Thermostability, Tunability, and Tenacity of RNA as Rubbery Anionic Polymeric Materials in Nanotechnology and Nanomedicine—Specific Cancer Targeting with Undetectable Toxicity

Abstract

RNA nanotechnology is the bottom-up self-assembly of nanometer-scale architectures, resembling LEGOs, composed mainly of RNA. The ideal building material should be (1) versatile and controllable in shape and stoichiometry, (2) spontaneously self-assemble, and (3) thermodynamically, chemically, and enzymatically stable with a long shelf life. RNA building blocks exhibit each of the above. RNA is a polynucleic acid, making it a polymer, and its negative-charge prevents nonspecific binding to negatively charged cell membranes. The thermostability makes it suitable for logic gates, resistive memory, sensor set-ups, and NEM devices. RNA can be designed and manipulated with a level of simplicity of DNA while displaying versatile structure and enzyme activity of proteins. RNA can fold into single-stranded loops or bulges to serve as mounting dovetails for intermolecular or domain interactions without external linking dowels. RNA nanoparticles display rubber- and amoeba-like properties and are stretchable and shrinkable through multiple repeats, leading to enhanced tumor targeting and fast renal excretion to reduce toxicities. It was predicted in 2014 that RNA would be the third milestone in pharmaceutical drug development. The recent approval of several RNA drugs and COVID-19 mRNA vaccines by FDA suggests that this milestone is being realized. Here, we review the unique properties of RNA nanotechnology, summarize its recent advancements, describe its distinct attributes inside or outside the body and discuss potential applications in nanotechnology, medicine, and material science.

Graphical Abstract

1. INTRODUCTION: DEFINITION, CHALLENGES, AND BREAKTHROUGH OF RNA NANOTECHNOLOGY

1.1. Definition of RNA Nanotechnology

Ribonucleic acid (RNA) nanotechnology is the bottom-up self-assembly of RNA structures at the nanometer scale (Figure 1).⁴³ In RNA nanoparticles, their major frame is mainly composed of RNA. The particles often include the associated scaffold, ligands, and regulators.²⁷ While classical RNA biology studies focus on intramolecular interactions and folding into 2D or 3D structures, RNA nanotechnology instead focuses on inter-RNA interactions and quaternary (4D) structure (Figure 2).⁷² With the development of RNA nanotechnology, the field also expands to other components such as exosomes, cholesterols, fluorophore, chemicals, and drugs to form more complex structures.

Figure 1.

Studies on the phi29 pRNA ring leading to the method for the assembly of the RNA nanoparticles via bottom-up assembly by intermolecular interactions. (A) Hand-in-hand interaction to build dimer, trimer, pentamer, and hexamer. (B) Foot-to-foot interaction to build dimer, trimer, tetramer, pentamer, and hexamer. (C) Branch extension using phi29 pRNA as bases to build difference size of 3WJ-pRNA, X-pRNA, and branched hexamers.^13,23–27 Adapted with permission from ref 13. Copyright 2010 Nature. Adapted with permission from ref 27. Copyright 2013 Shu et al. under CC By-NC 4.0 https://creativecommons.org/licenses/by-nc/4.0/.

Figure 2.

Demonstration of versatile structures of RNA nanoparticles. (A) Phi29 packaging RNA (pRNA) first demonstrated control of nanoparticle design.⁴³ Reproduced with permission from ref 13. Copyright 2013 Shu et al. under CC By-NC 4.0 https://creativecommons.org/licenses/by-nc/4.0/. (B) RNA nanoparticles were constructed into various polygon shapes and confirmed through AFM imaging for controlled formation. Reproduced from with permission from ref 1. Copyright 2018 American Chemical Society. Reproduced from with permission from ref 4. Copyright 2009 American Chemical Society. Reproduced with permission from ref 38. Copyright 2017 Oi et al. Reproduced from with permission from ref 52. Copyright 2019 American Chemical Society. Reproduced from with permission from ref 53. Copyright 2017 Elsevier. Reproduced from with permission from ref. Copyright 2014 Oxford University Press under CC By 4.0 https://creativecommons.org/licenses/by/4.0/. Reproduced from with permission from ref 55. Copyright 2011 American Chemical Society. (C) RNA nanoparticles were constructed into 3D shapes using various construction techniques and naturally occurring RNA motifs. Nanoparticles were confirmed by Cyro-EM. Reproduced with permission from ref 5. Copyright 2010 Macmillan Publishers Ltd. Reproduced with permission from ref 58. Copyright 2016 Wiley-VCH. Reproduced with permission from ref 59. Copyright 2016 Wiley-VCH. Reproduced with permission from ref 60. Copyright 2020 Guo et al. under CC By 4.0 https://creativecommons.org/licenses/by/4.0/. Reproduced with permission from ref 61. Copyright 2018 Wiley-VCH. Reproduced with permission from ref 62. Copyright 2015 Elsevier. (D) RNA nanoparticles with repeating structures were created into complex molecules and confirmed by AFM. Reproduced with permission from ref 2. Copyright 2017 The Royal Society of Chemistry. Reproduced with permission from ref 38. Copyright 2014 American Chemical Society. Reproduced with permission from ref 50. Copyright 2017 American Chemical Society. Reproduced with permission from ref 66. Copyright 2018 American Chemical Society. (E) Complicated RNA nanostructures were constructed through the versatile pRNA 3WJ that is able to stretch its natural angles to 60°, 90°, and 108°. Reproduced with permission from ref 67. Copyright 2020 American Chemical Society. Reproduced with permission from ref 67. Copyright 2021 Informa UK Limited.

Nanotechnology involves the creation and application of nanoscale materials using a bottom-up assembly or a top-down approach.²³ RNA macromolecules are negatively charged polymers that hold self-assembly properties into perfectly controllable nanoscale sizes that can serve as powerful building blocks for bottom-up assembly to build nanoarchitectures and nanodevices.²⁷ Building blocks utilized in nanotechnology assemblies must meet certain standards to be considered an ideal material. The building blocks should have the following properties:

1. Have diverse structural variety.⁷³

2. Build structures with controllable shape, size, and stoichiometry.^{26,44,50,58,74}

3. Automatically assemble when mixed.^4,75–77

4. Be thermodynamically, chemically, and enzymatically stable with a long shelf life.^20,41,46

5. The product must have a nanoscale structure.^30,50,58,74

6. The product can be purified into a homogeneous and uniformed state.³⁰

7. The product can be characterized by chemical or physical means to define their properties.^59,78

Therefore, RNA nanotechnologists have evolved from the basic studies on RNA structure, function, folding, chemistry, and biophysics. In the past several decades, research on the folding and structure of RNA motifs and the development of three-dimensional calculations from traditional intramolecular interactions to intermolecular interactions have laid a solid foundation for the future development of RNA nanotechnology.^{31,32,42,79–88} For example, nanotechnology research relies on each product under their investigation having a defined molecular shape and size that can be characterized by physical and chemical procedures. Therefore, in these days, there is no clear career path resulting in a RNA nanotechnologist but rather an interdisciplinary arena utilizing disciplines from different fields.

Traditionally, nucleic acids including both DNA and RNA were considered only as biological molecules. This has been changed since 1982, when Seeman proposed the concept of using DNA as the material to build DNA branched junctions.^89,90 Two other pioneering groups focused on RNA structure based on natural derived RNA motifs. Guo focused on the controlled assembly of packaging RNA of the phi29 DNA packaging motor (Figure 3),⁴³ while Jaeger, Leontis, and Westhof focused on the tecto-RNA.^91–93 The studies of RNA structure and function have been expanded to become an exciting field by overcome several challenges. First, in the past, the sensitivity of RNA to enzymatic degradation by RNases was the biggest obstacle in the production of RNA as a construction material.²⁰ It was found that simple chemical modifications such as 2′-modification⁹⁴ or 2′−4′ linkage⁹⁵ has made the resulting RNAs resistant to degradation without significantly changing their folding properties and, in some cases, even without leading to the change of their biological functions.²⁰ The powerful products formed by stable RNA has greatly facilitated the development of RNA nanotechnology and in vivo therapy into a reality.^{20,94,96–98} Second, the increased thermostability of nucleic acid nanoparticles is needed to ensure they remain assembled at low concentrations commonly seen during in vivo circulation in therapeutic applications.^41,42,46,50 Compared with DNA, RNA is unique in its higher thermodynamic stability (discussed in detail below), typical and atypical base-pairing ability, base stacking, and a variety of loops and bulges are suitable for intramolecular and intermolecular interactions.^41,46 Additionally, RNA nanoparticles have been designed with increased thermostability to become highly stable.^27,42 Finally, RNA nanoparticles have demonstrated no detectable toxicity in mice and have tunable and controllable immune response by varying the size, shape, and sequence of the RNA nanoparticle.^1,36,99 Recently, new design strategies in RNA nanoparticles have been developed that address all the above concerns. As a result, RNA nanoparticles now can be applicable for clinical applications.

Figure 3.

Important findings of the novel structure, function, stoichiometry, and assembly of the pRNA of bacteriophage phi29 DNA packaging motor that led to the proof-of-concept and the beginning of RNA nanotechnology. (A) Discovery of a novel small RNA to gear the phi29 DNA packaging motor.^16–19 Adapted with permission from ref 18. Copyright 2013 Zhang et al. under CC By-NC 4.0 https://creativecommons.org/licenses/by-nc/4.0/. Adapted with permission from ref 19. Copyright 2002 The American Society for Biochemistry and Molecular Biology. (B,C) Discovery that the stoichiometry of the pRNA ring is the common multiple of 2 and 3 and is interlocking by hand-in-hand interaction. Reproduced with permission from ref 43. Copyright 1998 Cell Press. (D) The use of Yanghui’s triangle (a Pascal’s triangle, the traditional expression binomial distribution) to determine that stoichiometry of the pRNA ring is between 5 and 6. Adapted with permission from ref 70. Copyright 1997 American Society for Microbiology. In a combination of the common multiple of 2 and 3, it is concluded that the RNA ring is a hexamer.

Many nanotechnology-based platforms have been previously established and applied in diverse systems such as lipid-based particles,^100–102 viral nanoparticles,^103,104 synthetic inorganic,^105–109 and polymeric particles,^110–113 and as mentioned above, nucleic acid nanoparticles. However, not all nanoparticle materials are optimal for clinical use due to inherent challenges such as immunological and pharmacological constraints. Some nanomaterials used in clinical treatments are recognized by the immune system as invaders, resulting in uncontrolled immunostimulation or immunosuppression.^114,115 To translate potential nanotechnology-formulated drugs from the bench to the clinic, a series of challenges must be addressed and resolved such as assessments of pharmacology, safety, and efficacy profiles. One of the first steps in evaluating a nanoparticle drug delivery platform is evaluating whether they are compatible with the immune system via evaluations of immune safety and potential toxicity through biodistribution, hematological, and histopathological studies.¹¹⁶ Nanoparticle size plays an important role in regulating an immune response by adjustable manufacturing protocols.¹¹⁷ All of these factors regulate the success of nanotechnology in drug delivery and translation from clinical trials to the bench side. Macromolecules or nanoparticles can enter the cell by membrane fusion or by endocytosis similar to the process of cell entry process of viruses during infection. The multivalency (combination of therapeutic drugs, targeting ligands, and detection components by nanoparticle branching) of RNA nanoparticles allows for targeted delivery to enhance local concentration while also allowing specific cell entry or receptor mediated endocytosis.^11,77 Furthermore, the RNA nanoparticle’s overall negative charge prevents nonspecific binding to negatively charged cell membranes, thus resulting in repulsion from nontargeted cells and immune cells. Research has shown that RNA nanoparticles can harbor a variety of ligands, miRNA, siRNA, and chemical therapeutics to increase efficiency in tumor targeting, drug delivery, and tumor inhibition.^7,9–12,15 The advantageous size of RNA nanoparticles (10–40 nm) is large enough to avoid nonspecific diffusion into cells but small enough to avoid nonspecific entry into organs. The lack of proteins eliminates issues with antibody induction, allowing for repeated treatment of chronic diseases. RNA nanotechnology has shown great potential as a drug delivery platform, as nanoparticles can be designed that strongly bind to tumors with little to no accumulation in healthy organs, delivering payloads of therapeutics directly to tumor cells. This favorable size leads to favorable pharmacokinetics (PK) with fast tumor targeting, fast clearance in circulation and organs, but longer terminal half-life in tumors.^12,118 Repeat IV injections of up to 30 mg/kg pRNA nanoparticles to immunocompetent C57B/6 mice did not result in any toxicity.¹¹⁸

RNA nanoparticles self-assembled from RNA fragments of less than 80 nucleotides, allowing for chemical synthesis.^4,42,75–77 RNA nanoparticles are classified as chemical drugs and are not biologically produced in a living system.²⁷ As a result, their structure, shape, and physical/chemical properties are well characterized, homogeneity can be achieved, and quality control is simpler (Figure 1).¹¹⁹ These features facilitate quality control and FDA drug approval. The ultrahigh thermostability of RNA nanoparticles including 3WJ and its derivatives (Figure 4)^{31,32,42,80–82,84–88} enable it to make a hydrophobic drug soluble for high loading delivery to specific cell cytosol without toxicity or liver accumulation.¹²⁰

Figure 4.

Discovery of the ultrahigh thermostability of the three-way junction (3WJ) of phi29 pRNA that leads to the possibility of constructing stable RNA nanoparticle for in vivo delivery of RNA nanoparticles. (A) 2D structure of pRNA and 3WJ, as well as annealing curves of different nature derived 3WJs. Adapted with permission from ref 42. Copyright 2011 Nature Publishing Group. (B) Structure and melting curves of triangle RNA nanoparticles built from phi29 3WJ, demonstrating that pRNA 3WJ displays the highest T_m. Adapted with permission from ref 50. Copyright 2014 American Chemical Society.

As an alternative to other nanomaterials, RNA has rapidly grown and emerged into a nanotechnology platform due to its structural and functional diversity, which can be proven by the explosive publications on RNA nanostructures in the past 15 years (Figure 5). These publications originate from different fields such as chemistry, biochemistry, structural biology, microbiology, cancer biology, cell biology, biophysics, pharmacy, and nanomedicine. RNA nanotechnology is an interdisciplinary and innovative research field that brings together many professionals of diverse backgrounds and skills. RNA nanotechnology is rapidly developing and a dynamic scientific field. Therefore, this review is not able to cover all the applications related to nanotechnology in the modern world and will focus on the stability and intrinsic elastic property of RNA nanoparticles to allow for vast applications in disease treatment.

Figure 5.

Growth in the number of publications in RNA nanotechnology. A Pubmed search using keywords “RNA nanoparticle,” “RNA nanostructure,” RNA-nanoparticle,” and “RNA-NP” was completed to show the growing interest and work completed in the RNA nanotechnology field since its conception in 1998.

1.2. History of RNA Nanotechnology

The earlier concept that RNA was categorized into mRNA, tRNA, and siRNA was subsequently rewritten with the discovery of ribozymes¹²¹ and self-replicating viral RNA. In 1987, Guo predicted that there were a large assortment of small RNA species, known as sRNAs, in cells with obscure and novel functions but have not been identified.¹⁶ The advancement of the field of RNA nanotechnology is a collaboration of the scientific community, and many of them contributed to a different section for the advancement of this field.^{2,26,31,32,42,73,80–82,84–88,91–93,119,122–128}

Looking back into history, the concept of RNA nanotechnology was nurtured from the study of the pRNA of bacteriophage phi29 DNA packaging motor (Figure 3).¹⁶ This pRNA displays many novel geometrical quantitative features, thus, its study has led to the emergence of RNA nanotechnology.²⁷ Binomial distribution studies applying the theorem of “Yanghui’s Triangle” (“Pascal’s Triangle”) using truncated pRNA led to the finding that the number of pRNA in one motor is approximately six (Figure 3D).⁷⁰ Subsequent studies revealed that the number of RNA in the motor is the common multiple of 2 and 3 because the use of either two or three complementary pRNA mutants leads to the assembly of infectious phi29 virions (Figure 1).⁴³ A formation of the hexameric RNA ring on the motor via hand-in-hand interaction was eventually elucidated,¹²⁹ thus confirming that the stoichiometry of the RNA is not five but six because the number of six agrees with the results of the common multiple of 2 and 3.¹³⁰ Single-molecule imaging in combination with the photobleaching confirmed the six copies of RNA per motor.¹³¹ The intriguing study of the pRNA hand-in-hand interaction to form RNA dimer, trimer, and hexamer that is the common multiple of 2 and 3 proved the concept of RNA nanotechnology in 1998.⁴³ In this paper, it was reported that the engineered RNA can from dimer, trimer, and hexamer via bottom-up self-assembly (Figure 3).

The concept of DNA nanotechnology was proposed by Ned Seeman in 1982⁸⁹ and was developed rapidly. Why is there a need for RNA Nanotechnology? The publication of the proof of concept of RNA nanotechnology has attracted the interest of the DNA nanotechnology pioneers who intend to work on RNA nanotechnology. However, the researchers later on found out that RNA nanotechnology is very different from DNA nanotechnology. They found that many aspects RNA folding and assembly did not follow the laws of DNA nanoparticle construction. Now that the RNA as the third milestone in pharmaceutical drug development is realized by the approval of several drugs by the FDA for clinical application, and the novel effect of the mRNA vaccines for Covid-19, leads to a clear signal that RNA is stable in vivo and can be applied to the human body. Many thanks to the RNA chemists, biochemists, computational scientists, and structure biologists who have dedicated their livelihoods to provide discoveries, theorems, principles, criteriums, and building blocks for the development and prosperity of the RNA nanotechnology field!

2. RNA AS A NATURAL, VERSATILE, ANION POLYMER AND LEGO BLOCK TO BUILD SELF-ASSEMBLING NANOARCHITECTURES

Polymers are one of the most common and well-studied materials in both material science and daily life. RNA by definition is considered as a biopolymer due to its composition of repeating structures of nucleotides bound together as one structure. From this, RNA has been coined as “polyribonucleic acid.”¹³² The value of RNA as a polymer is manifested in the vital use as building blocks in material and life sciences. However, RNA’s polymeric nature is commonly overlooked in material and technological applications due to an impression that RNA is seemingly unstable. However, recent efforts have proven the high thermodynamic stability of RNA nanoparticles and significantly increased the enzymatic stability.^20,44,46 By virtue of the distinctive physicochemical properties, recently, RNA has been extensively explored in material science. RNA is a biopolymer relevant to but distinct from DNA. Both RNA and DNA can self-assemble into controllable nanostructures;^30,75 however, RNA is much more thermodynamically stable than DNA (Figure 1).⁴¹

2.1. RNA Holds the Key Feature of Polymers

The word (poly)-(mer) means (many)-(parts) and typically refers to macromolecules consisting of a large number of repeating elementary units covalently joined together. It is with these “mers”, which control structural morphology of the interchain interactions, assembled into the macromolecular shape that dictates the final polymer properties. Given the large variations in polymer structures, there are multiple ways to describe polymer structure. At the simplest level of classification, polymers that are formed by one type of monomer are called homopolymers, and macromolecules containing different types of monomers in one single polymer chain are called heteropolymers. Heteropolymers can be further described by the arrangement of the monomers along the polymer backbone, including alternating, random, and block formations. Beyond this simple linear arrangement, RNA polymers can also adopt elaborate structures.³⁰ Taken together, these large arrays of potential structures and chemical diversity offer a large variety of available properties, leading to a complex selection space from which to design a polymer for a final application. RNA is unique in that it is between homopolymer and heteropolymer because RNA is composed of four nucleotides: adenosine (A), cytidine (C), guanosine (G), and uridine (U). Thus, RNA is a homopolymer of nucleotides but is also a heteropolymer of A, C, G, and U. Such a unique property makes RNA special among other polymers.³⁰ Furthermore, the combination of RNA strands into an RNA nanoparticle classifies RNA as a heteropolymer.

Biopolymers, the building blocks of living systems, are composed of repeating biological structures. Ribonucleic acid is one of the five important biopolymers along with DNA, proteins, lipids, and carbohydrates. Each RNA nucleotide is composed of a ribose sugar, a nucleobase, and a phosphate group. The presence of the phosphate group at each nucleotide gives RNA a negative charge while also covalently linking together the nucleotides through 3′−5′ phosphodiester bonds. The phosphodiester bonds provide directionality to the sugar–phosphate backbone that defines RNA as a polynucleic acids. The negative charge carried on the backbone makes RNA a polyanionic macromolecule at physiological pH. The negative charge of RNA has proven beneficial, as it results in repulsion from negatively charged cell membranes,²⁷ thus reducing nonspecific binding and side effects in medical applications.

Polymers are characterized predominantly by their molecular weight, which is usually at least 100 times greater than those of typical small molecules like water. Interestingly, given the synthetic route of most polymer systems, the concept of a single chain with a single-polymer molecular weight is a rarity. As such, polymers must be characterized by the distribution of chains that exist throughout the macromolecule. It is possible to characterize this distribution by looking at both the average molecular weight based upon the number of chains present or based upon the weight of each chain. One can get a sense of the breadth of the variability in chain length by looking at the polydispersity index or the ratio between the weight-average and number-average molecular weight. Owing to the inherently random nature of most polymerization mechanisms, chain polydispersity and controlled shaping remains one of the limiting hurdles of chemical polymer production. Large chain polydispersity is typically correlated with increased variable performance and overall reduced control of the final properties. As such, there has been much effort toward developing synthetic building blocks and schemes that can recreate monodisperse polymer systems. These, however, are typically very expensive, time-consuming, and difficult to control. In contrast, RNA is synthesized by either transcription or solid-phase chemistry with minimal effort. Both approaches are based on stepwise reactions and will generate a polymer with a clearly defined sequence, structure, and molecular weight. This fact leads to a polydispersity index of a specific RNA of 1 in contrast to traditional heterogeneous synthetic polymers which have polydispersity indexes greater than 1.¹³³

Structurally, the average molecular weight has a greater impact upon the observed mechanical properties of the system. However, as polydispersity increases, the lower molecular weight chains can act as a lubricating/solvating system, reducing the observed mechanical strength and integrity of the system. The weak chemical interactions that exist between polymer chains also impacts the observed structural strength and stability of the polymer. As increases in temperature result in increased molecular motion, these weak associations begin to destabilize and reduce mechanical integrity. Such RNA base-pairing and structuring is sensitive to heating and behaviors similar to thermoplastics, as their mechanical strength or TM weakens with increased temperature. However, this property, in combination with the elasticity and rubber-like property of RNA (discussed in detail below),³³ also means RNA polymers can be intentionally deformed at higher temperatures and regain their structural properties once cooled. The higher thermostability of RNA compared to DNA makes it more stable at higher temperatures, with some RNA nanostructures demonstrating boiling resistance.⁴⁴ The high stability at lower temperatures and programmable dissociation or misfolding at higher temperatures make the plasticity of RNA controllable at will. The boiling resistance of RNA suggests a potential of RNA in molding technology.

Chemical polymers have a wide range of applications in biomedical field and industry such as information technology and nanotechnology. Specific applications include:¹³² (1) sutures for surgery, (2) dental devices, (3) orthopedic fixation devices, (4) scaffolds for tissue engineering, (5) drug formulation (e.g., controlled release of drugs, excipients for drug stabilization, or solubilization), (6) biodegradable vascular stents, (7) biodegradable soft tissue anchors, (8) implantable biomedical devices, (9) flexible transparent displays. (10) field effect transistors, (11) solar cell panels, (12) printing electronic circuits, (13) organic light-emitting diodes, (14) supercapacitors, etc. However, as the field of RNA nanotechnology continues to develop, we see increased application of RNA biopolymers moving into these fields, often many times unexpectedly.^134,135 Usually, an ideal chemical polymer for biomedical and industrial applications should have the following properties: (1) nontoxic,^12,118 (2) no or low immunogenicity,³⁶ (3) biodegradable and metabolizable, (4) the mechanical properties must fit the needs for the specific application, (5) easily sterilized, (6) processing, (7) electrical conductivity,¹³⁴ and (8) thermostable.⁴⁶ RNA as a polymer favorably displays these properties, as described below in the following sections.

