cancer [7].

RCR product-templated nanostructures

DNA and RNA nanotechnology based on Watson–Crick or noncanonical base pairing has generated numerous exquisite 2D and 3D nanostructures with nanometer precision [5,6]. The facile synthesis of highly programmable nucleic acids with extremely high molecular weight makes RCR an attractive method to prepare building blocks for constructing DNA nanostructures. Hong et al. demonstrated the synthesis of Y-shaped DNA nanostructures for siRNA delivery by combining RCA reaction with restricted digestion to generate intermediate short ssDNA oligos [11]. In addition to using digested RCR products for assembling DNA nanostructures, applying intact RCR chains is also very popular. For instance, Hamblin et al. designed a DNA ‘ladder’ with RCA generated approximately 1400–15000 bp long ssDNA as the ‘strut’ and some other DNA oligos as ‘rungs’, resulting in very high aspect ratio of the final nanotube assembly [12]. Another version of RCR-based nanoassembly is the modified DNA ‘origami’, in which the generally used virus genomic DNA was replaced with RCR-generated DNA [13] or RNA [14] as the scaffold. In this strategy, the staples required to fold the scaffold were reduced from hundreds to only a few, which significantly simplified the folding process. In the proof of principle study by Weizmann and coworkers [13], intact and digested RCR products were both used to synthesize scaffold and staples, respectively. High aspect ratio and rigidity of the folded nanoribbon enabled its efficient cellular uptake as well as endosome escape, making the nanoribbon a multifunctional platform for delivering various types of cargos, such as fluorescent dyes, protein and siRNA.

RCR-generated macroscopic hydrogels

Aside from its role as building blocks in micro-/nanostructures, RCR products are also a type of hydrophilic and biodegradable polymer that can be harnessed to prepare bulk hydrogels for controlled drug release. Using a modified version of RCA, Lee et al. demonstrated the synthesis of bulk DNA gels [15]. On the basis of conventional RCA that used only one primer for amplification, a multi-primed chain amplification process was added using two additional primers complementary to each of the proceeding ssDNA product. Rheological properties of the synthesized gel were affected by the reaction time of RCA as well as the presence of water. Model drugs DOX and insulin were loaded into the DNA gel by intercalation and entrapment, respectively, and the DNA gel functioned as a drug reservoir for continuously releasing the drugs. An RNA version of the hydrogel was also demonstrated with two complementary templates for the RCT reaction, generating two complementary RNA chains [16]. After water evaporation, a thin layer of RNA membrane was obtained with its morphology affected by the nucleotide complementation. This RNA membrane worked as enzyme (RNAse or Dicer) responsive carriers for controlled release of DOX and siRNA.

Future perspective

RCR products are versatile materials that could be used as building blocks in drug delivery systems spanning from nano- to micro- and to macro-scopic scales. The development of DNA and RNA nanotechnology to precisely control nucleic acid nanostructures makes RCR a valuable technique for acquiring periodically repeated building blocks. The ability of incorporating thousands of copies of functional DNA/RNA moieties, such as aptamer, DNAzyme, ribozyme, siRNA or miNRA into a single chain makes RCR a powerful platform to fully utilize these DNA/RNA tools. Additionally, stimuli responsive nucleic acid moieties [17–19] can be easily installed for designing RCR-based controlled drug release system. Moreover, in contrast to the micro-/nano-scale nucleic acid assemblies based on synthetic oligos, the macroscopic characteristics of RCR product render it a useful tool even in the construction of macrodevices, such as controlling the shape of hydrogel assembly at the scale of millimeters [20].

Despite these advantages, challenges remain for practical application of RCR techniques. Although RCR is a facile reaction, the high cost of materials such as the phosphorylated DNA template and polymerases limited their mass production. Also, the intra- and inter-molecular interactions tend to aggregate the RCR products, making the formulation unstable and heterogeneous during storage [1]. Furthermore, the genetic nature of nucleic acids often raises concerns about its safety such as the immunogenicity or the chance of accidental incorporation into human genomes.

Financial & competing interests disclosure

This work was supported by a grant from NC TraCS, NIH's Clinical and Translational Science Awards (CTSA, 1UL1TR001111) at UNC-CH, the NC State Faculty Research and Professional Development Award, and the start-up package from the Joint BME Department of UNC-CH and NC State to Z Gu. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.