As a result of these developments, we are at another unique turning point in the evolution and design of RNA. We once again turn toward nature as our inspiration for the development of advanced materials. By applying what we have learned about the complexity and versatility of nucleic acid structures, it may be possible to create materials with revolutionary properties previously considered unobtainable by polymers. Recent advances in RNA chemistry, RNA biology, as well as the emerging field of RNA nanotechnology, has shown that RNA biopolymers not only share the common characteristics of other macromolecules but also possess a range of unique properties which are advantageous for applications in nanotechnology as well as biomedical and material sciences.

2.2. RNA as a Natural Biopolymer with Flexibility to Assemble into Nanostructures Using Unique Building Blocks

RNA is a naturally occurring biopolymer with specific superiority when compared to other biomacromolecules.¹³² While RNA molecules are commonly single-stranded and self-folded, they rely on Watson–Crick base-pair interactions between adenine and uracil (A:U) and guanine and cytosine (G:C), the wobble base-pairing, or other noncanonical base-pairing, and sugar puckering.^136,137 Noncanonical base-pairing is unique within RNA that allows for pairing between guanine and uracil (G:U) and cytosine and adenine (C:A) through hydrogen bonds provides additional basepairing above DNA that generates diverse structuring and stability in RNAs.^138–140 Additionally, nucleic acids undergo sugar puckering, with RNAs found in nature containing puckering.^141,142 Sugar puckering has been found to increase the stability of RNAs and minimalize their folding energies, including in tRNAPhe shown by Thiyagarajan et al., and the discovery of an axial C–H···O interaction between the C2′-H2′(n) group and the O4′(n+1) atom contributing to the stabilization of RNA helical regions by Auffinger et al.^142,143 Additionally, sugar puckering has been shown to create stabilization of chemically modified RNA sequences, including 2′-fluoro and locked nucleic acids.^144,145 In general, under physiological conditions, RNA helix displays A-type configuration, whereas the DNA helix is predominantly B-type. The 2′-OH in RNA ribose locks the ribose into a 3′-endo chair conformation which does not favor a B-helix. Although the difference in the stacking interaction is small between DNA and RNA, the sum over numerous base-pairs can make a large difference to the helix stability. Furthermore, RNA contains unique secondary, tertiary, and quaternary interactions that lead to multiplicity in structure and versatility in function. A substantial number of self-folding RNAs with multiple energy levels are possible, including loops, bulges, stems, kissing hairpins,¹⁴⁶ pseudoknot,⁸⁷ paranemic motifs,⁷⁶ kink-turns,¹⁴⁷ three-way and multihelix junctions,¹⁴⁸ bulges and loops,¹⁴⁹ and formation of triplexes and quadruplexes.^150–154 Such triplex and quadruplex structures form under noncanonical base-pairing of three or four nucleotides, respectively. G-quadruplexes are commonly formed in G-rich RNA sequences. These structuring components are essential to guide and drive RNA molecules to assemble into desired structures with high diversity.

RNA has intrinsically defined nanoscale features and may serve as powerful building blocks for the bottom-up fabrication of nanostructures and nanodevices. The interest in RNA nanotechnology has increased in recent years as recognition of its potential for applications in nanomedicine, including the treatment of cancer, viral infection, and genetic diseases, has grown (Figure 5). RNA can be designed and manipulated with a level of simplicity that is characteristic of DNA, while displaying flexibility in structure and diversity in function (including RNAs with enzymatic activity) that is similar to that of proteins.²⁷ Although RNA nanotechnology resembles DNA nanotechnology in many ways, the base-pairing rules for constructing nanoparticles is different. The large variety of loops and motifs found in RNA allows it to fold into numerous complicated structures, and this diversity provides a platform for identifying viable building blocks for various applications. The thermal stability of RNA also allows for the production of multivalent nanostructures with defined stoichiometry. An RNA oligo composed of up to 120 nucleotides can be chemically synthesized and, by variance of nucleotide composition creates ~10⁷² various sequences (as expressed as = 4¹²⁰) that will result in a diverse array of different structures. Moreover, the single-stranded stem-loops found in RNA have proven to serve for intra- and/or intermolecular interactions, and these can be used to make “dovetail” joints between different building blocks, thus removing the need for an equivalent of dowels in RNA nanostructures and nanomachines.¹⁵⁵ Loops and motifs also allow for the construction of a more complicated secondary structure. Furthermore, RNA molecules such as aptamers,¹⁵⁶ ribozymes,¹⁵⁷ and short interfering RNA (siRNA)¹⁵⁸ can have special functionalities. For example, RNA has a flexible structure and possesses catalytic functions that are similar to proteins.

With these advantages, RNA has been successfully demonstrated to serve as a unique polymeric material to build a variety of nanostructures. RNA nanoparticles can be constructed into multivalent structures via bottom-up assembly using unique RNA structure as building blocks, such as RNA polygons,³⁶ polyhedrons,¹⁵⁹ rings,¹⁶⁰ jigsaw puzzles, arrays,⁴⁴ bundles, membrane, microsponges,^161,162 and filaments. The RNA oligos used in the preparation of nanoparticles can be produced by automated chemical synthesis and one-pot self-assembly within seconds after mixing the RNA segments (Figure 2).^{30,42,163,164} The unusually stable three-way junction (3WJ) motif of packaging RNA (pRNA) was generated as described in our previous study (Figure 4A).⁴² The chemical and thermodynamic stability of RNA nanoparticles is further enhanced through RNA base modification, which makes the RNA nanoparticle resistant to RNase degradation without affecting its folding properties or biological functions (as discussed below).¹⁶⁵

It is well established that chemical and physical properties of nanoparticles play an important role in their in vitro and in vivo interactions, thus contributing to pharmacokinetics and biodistribution properties. The role of the RNA nanoparticle rubbery property on key functions including blood circulation time, biodistribution, antibody-mediated targeting, endocytosis, and phagocytosis has been well demonstrated. The application of various RNA nanostructures as a carrier in cancer therapy has been extensively studied. As discussed in detail below in the following sections, RNA nanoparticles have demonstrated to be of high thermodynamic stability; controllable size, shape, and stoichiometry; possess favorable tumor inhibition efficiency; and limited toxicity and immunogenicity. The rubbery property of RNA nanoparticles, as discussed in detail below, allows for high tumor accumulation; low retention by the tumor, kidney, and liver; and rapid renal clearance through glomerular filtration.³³

3. THE ELASTICITY AND TUNABLE PROPERTY OF RNA

Because of its wide use in industrial and daily life, rubber has become a model material for elastomers.^166,167 Rubber obtained from Hevea Brasiliensis, as well as synthetic rubber, structurally consists of long linear chains that allows for stretching of the long chains, creating a unique material.^168–170 Because of mammoth economic and technological importance of rubber-like materials, investigation and search for new elastomers in the burgeoning field of nanotechnology has been of enduring interest to scientists and engineers.¹⁷¹ Nucleic acid nanostructures allow for bottom-up assembly and has appeared to be able to stretch and deform into many confirmations using a common core motif.^50,58,74,172 The versatility of RNA is highly evident in the diversity of structural repertoires available in nature, which include simple structures as helical stems and single-stranded hairpin loops to more complex multiway junctions and pseudoknots. The mechanics of RNA bending is also manifested by its functional role in diverse cellular processes, such as RNA transcription, termination, RNA interference, RNA splicing, 5′ or 3′ end processing, and many others. These features of RNA nanoparticles resembles rubber and other elastomers, as they are able to extend under applied external forces and return to their original form upon relaxation of the force (as discussed in detail below).¹⁷¹

The realization of nanotechnology to clinical applications relies on the demands of developing a nanomaterial that has tunable physical properties ensuring specific targeting and delivery of therapeutics to the desired disease site. To overcome such roadblocks, a scientist must consider the design to include beneficial qualities to desired nanoparticles. For example, nanoparticles that are soft and elastic in nature are able to circulate through the bloodstream and squeeze through leaky blood vessels by temporarily deforming shape under the EPR effect to then spring back to the original shape within the tumor, allowing for higher retention times within the tumor. Additionally, nanoparticles must display a high stability to be able to remain intact at elevated in vivo temperatures and rapid dilution as it is injected into the bloodstream. Finally, the tuning of size and shape allows for the construction of nanoparticles to conform to the needs of retention time, avoidance of accumulation in healthy organs, and the incorporation of therapeutic modules to perform a variety of tasks in vivo.

3.1. The Elastic Nature of RNA

From a mechanical perspective, typically persistence length is used as a measure of the flexibility of DNA and RNA duplexes. The bending persistence length of dsRNA is slightly larger than dsDNA of similar length. However, RNA structural motifs can be inherently flexible or rigid at length scales much shorter than the persistence length. Several studies using atomic force microscopy (AFM) and tweezers has revealed that the elasticity of a micrometer long (few thousand bp) dsRNA is 10–20% greater than that of dsDNA, which exhibits a stretch modulus of ~1000pN.^173,174 Interestingly, shorter (~150 bp or less than 100 nm) dsDNA or dsRNA structures are much more elastic with a stretch modulus of ~100 pN, which is 10-fold smaller compared to longer micrometer dsDNA.^175,176 The mechanical properties of nucleic acids play an important role in the assembly of nanostructures, and it is imperative to study even shorter lengths of nucleic acids. Through single-molecule force spectroscopy, molecular dynamics (MD) simulations, and nuclear magnetic resonance (NMR), Chiu et al., studied ~10 nm long dsDNA and examined how the structure and elasticity of dsDNA is altered by the presence of rNMPs located at specific positions within the DNA strands (Figure 6A).⁵⁷ Figure 6B shows a typical force versus distance curve acquired during the loading and unloading process of an AFM tip approaching and retracting from the gold surface where the DNA duplexes are attached. The stretch modulus (M) is then defined as

$$M\equiv \text{stretching force}/\text{strain}=\frac{F_{\text{st }}}{\delta /L_0}=\text{elastic modulus}\times \text{cross sectional area of the molecule}$$ (1)

From the measurements, it is possible to obtain the stretching force F_st, the DNA extension δ, and the DNA initial length L₀ to calculate the stretch modulus M.

Figure 6.

Elastic nature of DNA affected by ribonucleotides. (A) Schematic of DNA stretching experiment using AFM. The AFM tip is streptavidin modified and interacts with the biotin end labeled DNA. DNA thiolated ends are attached to the gold surface. (B) Top image shows the force–distance curve obtained in typical AMF force measurements. Bottom image shows schematic of force–distance curves for the determination of L₀, δ, and F_st. (C) Chemical structures of an rNMP and a dNMP at the base G. The ribose and deoxyribose sugars are shown in red and blue, respectively. (D) The two sequences of the rGMP-embedded DNA molecules, and the DNA-only controls, dG-DNA, used in AFM force measurements. The dNMPs are shown in blue and the rGMPs are shown in red. (E,F) Summary of values L₀, δ, and F_st calculated from the force–distance curves. (G) Example of duplexes for analysis by AFM. Two duplexes containing five rNMPs are shown. The repeated sequence module is indicated by the black bracket. Adapted with permission from ref 57. Copyright 2014 The Royal Society of Chemistry.

Two different sequences (30 nt) were constructed for this study, denoted dG-DNA and rG-DNA (Figure 6C,D). For both sequences, rGMP (or rG) was inserted on one strand and then annealed to a complementary strand to generate the rG-DNA duplexes; for dG-DNA constructs, the aforementioned specific rGs were replaced by canonical dGMP (or dG). The presence of rGs did not markedly influence the duplex stability or interfere with the formation of the B-form helix. One end of the DNA strand was attached to a gold substrate through thiol chemistry, while the other end was attached to the AFM tip through biotin–streptavidin interactions. AFM-based single-molecule force spectroscopy was then used to stretch the dG-DNA and rG-DNA constructs and evaluate the elastic properties (Figure 6E,F). For a particular sequence (#1), the presence of rG intrusions reduced the stretch modulus of DNA by ~32%, indicating softening of the DNA complex. For another sequence (#2), the rG intrusions stiffened the DNA as evidenced by an increase in stretch modulus by ~23%. These data show that the rNMPs embedded in DNA have significant impact on the DNA mechanical properties. Given that both sequences have the same number of rG intrusions, the results demonstrate that the perturbation of elasticity is location and sequence dependent. MD simulations indicated that the perturbations are from local structural distortions (α- and γ-torsional angles of the sugar–phosphate backbone) arising from hydrogen bonding between the OH group of rG and electronegative sites of either the phosphate backbone or vicinal base. NMR measurements further confirmed the local perturbations of the duplex backbone. The structural perturbations revealed by MD simulations and NMR spectroscopy were likely linked to the elastic alterations caused by rNMPs in DNA but could not anticipate the direction and degree of the alterations. AFM is thus an excellent system to examine how rNMPs impact DNA elastic properties. The rNMPs substitutions in DNA clearly changed the mechanical properties of DNA in a way that can result in largely increased or even decreased DNA elasticity.

In another study, the Guo lab investigated elastic properties of RNA, 2′-fluoro (2′-F) RNA, and DNA architectures.³³ They constructed same-sized nanosquares of RNA and DNA and used single-molecule optical tweezers to study their elastic behavior. The nanosquares were self-assembled and sandwiched between the two dsDNA strands of 2500 bp in length. One end of the construct was labeled with biotin, while the other end was labeled with digoxygenin, allowing for trapping in two separate laser foci of optical tweezers through conjugation to streptavidin-coated polystyrene beads and antidigoxigenin, respectively. A single molecule was selected by bringing the two beads together and moving them apart. Force ramping experiments were performed along vertices of each square. A loading rate of 5.5 pN/s was applied to gradually increase the force by moving one of the trapped particles away from the other. The stretching led to a conformation change of the square, as demonstrated by a sudden jump event in force–extension curves recorded during the experiments (Figure 7). The conformation changes of nanosquare by stretching comes back to the original conformation upon slowly relaxing the force.

Figure 7.

Elastic stretching of RNA nanoparticles by a customized dual trap optical tweezer by Guo and colleagues. (A) Schematic diagram of dual trap optical tweezers with tethered RNA construct sandwiched between two dsDNA handles via affinity linkers. (B) A typical force–extension curve for stretching (red) and relaxing (black) of a square nanoparticle. Inset: Magnified view of conformational change of the square nanoparticle. (C) Histograms for change in extension measured from force–extension curves for nanosquares of RNA, 2′-F RNA, and DNA left to right, respectively. Adapted with permission from ref 33. Copyright 2020 American Chemical Society.

From the force–extension curve, a change in extension was quantified that reveals the stretchiness of the particles. The change in extension was obtained by subtracting the distance during extension from distance during relaxation from the feature in the force–extension curve. It was found that among the RNA, 2′-F RNA, and DNA nanosquares, the change in extension for RNA was the longest followed by 2′-F RNA ahead of DNA (Table 1). This could be due to the difference in the structure of helix and different number of nucleotides per helical turn of RNA and DNA. This result demonstrates a highly flexible and more stretchy behavior of the RNA nanoarchitecture compared to DNA. Kinetics of change in conformation was analyzed by fitting the force histograms of each square by an equation proposed by Dudko et al.¹⁷⁷ It was found that the change in activation energy is smallest for RNA. RNA also has a faster conformation change as demonstrated by the conformation change rate constant and smallest distance to the transition state from the state before change in conformation (Table 1). The lower change in activation energy required to change the conformation of RNA may have attributed to more flexible nature of the RNA nanosquares under tension.

Table 1.

Conformation Change Force (pN), Conformation Change Rate Constant (k_cc, per s), Conformation Change Energy Barrier (ΔG^†, kcal per mol), and Distance to the Transition State (x^†, nm) for DNA, 2′-F, and RNA Squares^a

	force	ΔX	Koff	x†	ΔG†
DNA	44.2 ± 2.3	2.99 ± 0.03	0.0016 ± 0.0007	0.132 ± 0.008	13.5 ± 0.5
2′-F	44.3 ± 4.3	3.2 ± 0.3	0.003 ± 0.002	0.12 ± 0.02	12.3 ± 0.3
RNA	44.6 ± 4.1	4.1 ± 0.1	0.005 ± 0.003	0.10 ± 0.02	11.4 ± 1.4

^aAdapted with permission from ref 33. Copyright 2020 American Chemical Society.

Additionally, Zonghe Xu et al. examined the structural strength of RNA nanoparticles and their ability to resist stretching by modeling force exertions upon the pRNA-3WJ along its helices.¹⁷⁸ The pRNA-3WJ was examined under SMD simulations and AFM force pulling to examine the forces required to deform and destroy the 3WJ. It was found that magnesium coordination sites along the portal axis created an ultrastable motif that was stronger than many mechanically stable proteins. The magnesium coordination sites proved valuable for creating a stable structure that linked helices across different axes together, allowing for deformation resistance. These simulations and force measurements further support the elastic nature of RNA, and the material shows the ability to resist immediate unfolding as mechanical strain is placed upon the pRNA-3WJ, allowing the material to stretch under force and return to its original structure upon release. Salt ion coordination sites are common within folded RNA structures, along with complex tertiary structuring, which may allow for a common strong mechanical strength and elasticity that is observed here in the pRNA-3WJ.^18,178–181

The results discussed here highlight the versatility of RNA to either increase the elasticity to tune the shape of propagating strands in nanoparticle assembly or increase the stiffness to fold RNA into rigid structural motifs that are distinct from those adopted by DNA.¹⁸² The pRNA-3WJ motif has proven RNA structuring provides a stronger more flexible platform compared with DNA. Additionally, similar studies looking at the elasticity of helical RNA and DNA have been completed, especially using magnetic tweezers, concluding the same findings as described above.^183–186

3.2. Tunable Nature of RNA Motifs

RNA structural motifs display distinct and reproducible folding based on primary sequences, allowing for their use in construction of nanoarchitectures.^93,179,187 Utilizing the crystal structure of natural RNA structural motifs, nanostructures can be built utilizing the motifs themselves as building blocks. Many diverse RNA nanostructures have been constructed utilizing this method with diverse end goals as the basis for construction as discussed in detail below.^{26,73,120,188} RNA is known to exhibit phenomenal flexibility, leading to the ability of the user to tune RNA motifs to their liking to construct diverse architectures from identical structural motifs. Additionally, by utilizing databases that collect information on RNA motifs, i.e., sequence, crystal structure, and junction angles, a particular motif can be chosen based on the desired construction parameters.^189–193 This has been displayed utilizing the pRNA-3WJ from bacteriophage phi29.

Here we discuss the tunable features of RNA using the pRNA-3WJ motif as an example.^1,44,50 On the basis of the crystal structure of the pRNA-3WJ motif (PDB 4KZ2),⁴² the naturally preserved angle between the helices H1 and H2 (denoted ∠AOB angle) is ~60° (Figure 8A,B). It was recently demonstrated that this ∠AOB angle is highly tunable and can be stretched to form different 2D polygonal structures: triangle (∠AOB = 60°), square (∠AOB = 90°), and pentagon (∠AOB = 108°).¹ Each of the polygons were designed using the pRNA-3WJ motif at the vertices, with the ∠AOB angle located at the interior of the polygons. The vertices were then connected using RNA duplexes, the length of which controlled the overall adopted geometry of the polygons (Figure 8C). RNA nanoparticles self-assembled upon mixing the component strands of each polygon in a single step with >90% efficiency, as demonstrated in gel shift assays. The polygons adopted a planar conformation as shown in the schematic and validated by AFM imaging (Figure 8D,E). The pRNA-3WJ has demonstrated its ability to stretch further into 3D shapes showing ultrahigh stability and flexibility of the core motif (Figure 2E).

Figure 8.

Tunable shape of RNA nanoparticles. (A) Secondary structure of the pRNA-3WJ shown with Leontis–Westhof nomenclature. Three distinct angles of 60°, 120°, and 180° are formed. (B) The 60° angle AOB was used in construction of multishaped nanoparticles. Side view of the pRNA-3WJ shows the flat structure of the pRNA-3WJ. (C) Triangle, square, and pentagon shaped nanoparticles were constructed using the inner 60° angle AOB of the pRNA-3WJ. Increasing the number of short strands (48 nt) and increasing the length of the internal strand (66 to 88 to 110 nt) induces a geometry shift from triangle to square to pentagon, stretching the 3WJ inner angle from its native state of 60° to 90°(square) to 108° (pentagon). (D) 3D models of the triangle, square, and pentagon nanoparticles. Located at each vertex of the nanoparticles is the pRNA-3WJ. Stable nanoparticles can be assembled even when stretching the native angle to almost double its native state. (E) AFM images corresponding to the above 3D models. AFM images confirm the presence of the respective geometries. Adapted with permission from ref 1. Copyright 2014 Oxford University Press under CC By 4.0 https://creativecommons.org/licenses/by/4.0/.

Detailed thermodynamic analyses were carried out to determine whether the stretching of the ∠AOB angle adversely affected the biophysical properties of these polygons. Equilibrium dissociation constant (K_D) extracted from temperature gradient gel electrophoresis (TGGE) assays revealed values of 18.8, 20.3, and 22.5 nM for triangle, square, and pentagon, respectively. Melting assays indicated a T_M values of 56, 53, and 50 °C, respectively, for triangle, square and pentagons.^{1,14,42,44,50,118} The results demonstrated that the pRNA-3WJ motif is highly dynamic and the helical angles can be tuned to fold into specific geometric arrangements while retaining favorable thermodynamic properties and genuine functionalities.

While the pRNA-3WJ has been studied in detail and used extensively for nanoarchitecture construction, other RNA tertiary motifs have also been used for the same “architectonic” approach to construct RNA nanoparticles.^{52,91,92,194,195} The HIV kissing loop structural motif has been used to construct tectoRNA nanoparticles in the shapes of squares.^52,93 Further organization of these squares led to the assembly of RNA square array structures. Additionally, these same tecto squares were constructed using tRNA-5WJ and UA_h 3WJ motifs.^2,196 Expanding dimensionality using these structures, the two-dimensional construction was translated to three-dimensional, resulting in cubic structures based on these tecto squares.^92,93,197 Other examples include the construction of RNA nanorings based on the RNAI/IIi kissing loop. The assembly angles of this structural motif makes it ideal for hexagonal shaped nanoparticles. These hexagonal nanorings have been functionalized with siRNA and used for gene therapy.^3,4 It is, however, important to note that many factors such as sequence, local structural environment, and metal ion concentrations can significantly influence RNA elasticity. The balance of the ion-mediated electrostatic force and the nonelectrostatic (chemical) forces determine the equilibrium structure of the RNA. Furthermore, the stiffness of the structure directly impacts folding stability and kinetics. This is exemplified by small-angle X-ray scattering (SAXS) experiment for the P4–P6 domain of Tetrahymena thermophile ribozyme, which showed that a rigid J5/5a hinge causes a higher energy barrier and slow folding rate for the formation of the P4–P6 tertiary structure.¹⁹⁸

There are many examples of RNA nanostructure construction using the tunable nature of RNA motifs. RNA structural motifs have distinct advantages over alternate nanoparticle construction techniques: they fold independently, they are programmable by sequence, they can be arranged to construct nanoparticles, they can be decorated to increase functional uses, while the structural motifs retain their authentic folding and drive assembly of the nanoparticle.

3.2.1. Tuning RNA Nanoparticle Size and Properties.

In nanotechnology, finely tuning size, shape, and physicochemical characteristics of nanoparticles is of paramount importance as these properties can affect the biodistribution, circulation time, toxicity, and delivery efficacy in vivo.^199–203 Many polymer-based nanoparticle systems rely on aggregation to form, which can result in broad size distributions and large supramolecular complexes that cause unwanted toxicity.^110,204 RNA nanoparticles, on the other hand, assemble according to precise canonical and noncanonical base-pairing. This precision in nanoparticle formation allows for complete control over size and shape.^1,44

The ease with which RNA nanoparticle size can be controlled was demonstrated by the construction of multiple variants of square shaped RNA nanoparticles (Figure 9A).⁴⁴ Utilizing the pRNA-3WJ at each corner of the RNA square, different length RNA duplexes were used to fabricate RNA squares of 4.0, 11.2, and 24.9 nm as determined by dynamic light scattering (Figure 9B). AFM imaging also revealed clear size differences among the squares. As hypothesized, based on nearest neighbor thermodynamic parameters, the size difference lead to diverse thermodynamic properties. The elastic and tunable properties of RNA allow multiple nanoparticles to be generated from one motif.

Figure 9.

Tunable size and properties of RNA nanosquares. (A) The size of nanoparticles can be easily increased while retaining the same geometry and folding built around the pRNA-3WJ. Duplex length between pRNA-3WJ units are simply increased or decreased to generate variously sized square nanoparticles. By modifying the core strand during construction, using DNA or 2′-F RNA, the physicochemical properties of the square nanoparticles can be tuned. (B) AFM images corresponding to the designed 5, 10, and 20 nm square nanoparticles. An apparent size increase is observed when moving from small to medium to large square. All three square sizes were mixed and analyzed by AFM (far right). Three distinct size nanoparticles can be seen. Below AFM images are DLS measurements showing an increase in size, as predicted by computer modeling. Adapted with permission from ref 44. Copyright 2014 American Chemical Society.

2′-F modified RNA has been shown to impart higher thermal and chemical stability on RNA nanoparticles without affecting the folding of the nanoparticles.^{1,41,42,44,50} Furthermore, it has been shown that modifying RNA nanostructures with different oligonucleotide stands, DNA, or 2′-F strands leads to diverse and controllable properties.⁴⁴ Control over properties can be applied by simply substituting one strand of RNA during construction with DNA or 2′-F modified RNA. It was shown that the addition of 2′-F RNA as the core strand drastically increased both the melting temperature and the nanoparticle’s resistance in serum (Figure 4B). This is a straightforward method to control the way the RNA nanoparticle reacts to their environment. Not all applications require the same nanoparticle with the same properties, and by implementing the elastic nature of RNA nanoparticles to tune stability, RNA nanoparticles can be generated for diverse applications.

3.3. Application of the Rubbery and Tunable Property of RNA to Generate RNA Architectures with Medical Applications

As a result of the optical tweezer studies in the Guo lab elucidating the rubber-like property of the developed RNA nanoparticles, future in depth studies were completed examining the in vivo behaviors of the nanoparticles.³³ Developed multiway junction RNA nanoparticles were fluorescently labeled and their biodistribution profiles were compared to the same size iron and gold nanoparticles (Figure 10).³³ As a result of the flexibility in size and shape of the nanoparticles, a significantly higher nanoparticle accumulation was seen by the RNA nanoparticles over the metal nanoparticle counterparts when compared to liver and kidney accumulation. This data proved the concept that RNA nanoparticles are able to deform their shape to slip through tiny holes in leaky tumor vasculature to enhance the EPR effect normally seen. Additionally, the clearance from the body of the RNA nanoparticles was examined.³³ Interestingly, high amounts of RNA nanoparticles were found in urine samples of the mice at early time points (<1 h), indicating the RNA nanoparticles were rapidly clearing through the kidneys and preventing damage to healthy tissues. Most notable from the studies was that the 10 nm RNA nanoparticles were able to clear through the kidneys that normally are limited to a 5.5 nm size cutoff. This data further demonstrated the elastic nature of the RNA nanoparticles to deform to squeeze through the kidneys and returning to their original shape and remaining stable in the urine.

Figure 10.

Demonstration of the rubbery properties of RNA nanoparticles by comparing retention times in tumors, kidneys, and livers. (A) The Cy5.5-labeled nanoparticles were detected in ex vivo organs 8 h postinjection in mice bearing KB xenografts. (T, tumor; H, heart; S, spleen; Lu, lung; K, kidney; Li, liver. Color scale: radiant efficiency [p s⁻¹ cm⁻² sr⁻¹]/[μW cm⁻²].) (B) Quantitative analysis of whole body biodistribution to quantify the tumor-to-liver and tumor-to-kidney ratios using images in (A). (C) Quantitative analysis of the biodistribution based on the tumor-to-liver and tumor-to-kidney ratio, as quantified from the homogenized organ sample. Adapted with permission from ref 33. Copyright 2020 American Chemical Society.

4. IN VITRO RNA THERMOSTABILITY

The common misconception of RNA being unstable comes from its susceptibility to enzymatic cleavage. This instability has since been overcome through the incorporation of modified nucleic acids. However, RNA duplexing is relatively thermodynamically stable or thermostable. Thermostability is related to the level of energy or heat that is required to dissociate folded RNA, which can be expressed as the melting temperature (T_m) and is also in direct relationship to the concentration of assembled RNA nanoparticles. Additionally, the presence of metal ions and solution conditions play great roles in the folding of RNAs.²⁰⁵ Because of the base stacking, Watson–Crick base-pairing, noncanonical base-pairing, and loop and bulge region of RNA, allowing for tertiary interactions; RNAs are able to fold into relatively stable complexes. As mentioned above in the Introduction, RNAs have been shown to be more thermostable than its counterpart DNA. The thermodynamic stability of RNA has allowed this field to grow into what it is today, as many RNA nanoparticles have proven to remain thermostable in vivo, allowing for drug therapy applications.

Within the design and structuring of RNA nanoparticles, one must take into consideration the thermodynamics and kinetics to lead to a successful nanoparticle design.^206–209 Thermodynamics governs the assembly of multiple RNA oligo strands into a nanoparticle complex and their assembled stability. Furthermore, looking at the kinetics or speed of interactions between the components of RNA nanoparticles, allows for information to be gained in the stability of RNA nanocomplexes, especially while looking at dissociation rates.^30,210–213 Here we focus and discuss the thermodynamics and kinetics behind the construction of RNA nanoparticles and take an in-depth look at 3WJ-based nanoparticles and how they play into design considerations due to their well-studied properties.

One of the most well-studied RNA nanoparticles in terms of thermodynamic studies is the three way junction (3WJ) as the central core motif within packaging RNA (pRNA) on the phi29 bacteriophage DNA packaging motor and will be discussed here (Figure 8).⁴² The individual strands of the pRNA-3WJ assemble into the complex with very high affinity to each other, in the absence of supporting metal ions nor heating cycles that are typically needed for association of nucleic acid strands. The high yield of formation of the pRNA-3WJ and its stability makes it an ideal candidate to serve as a scaffold for a therapeutic platform.^11,12,15,39

4.1. Energetics in the Formation of RNA Nanoparticles

By nature, RNA nanoparticles are typically composed of several single-stranded, short RNA oligo building blocks that self-assemble into complex 3D nanostructures upon incubation of each of the component strands.^30,41 To construct such RNA nanoparticles, during the design process, sequences must be carefully selected to ensure that association of the multiple RNA strands takes place in a spontaneous manner and results in the production of a stable RNA particle. The assembly of single-stranded RNAs (ssRNA) into nanoparticles is governed and described by thermodynamics or the changes in heat and energies of the ssRNAs to nanoparticles. Thermodynamics allow for the prediction of spontaneous nanoparticle assembly and degree of stability the ssRNAs to stay assembled as nanoparticle complexes. Gibbs free energy, a central energy calculation in thermodynamics, known as available energy, is used to understand the complex’s stability while looking at the change in energy over a reaction. Assembly of a nanoparticle is controlled by Gibbs free energy and is composed of enthalpy and entropy.⁴¹

4.1.1. Gibbs Free Energy.

As stated, Gibbs free energy is the available work or available energy of a system and is comprised of enthalpy and entropy while being temperature dependent. Changes in Gibbs free energy (ΔG) combines the changes in enthalpy (ΔH) and entropy (ΔS) as described in the equation below:

$$\Delta G^{°}=\Delta H^{°}-T\Delta S^{°}$$ (2)

where T is the temperature of the reaction.⁴¹ Combining enthalpy and entropy at a given temperature of the reaction provides a ΔG° in which the spontaneity of the reaction is determined by its value. If a change in free energy is negative, the chemical reaction, or in this case the assembly of RNA nanoparticles, is spontaneous, while a positive value does not spontaneously occur and requires input of energy for the assembly. Thus, it is important to ensure during the design of RNA nanoparticles, they assemble in a spontaneous matter so that the nanoparticle remains folded and stable during future applications. Additionally, the more negative the ΔG of the formation of a nanoparticle, the more readily it will fold and indicates a stable nanoparticle, as the dissociation constant of the nanoparticle can be calculated from the change in Gibbs free energy. While we discuss the calculation and modeling of energy levels of RNA folding below, an increase in Watson–Crick base-pairing in a structure will lead to a more stable structure that folds more spontaneously. Addition stability can be gained by increasing the guanine and cytosine (G:C) contents of the strands.

It is important to note the inverse relationship of enthalpy and entropy in relationship to ΔG. This means that a release of heat energy (−ΔH) and an increase in disorder (+ΔS) are beneficial to a spontaneous reaction. Generally, the formation of RNA structures goes from a disordered to ordered system, resulting in a −ΔS. Thus, spontaneity can be created if the change in enthalpy outweighs a negative change in entropy. The assembly of dsRNA is generally a spontaneous reaction that relies on the release of internal energy of ssRNA to form a much more energy conserving dsRNA structure (Figure 11). Unpaired RNA will fold upon itself forming short dsRNA regions with loops and bulges; however, base-pairing with a second RNA strand conserves energy, thus creating a negative change in enthalpy. The entropically unfavored process of going to a more structured system during RNA folding is typically far outweighed by the larger negative change in enthalpy during assembly.^41,214–218

Figure 11.

Example of heat energy required for RNA nanoparticle assembly. RNA strands preferentially folding into the lowest energy conformation (low enthalpy). RNA strands are initially heated to remove structuring and slowly cooled, allowing for the formation of RNA nanoparticles. The folding of RNA is an exothermic process (−ΔH), releasing heat until reaching the most stable complex. Adapted with permission from ref 41. Copyright 2014 American Chemical Society.

4.1.2. Enthalpy in Nanoparticle Assembly.

Enthalpy, a state function, defined as the heat energy of a system, or in this case an RNA molecule or RNA structure. Heat is the combination of the internal energy with the system’s pressure and volume; when considering RNA nanoparticle assembly, the volume and pressure are held constant, thus we can consider enthalpy the internal energy of the RNA itself. Changes in enthalpy results in either the release or absorption of heat during a reaction, thus is often referred to as the heat of a reaction or heat of formation. The formation of RNA nanoparticles consists of ssRNAs base-pairing into complex structures and motifs.^30,41 The structuring of ssRNA generally is much less favorable and not nearly as stable as double-stranded RNA (dsRNA), as ssRNA has some of self-folding and base-pairing but is incomplete in structuring compared to dsRNA (Figure 11).^{30,41,217,219,220} These less favorable structures hold higher energy levels compared to complete Watson–Crick base-pairing; therefore, the assembly of RNA nanoparticles releases the internal energy as heat resulting in a negative enthalpy (−ΔH). Furthermore, this release of heat during a reaction or process is known as an exothermic reaction,²¹⁷ while the denaturing of helical RNA or folded complexes requires energy input and is known as an endothermic process.

4.1.3. Entropy in Nanoparticle Assembly.

Entropy is more difficult to visualize and comprehend than the previously discussed enthalpy and is described as the level of disorder or randomness of a system or molecule. The entropy of a system can be calculated by looking at the number of the same species of particles in different conformations at any given time;^213,221 this is a difficult task, but looking at the change in structuring over time during a reaction or assembly of RNA nanoparticles is more feasible. A change in level of disorder of a system is described by ΔS, where a positive value is a result of an increase in the disorder or randomness, and a negative ΔS indicates a change to more ordered structuring. It is important to note that while the entropy of a system can be negative and return to being more ordered, the entropy of the universe is constantly increasing and moving to a more disordered state. In looking at the assembly of RNA nanoparticles, moving from ssRNA to primarily dsRNA structuring results in a decrease in disorder, as ssRNAs are capable of self-folding into many unstable conformations, whereas RNA nanoparticles fold into predicted and stable structures. This results in a negative change in entropy.

4.1.4. Prediction and Calculation of Thermodynamic Parameters of RNAs.

Understanding and predicting thermodynamic parameters is a powerful tool in RNA nanoparticle design to ensure that resulting motifs and structures from multiple RNA strands fold easily while remaining stable. Significant research has been completed in understanding sequences of RNA strands and their base-pairing to complement sequences to calculate the thermodynamic contributions to RNA folding for sequence predictions. Thermodynamic parameters can be calculated via a van’t Hoff plot, in which the concentration of materials are varied while examining the difference in melting temperatures (T_m) of the duplexes (Figure 12).^41,214,222 The melting temperature of a duplexed nucleic acid structure is defined as the temperature in which 50% of the species exists as a duplex, while 50% exist as single strands.^41,42 The T_m of nucleic acids is dependent on strand concentrations and can be created into a linear relationship by plotting the natural log (ln) of concentration versus the inverse of T_m(1/T_m). As a result, the slope of the van’t Hoff plot is related to enthalpy and the y-intercept is related to the entropy and enthalpy as described in the equation below for a nonself complementary duplex:

$$\frac{1}{T_{\text{m}}}=\frac{R}{\Delta H^{°}}\ln C_t+\frac{\Delta S^{°}-R\ln 4}{\Delta H^{°}}$$ (3)

where R is the universal gas constant and C_t is the concentration of RNAs.

Figure 12.

Calculation of thermodynamic parameter for 3WJs formation. (A) Representation of temperature gradient gels to calculate T_m of each 3WJ. (B) Melting temperature profiles of RNA, 2′-F RNA, and DNA 3WJs by various concentrations. (C) Plots of T_m vs 3WJ concentrations (van’t Hoff analysis) to calculate thermodynamic parameters. Reprinted with permission from ref 41. Copyright 2014 American Chemical Society.

Studies were originally completed on DNA sequences with strands including single-nucleotide mutations and found that thermodynamic parameters rely on what is known as nearest neighbor laws, in which the thermodynamics are affected by the two consecutive nucleotides in a sequence rather than a single nucleotide.^223,224 These pioneering studies in DNA were then further extended to RNA^214,215,217 and DNA/RNA hybrids.^222,225 These studies allow for one to calculate by hand the thermodynamic parameters along with melting temperature for any given duplex structure and aid in the sequence design of RNA nanoparticles. However, the nature of folding of RNA with Watson–Crick base-pairing, noncanonical base-pairing, and base stacking creates multiway junctions, bulges, and loops, all of which play into the thermodynamic parameters. As such, Mathews et al. was able to experimentally derive these parameters for three-way and four-way junctions.²¹⁸ More recently, several additional studies have been published on the folding energies behind RNA and DNA sequences.^{209,226–228}

4.2. Thermodynamic Stability of 3WJ RNA Nanoparticles

4.2.1. Stability of the pRNA-3WJ.

RNA nanoparticles applied to in vivo applications require a high thermodynamic stability to ensure the nanoparticle does not dissociate into individual building blocks in harsh conditions and by dilution within the body during circulation. The previous studies described above looked at the folding of RNA duplexes and junctions based on single- and two-stranded systems. However, RNA nanoparticles are typically composed of several RNA oligoes and are more complex in structure than dsRNA used in many of the studies above.²⁷ Looking at the thermodynamics of a multistranded systems brings a level of complexity that is difficult to study; however, Binzel et al., used the pRNA-3WJ from the phi29 DNA packaging motor (Figure 13A) as a model system in identifying the thermodynamic stability of three short RNA strands folding together.⁴¹

Figure 13.

Overview of the Phi29 pRNA and the three-way junction (3WJ). (A) (i) Secondary structure of the phi29 pRNA monomer with the pRNA-3WJ outlined by the box, which connects the helical domain to the interlocking procapsid binding domains. (ii) Secondary structure and sequence of the pRNA-3WJ and the (iii) crystal structure of the pRNA-3WJ. (B) Assembly gel of the pRNA-3WJ from the three short RNA oligo strands showing efficiency in folding and particle homogeneity. (C) Melting profiles demonstrate the 3WJ shows the highest melting temperature (T_m) over any of its components of dimer species. (D) Thermostability comparison of three 3WJ with chemical modifications. Reproduced with permission from ref 41. Copyright 2014 American Chemical Society. Reproduced with permission from ref 30. Copyright 2016 Binzel et al. under CC By-NC 4.0 https://creativecommons.org/licenses/by-nc/4.0/.

The pRNA-3WJ proved to be stable at high temperatures as the T_m was found to be nearly 60 °C at 10 μM (Figure 13C,D).⁴¹ This high melting temperature demonstrated high cooperation between the three short oligos to form three separate helical regions joined by the central junction. Additionally, the assembly and melting profiles showed a single transition from ssRNA to pRNA-3WJ, hinting that all three strands assemble without any sign of dimer intermediate formation. The 3WJ formed with the highest stability over any of the dimer species with the highest melting temperature (Figure 13C). Thermodynamic parameters were calculated through a van’t Hoff plot as described above, resulting in elucidating a significantly negative ΔG (Figure 12 and Table 2).⁴¹ It is interesting to note that the RNA and 2′-F chemically modified 3WJs are more thermodynamically stable than the DNA 3WJ. This increased stability is seen in a more favorable (less negative) entropy, most likely due to the less rigid structuring of RNA over DNA to allow for flexibility in the RNA and slight structure variances. This results in a higher level of disorder. Finally, the pRNA-3WJ has been shown to remain stable at concentrations as low as 160 pM concentrations as assay by PAGE and radiolabeling.⁴² These results indicate the pRNA-3WJ is able to form into stable complexes suitable for in vivo applications.

Table 2.

Thermodynamic Parameters for 3WJs Formations^a

3WJs	1/Tm vs log (Ct)
	ΔG°37 (kcal/mol)	ΔH°37 (kcal/mol)	ΔS°37 (eu)	Tmb (°C)
2′-F RNA	−36 ± 0.45	−200 ± 5.7	−520 ± 17	72.1
RNA	−27 ± 0.58	−170 ± 13	−440 ± 39	60.4
DNA	−15 ± 0.71	−220 ± 25	−650 ± 83	35.2

^aParameters derived from 15% native TGGE. Reproduced with permission from ref 41. Copyright 2014 American Chemical Society.

^bT_m values for 3WJ strand concentrations of 10⁻⁶ M

4.2.2. Effects of Chemical Modifications on the Stability of RNA and 3WJ Nanoparticles.

There has long been concerns of the thermodynamic and enzymatic stability of RNA and RNA nanoparticles preventing the translation to the clinic. As shown above, 3WJ nanoparticles have overcome the thermodynamic hurdle, as their high melting temperature and low ΔG indicated high stability even at ultralow concentrations. While RNA is susceptible to RNases found within the body, it has been found that by modifying the backbone of nucleotides at the 2′ carbon location changes the conformation of the nucleotide and prevents RNase identification.^{94,165,229–231} Commonly used modifications to add enzymatic stability to RNAs include 2′-F and 2′-O-methyl and modified RNAs remain stability in serums for over 24 h, while natural RNA is degraded within the first 30 min.^94,232

Additionally, the 2′ chemical modifications to RNA have been found to further alter their thermodynamic stabilities. In looking at the pRNA-3WJ chemical modifications, we have significantly increased the melting temperatures. 2′-F modifications on pyrimidine nucleotides raised the melting temperature of the 3WJ from ~60 to ~70 °C (Figure 13D),^41,42,46 while substituting to locked nucleic acids (LNA) above 80 °C,⁴⁶ the limit of detection using temperature gradient gel electrophoresis. These results were further confirmed by examining the thermodynamic stabilities of RNA/2′-F and RNA/LNA hybrids. As the chemical modification content increased from one to three strands, the melting temperatures of the three-way junctions continued to increase. Additionally, the Gibbs free energy of formation for 2′-F pRNA-3WJ was found to be more negative than the native pRNA-3WJ, indicating a higher level of stability and of preferential folding.⁴¹ The high stability of the LNA modifications and the pRNA-3WJ allowed for the nanoparticles to remain stable within the body, as complete 3WJ assemblies were purified from urine of mice after being intravenously administered.⁴⁶ Alternatively, the thermodynamic stability was decreased as the DNA composition was increased within the pRNA-3WJs.^41,46

4.3. Assemblies of RNA Three-way Junction (3WJs) Driven by Entropy and Enthalpy

The pRNA-3WJ is composed of three short RNA oligos that coassemble into a junction. Initial assembly studies of the pRNA-3WJ led to the belief that all three RNA strands coassociate at the same time, as there was no evidence of a dimer species forming.^41,42 While it is highly rare for three components to associate together at the same time, the high speed and ease of formation of the pRNA-3WJ with its strong stability further supported the idea that three components strands associated together at the same time. Through assembly gels and assembly profiles elucidating melting temperatures showed a high yield of three-way junction formation with no sign of dimer intermediates formation as shown in a single transition.^41,42

With this hypothesis, Binzel et al., completed kinetic studies of the three component strands associating into the 3WJ by surface plasmon resonance (SPR).³⁰ It was found that the pRNA-3WJ follows a two-step reaction of association in which a dimer species is very temporarily formed. During the association of each of the three dimer associations, it was found that the 3WJ_bc dimer formed at a more rapid rate than the modeled 3WJ itself and the other two dimers (Figure 14). However, this dimer intermediate is normally not observed, as immediately upon its formation the third strand, 3WJ_a, interacts with the 3WJ_bc folding into the 3WJ. This is due to the dimer formation increasing the number of nucleotides available for the 3WJ_a to bind, thus significantly increasing its affinity to the forming 3WJ. This was evidenced by the fact that the second association step occurring at a rate too rapid to be observed by SPR due to a limit of detectable concentrations. This led to the development of a pseudo-one step association model of the pRNA-3WJ in which a three-way junction is immediately formed upon folding of the dimer species as described below and in Figure 15.³⁰

$$3\text{W}{\text{J}}_{\text{a}}+3\text{W}{\text{J}}_{\text{b}}+3\text{W}{\text{J}}_{\text{c}}\leftrightarrow 3{\text{WJ}}_{\text{a}}+3\text{W}{\text{J}}_{\text{bc}}\leftrightarrow 3\text{WJ}$$ (4)

Interestingly, from the SPR studies, it was found that the pRNA-3WJ stability and formation benefited from being composed of three strands. In kinetics, a fast association of a complex typically leads to a rapid dissociation, while the opposite is true of a slow reaction resulting in a slow dissociation is typically true within RNA. However, the pRNA-3WJ combines the fast association of the 3WJ_b and the 3WJ_c strands into a dimer, but the slow dissociation of the 3WJ_ab dimer. This results in a multistranded RNA motif that associated rapidly into a stable complex that is slow to dissociate. As one helical region of the 3WJ dissociates, the other two branches hold the structure together instead of fully unzipping and quickly refolds due to a high local concentration.³⁰

Figure 14.

Surface plasmon resonance (SPR) of pRNA-3WJ dimers. Dimer species that make up the pRNA-3WJ were examined through SPR at concentrations ranging from 20 μM to 78 nM (A–C). Dimers were formed on the chip for the first 660 s (association phase) followed by a dissociation phase until 2700 s. The 3WJ_ab dimer (A) shows the slowest association and dissociation, while the 3WJ_bc(B) shows the fastest formation and dissociation and the 3WJ_ac mix of strong formation and slow dissociation. Data was fit using a two-component pseudo-first-order Langmuir model. Reproduced with permission from ref 30. Copyright 2016 Binzel et al. under CC By-NC 4.0 https://creativecommons.org/licenses/by-nc/4.0/.

Figure 15.

Assembly mechanism of pRNA-3WJ. The pRNA-3WJ assembles through a two-step association mechanism in which the 3WJ_bc dimer first forms as shown in the first panel. This association step greatly increases the attraction of the 3WJ_a strand by doubling the nucleotides presented for binding resulting in a dropped ΔG°. This second reaction occurs at an unobservable rate and results in forming a stable nanostructure, in which each of the three strands are locked into the structure by two areas of base-pairing. Adapted from with permission from ref 30. Copyright 2016 Binzel et al. under CC By-NC 4.0 https://creativecommons.org/licenses/by-nc/4.0/.

In looking at the structuring of the ssRNA components of the pRNA three-way junction, they are of high entropy or high levels of disorder, as the 3WJ_a and 3WJ_c strands generally do not stain by ethidium bromide intercalating dye. This shows that the strands do not have defined structures and are in a transient state. When the three strands assemble together into the 3WJ, a change is seen to a defined structure and distinct band on the gel, indicating an unfavorable decrease in entropy. The formation of the 3WJ_bc dimer presents a defined motif for the binding of the 3WJ_a strand to bind to further demonstrating this decrease in entropy. However, the formation of the three-way junction produces a stable structure that has low internal energy, thus the release of enthalpy is able to overcome the unfavorable change in entropy. However, as mentioned above, the branched structuring of the pRNA-3WJ allows for flexibility between each of the helical regions as well as strand breathing. These slight structure variations allow for a higher level of entropy over dsRNA further improving the stability.

5. RNA STRUCTURAL MODELING AND COMPUTATION

The continuous progress in RNA field is not limited to its structure characterization, biophysical studies, or RNA probing. RNA structure prediction is very important to understand the physical mechanism of RNA functions and to design RNA-based therapies. The RNA structure prediction includes primary sequence, secondary structures, and three-dimensional (3D) structures. The recent progress in RNA folding theories and different predictive models have provided new avenues to understand the mechanistic views of RNA function and RNA motif sequence design that have been utilized in RNA nanotechnology. However, further advancement in these models likely require a fusion of physics or experimental data-directing models that require further development.

5.1. Modeling of RNA Energy Levels

Using calculated thermodynamic parameters of RNA and DNA, secondary structure modeling software has become available to predict base-pairing within single-stranded and double-stranded nucleic acids. This software allow researchers to predict folding and structuring of RNA nanoparticles for provided sequences as well as predict the thermodynamic stabilities. mFold, a public server, provides DNA and RNA secondary structure and base-pairing modeling that allows users to see the most energy preferred folding of an RNA strand along with the predicted folding energies and stabilities.²³³ Mathews and Turner have created predictions of secondary structures by minimizing the free energies following the methodology above.²³⁴ Additionally, UNAfold was developed to create and predict secondary structuring between two RNA strands building upon the development of mFold.²³⁵

Using only the RNA sequence significantly limits the capacity to predict and analyze the higher-order RNA structures and features. To accurately generate RNA structure models, quantitative mapping of thermodynamic parameters of RNA energetics is necessary. Through the use of entropy and enthalpy, RNA-based therapeutics for a number of disease states have been developed. One of the key parameters to the construct of RNA structure models is the nearest neighbor or Turner rule that states that RNA secondary stability can be predicted using the quantification of melting temperature of nucleic acid duplexes. However, many additional thermodynamics parameters are still required for large RNA species. The nearest neighbor model described above in section 4.1.4 is built on two principles: (1) that nucleotides present within motifs and their nearest neighbor control the free energy of secondary structure, and (2) that non-nearest neighbor bases make no contribution to the stability of these motifs. These results have created a promising updated model for anticipating the thermodynamic parameters for the stability of hairpin stem–loop structures.²³⁶ This approach has significantly contributed to our understanding of RNA structure and stability, however, additional quantitative approaches are still required. The RNA-MaP platform is one of these sources that measures the equilibrium binding for different RNAs. RNA-Map provides the thermodynamic models that quantitatively predict the RNA-PUM1/2 interactions as well as cellular PUM2-RNA occupancies.²³⁷ Statistical mechanics-based computational methods that can predict the thermodynamics stability based on entropy changes during loop–loop interactions in RNA complexes have also been developed. This method is not limited to prediction of structure for native RNA forms and can also incorporate alternative metastable structures. This approach has recently been used to predict the two binding sites for hTR dimerization.¹⁵⁵ Determining the cotranscriptional folding kinetics of riboswitches is an important problem and complex issue to use. The development of a new method that uses a helix-based transition rate model and increases the efficiency of conformational sampling. This method has several major RNA applications in understanding population kinetics, time extension curve, and kinetics pathways in different time scales.²³⁸

Ion binding is a key consideration for RNA folding, structure, and stability.²⁰⁵ Because of the strong electric field around the negatively charged RNA backbone, ion binding can be highly significant for RNA folding. Additionally, ions distribution around an RNA can fluctuate. The ion correlation and fluctuation effects are particularly important for multivalent ions such as magnesium ions around RNAs. TBI^239,240 and MCTBI^241,242 models were created to take into account the ion correlation and fluctuation effects in ion–RNA interactions. These events are sensitive and often specific for ion species, however, recent work suggests that adenine riboswitches can compete for different ion species.²⁴³ In contrast, very specific ion binding species like the Mg²⁺ ion are found in kissing loops of the adenine riboswitch. These models along with others like ligand binding models have increased our understanding of the RNA field for various purposes like RNA structure determination, RNA-RBP interaction, as well as RNA-based drug discovery. More work is still required to map and accurately predict RNA folding in vivo. In particular, addressing RNA folding as a cotranscriptional event and measuring the kinetics of these reactions represents a significant challenge for all predictive models.

5.2. 2D/3D Modeling/Prediction of Desired RNA Nanoparticles

Structural predictions of small RNAs has helped understand RNA dynamics, however, the issues arise with this approach when complex RNAs like multibranched loops and noncanonical interactions are considered. Knowledge-based methods have circumvented some of these bottlenecks, and a number of excellent databases now provide accurate information on RNA including whole structures, helices, pseudoknot loops, internal bulge loops, hairpin, junctions, as well as several other fragments.^244–246 Further advancements to measuring or predicting RNA structures have been provided from physics-based methods including energy and scoring functions.

There are many programs for predicting RNA 3D structures, like NAST,²⁴⁷ BARNACLE,²⁴⁸ FARFAR utilizing AMBER/CHARMM-like force field methods to predict the RNA structure even for noncanonical forms,^249,250 IsRNA,^251,252 SimRNA,^253,254 and others. NAST, IsRNA, and SimRNA are molecular dynamics simulation tools consisting of a knowledge-based statistical potential function applied to a coarse-grained modeling. NAST requires secondary structure information and accepts tertiary contacts to direct the folding. Its greatest strength is to allow modeling of large RNA molecules. When only the secondary structure information is considered, accurate prediction is limited, but when input information on tertiary contacts, prediction accuracy can be dramatically improved. BARNACLE generates reasonable RNA-like structures using secondary structure information for small RNA molecules (<50 nt) but not for longer than 50 nt due to an increase in complexity of the probabilistic model. BARNACLE allows for efficient sampling of RNA conformations in continuous space and with related probabilities. FARFAR is only applicable to small RNA and has variable accuracy relying on the framework found successful for proteins using atomistic models and empirical potential functions.

There are also other programs that are used to predict RNA structures. RNAcomposer treats secondary structural motifs in a more detailed way by using slightly distorted helical conformations and tertiary contacts in secondary structure motifs.²⁵⁵ CovaRna and CovStat were developed to explore long-range covarying RNA interaction networks using whole genome alignments.²⁵⁶ MC-sym builds all-atom structures using the 3D version of the nucleotide cyclic motif (NCM) fragments which are not only extracted from known RNA 3D structures but also built on the fly if necessary, but this method is limited to short RNAs requiring 2D structures as input due to the limited NCM fragments for large, complex NCM motifs, such as six-way junctions and kissing loops. There are several programs that combine 2D and 3D analyses together, such as iFoldRNA, Vfold, Nanotiler (Figure 16), RNA2D3D, and FR3D.^31,32,193 iFoldRNA uses 3 three-bead models to predict RNA 3D structures by applying discrete molecular dynamics simulations.²⁵⁷ The usage of coarse-grained representation and DMD approach significantly reduces the computational complexity and enhances the conformational sampling. It does not require secondary structure information and can rapidly predict structures for small RNAs. Vfold is a model used to predict RNA 2D and 3D structures, and the folding stability from the sequence using coarse-grained RNA conformational models and novel statistical mechanics-based model and motif-based template library modeling, respectively.^{87,88,258–261} The model distinguishes itself from other models by two unique features: physics-based modeling of conformational entropy for 2D structure prediction, and template-based multiscale modeling for 3D structure prediction. Vfold3D is a template-based coarse-grained structure prediction model, which can divide 2D structure into different motifs and assemble the 3D motif templates from known RNA structures to predict 3D structures.^260,261 VfoldLA is a new model for the assembly of RNA 3D structures using loop templates.²⁶² The model divides a 2D structure into helices and single-stranded loops and classifies the loops into four different types. The model assembles RNA 3D structures from loop templates according to the loop–helix connections. FR3D is a server for finding and superimposing RNA 3D motifs.¹⁹³ FR3D Align is a global pairwise alignment of RNA-3D structure using local superpositions.²⁶³ JAR3D is predicting RNA 3D motifs in sequences. R3D-2-MSA server is for accessing alignments from 3D structures. ASSEMBLE is another automated program using secondary or tertiary structure information from homologous RNAs to build a first-order approximation RNA 3D model. The computational modeling of RNA folding as transferred from predictive models to applications. RNA models have enabled successful design of RNA aptamers for potential therapeutic applications.^264,265 Furthermore, developed models have also allowed for insights into RNA folding and kinetic pathways for conformational switches.^266,267 While most RNA structure prediction softwares are for single-stranded RNAs, Cartaj²⁶⁸ allows for the prediction of RNA structures from multiple strands. Cartaj was developed to predict the structuring and characterize of 3WJs from inputted sequences, which is exceptionally useful in RNA nanoparticle design. Additionally, computation modeling tools have been generated to predict the docking and interactions of RNAs to ligands and small chemicals including RLDOCK,²⁶⁹ DOCK 6,²⁷⁰ and AnnapuRNA.^271,272

Figure 16.

Modeling of a highly stable functionalized truncated tetrahedral RNA nanostructure. (A) Hexagonal ring from which structure was built.^3,4 (B) Connecting H-shaped motif constructed from two 3-way junctions (derived from E. coli ribosome, PDB 4V4Q). (C) Secondary structure depiction showing connectivities of rings to each other using H-shaped motif. (D) Final 3D model, superimposed with cryo-EM envelope, of truncated-tetrahedron functionalized with 12 Dicer substrates, built in part using software NanoTiler.^31,32 Secondary structure depiction was generated by the software RiboSketch.⁵¹ Adapted with permission from ref 56. Copyright 2020 The Royal Society of Chemistry.

5.3. RNA Motif Sequence Design/Prediction

Building databases of the vast structures and motifs of RNAs found in nature for the structure prediction discussed above has led to the creation of computational programs and tools to allow for the sequence design to create novel RNA motifs and interaction. These tools are used for during RNA nanoparticle design in developing sequences to compose the nanoparticles. RASP,²⁷³ RAG (RNAAs-Graphs),²⁷⁴ and GraPPLE²⁷⁵ are computational programs to design single-stranded RNA motifs. RAG is an RNA topology resource developed by the Schlick group and used to classify/analyze topological characteristics of existing RNAs and design and predict novel RNA motifs. GraPPLE applies secondary graph representations to identify and classify noncoding RNAs from sequences by using graph properties that capture structural features of functional RNAs. Additionally, multistranded RNA motifs can be designed using RNAJunction and NanoFolder.^219,276 RNAJunction is a database containing structure and sequence information for known RNA helical junctions and kissing loop interactions.²⁷⁶ NanoFolder is another web-based software tool for RNA nanostructure by predicting the structure and sequence attributes of multistranded RNA constructs.²¹⁹

5.4. Folding Predictions vs Structural Building Blocks

Nowadays, scientists have discovered various functions of RNA, such as mechanisms of protein synthesis, enzymatic activities, gene regulation, and targeting. RNA functions highlight the biological significance of RNA folding because the function is determined by the structure. Hence, it is of great need to have a predictive model for RNA folding.

The recent development in statistical mechanical modeling of RNA folding has led to success in predicting RNA structures, including folding stabilities and folding kinetics for structures with increasing complexity. However, there are still several key issues unsolved in this field. Some examples are the computation of the entropy for RNA tertiary folds and the extraction of the energy/entropy parameters for noncanonical tertiary interactions from thermodynamic data and known structures. The early computational methods for RNA pseudoknot prediction were based on the genetic algorithm. These methods, such as the STAR and the MPGAfold models, predict the structures with the optimal free energy of the molecule, and in several instances result in the production of the optimal free energy structure.^277–280 The structural building blocks approach involves several steps. First, the fragment length should be decided. Then, a set of training data should be given. All possible fragments of the specified length should be compiled to the whole set. Next, a clustering method is used to divide these fragments into clusters and pick up the center of each cluster to be a building block. If these building blocks are good enough, they can be used to represent all original fragments within a tolerable limit and reconstruct the 3D structure of a whole RNA within some tolerance.

6. IN VITRO RNA STABILITY AND CHEMICAL MODIFICATION

The ever-growing utilization of RNA for therapeutic treatments has necessitated the discovery of novel methods to protect this RNA from degradation. As discussed in section 6.2 (see below), RNA exposed to cellular conditions is rapidly compromised by a number of mechanisms that are essential for preventing viral infection, cellular transformation and aberrant immune responses. Determining the rules regulating the stability of RNA generated in vitro and incorporating these lessons into the design of RNA nanoparticles, medicines, or vaccines is essential for the advancement of RNA nanoparticles as therapeutics.

6.1. RNA Structure

The additional 2′-OH side chain on RNA provides extra diversity to the structure and geometry of RNA molecules. Additionally, RNA modifications and noncanonical base-pairing further enhances this diversity. In particular, uracil nucleotides can form multiple noncanonical base-pairing arrangements in which Watson–Crick U-A base-pairing is replaced with U-G.¹³⁶ The utilization of RNA geometry has enabled the generation of noncanonical or unnatural structures that have unique biological properties and can be used for designing therapeutics. Examples include: three-way junctions (3WJ),^18,42 four-way junctions (4WJ),^5,14,118 kink turn (helix–internal loop–helix motif with a 50° bend in the helical axis),¹⁴⁷ hairpins, and kissing loops.^4,197,276 Each of these formations have been used to construct stable RNA nano structures of different sizes, shapes, and payload capacities. In particular, the 3WJ core of the phi29 motor RNA is highly thermostable and provides the ideal platform for the development of therapeutic RNA nanoparticles.⁴² To deliver intact in vitro RNAs, including nanomedicines or vaccines to target cells, three key stability factors must be considered: thermal (Thermo) (discussed above), enzymatic, and chemical stability. Collectively, these independent concepts directly contribute to the half-life and overall stability of in vitro RNA species. In the next sections, we will discuss how each of these processes affects RNA nanoparticles.

6.2. Enzymatic Stability

RNA is sensitive to enzymatic digestion, and this represents an important protective mechanism utilized by organisms to prevent viral infection. In particular, human body fluids, including blood and saliva, contain very high levels of hydrolytic enzymes that target and degrade exposed RNA. Protecting therapeutic RNAs from these enzymes and environments has been a key driver in the development of RNA nanoparticles and nanomedicine. A number of strategies have been established that mitigate the RNase-mediated degradation. The most prominent include the switching of reactive 2′-OH groups from pyrimidine bases to RNase-resistant chemical moieties: 2′-F, 2′-NH₂, and more commonly, 2′-O-methyl (Figure 17). This relatively minor modification removes the 2′-OH utilized by RNases for RNA turnover and replaces them with highly similar structures that are incompatible with RNase activity. The chemical modification of RNA has become well studied, and 2′-F modifications have been shown to maintain RNA structuring with a C3′ endo conformation with an A-form helix.²⁸¹ This removal of the reactive 2′-OH group provides protection from nuclease and base degradation. It has been shown that strategic chemical modification to siRNAs has provided stability while not affecting the ability of the siRNA to incorporate into RISC for mRNA recognition.^{95,231,282–284} Additionally, chemical 2′-F modifications have not shown significant change in RNA structuring or ability to fold into high structures.²⁰ Through the inclusion of such modifications, RNAs and RNA nanoparticles are produced with high half-lives against nuclease degradation allowing for in vivo applications.

Figure 17.

Construction of enzymatically stable RNA nanoparticles (A) Urea-PAGE denatured gel demonstrating the stabilities of unmodified pRNA Aa′ and 2′-F-modified pRNA Aa′ after incubation in the presence of RNase A (upper one; 1 mg/mL) and fetal bovine serum (lower one; 10%). Adapted with permission from ref 20. Copyright 2011 American Chemical Society. (B) Serum digestion over time of RNA square nanoparticles and pRNA 3WJ. Reproduced with permission from ref 44. Copyright 2014 American Chemical Society. Reproduced with permission from ref 46. Copyright 2018 Piao et al. under CC By-NC 4.0 https://creativecommons.org/licenses/by-nc/4.0/.

6.3. Chemical Stability

In vitro generated RNA is particularly sensitive to chemical degradation caused by changes in pH and/or chemical agents. Reactive side chains make RNA a highly dynamic and flexible molecule, however, these residues can be utilized to trigger RNA instability. In particular, the 2′-OH groups in RNA sensitize RNA to hydrolysis, as a number of chemical hydrosols can utilize this exposed group to catalyze further reactions. In addition, changes in pH also have significant effects on RNA stability. This is an important consideration, as bodily fluids including blood and saliva have different pH spectra to cells and organs.

6.4. RNA Chemical Modifications to Protect in Vitro Generated RNA

More than 100 different modifications have been discovered in RNA, and the vast majority of these modification mostly are present in structural RNAs including tRNAs and rRNAs. A smaller number have been identified in mRNA and regulate several important aspects of gene expression. Researchers are utilizing these modifications to improve current technologies including CRISPR-Cas9 and antisense oligonucleotide-based drug delivery. The use of nucleotide modification has emerged as one of most suitable methods to improve the stability and efficiency of oligonucleotide drug delivery. Several modifications either to the backbone, ribose sugar, or in the base composition significantly increases the drug like property, as well as the delivery methods.^285,286 Out of the 10 clinically approved oligonucleotide therapies, eight are dependent on nucleotide modifications for drug delivery approaches. In diseases like Huntington’s, where allele specific silencing of mutant HTT and ATXN3 transcripts are required, nucleotide modification such as abasic nucleotides have been an important tool.²⁸⁷ Ribose modifications including 2′-O-methyl (2′-OMe), 2′-O-methoxyethyl (2′-MOE), and 2′-F also increase the resistance of oligonucleotides to degradation, which provides elevated stability, half-lives, and drug efficacy.

Alnylam Pharmaceuticals has utilized nucleotide modification for developing siRNA patterns for its products that have increased their potency more than 500-fold as compared to unmodified siRNA in some cases.²⁸⁸ Other oligonucleotide-based drugs stimulate the immune reaction in a sequence or chemistry dependent manner.²⁸⁹ 2′-OMe modifications at key positions in these oligonucleotide-based drugs can abrogate such immune responses.^290,291 Similarly, 5′-triphosphate and other several other modifications have also been implemented as immuno-stimulatory in oligonucleotide-based therapeutics.^292,293

Recent studies have also utilized nucleotide modification to improve CRISPR-Cas9 efficiency for genome editing applications. The use of 2′-fluoro ribose and full ψ-modification of the single-guide RNA (sgRNA) improved the gene editing efficiency in five different cell lines.²⁹⁴ Others have incorporated 2′,4′-BNANC [N-Me] and locked nucleic acid (LNA) modifications at specific locations in sgRNAs to reduce off-target DNA cleavage by Cas9 in vitro. These findings suggest that nucleotide modification could be key to improving CRISPR-Cas9 platforms. The importance of nucleotide modification is highly amenable to a new of therapeutic technologies including ASOs, siRNAs, and CRISPR based drug platforms. Further understanding and discovery of additional nucleotide modifications could further expand the use of these structures in both experimental and clinical applications.

6.5. Conjugation of Chemicals to RNA and RNA Nanoparticles

RNA nanotechnology has gained greater popularity due to the ability to manipulate RNA at a molecular level by introducing various functional tools in an effort to engineer RNA functions for various nanotechnological applications. The functional tools can range from simple isotopes, fluorescent markers, hydrophobic drug molecules, imaging agents, and other molecular probes to alter RNA’s function that are essential to develop useful RNA nanotechnological applications. The site-specific conjugations can be incorporated into RNA through synthetic chemistry, enzymatic approaches, and post synthetic conjugation chemistries. A robust arsenal of nucleic acid chemistry methods has been developed and applied to the manipulation of RNA structure for decades. The nucleic acid chemists’ tool kits include methods for modification of the base, sugar, phosphate backbone, and termini of oligonucleotide sequences. The synthetic method most amenable to introduce modifications into RNA is solid-phase synthesis, wherein reactive functional groups are directly introduced during the synthesis. The commonly used functional groups for the introduction of chemical drugs, fluorescent markers, and other molecular probes are alkyne, thiol, aldehyde, amine, carboxylic acid, maleimide, NHS ester, and aldehydes in order to engineer RNA nanoparticles for applications in visualization, structural elucidation, localization, diagnosis, and nanomedicine. These functional groups can be modified with drug molecules or imaging markers harboring orthogonal functional groups with appropriate chemistries, they include click chemistry, N-hydroxysuccinimide (NHS) chemistry, thiol chemistry, periodate chemistry, imine formation, and 5′-phosphate activation. Together, these synthetic conjugation chemistries now provide the means to modify virtually any RNA of interest. This section describes the synthetic tools available for the introduction of chemicals into the RNA.

6.5.1. Click Chemistry.

Click chemistries are modular reactions, simple to perform, highly reliable selective organic reactions, high yielding with wide scope in the fields of drug discovery, combinatorial chemistry, target-templated in situ chemistry, nucleic acid research,²⁹⁵ diagnosis, and therapy.²⁹⁶ Click reactions are not disturbed by water, generate minimal and inoffensive byproducts, and are quick while producing an irreversible single product with high reaction specificity.²⁹⁵ Thus, click chemistry allows for the joining of substrates of choice, leading to cementing into applications in various fields ranging from chemistry to biology. Bio-orthogonal click reactions have attracted recent attention for the click chemistry due to the ease of biomolecules labeling under in vivo conditions and enabled the visualization of biomolecular processes, therapeutic cells, tumors, and bacteria. There are many reactions that fit into the concept of click chemistry; however, only a handful of reactions are used widely in the fields of nanotechnology and cancer biology (Figure 18). They include (i) copper-catalyzed azide–alkyne cycloaddition, (ii) strain promoted click chemistry, and (iii) inverse Diels–Alder reaction.

Figure 18.

Chemical modification of RNA nanoparticles by click chemistry. Common chemical modification pathways used in RNA nanoparticle chemistry.

6.5.1.1. Copper-Catalyzed Azide–Akyne Cycloaddition.

Copper-catalyzed azide–alkyne cycloaddition (CuAAC) click chemistry has many advantages for the preparation of 1,2,3-triazole derivatives due to its high reaction efficiency, remarkable chemo- and regioselectivities, and excellent functional group compatibility under mild reaction conditions.²⁹⁷ The reactivity of azide as a 1,3-dipole is used to facilitate a [3 + 2] azide–alkyne cycloaddition to generate a stable 1,2,3-triazole linkage. The reaction rate of the Cu(I)-catalyzed cycloaddition can be approximately seven orders of magnitude higher than that of uncatalyzed cycloaddition and can be further accelerated in the presence of certain ligands for Cu(I).²⁹⁸

6.5.1.2. Strain Promoted Click Chemistry.

Cu-free strain-promoted [3 + 2] azide–alkyne cycloaddition (SPAAC) to form triazole products is achieved by incorporating the alkyne within a strained cyclooctyne system. Although the strain-promoted cycloaddition reaction rates are significantly slower than that of CuAAC, it has improved the biocompatibility. To improve the reaction kinetics, electron-withdrawing fluorine atoms are introduced into cyclooctyne system to obtain difluorinated cyclooctyne (DIFO), the kinetics of DIFO are dramatically increased and are comparable to those of CuAAC in labeling biomolecules.²⁹⁹

6.5.1.3. Inverse Diels–Alder.

The inverse electron demand Diels−Alder (IEDDA) reaction generally capitalizes on the coupling between electron-withdrawing dienes and electron-donating dienophiles. The unique IEDDA transformation, particularly the tetrazine cycloaddition, has been shown to outperform most of the other bio-orthogonal click reactions due to its rapid and tunable reaction kinetics, catalyst-free nature, outstanding orthogonality, superior chemoselectivity, and excellent biocompatibility. Since the discovery of the IEDDA reaction between 1,2,4,5-tetrazines and strained alkenes, the IEDDA toolbox has been significantly expanded to include other reactive dienophiles, such as transcyclooctene, cyclooctyne, cyclopropene, cyclobutene, norbornene, and isonitrile. In particular, the IEDDA reactions based on transcyclooctene have been reported to possess tunable reaction rates. Because of this, the IEDDA reaction has probably the fastest reaction kinetics among all of the bio-orthogonal click reactions. Its reaction kinetics can reach as high as 10 000-fold that of CuAAC, enabling IEDDA ligation to occur at very low concentrations of coupling reagents. Therefore, tetrazine-transcyclooctene based IEDDA ligation has been increasingly exploited in nuclear medicine, particularly for pretargeted in vivo radiolabeling and imaging of tumors.^300−302

6.5.2. Chemical Conjugation via Reactions Involving the Bonds of NHS, Maleimide, EDC, and Disulfide.

The popular N-hydroxysuccinimide ester (NHS), maleimide, N-(3-(dimethylamino)propyl)-N′-ethylcarbodiimide (EDC), and disulfide reactions often used for conjugation of various fluorescent markers or other molecular probes to the oligonucleotides (ON) bearing amine, thiol, carboxylic acids, or hydroxyl reactive groups. These groups offer several options for forming reversible or irreversible linkages for successful postsynthetic conjugations.³⁰³

6.5.2.1. Thiol Chemistry.

Thiol allows for postsynthetic conjugations with molecular tags having maleimide,³⁰⁴ vinyl sulfone, or pyridyl disulfide functional groups.³⁰⁵ RNA/DNA bearing thiols react with molecular probes with maleimide or vinyl sulfone and form a stable thioether bond, whereas pyridyl disulfide gives a reversible disulfide linkage.^303,305 The thiol groups are also often used for the conjugation of gold through an irreversible covalent linkage.³⁰⁶

6.5.2.2. NHS Chemistry.

The NHS ester is an activated ester of a carboxylic group which readily undergoes trans amidation reaction with primary amines and forms a stable irreversible amide bond and is one of the widely used methods for postsynthetic modifications. The amine functionality is first introduced during automated ON synthesis by using various protected primary amine moieties that are later deprotected to expose the free amine for further conjugation with molecular tags having an NHS ester reactive group.³⁰⁷

6.5.2.3. EDC Reactions.

EDC rapidly reacts with carboxylates or phosphates to form an active complex, able to couple with primary amine-containing compounds. Oligonucleotides containing a carboxyl group reacts with EDC and forms an active ester which subsequently couples with primary amines or hydroxyl groups to give stable amide or ester bonds, respectively. The incorporation of imidazole with EDC activates the phosphate group to a highly reactive phosphodiester intermediate phosphoeimidazolide.³⁰⁸ The reactive phosphoeimidazolide will rapidly couple to amine-containing molecules to form a phosphoramidate bond.

6.6. Conjugation of Fluorescent Dyes, Targeting Ligands, and Drugs to RNA Nanoparticles

RNA has emerged as a powerful building block for fabricating nanoparticles due to its diversity in structure and function and high thermodynamic stability that enabled the addition of various functional modules to the RNA nanoparticles to obtain diverse nanoparticles for different nanotechnological applications.

Many chemistries employed to instill various molecular probes into the RNA for various RNA nanotechnological applications. Among them, the NHS, EDC, and disulfide formation chemistries gained more popularity due to their ease of reaction and higher yields. The functional groups needed for these reactions, such as thiol, carboxylic acid, amine, and maleimide, can be introduced into RNA by incorporating modified nucleotides or 5′/3′ terminal modifiers into the RNA strand. The modifications on the nucleotide can be done at nucleobase (U/G/C/T) or to the ribose, particularly at 2′-position.

6.6.1. Incorporation of Fluorescent Marker into RNA Nanoparticles.

Many strategies have been developed for postsynthetic labeling of RNA molecules with fluorescent dyes. Solid-phase synthesis of RNA makes it possible to incorporate various reactive functional groups that enables the incorporation of fluorescent markers into the RNA. The RNA with functional group reacts with molecular tags bearing an orthogonal reactive group under appropriate reaction conditions. NHS activated fluorescent compounds can be directly coupled to an NH₂ group of single-labeled RNA fragments, likewise, alkyne with azide, thiol with maleimide, etc. Using similar approaches, it was possible to efficiently synthesize RNA directly and incorporate traditional coupling reactive groups, such as amino −NH_2,-COOH, maleimide, thiol, alkyne, and azide,³⁰⁹ for constructing polyvalent RNA nanoparticles.³⁰³ Fluorescent dyes (FITC, Cy5, Cy3, AF-647) can be coupled to RNA,³⁰³ and the resulting derivatives are shown to be efficient for the construction of single-labeled RNA nanoparticles used for multiple purposes such as diagnosis, biodistribution studies, etc.²⁵

6.6.2. Conjugation of Targeting Ligands to RNA Nanoparticles.

Typically, RNA nanoparticles can be conveniently functionalized with tumor targeting ligands to achieve selective targeting and specific delivery to tumor. Tumor-specific delivery of drugs is essential to avoid toxic effects from anticancer drugs. Cancer cells express specific receptors and uniquely bind to small molecule ligands, antibodies, or aptamers. For example, liver cells overexpress asialoglycoprotein on hepatocyte cell surface and are specific to galactosamine molecules. Likewise, folate receptors specific to folic acid, EGF-alpha specific to EGF-aptamers. The ligands can be conjugated RNA nanoparticles using different chemistries that include click chemistry, NHS ester,⁷⁸ EDC,^310,311 thiol-maleimide, disulfide formation,³⁰³ or readily available corresponding phosphoramidites.

6.6.3. Conjugation of Chemical Drugs to RNA Nanoparticles for Targeted Drug Delivery.

Chemotherapeutics play a significant role in cancer treatment. Many chemotherapeutic drugs such as paclitaxel, docetaxel, camphotecin, and their derivatives have been used clinically in cancer treatment. However, the small molecule hydrophobic drugs often exhibit adverse side effects due to their accumulation in vital organs due to the lack of targeting ability and low water solubility that eventually affect their biodistribution and pharmacokinetic profiles in vivo. Guo. et al., very well adopted and utilized the RNA modification techniques for the conjugation of drugs, targeting ligands, and fluorescent molecules and also studied the effects of chemical drugs conjugation to RNA nanoparticles on in vivo biodistribution, pharmacokinetics, and therapeutic efficacies.^5,6,312 The RNA nanoparticles with cell-specific targeting ligands could deliver the hydrophobic drugs to cancer cells and showed improved therapeutic efficacies with no or fewer side effects. For example, poor solubility of camphotecin (CPT) in aqueous solution has improved dramatically by conjugating 7-CPT molecules to 3WJ-RNA nanoparticles as prodrugs using click reaction, and folic acid is used as a cancer targeting ligand. The tumor growth is effectively inhibited by CPT-RNA nanoparticles in the KB tumor xenograft.⁶ Although pRNA-3WJ nanoparticles improved the water solubility of the CPT molecules, the drug loading onto the RNA nanoparticles was limited due to their diminished thermodynamic stabilities. To improve drug loading, four-way junction RNA nanoparticles were constructed and conjugated to hydrophobic 24 paclitaxel molecules (Figure 19A). The PTX conjugation to 4WJ RNA nanoparticles not only improved the solubility of hydrophobic PTX molecules but also increased the therapeutic efficacy demonstrated using a breast cancer mouse xenograft model. The water solubility of the PTX was improved 10 000-fold compared to PTX alone (Figure 19C).⁵ This significant enhancement demonstrates that RNA can greatly solve the water-insolubility problem of hydrophobic chemical drugs. Three-dimensional pyramid- shaped RNA nanocages are designed to protect prodrugs cleavage from proteases present in the serum. A demonstration that the nanocages were conjugated to hydrophobic PTX molecules and released the PTX molecules in a temporal manner upon light exposure and was found to be effective toward breast cancer cells.³¹³ RNA also used as a building block for dendrimer construction that allows for the conjugation of multiple drug molecules. The dendrimers are demonstrated for the shielding effect of PTX with cells and plasma proteins thus reduced the probability of PTX release before reaching the tumor. The RNA nanoparticles could internalize into cells, release the PTX molecules, and exhibited cytotoxicity.⁷⁷

Figure 19.

Conjugation of paclitaxel (PTX) to RNA four way junction (4WJ) nanoparticles for cancer therapy. (A) Schematic of RNA-6 PTXs chemical conjugation and 4WJ-X-24PTXs self-assembly. (B) Conjugating six PTXs to an RNA, evaluated by denaturing PAGE. (C) HPLC chromatogram (absorbance 260 nm) with an inserted gel image of RNA-6 alkynes (blue) and RNA-6 PTXs (red). (D) Turbidity changes of PTX and RNA-6 PTXs in aqueous solution at 2-fold serial dilution. (E) Representative organ images showing specific tumor targeting of Alexa Fluor 647 labeled 4WJ-X-EGFR_apt nanoparticles 8 h postinjection into mice bearing MDA-MB-231 xenograft (T, tumor; H, heart; S, spleen; L, lung; K, kidney; Li, liver). (F) Intravenous treatment of nude mice bearing orthotopic MDA-MB-231 xenografts with 4WJ-X-24 PTXs-EGFR_apt nanoparticles. Reproduced with permission from ref 5. Copyright 2020 Guo et al. under CC By 4.0 https://creativecommons.org/licenses/by/4.0/.

Spherical nucleic acids (SNAs),³¹⁴ a class of nanoparticles which consist of a nanoparticle core functionalized with oligonucleotide sequences, include either RNAi therapeutics or reporter molecules. The unique three-dimensional SNA structures are resistant to nuclease degradation and internalize into cells even in the absence of transfection agents. The SNAs have been extensively studied against glioblastoma (GBM) models, the most aggressive and prevalent form of malignant brain cancers. The SNAs could effectively penetrate through blood–brain and blood–tumor barriers and robustly reduced tumor progression.^314,315 The SNAs are also used as a diagnosis tool and are capable of simultaneously detecting two distinct mRNA targets inside a living cell. The SNA conjugates consisting of oligonucleotide sequences which are hybridized to a reporter molecule with a distinct fluorophore label and each complementary to its corresponding mRNA target. Compared to single-component nanoflares, these structures allow for determination of precise relative mRNA levels in individual cells, improving cell sorting and quantification.³¹⁶ The exosome-encased SNAs are another class of cargos to deliver RNAi therapeutics into cancer cells, where the exosome-encased SNAs can be selectively cultured and isolated from cancer cells of interest and utilized them as a therapeutic vehicle against the same cancer type is a new way of treating cancer.¹⁰⁷

6.7. Conjugation of RNA Polymer with Other Nanoparticles

The conjugation of RNA polymer with the nanoparticles such as quantum dot, iron oxide nanoparticle, or gold nanoparticle have been successfully demonstrated by a series of studies using the siRNA or phi29 pRNA.³¹⁷ The siRNA was conjugated to various nanosized imaging agents for the multifunctional nanoparticles which show the therapeutic agents and diagnostic agents.^{105,108,317–320} Bhatia’s group demonstrated the siRNA and tumor-homing peptides (F3) were conjugated to the PEGlyated quantum dot (QD) core as a scaffold.^318,319 The amine group-modified QD was conjugated to F3 peptide and thiol modified-siRNA via sulfo-LC-SPDP (sulfosuccinimidyl 6-(3′-[2-pyridyldithio]-propionamido) hexanoate) and via sulfo-SMCC (sulfosuccinimidyl 4-(N-maleimidomethyl) cyclohexane-1-carboxylate) as heterofunctional cross-linkers, respectively. The QD-siRNA/F3 conjugate nanoparticles were efficiently delivered to HeLa cells and released from their endosomal entrapment, which provided knockdown of EGFP signal. They showed the siRNA–nanoparticle conjugates can be performed to dual purpose such as treat and image.

Moreover, the siRNA was conjugated to the iron oxide nanoparticle, which has the magnetic properties for biomedical application agent.^108,320 The siRNA was conjugated to iron nanoparticle for dual-purpose (1) in vivo transfer of siRNA and (2) accumulation of siRNA in tumor through magnetic resonance imaging (MRI) and near-infrared in vivo optical imaging (NIRF). For the coupling between siRNA and magnetic nanoparticles, the amine groups of iron oxide nanoparticles were activated with the m-maleimidobenzoyl-N-hydroxysuccinimide ester (MBS). Then, the reduced thiol group of RNA was reacted. Also, the near-infrared Cy5.5 dye (NIRF) and membrane translocation peptides were also conjugated to magnetic nanoparticles. The prepared RNA polymer–nanoparticle conjugates show the possibility of in vivo measurement of siRNA-magnetic particle uptake utilizing MRI, simultaneously, this conjugate can be seen in near-infrared optical imaging (NIRF). The conjugation of gold nanoparticle and RNA is also reported to increase the availability of tethered RNA splicing enhancer.¹⁰⁶

In 2007, Guo’s group developed a pRNA of the phage phi29 DNA-packaging motor conjugation to the gold nanoparticles for the single particle study of phage assembly.¹⁰⁵ To introduce the thiol (−SH) group to pRNA, SH-tagged DNA oligonucleotide was incorporated into 3′ end of the pRNA. Then, the gold nanoparticle was conjugated to the thiol-modified pRNA. The prepared pRNA/gold nanoparticle conjugates were bound to procapsid by in vitro phage assembly. In this study, they demonstrated the RNA polymer was easily conjugated with gold nanoparticle that can be provided more functionality for imaging.

Recently, the pRNA-3WJ was directly conjugated to the quantum dot for the resistive biomemory application.¹³⁴ For the conjugation between pRNA-3WJ and quantum dot (QD) specifically, they introduced a sephadex G-100 resin-recognizing RNA aptamer in the biotin-tagged pRNA-3WJ (SEP_apt/3WJ/Bio) motif. First, the SEP_apt/3WJ/Bio was bound to the G100 resin. Then the streptavidin-tagged QD (STV/QD) was bound to SEP_apt/3WJ/Bio through streptavidin–biotin binding on Sephadex G100, and the STV/QD- SEP_apt/3WJ/Bio conjugates were dissociated with elution buffer. The elution buffer split the STV/QD-Bio/3WJ_b and 3WJ_a, SEP_apt/3WJ_c fragments. The STV/QD-Bio/3WJ_b fragment was purified and reassembled with pRNA 3WJ_a and thiol-modified pRNA 3WJ_c for the resistive biomemory constitution. Thus, the RNA polymer can easily conjugated with other nanoparticles and this can be extended their functionality for various applications such as cellular imaging and therapeutics and bioelectronics.

6.7.1. Conjugation of RNA to Graphene.

Graphene is a one-atom thick two-dimensional honeycomb lattice made of sp2-bonded carbon atoms. It can be used as a fundamental building block for constructing other graphitic structures such as fullerenes, nanotubes, and graphite. This newly discovered carbon-based material has been the focus of much recent research in material science as well as nanotechnology. Most notably, British scientists Andre Geim and Konstantin Novoselov at the University of Manchester received the Nobel Prize in Physics in 2010 for their groundbreaking experiments with graphene.

Interestingly, recent studies have demonstrated that RNA can also be conjugated to graphene or immobilized on graphene for fabricating novel functional materials or the delivery of short interfering RNA (siRNA) to cancer cells. Hu et al.³²¹ covalently immobilized a RNA aptamer on graphene oxide and thus constructed polydispersed and stable RNA–graphene oxide nanosheets. The RNA attached to these nanosheets was resistant to nuclease degradation. Nanosheets manufactured with an aptamer to microcystin-LR specifically and competitively absorbed trace amounts of the peptide toxin from drinking water. Sharifi et al.³²² also used RNA as a surfactant to exfoliate flakes of graphene from nanocrystalline graphite to produce transparent and conducting RNA–graphene-based thin films, which can be used in a variety of electronic applications. Further research in RNA–graphene nanocomposites may open a new avenue toward many applications of graphene-based conductive materials.

7. IN VIVO RNA STABILITY, MODIFICATION, PROCESSING, AND REGULATION

Regulating the half-life and productivity of RNA is essential for life, and as RNA is sensitive to degradation, cells use a number of processes to protect this valuable information. By adding structures to both the 5′ and 3′ end of mRNAs, these RNAs become much more difficult to enzymatically process and are resistant to exonuclease activity. Changes in the position or removal of these structures result in the exposure of these RNAs and their rapid degradation. The overall stability of endogenous RNA can also be changed by processes that remove regulatory elements including alternative polyadenylation (APA) or splicing and/or result in the retention or exclusion of destabilizing or stabilizing exons or introns. The regulation of RNA stability of endogenous RNA is complex and controlled by a large number of competing processes, well beyond the scope of this review, here we have focused on discussing the main regulators of RNA stability in vivo.

7.1. RNA Regulation

RNA is a key molecule for the transmission of protein coding, structural, and signaling information. This fundamental role in cellular activity requires that the half-life of RNA be tightly controlled. A number of topics within this review have focused on discrete modifications that contribute to RNA stability. In this section, we will discuss the mechanisms widely utilized by cells to protect mRNA and examine how differences in each of these approaches can significantly alter RNA stability. As RNA is such an important and phenotype changing tool, endogenous RNAs are constantly screened by a plethora of enzymes including RNases and RNA exonucleases. These mechanisms are highly efficient, and almost all endogenous RNAs require protection from these enzymes to enable them to function correctly. In this section, we will focus on the most common RNA structures that contribute to RNA stability in vivo.

7.1.1. 5′ Capping of mRNA.

Exposed ends of RNA make ideal substrates for RNA exonucleases. For mRNA that is generally long and is required intact for function, rapidly protecting the end of the RNA is essential. A number of mechanisms for protecting the 5′ end of the mRNA are utilized by cells, however, the majority of mRNAs undergo 5′ capping.^323,324 The addition of the 5′ cap to the nascent mRNA is a multistep process that first requires the guanylylation of the phosphate groups at the 5′ end of the RNA^325,326 followed by a methylation step.³²⁷ Together, this generates a highly stable m7Gppp structure that is highly resistant to exonuclease activity and provides real protection for the RNA. In addition to stabilizing the mRNA, the m7Gppp cap attached to the 5′-end also aids protein translation by serving as a docking site for translation initiation factors.³²⁸

This mechanism of enhancing mRNA stabilization and enabling protein translation by 5′-capping has been utilized in mRNA vaccine development (see section 13, Construction of stable and translatable mRNA for mRNA vaccines). In mRNA vaccine production, the in vitro synthesized mRNA is attached with a cap at the 5′-end by mandatorily using an mRNA capping enzyme. SARS-CoV-2, the virus that causes COVID-19, displays a spike protein on the viral surface. Once the mRNA for the spike protein reaches immune cells, the cells make the protein spike antigen. The mRNA 5′-end-capping is a critical step to ensure that the in vitro synthesized mRNA can be stable after injection into patients and can be recognized by the host cell’s translation machinery to translate the mRNA into the SARS-CoV-2 protein antigen. However, the production of this capping enzyme in vivo is due to the fact that this capping enzyme is composed of two heterogeneous subunits D1 and D12.³²⁹ A single subunit cannot accomplish the capping function. The coexpression of these two subunits in cells has resulted in inclusion body and aggregation, thus the expressed proteins do not have the capping activity. In 1991, Guo invented a special technology for the production of the active vaccinia virus capping enzyme in vivo.³³⁰ The technology includes the methods of expression without producing an inclusion body, assembly of the two subunits D1/D12 mRNA capping enzyme, and the method of purification. This enzyme has now been manufactured with GMP grade to cap the mRNA used in Covid-19 mRNA vaccine development.

Removing the 5′ m7Gppp cap from RNA requires the specialized activity of “de-capping” enzymes, specifically DCP1 and DCP2.³³¹ These enzymes remove the cap structure and expose the 5′ unprotected end of the mRNA to exonuclease degradation.^332,333 This reaction is highly efficient and results in the rapid loss of the RNA.

7.1.2. 3′ Poly-Adenylation (Poly-A).

As discussed above, protecting the vulnerable ends of RNA is essential. The 3′ end of the mRNA utilizes a significantly different mechanism to protect the RNA from degradation. Poly adenylation is the addition of tens to hundreds of untemplated adenosine groups to the exposed 3′ end of the RNA.^334–337 The addition of these bases is catalyzed by Poly-A polymerase following transcriptional termination and cleavage of the RNA by the cleavage initiation^338,339 and stimulation complexes.^340,341 This long stretch of adenosine bases is highly resistant to degradation and provides an additional platform for protein binding and translation control.³⁴² Similarly to the 5′ cap, the removal of the poly-A tails requires the specialized activity of the family of PABP. Rather than the complete removal of the poly-A tail,³⁴³ PABP-mediated shortening of the tail to around 10–15 bp is sufficient to trigger the rapid degradation of the RNA.³⁴⁴

Poly-A addition and position are each important for controlling RNA stability in vivo. The 3′UTR of genes provides important RNA regulatory and stability information,³⁴⁵ and changes to the length of this region can significantly alter RNA stability. The most common form of 3′UTR length alteration is caused by alternative polyadenylation (APA) via the use of noncanonical Poly-A sites³⁴⁶ or changes in chromatin structure that prevent Poly-A site usage. A number of mechanisms have been shown to affect or trigger APA, however, the alteration in 3′UTR length is common throughout development^347,348 and tumorigenesis³⁴⁶ and frequently leads to improved RNA stability.

7.1.3. Splicing.

mRNAs are naturally comprised of introns and exons that enable the generation of many protein isoforms from a single gene.³⁴⁹ This flexibility is a key adaption of higher eukaryotes³⁵⁰ but can also significantly contribute to changes in RNA stability and function. Exon choice is a complex biological problem regulated by a number of processes, including spliceosome composition, chromatin status, and transcriptional processivity. The distribution and retention or exclusion of unstable exons and introns is a key contributor to RNA stability changes in a number of human pathologies.³⁵¹

In particular, the retention of intronic regions within mRNAs has been demonstrated to significantly lower the protein translation potential and overall stability of the RNA.³⁵⁰ Retained introns are unique and generally share features that make their processing and exclusion complex.³⁵² These types of RNAs have been found in a number of human diseases including cancer^353,354 that have mutations within splicing and/or RNA processing factors.³⁵⁵ Similarly, changes in the splicing of exons within the 3′UTR of genes can remove or add regulatory elements that alter the overall stability of the transcript.

7.2. Cellular RNA Regulatory Mechanisms

7.2.1. RNA Structure and Destabilizing Elements.

RNA can fold into complex structures in a number of ways. Self-assembling RNA structures are frequently formed by riboswitches and ribozymes that fold into different confirmations depending on the availability of metal or metabolic cofactors. Many of the best characterized RNA folding processes are facilitated by RNA–protein interactions and include the structures formed by ribosomal RNAs, snRNPs, microRNAs, telomerase RNA (TERRA), and other long noncoding RNAs. Understanding the complex world of RNA structure and stability is important for discerning the biological functions of these RNAs.^356,357

Messenger RNA (mRNA) is the template used for protein production by cells, and the amount of protein generated by cells depends on the availability of RNA. This makes regulating the levels of RNA and the capacity of this RNA to be translated as important for preventing metabolic exhaustion and/or cellular transformation. Each mRNA contains numerous post-transcriptional regulatory sites that enable RNA–RNA or RNA–RNA binding protein (RBP) interactions (Figure 20). These binding events are mediated via either structural features or sequence motifs that function to destabilize or regulate the translation potential of the RNA.

Figure 20.

Role of RBP (RNA-binding protein) in RNA processing (A) mRNA capping. During initiation of transcription, TFIIH phosphorylates the RNA pol carboxy-terminal domain (CTD). Human mRNA capping enzyme (RNGTT) and cap methyltransferase (RNMT) are recruited to phosphorylated CTD, adjacent to 5′ end of nascent RNA. (B) mRNA splicing: A group of small nuclear ribonucleoproteins (snRNPs) bind to intron that results in folding of introns and bringing the 5′ and 3′ closer to form a loop. The ends of exons also move closer, and introns detach to make a mature mRNA. (C) Polyadenylation: Cleavage occurs at poly A site that usually lies in upstream of poly A signal (AAUAAA). Poly A polymerase then adds the several hundreds of A nucleotides to 3′ end of mRNA. (D) RBP mediated mRNA processing: RBP regulates the mRNA stability directly or indirectly at different steps like capping, splicing, and polyadenylation.

7.2.2. RNA–RNA Regulation of Stability.

MicroRNA (miRNAs) were first identified in the early 1990s and regulate gene expression via a number of mechanisms, including translation regulation and/or RNA degradation.³⁵⁸ Following miRNA processing, mature miRNAs are loaded into the RNA-induced silencing complex (RISC) complex via their interaction with the Argonaute 2 (AGO2) protein. As part of this complex, miRNAs search for complementary RNA sequences within the 3′ untranslated region (UTR) of cellular RNAs. Once a complementary sequence is found, which is generally referred to as the seed sequence, the RISC complex, via the RNA cleavage capacity of AGO2, can trigger the degradation of the substrate. In addition, miRNAs can also degrade mRNAs via slicer independent mechanisms by triggering the deadenylation and/or decapping of RNA substrates. miRNAs and RBPs can also function synergistically or cooperatively to control the mRNA degradation.³⁵⁹

Long noncoding RNAs (lncRNAs) contribute to RNA stability by base-pairing with RNA, resulting in RNA destabilization and turnover. The RNA duplex formed between the lncRNA and interacting RNA can alter the availability of regulatory elements and provide binding sites for other trans-acting factors. A new and emerging group of RNA–RNA regulators that have recently been identified are small vault RNAs (svRNAs). These barrel shaped structures are formed by specific protein and noncoding vault RNAs (vRNAs) interactions that result the production of svRNAs. These RNAs are able to interact with Ago proteins to cleave the target mRNAs in a sequence specific manner.^360,361

RNA stability represents the cumulative effect of many independent RNA factors that enables cells to balance the level of mRNA in response to both internal and external stimuli/signals. These processes are largely depend on mRNA degradation via the mechanisms of miRNAs, svRNAs, lncRNAs, and likely other poorly understood RNA species.

7.2.3. RNA–RBP Regulation of Stability.

A plethora of RBPs directly or indirectly regulate the stability of RNA. Each of these interactions is important of modulating RNA fate. The function of these RBPs vary greatly, from generic RNA destabilizing enzymes to highly specialized RBPs with complex regulatory activity.³⁶²

Protein-mediated mRNA decay is commonly initiated by a number of mechanisms, including 5′ cap removal, endonucleolytic cleavage, and/or poly(A) shortening. Of these, poly(A) tail shortening occurs at that greatest frequency and removes the protective structure at the 3′ end. Following the truncation of the poly A tail to 10–15 nucleotides, these transcripts are decapped by the Dcp1p-Dcp2p complex. This enables the subsequent recruitment of exonucleases that rapidly degrade the poorly protected RNA. A large group of RBPs function in this capacity, however, the best characterized is the Poly A binding protein (PABP). PABP is a highly conserved protein that contains four tandem RNA recognition motifs (RRM 1–4). PABP plays a complex but important role in regulating mRNA stability. PABP contributes to mRNA stability via a number of mechanisms including promoting poly(A) shortening on RNAs that have long A stretches that may contribute to excessive poly(A) addition. Cytoplasmic polyadenylation element-binding proteins (CPEBs) are another class of generic RBPs that regulate the poly(A) tail length and subsequently mRNA stability. However, unlike PABP, CPEBs recruitment can have vastly different effects on RNA stability. Much of these different activities of CPEB is regulated by cofactor RBPs. Contrastingly, CPEB binding has been shown to both increase and decrease poly(A) tail length. In this way, the CPEBs act as a molecular scaffold for regulating RNA stability.

A significant number of RBPs function by recognizing a conserved regulatory motif within RNA. The site and position of the RBP interaction has real consequences on the overall fate and protein translation of the target RNA.³⁶³ In the following section, we will focus on a select subset of RBPs that have key roles in modulating RNA stability. The CCR4-NOT family of RBPs functions to remove the poly(A) tails from mRNAs. This activity exposes the RNA to additional exonucleases that can trigger the instability and ultimately the degradation of the RNA.³⁶⁴ The CCR4-NOT complex is frequently utilized by other RBPs that lack enzymatic activity themselves, including the Pumilio, Nanos protein families and miRNAs within the RNA-induced silencing complex (RISC).³⁶⁵ This is a common feature of many enzymatic RBPs, which separate RNA cleavage from target recognition, thus enabling greater diversification in regulation.

Selecting RNAs for destabilization is an important process, and errors that result in the aberrant degradation of RNA either through incorrect substrate choice or failure to remove a targeted RNA can have toxic outcomes. Many RBPs favor AU-rich elements (ARE), as these nucleotides are over-represented in the 3′UTRs of genes and allow favorable binding conformations. AUF1 (hnRNP D) was the first AU-rich element-binding protein (ARE-BP) to be identified and has long been linked to mRNA destabilization.³⁶⁶ AUF1 binds directly to long AU-elements within 3′UTRs, as this enables AUF1 dimers to assemble into larger and more stable tetramer complexes. This results in the recruitment of additional proteins such as PABP, eIF4G, and HSPs that target the mRNA for degradation. However, emerging studies from a number of cell types suggests that AUF1 has context-dependent roles in RNA regulation that may alter the outcome for target RNA.^367,368 As our capacity to map RNA–RBP interaction grows, our understanding of the complex mechanisms utilized by RBPs to regulate RNA stability has expanded. One such example is the family of Pumilio (PUM) proteins. The PUM proteins are highly conserved and required for organismal development, pluripotency, and differentiation. These proteins form variable complexes with additional RBP families, including NANOS and BRAT, that can modify their specificity. The PUM proteins and the complexes that they form have no enzymatic activity, however, they utilize a number of cofactors to cleave the 5′ cap or poly-A tail to trigger RNA decay. Although most RNA destabilizing proteins are generic RNA regulators, key RNA control events are most frequently controlled by specific RBPs and their binding sites. This targets these RNAs that have the potential to be highly toxic or transformative for degradation through the use of high-affinity regulation rather than nonspecific and potentially less-effective mechanisms.

7.3. Cellular RNA Modification

RNA, much like DNA, is a modifiable canvas that can be utilized to convey important information along with the RNA template. Currently, more than 130 RNA chemical modifications have been discovered, and to date the functions or molecular reason for most of these RNA modifications are still unknown. However, the sheer number and diversity of these modifications makes their investigation essential. A number of evolutionarily conserved RNA modification events have been identified suggesting these molecular events may be coordinated during different environmental conditions.^369–371 Defining how these modifications work and their molecular consequence has opened up new avenues of research for RNA biologists and could further be mined for therapeutic opportunities. In the majority of instances, these RNA modifications have been best explored in abundant RNAs including tRNAs (tRNAs) and rRNAs (rRNAs). In the following section, we will examine a subset of RNA chemical modifications, describe what is known about their roles in regulating RNA fate, stability, and processing.

Hundreds of different RNA editing and RNA chemical modification events are found in cells and clearly have important biological functions. These findings highlight the malleability and adaptability of RNA as a template for cellular information and provide real opportunities to chemically modify RNA for improved therapeutic efficacy.

7.3.1. RNA Modification and mRNA Stability.

Regulating mRNA stability and protein translation potential is crucial, and various RNA chemical modifications have specific functions that alter mRNA stability. Over the next subsections, we have summarized key findings on a subset of RNA modifications including: N⁶-methyladenosine (m⁶A), N⁶,2′-O-dimethyladenosine (m⁶Am), 8-oxo-7,8-dihydroguanosine (8-oxoG), pseudouridine (Ψ), 5-methylcytidine (m⁵C), and N⁴-acetylcytidine (ac⁴C) (Figure 20).

7.3.1.1. N⁶-Methyladenosine (m⁶A).

m⁶A is one of the most common mRNA modifications. In most organisms, m6A modification is controlled by methyl-transferase complexes (humans: METTL3, METTL14, KIAA1429, and WTAP). These enzymes catalyze the m⁶A modification in nascent RNAs and transfer a methyl group to the N6 position of the adenosine group following the motif (RRACH). These m⁶A modifications affect a number of RNA processes including transcription, protein translation, RBP-binding, pre-mRNA splicing, and mRNA stability.³⁷² m6A changes on RNA have important biological functions, and m6A has been linked to the maternal to zygotic transition (MZT), neurogenesis, pluripotency, and acute myeloid leukemia (AML).³⁷³

7.3.1.2. N⁶,2′-O-Dimethyladenosine (m⁶Am).

Rather than a single methylation, two modification events occur on a single adenosine, one at the 2′-OH position and the second at N⁶ position is required to generate, N⁶,2′-O-dimethyladenosine (m⁶Am). Although m⁶A and m⁶Am originate at the same adenosine, they exhibited different functions. mRNAs that contain m⁶Am modifications are comparatively more resistant to RNA decapping by the microRNA-mediated degradation by the DCP2 enzyme.^374,375

7.3.1.3. 8-Oxo-7,8-dihydroguanosine (8-oxo-G).

8-oxo-G is a form of oxidized RNA that is generated by a number of oxidizing agents including reactive oxygen species (ROS). The oxidation of these mRNAs changes the overall stability and translation potential of the RNA. Although the mechanisms regulating this process are unclear, 8-oxo-G containing RNAs are very sensitive to a collection of RBPs including AUF1, PCBP1, and YBX1.^376,377

7.3.1.4. Pseudo-Uridine (Ψ).

Ψ is a product of C–C glycosidic isomerization of a uridine base. Several Ψ have been discovered in a number of different RNA species, including tRNA, rRNA, snRNAs, and mRNAs. The chemical properties of uridine and Ψ are sufficiently different to alter the local secondary structure and subsequently affect many RNA processing events, such as splicing, stability, translation, and localization. A number of studies have linked pseudouridine (Ψ) to mRNA stability,³⁷⁸ however, the mechanisms of how this RNA modification contributes to RNA stability requires further investigation.

7.3.1.5. 5-Methylcytidine (m⁵C).

Using an analogous process to m6A modification, a number of proteins, including NOP2/Sun RNA methyl transferase 2, can methylate RNA at 5 position of cytosine. Although the substrates of m⁵C remain poorly characterized, RBPs that can read these modifications, ALYREF and YBX1, have been identified. To date, proteins able to erase m⁵C from RNA have as yet not been found. A growing body of work has shown that m⁵C increases the overall stability of RNA and may have important biological functions.^379,380

7.3.1.6. N⁴-Acetylcytidine (ac⁴C).

ac⁴C in tRNA, mRNA, or rRNA are produced by N-acetyltransferase 10 (NAT10). These modifications are found in both coding and noncoding RNAs. For coding RNAs, ac⁴C is over-represented near the translation initiation codon. These ac⁴C sites typically increase the half-life of mRNA and may help to promote the protein translation. Recent studies suggest that NAT10 may function as the acetyl-transferase of ac⁴C,³⁸¹ however, currently, the reader and eraser function of this modification remains unknown.

7.3.1.7. RNA Modification and Translation Regulation.

An enormous number of RNA modifications have been identified, each of which is likely to have important cellular and developmental roles. Up until recently, mainly biophysical and biochemical methods were used to determine the role of these modifications in RNA stability and protein translation. Recently, the effects of nucleotide modification on the translation system was assessed using the “PUR-Express translation experimental system” in which m⁵C, m⁶A, Ψ, or 2′-O-methylated nucleotides were incorporated into bacterial ErmCL mRNA codons. This study identified RNA modifications that resulted in the premature termination of translation based on the type and the positions of RNA modification.³⁸² Additional work has also shown that 2′-O-methylation sterically impairs the codon–anticodon helix interaction, as well as A site tRNA accommodation that is necessary for efficient protein synthesis.³⁸³

7.3.1.8. RNA Editing and Detection.

RNA modifications clearly have important and under-appreciated roles in regulating RNA fate and organismal biology. Part of the problem with studying these processes is the limited number of experimental approaches available to reproducibly and accurately identify them. However, recent technological advancements now allow the global detection of various RNA modifications in the transcriptome.^384–386

RNA editing is an important post-transcriptional modification that results in the deamination of cytosine to uridine (C to U) or adenosine to inosine (A to I). This reaction is catalyzed by the family of adenosine deaminase (ADAR) enzymes. In human cells, three ADAR proteins (ADAR1–3) regulate this process. RNA editing is found in both coding and noncoding RNA and contributes to a number of biological functions.³⁸⁷ These RNA editing significantly alter the chemical structure of the base-pair and local RNA folding, resulting in changes in RNA stability, RNA-binding protein interactions, and protein translation. Unlike the majority of the RNA chemical modifications, the role of ADAR-mediated deamination of RNA as an important mechanisms for “self” determination is well understood. Via this mechanism, mammalian cells can identify viral RNA, as this is not ADAR-edited and trigger an appropriate immune response. This is likely an important archaic mechanism to prevent viral integration and cellular transformation.

7.3.1.9. RNA Modifications and RNA Fate.

An ever growing number of RNA modifications have been identified on human RNA, and their role in altering and/or regulating the fate of RNA remains poorly characterized (for review, ref 388). One of the developing trends that is beginning to emerge is the capacity of different RNA modifications to attract a shared subset of RNA binding proteins that alter RNA fate and stability. A clear example of this is the Y box binding protein 1 (YBX1) that binds to RNA containing 8-oxo-G or m⁵C modifications. However, this interaction results in either destabilization (8-oxo-G)³⁷⁶ or stabilization (m⁵C)³⁸⁹ of the RNA via YBX1 activity, highlighting the important descriptive role that RNA modifications have in regulating RNA stability. RNA modifications can also have very different effects on RNA stability, as they can attract either decay factors or post-transcriptional regulators. The m6A RNA modification provides the best example of this. m6A modified RNA is actively bound by the family of YTH domain containing proteins (YTHDF1–3),³⁹⁰ which trigger RNA decay and/or destabilization. In contrast, m6A modifications can also be bound by more traditional RBPs, including FMRP³⁹¹ and HuR, resulting in RNA stabilization. Although our understanding of how RNA modifications label RNA and contribute to stability changes is still developing, it is clear that these changes in RNA chemistry have real impacts on RNA fate and stability.

7.3.1.10. Point of View on mRNA Modification and Stability Concerning Nanotechnology and Clinical Applications.

In vitro and in vivo RNA sensitivity to RNase degradation has long been a stumbling block in RNA technology. This has made many scientists flinch away from RNA nanotechnology. Concerning the in vivo degradation of RNA by serum RNases has delayed the arrival of RNA as the third milestone in pharmaceutical drug development for decades. Although challenges of RNA stability in vivo have now been overcome, and the third milestone in drug development using RNA has arrived, the topic of RNA stability is still an important topic. The major problem of “RNA degradation” is caused by RNase (ribonucleases). RNases are a large group of hydrolytic enzymes,^{20,42,392–394} such as RNase A and RNase H, etc. Fortunately, we now know that the major serum RNases cleave the phosphodiester bond between the 3′-phosphate group of a pyrimidine nucleotide (C and U but not purine G and A) and the 5′-ribose of its adjacent nucleotide. One amino acid of the RNase, for example, the His12 of RNase A, acts as a general base, accepting the 2′-OH proton of the ribonucleic sugar ring. That is to say, the 2′-OH group and the pyrimidine base are involved in RNase cleavage. On the basis of the understanding of this mechanism, for decades scientists have published tens of papers and filed tens of patents to develop chemical methods to modify the 2′-OH group of only the pyrimidine C and U in RNA. These RNAs with 2′-F, 2′-NH2, or 2′-O-methyl modifications are highly resistant to RNase degradation. This is to say, when we considered the RNA degradation problem in vivo, we only need to consider the modification for pyrimidine nucleotides but not purines. In RNA therapeutics, for example, siRNA, you only need to modify the 2′ location of the C or U of the sense strand, but not both strands. As reviewed above and summarized in Figure 21, chemically modified pyrimidines are available. It will provide useful information for the consideration of mRNA vaccine development, cancer mRNA vaccines, and RNAi therapeutics.

Figure 21.

Modifications of RNA that alter mRNA stability. RNA modifications are generated by writers that directly change the RNA. Readers are RNA-binding proteins that can identify RNA modifications, and erasers are proteins that can remove the RNA modification for the substrate. For each RNA modification, proteins that function in these roles have been included, where identified. Ac⁴C = N⁴-acetylcytidine: writer NAT10 = N-acetyltransferase 10; reader/eraser unknown. m⁵C = 5-methylcytosine: writer NSUN2 = NOP2/Sun RNA methyl-transferase 2; reader YBX1 = Y-box binding protein 1, eraser unknown. Ψ = pseudouridine: writer PUSs = pseudouridine synthases; reader/eraser unknown. 8-oxo-G = 8-oxo-7,8-dihydroguanosine: writer reactive oxygen species (ROS); readers AUF1= AU-rich element RNA-binding protein 1, eraser unknown. m⁶Am = N⁶,2′-O-dimethyladenosine: writer PCIF1 = PDX1 C-terminal inhibiting factor 1; reader unknown; eraser FTO = α-ketoglutarate dependent dioxygenase. m⁶A = N⁶-methyladenosine: writer METTL3 = methyl-transferase like 3, readers IGF2BP = insulin-like growth factor 2 mRNA-binding proteins, G3BP1 = Ras GTPase-activating protein-binding protein 1, YTHDF = YTH N⁶-methyladenosine RNA binding protein; eraser ALKBH5 = AlkB homologue 5, RNA demethylase.

8. PASSIVE DELIVERY OF THERAPEUTICS TO TUMORS WITHOUT TARGETING LIGANDS

The use of nanoparticles as a delivery platform for therapeutics has been proposed to reduce systemic toxicity and enhance patient tolerances. Over the past three decades, there has been improvement in nanomedicines developed for tumor therapy, with formulation designs being produced at an incredible rate.³⁹⁵ However, only 10 nanomedicines for use in cancer therapeutics have been approved by the U.S. Food and Drug Administration and European Medicines Agency.³⁹⁶ Among the nanomedicines that have reached phase III clinical trials, as few as only 14% have demonstrated efficacy.³⁹⁶ Nanoparticle delivery from circulation into the tumor microenvironment plays a key role in determining the therapeutic efficacy of cancer nanomedicines.³⁹⁷ The molecular and cellular mechanisms of nanoparticle delivery into solid tumors under in vivo physiological conditions needs to be understood for establishing the basis of therapeutic principles in the field of cancer nanomedicine.³⁹⁸ Nanotechnology offers many advantages in the fight against cancer; one of its most crucial advantages has been relatively higher accumulation in tumors due to specific pathophysiological properties, known as the enhanced permissive tumor retention (EPR) effect. In addition, RNA nanotechnology possesses a unique rubbery property to further facilitate the accumulation in tumor site.

8.1. Rubbery Property of RNA Nanoparticles

Rubber is a fascinating material used in both industry and daily life. The development of elastomeric materials in nanotechnology is imperative due to their economic and technological potential. RNA exhibits features of elasticity resulting in the rubbery property of RNA nanoparticles. The rubbery property of RNA nanoparticles has been found to be responsible for the high passive delivery in cancer therapy. Elasticity³⁹⁹ is an important property of nucleic acid polymers, allowing them to behave similary to rubber at nanoscale level in different dimensions. Elasticity is the physical property (genotype), while rubbery is a general descriptive term, used here to refer to the biological effect (phenotype).

The rubbery property of RNA nanoparticles directly affects the cancer homing, renal excretion, organ accumulation, circulation time, and drug biodistribution and is demonstrated in vivo by comparing RNA nanoparticle retention times in the tumor, kidney, and liver. The rubbery behavior of the RNA nanoparticles is further supported by the presence of 10 nm RNA nanoparticle in urine of mice after systemic injection, which is larger than the glomerular filtration capacity (5.5 nm).³³ The effect of the rubbery property of RNA nanoparticles as a favorable factor for their accumulation in the cancer vasculature, tumor uptake, organ retention, and kidney clearance was further evaluated by injecting NIR fluorescent dye-labeled RNA nanoparticles (3WJ RNA (6 nm), 4WJ RNA (10 nm), 6WJ RNA (12 nm)) and other inorganic nanoparticles of similar size such as iron (10 nm) and gold (10 nm) into mice intravenously (Figure 10).³³ The in vivo organ distributions of the various RNA nanoparticles and their inorganic nanoparticle counterparts were examined 8 h post injection. The fluorescence intensity of entire organs ex vivo as well as homogenized organ samples were quantified. RNA nanoparticles of different stoichiometries and the iron or gold nanoparticles exhibited large differences in organ and tumor retention (Figure 10A). After normalizing the radiation efficiency by organ weight, 4WJ-RNA nanoparticles exhibited a higher tumor-to-liver and tumor-to-kidney ratios (Figure 10B).³³ The homogenized organ sample data are consistent with the ex vivo results (Figure 10C), which further demonstrates the favorable biodistribution profile of RNA nanostructures. Interestingly, the tumor-to-organ fluorescence ratio of the 10 nm RNA nanoparticles was much higher than that of the iron and gold nanoparticles of the same size. These results support the hypothesis that RNA nanoparticles display rubbery properties, leading to a stronger EPR effect and enhanced vessel extravasation in tumor targeting. Compared to gold and iron nanoparticles of the same size, RNA nanoparticles displayed stronger cancer-targeting outcomes but less RNA accumulated in healthy organs.³³ Generally, the upper limit of renal excretion is 5.5 nm; however, the 5-, 10-, and 20 nm RNA nanoparticles passed renal filtration and resumed their original structure in urine (Figure 22). These findings explain two previously unanswered questions: (1) How RNA nanoparticles have unusually high tumor-targeting efficiency? Their rubber- or amoeba-like deformation property enables them to squeeze out of the leaky vasculature to improve the EPR effect (Figure 23). (2) How can RNA nanoparticles remain nontoxic? They can be rapidly cleared from the body via renal excretion into urine with little accumulation in the body. Considering their controllable shape, size, and rubber-like properties (Figures 8 and 9), the RNA nanoparticles are endowed with great promise for industrial and biomedical applications, especially in cancer therapeutic delivery.

Figure 22.

The rubbery properties of RNA nanoparticles leads to fast renal clearance. (A) Renal excretion is a process in which RNA nanoparticles are subjected to kidney glomerular filtration. The RNA nanoparticles display rubber and amoeba like properties, compelling vessel extravasation to enhance fast renal excretion. Such mechanisms can explain why RNA nanoparticles remain nontoxic because they can be rapidly cleared from the body via renal excretion into urine with little accumulation in the body. (B) Near-infrared AF647 labeled dsRNA, RNA 3WJ, and 4WJ were found in mice urine assayed by 12% native gel 0.5 h post IV injection. Adapted with permission from ref 33. Copyright 2020 American Chemical Society.

Figure 23.

The rubbery property of RNA nanoparticles enhanced tumor accumulation. RNA nanoarchitecture is stretchable and shrinkable by optical tweezer manipulation with multiple extension and relaxation repeats like rubber. Such effects enhanced high tumor targeting efficiency due to their rubber or amoeba-like deformation property. The rubber property of RNA nanoarchitecture enables it to squeeze out of the leaky vasculature to improve the EPR effect. Adapted with permission from ref 33. Copyright 2020 American Chemical Society.

8.2. EPR Effect

The EPR effect paradigm can explain drug extravasation by the leakage of nanoparticles through aberrant vascular gaps between adjacent endothelial cells in a tumor. The mechanism by which nanoparticles deliver drugs is based on their escape from circulation and entrance into the tumor microenvironment, which play key roles in determining the therapeutic efficacy of cancer nanomedicines. Nanotechnology provides many advantages for the fight against cancer; one of its most crucial advantages has been the relatively higher accumulation of nanoparticle-delivered medicines in tumors due to specific pathophysiological properties, collectively known as the EPR effect. First described by Matsumura and Maeda in 1986 (Figure 24A),^47,400 the importance of EPR has only grown. The factor enabling this process is the abnormal vasculature of cancer, in which blood vessels are described as “leaky.” The vasculature around tumors is chaotic and dense, causing leaky blood vessels. Leakiness and rapid vessel growth enable nanoparticles to extravasate into the tumor region with appreciable ease.

Figure 24.

Comparison of the active transcytosis and passive EPR effects for effective nanoparticle extravasation into solid tumors. (A) Passive delivery of nanoparticles from circulatory blood vessels into the tumor microenvironment through leaky and/or fenestrated vessels. (B) Active transcytosis involving adsorptive transcytosis, nanoparticle drug uptake, intracellular transport, and exocytotic release of nanoparticles from endothelial cells into the tumor microenvironment and receptor-mediated transcytosis, in which nanoparticles bind with receptor-specific ligands to improve the uptake of drugs by endothelial cells.⁴⁷

Since the late 1980s, the EPR effect has been the central paradigm supporting the use of nanomedicines to treat solid tumors, in which enhanced permeability due to the leaky vasculature and fenestration of tumor blood vessels and increased drug accumulation results from a collapse of the lymphatic drainage system near the tumor.⁴⁰⁰ The EPR effect paradigm can explain drug extravasation from the leakage of nanoparticles through aberrant vascular gaps between adjacent endothelial cells in the tumor. The EPR effect has been demonstrated to be the major pathway by which nanoscale particles invade tumor sites, strongly facilitating RNA nanoparticles delivery and accumulation in the cancer vasculature (Figure 25).^401,402 The rapid growth of a tumor results from the underdevelopment of blood vessels, which are incomplete and leaky.⁴⁰³ There is some debate as to whether or not all human tumors undergo the EPR effect in the same manner as observed in animals; however, recent studies have suggested that at least a portion of human cancers contain leaky blood vessels, providing holes on the order of tens of nanometers for nanoparticle penetration.^403,404 Some pathophysiological factors secreted in tumor tissues, such as nitric oxide, bradykinin, and vascular endothelial growth factor (VEGF), further promote the extravasation of nanoparticles.⁴⁰⁵ In addition, tumor tissues lack lymphatic drainage. These characteristics account for the favorable retention of RNA nanoparticles in tumors. Passive targeting is highly dependent on particle size. Generally, the size of RNA nanoparticles is approximately 6–15 nm, which is less than the cutoff size of the tumor vasculature (200–800 nm).⁴⁰⁶ This size discrepancy facilitates passage through tumor vasculature while overcoming the traditionally poor access to tumors of other nanoparticles and proteins. The favorable 6–15 nm size of RNA nanoparticles facilitates their passing through narrow cavities that can otherwise become mechanical barriers, overcoming interstitial pressure-related dispersion. The EPR effect of RNA nanoparticles has been extensively exploited as a novel drug delivery platform. The size of RNA nanoparticles results in a strong EPR effect, while entrapment in the liver and spleen by the mononuclear phagocytic system (MPS) is avoided. RNA nanoparticles strongly and specifically bind to cancers, with few entering the liver, lung, or any other vital organs or tissues (Figure 25).^13,14,42,118 RNA is an elastomer that can be the basis for rubbery RNA nanoparticles, thus these nanoparticles induce the EPR effect and enhance vessel extravasation for tumor targeting by changing shapes in order to slip through the leaky vessels.⁴⁰³

Figure 25.

Biodistribution profiles of RNA nanoparticles to various cancer animal models. RNA nanoparticles demonstrate high tumor accumulation with no detectable accumulation in healthy organs. Adapted with permission from ref 9. Copyright 2015 American Chemical Society. Adapted with permission from ref 10. Copyright 2016 The American Society of Gene and Cell Therapy. Adapted with permission from ref 11. Copyright 2015 Oncotarget under CC By 3.0 https://creativecommons.org/licenses/by/3.0/. Adapted with permission from ref 12. Copyright 2015 Cui et al. under CC By 4.0 https://creativecommons.org/licenses/by/4.0/. Adapted with permission from ref 13. Copyright 2013 Shu et al. under CC By-NC 4.0 https://creativecommons.org/licenses/by-nc/4.0/. Adapted with permission from ref 14. Copyright 2012 Elsevier Ltd. Adapted with permission from ref 15. Copyright 2015 American Chemical Society.

The amoeba-like deformation property of RNA nanoparticles allows them to squeeze through leaky vessels of the tumor vasculature to improve the EPR effect. This outcome may explain the recent findings described by Warren Chan and colleagues, who recently performed an analysis of the published nanomedicine literature over a ten-year period (2005–2015). Their results demonstrated that an average of 0.7% of the nanoparticles administered systemically reached solid tumors.⁴⁰⁷ In addition, their data revealed that fewer than 1% of the administered nanoparticles accumulated in solid tumors in animal models.⁴⁰⁸ Similarly, Andresen and colleagues evaluated liposome accumulation within various tumor types and found heterogeneity in the nanoparticles accumulated in different tumors. These data further demonstrated that the EPR effect cannot be considered a general principle for the delivery of all types of nanoparticles (such as liposomes) into solid tumors.⁴⁰⁹

Therefore, facilitating passage through tumor vasculature must be accomplished while overcoming the traditionally poor accessibility to tumors by protein Abs, and the nanoparticles must also pass through narrow cavities by overcoming mechanical barriers, interstitial pressure-related dispersion, and highly immunosuppressive microenvironments.^27,410,411 The EPR effect of RNA nanoparticles has been extensively exploited as a novel drug delivery platform.^27,42 The size of the RNA nanoparticles results in a strong EPR effect and prevents nanoparticle entrapment in the liver and spleen by the mononuclear phagocytic system (MPS).^26,48,412 RNA nanoparticles strongly and specifically bind to cancers with limited entry into the liver, lung, or any other vital organs or tissues (Figures 25and 26).^13,14,42,118

Figure 26.

Highly specific efficient cancer targeting and fast body clearance of pRNA-X nanoparticles. (A) The pRNA-X (harboring folate and Alexa-647) nanoparticles specifically targeted folate receptor positive tumor xenografts upon systemic administration in nude mice, as revealed by whole body imaging (A) and internal organ imaging (B). Control: PBS treated mice. Scale bar: fluorescent intensity. Reproduced with permission from ref 14. Copyright 2012 Elsevier Ltd.

As described above, the infrequent gaps in tumor vessels cannot account for the EPR effect. Warren Chan and colleagues used the so-called zombie model to determine the uptake mechanism leading to nanoparticle accumulation.⁴⁰⁸ The results show that gaps in the tumor vasculature accounted for only 3–25% of number of extravasated nanoparticles.⁴⁰⁸ These observations indicated that the EPR effect cannot fully explain the poor therapeutic efficacy of nanomedicines, which has been attributed to insufficient accumulation of nanoparticles in tumors and poor pharmacokinetics. In addition, the data from a study on active transcytosis revealed that 97% of nanoparticles are transported from blood vessels into solid tumors by endothelial cells (involving active uptake, the formation of fenestrae, intracellular transport, and subsequent exocytosis) (Figure 24B).⁴¹³ This process employs endocytosis, receptor-mediated endocytosis, and exocytosis to accomplish its task.⁴¹⁴ In addition to the importance of adsorptive active transport mechanisms (adsorptive transcytosis, Figure 24B), cancer cell-specific ligand–receptor binding is likely to be important in the early steps of active transcytosis for nanoparticle extravasation.

In conclusion, on the basis of observation, most if not all cellular uptake can be attributed to one of these two methods, and further elucidation will allow for easier delivery of anticancer drugs via RNA nanoparticles.

8.3. Application of Passive Delivery Mechanisms

Micelles are small particles comprised of inward facing hydrophobic tails and outward facing hydrophilic tails, allowing for ease in penetrating cancerous cells.⁴¹⁵ Micelles, when utilized to deliver siRNA and miRNA structures, confer advantages previously discussed in the EPR section.⁴¹⁵ In a recent study, the pRNA-3WJ acted as a scaffold to construct RNA micelles, with an 8-nucleotide locked nucleic acid (named anti-miRNA-21) complementary to the seed region of miRNA-21 conjugated to the micelle shell and used to inhibit cancer (Figure 27).⁴⁸ Cholesterol was conjugated to the helical end of a branched pRNA-3WJ motif to facilitate the insertion of the hydrophilic RNA into the hydrophobic micelle (Figure 27).^21,48 This anti-miR-21-harboring RNA micelle was internalized into cancer cells and specifically inhibited the function of oncogenic miR-21 to trigger cancer cell apoptosis and inhibit tumor growth in xenograft models. Researchers further extended the multifunctionality of micelles by loading the hydrophobic therapeutic paclitaxel into them to increase their solubility and create a nanodelivery platform that efficiently targeted tumors for drug delivery.^21,48 The advantageous features described above, including multivalence, precise controlled assembly, specific tumor targeting, effective cancer cell permeability, and low (or no) nontarget organ toxicity, indicate that pRNA-3WJ loaded micelles show great promise as reliable platforms for drug delivery in clinical applications.

Figure 27.

Design, construction, and delivery methods of RNA micelles. 3WJ motif of pRNA from bacteriophage phi29 DNA packaging motor to form 2D structure of 3WJ/FA/anti-miR21. pRNA-3WJ micelles formation by hydrophobic force. Adapted with permission from ref 48. Copyright 2019 American Chemical Society. RNA nanoparticles and micelles target tumor cell surface receptors and undergo receptor-mediated endocytosis (RME). In initiating the process, an external ligand binds to a specific membrane spanning receptor, triggering lateral diffusion across the membrane until a clathrin-coated pit forms (denoted as step 3). The complex is then absorbed through the membrane (denoted as step 4), where it is uncoated and begins to fuse with an early endosome (denoted as step 5).⁶⁵

In another study, antisurvivin siRNA and paclitaxel were loaded into polymeric micelles and targeted to ovarian cancer cells. This study concluded that a platform delivering both anticancer drugs and gene-silencing RNA was viable for use in chemoresistant tumors; in addition, these authors, through their previous work, have shown other viable methods for siRNA delivery via micelles.⁴¹⁶ One study sought to determine whether when micelle-delivered miRNA-34a and doxorubicin show additive potency to downregulate genes and prohibit cancerous growth. The results from this study showed that the miRNA was able to downregulate Bcl2n, survivin, and notch1 by 65%, 55%, and 46%, respectively, thereby showing effective delivery of intact miRNAs and adding to the list of theragnostic strategies.⁴¹⁷

9. ACTIVE DELIVERY OF THERAPEUTICS BY RNA NANOPARTICLES TO DISEASE SITES

In addition to passive uptake and accumulation, active tumor targeting effects also play vital roles in targeted drug delivery. In contrast to the EPR effect, which facilitates nanoparticle accumulation in the tumor region, active targeting ligands promote particle uptake into cells after they are delivered to the tumor via the EPR effect. Cancer cells overexpress different receptors, leading to various targeting ligand designs. Utilizing multispecific RNA nanoparticles, we conjugated drugs and targeting ligands to one particle, further increasing the drug internalization prospects and reducing nonspecific toxicity (Table 3).

Table 3.

RNA Nanoparticles Functionalized with Ligands and RNA Aptamers for Targeted Delivery

targeting group	receptor	RNA NP	cell line	disease	ref
FA	FR	3WJ, 4WJ	KB	cancer	5,42
FA	FR	3WJ	HT29	CRC	15
FA	FR	3WJ	MGC803	gastric cancer	12
FA	FR	RNA-exosome	HT29	CRC	8
GalNAc	ASGP-R	6WJ	HepG2	HCC	49
EGFRapt	EGFR	3WJ, 4WJ	MDA-MB-231	TNBC	5,9
EGFRapt	EGFR	RNA-exosome	MDA-MB-231	TNBC	8
CD133apt	CD133	3WJ	MDA-MB-231	TNBC	39
PSMAapt	PSMA	3WJ	LNCaP	prostate cancer	10
PSMAapt	PSMA	RNA-exosome	LNCaP	prostate cancer	8
EpCAMapt	EpCAM	3WJ	HCA-7 colony 29	CRC	34

9.1. RNA Nanoparticles Functionalized with Targeting Groups as Active Delivery Platforms

After delivery of the nanoparticle drugs to tumor regions, active targeting is effective for the transport of small molecule drugs, antibodies, and nucleic acids to targeted cancer cells, avoiding entry into adjacent normal cells and thereby enhancing therapeutic efficiency while reducing toxic effects. Active targeting is required to enhance the quantity of drug delivered to realize the accumulation of the highest dose of antitumor agents in the target cell over possible nontargeting drugs or passive delivery systems. To achieve this purpose, nanoparticles are conjugated or decorated with ligands that specifically bind to receptors overexpressed on the tumor cells. A successful example was demonstrated in 1980 with antibodies grafted on the surface of liposomes,⁴¹⁸ followed by studies using the same concept with other various kinds of ligands, such as peptides, nucleic acids, and aptamers.^111,419

Small molecule ligands and RNA ligands are known as active targeting molecules. Drugs that utilize them are proving to be invaluable due to their specificity and aversion to systemic negative effects.⁴²⁰ RNA ligands are also known as RNA aptamers and are quickly gaining traction as viable tools.^421,422 RNA aptamers have been rapidly developed, offering an ever-expanding array of targeting ligands. RNA aptamers are described as RNA oligonucleotides that vary from 15 to 100 nt, forming tertiary or even fourth-level structures, which affords them a large surface area allowing for high affinity binding.⁴¹⁶ Their advantages over traditional molecules include high binding affinity and selectivity and are readily synthesizable and modifiable.⁴²³ The method for binding involves a combination of van der Waals forces, base stacking, hydrogen binding, and electrostatic interactions.^27,119

9.1.1. RNA Nanoparticles Harboring Small Molecules Ligands As Targeting Group.

For small molecule ligands, folate (FA) and N-acetylgalactosamine (GalNAc) have been attached to RNA nanoparticles for targeted drug delivery. Folate receptors (FRs) bind folate reducing folic acid derivatives and mediate the delivery of tetrahydrofolate to the interior of cells. FA is necessary for DNA nucleotide synthesis and cell division in normal cells, whereas cancer cells express 100-fold more folate receptors than normal cells.⁴²⁴ The high affinity of FA for FRs provides a valuable opportunity for the specific targeting and delivery of nanoparticles to cancer cells. Although FA exhibits many advantages compared to other ligands, it is difficult to overcome their major weak point: poor water solubility. As RNA nanoparticles have proven to solubilize hydrophobic molecules, FA was conjugated to multifunctional RNA nanoparticles derived from the three-way junction (3WJ) of bacteriophage phi29 motor pRNA for active targeting of cancer cells.^99,118,425 FR was detected at a high level in human colorectal cancer (CRC) lung and liver metastasis tissues.¹⁵ and these RNA nanoparticles proved to actively target, bind, and enter the CRC cells (Figure 28A).¹⁵ Additionally, FA conjugated RNA nanoparticles targeted metastatic cancer cells in a clinically relevant mouse model of CRC metastasis.

Figure 28.

RNA nanoparticles harboring targeting ligands selectively bind tumor cells. (A) 3WJ RNA nanoparticles conjugated with folate (FA) (red) demonstrated high tumor cell binding and internalization by confocal microscopy. Reproduced with permission from ref 15. Copyright 2015 American Chemical Society. (B) 3WJ RNA nanoparticles harboring EGFR targeting RNA aptamers (red) demonstrated selective targeting and cell internalization into triple negative breast cancer cells by confocal microscopy. Reproduced with permission from ref 9. Copyright 2015 American Chemical Society. (C) 3WJ RNA nanoparticles harboring PSMA targeting RNA aptamers (red) demonstrated selective targeting and cell internalization into prostate cancer cells by confocal microscopy. Reproduced with permission from ref 10. Copyright 2016 The American Society of Gene and Cell Therapy.

More recently, RNA nanoparticles functionalized with paclitaxel and miR122 as therapeutic groups and GalNAc as targeting ligand have been developed and used for hepatocellular carcinoma treatment.⁴⁹ GalNAc is the galactosamine derivative that has a high binding affinity to the asialoglycoprotein receptors (ASGP-R) expressed on the cell surface of hepatocytes. 6WJ RNA nanoparticles conjugated with GalNAc showed increased binding to and internalization into HepG2 cells compared to 6WJ nanoparticles without ligand.⁴⁹ Furthermore, biodistribution study was conducted and RNA nanoparticles with GalNAc showed predominant accumulation in tumor site instead of health organs in tumor induced HCC xenograft model.

9.1.2. RNA Nanoparticles Harboring RNA Aptamer As Targeting Group.

As for RNA aptamers, RNA nanoparticles have also achieved tumor accumulation using different aptamers as targeting groups. The epidermal growth factor receptor (EGFR) family consists of four members. HER2 is a member of the EGFR family of receptor tyrosine kinases detected in breast cancer with a more aggressive behavior and worse prognosis when untreated.^426,427 The HER2-specific antibody (trastuzumab) is the standard of care for HER2-positive breast cancer patients, markedly improving disease-free and overall survival. HER2-specific RNA aptamers capable of delivering therapeutic small-interfering RNAs (siRNAs) to HER2-expressing breast cancer cells were also investigated.⁴²⁸ Our previous studies revealed that the estrogen receptor (ER) coactivator mediator subunit 1 (MED1) plays a critical role in tamoxifen resistance in ER-positive breast cancer cells through cross talk with HER2.²⁸ We assembled a three-way junction (3WJ) pRNA–HER2_apt–siMED1 nanoparticle to target HER2-overexpressing human breast cancer via an HER2 RNA aptamer to silence MED1 expression in a HER2-overexpressing (BT474) cell xenograft model.²⁸ Intravenous injections of RNA-paclitaxel nanoparticles (Alexa Fluor 647-labeled 4WJ-X-EGFR_apt) with a specific cancer-targeting ligand (EGFR aptamer) dramatically inhibited TNBC (MDA-MB-231 cell) xenograft breast cancer growth, with nearly undetectable toxicity and immune responses in mice (Figure 19).⁹ Constructed pRNA-3WJ RNA nanoparticles harboring the EGFR_apt demonstrated cellular targeting and binding by fluorescence microscopy (Figure 30B).^9,39 Additionally, ex vivo images of tumors and healthy organs harvested from mice 8 h post injection showed that 4WJ-X-EGFR_apt nanoparticles strongly accumulated in tumors, with low or no accumulation in the liver or other organs.⁵

Figure 30.

RNA nanotechnology for decorated EVs for tumor regression. (A) Intravenous treatment of nude mice bearing LNCaP-LN3 subcutaneous xenografts with PSMA_apt/EV/siSurvivin or PSMA_apt/EV/siScramble. (B) Intravenous treatment of nude mice bearing breast cancer orthotopic xenografts with EGFR_apt/EV/siSurvivin and controls. Adapted with permission from ref 8. Copyright 2017 Pi et al.

Prostate-specific membrane antigen (PSMA) was detected in prostate cancer cells and the neovasculature of different kinds of malignant neoplasms, including breast,⁴²⁹ brain,^430,431 lung,⁴³⁰ and other cancers⁴³² but not in normal vascular endothelium. Our previous study demonstrated ligand–receptor binding of RNA nanoparticles with PSMA aptamers (Figure 30C) and PSMA-RNA nanoparticle displaying EVs. 3WJ nanoparticles with PSMA aptamer as targeting ligand showed significant tumor accumulation in biodistribution study. Fully functionalized 3WJ-anti-miR21-PSMA_apt RNA nanoparticles showed tumor inhibition effect in LNCaP xenograft mouse model. The therapeutic effect of PSMA aptamer-displaying EVs for prostate cancer treatment was evaluated using LNCaP-LN3 tumor xenografts. The PSMA aptamer displaying EVs conjugated with survivin siRNA (PSMA_apt/EV/siSurvivin) completely inhibits prostate cancer growth in nude mice bearing LNCaP-LN3 subcutaneous xenografts.⁸

9.2. RNA Nanoparticles Facilitate Endosomal Escape

RNA nanoparticles functionalized with targeting groups, including small molecule and RNA aptamers, could target deliver cargos to disease sites and internalize into the cells. However, loaded cargos such as macromolecules still need to overcome another barrier, which is the endosome release in order to exhibit their regulation function in cytosol.

Exosomes have gained attention for their potential use as a direct delivery method for RNA payloads into the cytosol.^22,433 Exosomes can contain lipids, proteins, and, most importantly, nucleic acids. This ability to carry nucleic acids while targeting cells naturally and being readily modifiable by various methods makes them potent tools for cancer drug delivery.^434–438 The exosome is an ideal platform for safe drug delivery to pass through the blood–brain barrier as well as to target other tissues while avoiding both immune system response and rapid renal clearance. However, exosomes lack specificity for entry into cancer cells, which results in off-target effects when loaded with siRNA or miRNA. To address this challenge, Guo et al., developed a novel platform for the targeted delivery of exosomes using cancer cell receptor-specific RNA aptamers (Figure 29).⁸ The arrow-shaped RNA (Figure 29A) was conjugated with membrane-anchoring cholesterol at the arrow tail of Phi29 3WJ nanoparticles to create exosomes that display RNA aptamers that serve as specific targeting ligands for receptors overexpressed on cancer cells (Figure 29). The 3WJ nanoparticles were bound to tumor cells, and the siRNA was delivered into cytosol through receptor binding and cell membrane fusion (Figure 29B–D). These 3WJ vesicles were also conjugated with siRNA (either targeting survivin, which inhibits apoptosis, or a scrambled siRNA version that serves as a control) and loaded into the exosomes to be injected into xenograft mouse models. Vesicles with functioning siRNA were shown to specifically target tumors, reducing survivin expression and tumor size (Figure 30).⁸ This novel platform enables RNA nanoparticles to either load their cargo inside exosomes or on the exosome surface.

Figure 29.

RNA nanotechnology for decorating native exosomes (EXs) or extracellular vesicles (EVs) (A) Schematic of RNA nanoparticle loading and decoration of exosomes for drug delivery and tumor targeting, respectively. Adapted with permission from ref 8. Copyright 2017 Pi et al. (B) Mechanism of exosome binding and fusing with cell membrane for direct cytosol delivery of loaded cargos. Reproduced with permission from ref 22. Copyright 2019 Elsevier BV. (C) Decorated or loaded exosomes (red) were imaged for endosome colocalization by staining for EEA1 for early endosome and LAMP1 for lysosomes (green). Separate signals demonstrated exosomes avoided endosome entrapment. Reproduced with permission from ref 22. Copyright 2019 Elsevier BV. (D) Confocal imaging distinguishes between direct fusion and back fusion. Decorated ligands (red) and loaded cargo (green) of exosomes were imaged by confocal microscopy showing cell membrane incorporation and cytosol delivery, respectively. Reproduced with permission from ref 22. Copyright 2019 Elsevier BV.

To elucidate the mechanism underlying the efficient therapeutic effect of RNA nanoparticle decorated exosomes, another study was conducted by Zheng et al. using exosomes displaying a 3WJ-FA RNA nanoparticle on the surface and encapsulating siRNA inside as the model. When compared to exosomes lacking FA receptors, the FA-decorated exosome was able to deliver the siRNA payload while triggering noticeable suppression of the targeted genetic pathway both in vitro and in vivo.²² This efficient cancer suppression of FA-displaying exosome is contributed to both the targeting effect of FA and the endosomal escape induced by exosome. Cytosol delivery of siRNA using FA-displaying exosome mediated by direct fusion mechanism was confirmed using fluorescence colocalization. While this is a recent development, the ability to subtly modify positions of ligands for enhanced targeting can lead to novel applications for previously developed exosome and ligand complexes.

In addition to utilizing RNA nanoparticle decorated exosomes, other methods could also be combined with RNA nanotechnology to avoid endosome trapping. Photochemical internalization (PCI) is a drug delivery method that utilizes light-triggered molecules to facilitate the endosomal escape.^439,440 After the activation by light, endosomal localizing photosensitizers generate reactive oxygen species (ROS), which causes the disruption of endocytic membranes. Photosensitizer molecules have also been used in photodynamic therapy (PDT), which produce oxidant species and cause cell death after the exposure to light. These photosensitizer reagents can be conjugated to RNA nanoparticles to improve the delivery efficiency to cytosol and increase the cytotoxicity to cancer cells while reducing side effects.

10. FABRICATION OF RNA NANOPARTICLES

RNA nanotechnology is the application of utilizing RNA motifs and structures for the development of controllable and defined nanostructures. RNA nanoparticles form into predefined structures that are implemented during the design process through predictive folding and computation methods described above. Nanoparticle component strands are designed with precise sequences to ensure proper folding into the desired nanoparticle structure through inter- or intramolecular interactions. These component strands are synthesized through either solid-phase chemical synthesis or biological polymerase reactions. Upon synthesis, many RNA nanoparticles self-assemble due to the high stability of RNA nanoparticles, at which the nanoparticles can be fully characterized. Here we provide detailed methods of RNA nanoparticle construction and characterization to allow for future applicational use.

10.1. Synthesis of RNA Nanoparticle Component Strands

The construction of these RNA component strands are accomplished in two manners: (1) chemical synthesis and (2) enzymatic synthesis.^26,120 While chemical synthesis allows for large-scale production of RNA oligos and provides the availability to incorporate many different chemical modifications, it is limited in the number of nucleotides it can incorporate into a strand due to the chemical reaction efficiencies.¹²⁰ Therefore, RNA production by enzymatic synthesis is valuable in that it can theoretically produce RNA strands of any length to overcome the limitations of chemical synthesis.

10.1.1. In Vitro Transcription of RNA Oligo Strands.

Researchers have developed in vitro techniques of RNA synthesis to allow for the reliable production of RNA using the T7 phage RNA polymerase, which is discussed in detail below.^441–447 RNA enzymatic production requires several bottom-up construction steps, including the construction of a DNA template, in vitro transcription, and purification of RNA products. The T7 phage RNA polymerase is the most commonly used polymerase due to its high rate of transcription and a very low error rate when reading template DNA.^442,448 The T7 RNA polymerase (RNAP) protein has been isolated from the T7 bacteriophage and expressed into plasmids that have been stably place in bacteria such as Escherichi coli. This has allowed for easy purification of the polymerase to be used for in vitro applications.^{441,444,446,449,450} Large quantities of the T7 RNA polymerase are easily created by scaling up bacteria production and purification of the polymerase through common molecular biology techniques.

To complete enzymatic RNA synthesis, a DNA template is required for the polymerase enzymes to read and transcribe into RNA. The DNA sequence must include the polymerase promoter sequence to provide a binding domain and transcription start sequence.^442,443 Upon the construction of a DNA template, transcription of DNA into RNA is completed through a simple biological reaction. T7 polymerase requires the presence of magnesium ions, thus a reaction buffer is needed to initiate the transcription.^13,451 In addition to mixing of the RNAP, buffer, and DNA template, ribonucleic acid nucleotides are needed for the polymerase to add one by one onto the RNA strand being synthesized. Within the T7 polymerase system, RNA is synthesized at an estimated rate of ~230 nuc/s.⁴⁴⁵ As the polymerase reaches the end of the DNA template, it falls off ending transcription until it is able to identify a new DNA template strand in solution. The reaction will occur until the polymerase runs out of activity, RNA nucleotides are depleted, or buffer and temperature conditions change enough to prevent polymerase activity.

10.1.2. Inclusion of Chemical Modifications into Enzymatic Synthesis of RNA Oligos.

The chemical modifications (i.e., 2′-fluoro and 2′-OCH₃) have become very prevalent in the field of RNA nanotechnology and have been used in FDA approved RNA based therapies.^27,120,452 Native T7 RNA polymerase is not able to incorporate these modified nucleotides during transcription without modification to the transcription reaction or T7 polymerase itself. Researchers have developed a single amino acid mutated (Y639F) T7 RNA polymerase variant that is able to incorporate 2′-fluoro modified nucleotides.⁴⁵³ Additional RNA polymerases have been modified in similar ways such as the Syn5 polymerase to allow for these important modifications on RNA nanoparticles.^454,455 Therefore, by using a modified in vitro transcription, enzymatic production of RNA nanoparticles with modifications for clinically relevant approaches can be completed.

10.1.3. Chemical Synthesis of RNA.

Solid-phase synthesis is a regular employed method for the synthesis of RNA using the phosphoramidite chemistry.^456,457 The process has been automated, and the RNA oligonucleotide synthesis is carried out by a stepwise addition of nucleobase phosphoramidites to the 3′-terminus of the growing chain until the desired sequence is synthesized. Each addition is referred to as a synthetic cycle and consists of four chemical reactions. (step i) Deblocking: The 4,4′-dimethoxytrityl (DMT) group present on solid support is removed with a solution of 2% trichloroacetic acid (TCA) in solvent dichloromethane which leaves a free 5′-terminal hydroxyl group. (step ii) Coupling: The nucleoside phosphoramidite in acetonitrile is activated by an acidic azole catalyst, the activated phosphoramidite is then brought in contact with the starting solid support to react with the 5′-hydroxy group to form a phosphite triester linkage. (step iii) Capping: The capping step is performed by treating the solid support-bound material with a mixture of acetic anhydride and 1-methylimidazole to block unreacted 5′-hydroxy groups and is acetylated by the capping mixture. (step iv) Oxidation: The treatment of the support-bound material with iodine and water in the presence of a weak base, usually pyridine, oxidizes the phosphite group into a stable phospho-diester bond. The steps i–iv will be repeated until the desired sequence synthesis is completed in the 3′ to 5′ direction.

10.1.4. Synthetic Methods for RNA labeling.

The site-specific incorporation of labels and reactive groups can be introduced through synthetic chemistry. Solid-phase synthesis is the widely used method to incorporate the labels/functional tools by incorporating modified nucleotides/terminal modifiers having various functional groups such as alkyne,^458–463 thiol,^464–470 aldehyde,^471–474 amine,^475–480 carboxylic acid,^481–486 and biotin^{465,487–490} into the extending RNA site specifically. These functional groups can be modified with drug molecules or imaging markers with orthogonal functional groups with appropriate chemistries, and they include N-hydroxysuccinimide (NHS) chemistry, copper-catalyzed click reaction, copper-free click chemistry, thiol chemistry, periodate chemistry, imine formation, and 5′-phosphate activation.^{104,491–500} Among them, the NHS, thiol as well as copper-catalyzed and copper-free click chemistries are widely used for both internal and terminal modifications, whereas 5′-phosphate activation and periodate chemistry are mainly employed for terminal modifications. The functional tools are further utilized for labeling and conjugation of various molecular probes and drug molecules in order to engineer RNA nanoparticles for applications in diagnosis and medicine.

10.1.5. Synthetic Methods for RNA Ligation.

The synthetic chemistry methodologies extended for ligation of RNA to another RNA/DNA strand by installing appropriate functional groups within two nucleic acid molecules. This process can be enhanced by bringing the two molecules together with a “splint”.⁵⁰¹ Various functional group combinations are used in joining two nucleic acid molecules, and they include 5′-phosphate/3′-OH,^502,503 3′-disulfide/5′-SH,^504–506 2′-NH₂/5′-aldehyde,^507–509 and 3′-O-propargyl/5′-azide.^473,510,511 The functional groups are activated using appropriate activators to obtain the ligated full-length nucleic acid products. The chemistries also extended to ligate ribozymes, Morpholino-linkages and sulfur chemistries are often employed to ligate small ribozymes and two pieces of the Varkud satellite (VS) ribozymes respectively.^{157,512–515} Click chemistry is also utilized in ligating ribozymes and they are functionally indistinguishable from a transcribed form of the ribozyme.

10.2. Methods for the Construction of RNA Nanoparticles

10.2.1. Bottom-up Assembled RNA Nanoparticles.

As a significant portion of the genome is composed of noncoding RNA, RNA molecules are widely present in living organisms with various tertiary structures that carry out different biological functions. Inspired by the naturally existing complex RNA architectures, artificial RNA nanoparticles have been constructed utilizing the RNA motifs derived from natural structural components (Figure 31), such as tRNA. Composing RNA motifs were first synthesized and then mixed together to form the tertiary structure. 2D square-shaped RNA particles were constructed utilizing three RNA components: the right-angle motif, three-way junction motif, and tRNA-motif through the kissing loop interaction (Figure 2).⁵² 3D structures such as polyhedrons were then developed utilizing different tRNA motifs through kissing loop and tail–tail interactions.⁵⁹ More recently, a library of RNA motifs has been built to construct more complex RNA nanoparticles (Figure 2).²

Figure 31.

Schematic diagram representing the different stem and loop structure. (A) Different stem and loop forming hairpin loop, bulge loop, multibranch loop, internal loop and helices. (B) G-quadruplex. (C) Three stem junction and (D) four stem junction.

Besides utilizing RNA motifs, another approach to construct RNA nanoparticles utilizing naturally derived RNA motifs through completing one-pot assembly of single RNA strands (Figure 32A). RNA strands are synthesized and mixed together directly and form RNA nanoparticles through base-pairing. The pRNA-3WJ motif is derived from phi29 bacteriophage packaging RNA and has been extensively used as the building block to construct various RNA nanoparticles. As discussed in previous sections, the phi29 pRNA-3WJ is composed of three short oligo strands that fold into a highly thermodynamics core (Figure 4A). This high stability allows for its usage in forming more complex nanostructures. 2D RNA structures such as triangles, squares, and pentagons have been constructed utilizing the phi29 pRNA-3WJ motif (Figure 2 and 8).⁷ Similarly, 3WJ motifs were used to construct 3D shaped nanostructures such as prisms, tetrahedrons, and dendrimers by connecting 3WJs through different design. Such structures were confirmed by cryogenic electron microscopy (Cyro-EM) and atomic force microscopy (AFM) studies to display 2D and 3D structures that accurately reflected the predicted structures.⁴¹²

Figure 32.

Construction of RNA nanoparticles. (A) Bottom-up assembly, from single strands and from RNA motifs. Adapted with permission from ref 1. Copyright 2014 Oxford University Press under CC By 4.0 https://creativecommons.org/licenses/by/4.0/. Adapted with permission from ref 2. Copyright 2017 American Chemical Society. (B) Schematics of the design and structure of 2-helix RNA origami (2HO-RNA) containing one kissing loop (in green) (left). Sequence design of 2HO-RNA tethered with two thrombin RNA aptamers (in blue and purple rectangles) (right). Reproduced with permission from ref 35. Copyright 2019 Wiley-VCH. (C) Design and synthesis of multiple-component RNAi microsponges (Multi-RNAi-MSs) using various circular DNAs via RCT approach (left). SEM image and confocal microscopy image of Multi-RNAi-MSs (scale bar: 1 μm). Reproduced with permission from ref 40. Copyright 2016 Roh YH, et al.

In addition to utilizing natural RNA motifs, RNA nanoparticles can also be designed and constructed from scratch with the assistance of computational modeling. RNA hexagonal nanorings have been constructed using a loop–loop interactions aided by molecular dynamic simulation and MFOLD to predict its flexibility, stability, and structure.^{3,4,86,449,516} 3D cubic RNA-based scaffolds have been designed in silico from scratch and constructed through in vitro transcription.⁶¹

The bottom-up assembly approach is particularly suitable for incorporation of various functional groups into RNA nanoparticles with defined number and position. During the synthesis of single RNA strands, functional RNA components such as siRNA, miRNA, and RNA aptamers can be directly incorporated.²⁶ After the synthesis of RNA strands, functional groups such as small molecule drugs and fluorescent dyes can be covalently conjugated to RNA strands with defined stoichiometry.⁶ Multifunctional RNA nanoparticles could easily be developed using RNA strands with different functionalization such as targeting ligand, therapeutic agent, and fluorescent dye for therapeutic and/or diagnosis purposes.

10.2.2. RNA Origami.

Recently, RNA nanostructures have been produced cotranscriptionally using the single-stranded RNA origami technique, where a single strand of RNA folds upon itself to form a designed structure (Figure 32B). The shape of the origami was tuned by modulating the angle of the kissing loop used for tile association. The real advantage lies in their great potential to be cloned and expressed in large quantities in vivo for potential applications in synthetic biology. Different RNA origami molecules could be constructed with functional RNA modules such as siRNA and RNA aptamer, riboswitch, and ribozyme.^517–520 A 3D scaffold RNA origami wireframe structure, RNA octahedron, with intrinsic functioning siRNA embedded was constructed. Efficient silencing of a target reporter (eGFP) induced by the release of siRNA (siEGFP) after dicer recognition and process were demonstrated in vitro.⁶²

10.2.3. Rolling Circle Transcription (RCT).

Rolling circle transcription was used for the isothermal enzymatic production of nucleic acids. RCT produces long RNA transcript repeats through the construction of circular DNA templates (Figure 32C).⁷⁵ These templates incorporate T7 polymerase promoter sequences that allow for the start of transcription into long concatemers. An advanced technique, complementary rolling circle transcription (cRCT), was also developed. With this approach, two complementary circular DNAs were used to replicate long RNA strands continuously. The resulting two complementary RNA strands bind together by hybridization and self-assemble into a robust and colossal RNA architecture.⁵²¹ Different RNA nanoparticles have been constructed with incorporation of different siRNA and mRNA using RCT approaches, including RNAi microsponges (Figure 32C), siRNA nanosheets, and tumor-targeting RNA nanovectors, for the treatment of different diseases.^75,163 In addition, the bottom-up construction of RNA nanoparticles has been translated into RCT through the incorporation of the phi29 pRNA-3WJ into a circularized DNA templates as well as self-cleaving RNA ribozymes for transcription.¹⁶⁴ As a result, as the concatemers were produced the short component strands were cleaved by the incorporated ribozymes and self-assembled into the pRNA-3WJ.

10.3. Characterization of RNA Nanoparticles

10.3.1. Physiochemical Properties of RNA Nanoparticles.

RNA nanoparticles are an ideal drug delivery system due to several of their properties. They can be easily assembled through bottom-up assembly or cotranscriptional assembly as described above. Their negative charge can avoid the nonspecific binding to the negatively charged cell membrane which will minimize off-target side effects. Their size, shape, and composition are highly programmable. RNA nanoparticles with various 2D and 3D structures as well as with different functional groups and motifs have been successfully constructed. Furthermore, with the utilization of modified phosphoramidites, RNA nanoparticles have become thermodynamically, chemically, and enzymatically stable to be used in in vivo applications without fast degradation, digestion, or dissociation as discussed in detail in previous sections.

With the development of RNA nanotechnology over the past decade, approaches for systematic characterization of RNA nanoparticles have been matured. Constructed RNA nanoparticles are first subjected to either agarose or polyacrylamide gel electrophoresis to examine their assembly efficiency by checking the size and folding properties. The size and zeta potential of RNA nanoparticles are then measured simultaneously by dynamic light scattering (DLS). Nanoparticle shape and general sizing can be evaluated by atomic force microscopy (AFM), transmission electron microscopy (TEM), scanning electron microscopy (SEM), and cryogenic electron microscopy (Cryo-EM).⁶³ Their thermal stability can be measured by temperature gradient gel electrophoresis (TGGE) for their melting profile and their annealing profile can be measured by thermal cycler using intercalating dyes to detect folded regions.⁷⁷ The chemical stability can be proven by the polyacrylamide gel with added urea for denaturing.⁴² As for the enzymatic stability, their half-life in serum could be measured by incubation with the 50% FBS at 37 °C for a series of time points followed by the quantification of intact nanoparticles at each time point.

10.3.2. Cellular Uptake and Toxicity Study of RNA Nanoparticles.

RNA nanoparticles can easily incorporate targeting ligands including chemical ligands and RNA aptamers to provide specific targeting against cell surface receptors, proteins, and even fluorescent chemicals. Chemical ligands, such as folate, can be attached to the end of the RNA strand during RNA synthesis using modified RNA phosphoamidites or post synthesis through chemical conjugation reaction. As for the RNA aptamers, such as EGFR RNA aptamer, the aptamer strand can be synthesized as the elongation of the scaffold strand.⁹ RNA nanoparticles with ligands are typically labeled with fluorescent dyes or radio-isotopes and then tested for their binding efficiency in vitro and in vivo.³¹² The in vitro binding and internalization studies are usually done by incubating labeled RNA nanoparticles with cells followed by the quantification using the flow cytometry and/or fluorescence confocal microscopy (Figure 28). Different RNA nanoparticles with ligands significantly increase cellular uptake compared to bare RNA nanoparticles, as discussed above.

Besides cell binding, the therapeutic and side effects of RNA nanoparticles will also be tested in cells before animal studies. To evaluate the cytotoxicity of RNA nanoparticles, cell proliferation assay, such as MTT, as well as cell apoptosis assay, such as fluorescence-activated cell sorting (FACS) analysis, have frequently been done.⁵ Normally, RNA nanoparticles themselves do not affect the cell viability, while RNA nanoparticles loaded with therapeutic groups show significant cytotoxicity and induce cell apoptosis. As for the determination of immunogenicity triggered by RNA nanoparticles, induction of cytokines, and chemokines in macrophage-like cells would be performed through enzyme-linked immunosorbent assay (ELISA).³⁶ The immunogenicity of RNA nanoparticles depends on its sequence, composition, and structure. The immunogenicity of planar RNA nanoparticles without specific sequences such as CpG is negligible.¹

10.3.3. In Vivo Behavior of RNA Nanoparticles: Biodistribution.

After the characterization of RNA nanoparticles in vitro and ex vivo, RNA nanoparticles would then be subjected to in vivo biodistribution studies to provide their relative accumulation in different vital organs and tumors.¹²⁰ RNA nanoparticles labeled with fluorescent dyes or radio isotopes allows for imaging of the whole body and major organs of animals as well as the quantification of tissue accumulation in the case of radio-isotopes. RNA nanoparticles were injected through a tail vein in mouse models and underwent whole body imaging as well as harvested organ and tumor were obtained at different time points. RNA nanoparticles with the targeting ligand have shown specific accumulation in tumor with little or no accumulation in vital organs (Figure 25).¹²⁰ To further investigate the influence of structures of RNA nanoparticles on their biodistribution profile, different parameters such as size, shape, and payload have been evaluated.³¹² In terms of size and shape, the larger the nanoparticle is, the more accumulation in liver is observed. RNA nanoparticles are used to deliver various cargos, so the influence of the different cargos to the nanoparticles have also been evaluated. RNA nanoparticles conjugated with different fluorescent dyes with different hydrophobicity were injected in mice model, respectively, and the results indicate that the more hydrophobic the dye is, the more accumulation in liver is observed.³¹² In general, it is determined the larger the nanoparticle and the more hydrophobic the payload results in increased liver accumulation.

11. APPLYING RNA NANOTECHNOLOGY TO THE TREATMENT OF DISEASES

11.1. Delivery of Small Molecule Anticancer Drugs

Small molecule anticancer drugs have been widely used for cancer treatment. However, problems such as insolubility and nonspecific toxicity of the drugs have limited their applications. RNA nanoparticles have been used as drug carriers to delivery small molecules anticancer drugs with two approaches: intercalation and covalent conjugation (Figure 33A) (Table 4). RNA nanoparticles loaded with doxorubicin (Dox) have been constructed using the intercalation approach. This was achieved by incubation of Dox with RNA nanoparticles harboring GC rich sequence in the intercalation buffer because Dox can preferentially bind to the double-stranded GC sequence.⁷ Endo28-3WJ-Sph1-Dox nanoparticles showed increased cellular delivery and cytotoxicity compared to free Dox with the potential for ovarian cancer treatment. More recently, RNA nanoparticles with covalent conjugated drugs through click chemistry have been developed. Compared to intercalation, covalent conjugation increases the system stability and can precisely control the drug loading amount. A pRNA-3WJ-PTX nanoparticle loaded with one PTX per nanoparticle was constructed with three RNA strands, and one of these RNA strands has one PTX attached at the end. This RNA strand is synthesized with one alkyne group at the 5′ end which is used to react with PTX-azide via click reaction. The pRNA-3WJ-PTX was then used as the building block for RNA micelles which showed cancer cell inhibition effect in vitro.²¹ This technology is further developed which allows RNA nanoparticles to carry multiple drugs by using RNA strands synthesized with multiple alkyne positions for conjugation. 3WJs loaded with seven copies of camptothecin (CPT) have been constructed by self-assembly of RNA-CPT strands with multiple CPTs attached.⁶ Fully functionalized 3WJ-CPT-FA RNA nanoparticles were injected in the tumor bearing mice and showed significant suppression of tumor growth. Using the same approach, 4WJ nanoparticles loaded with 24 copies of PTX have also been constructed and demonstrated great tumor inhibition effect in triple negative breast cancer (TNBC) mice (Figure 19F).⁵ These studies demonstrated the great potential of RNA nanoparticles as a novel drug delivery platform to target deliver small molecule drugs for cancer treatment.

Figure 33.

RNA as a drug for disease treatment. (A) RNA nanoparticles loaded with small molecule anticancer drugs through intercalation and covalent conjugation. Adapted with permission from ref 5. Copyright 2016 Elsevier. Adapted with permission from ref 6. Copyright 2019 Piao et al. under CC By 4.0 http://creativecommons.org/licenses/by/4.0/. (B) RNA nanoparticles loaded with siRNA through bottom-up assembly and RCT. Adapted with permission from ref 28. Copyright 2017 American Chemical Society. Adapted with permission from ref 29. Copyright 2019 Elsevier Ltd. (C) RNA nanoparticles attached with SEQ with different size (left) and RNA nanoparticles with different shape (right) showed immune stimulation to different extent. Adapted with permission from ref 36. Copyright 2017 Guo et al. under CC By-NC-ND 4.0 https://creativecommons.org/licenses/by-nc-nd/4.0/. Adapted with permission from refs 37. Copyright 2019 Hong et al. under CC By 4.0 https://creativecommons.org/licenses/by/4.0/. Adapted with permission from ref 38. Copyright 2018 American Chemical Society.

Table 4.

RNA Nanoparticles Functionalized with Therapeutic Groups for Disease Treatment

RNA NP	therapeutic group	targeting group	disease	ref
Small Molecule Drug
3WJ	CPT	FA	cancer	6
4WJ	PTX	EGFRapt, FA	TNBC	5
RNA micelle	PTX	N/A	cancer	21
RNAi Drug and RNA Aptamer
3WJ	MED1 siRNA	HER2apt	breast cancer	28
3WJ	delta-5-desaturase siRNA	EpCAMapt	colon cancer	34
3WJ	BCRAA1 siRNA	FA	gastric cancer	12
3WJ	anti-miR21	EGFRapt	TNBC	9
3WJ	anti-miR21	CD133apt	TNBC	39
3WJ	anti-miR21	PSMAapt	prostate cancer	10
6WJ	PTX, miR122	GalNAc	HCC	49
siRNA nanoball	VEGF siRNA	N/A	AMD	69
BRC	USE1 siRNA	N/A	lung cancer	29
2-helix RNA origami	thrombin aptamer	N/A	blood coagulation	35,67

11.2. Delivery of Therapeutics Oligos and RNA Aptamers

RNA interference (RNAi) agents, including siRNA, miRNA, and anti-miRNA, have been studied extensively for their applications in the treatment of various diseases. siRNAs are described as double-stranded RNAs that utilize endogenous RNAi channels to elicit gene silencing. Recently, siRNAs were developed for use in 22 separate therapeutics involved in the treatment of at least 16 diseases, with the first clinical trial in 2004.⁵²² miRNAs are described as ~22 nt long, with the capability to subdue genetic expression post-transcriptionally by binding with mRNAs. Their target mRNAs are not generally singular in nature, but the seed region or the targeting region in miRNA design can be modified to enable a single miRNA to target multiple mRNAs, thereby greatly increasing its efficacy.⁵²³ However, their application has been limited by their poor cell entrance ability and unfavorable biodistribution profiles.⁵²⁴ RNA nanotechnology offers a solution for the delivery problem by incorporating RNAi agents into the nanoparticles, which can increase the delivery efficiency without introducing undesirable toxicity. Here, we summarized the current status of using different RNA nanoparticles to deliver RNAi agents for disease treatments (Table 4). Using bottom-up assembly approach, two siRNAs including siMED1-1 and siMED1-2 have been incorporated into the two helixes of 3WJ (Figure 33B),⁷ and the third helix is incorporated with HER2 aptamer that was used for targeted delivery. This functionalized nanoparticle showed targeting effect and tumor inhibition effect both in vitro and in vivo. RNA nanoparticles with the incorporation of anti-miRNA 21 have been constructed with different aptamers including EGFR aptamer and CD133 aptamer for TNBC treatment as well as PSMA aptamer for prostate cancer treatment.^9,10,39 RNA nanoparticles with miR122 has also been constructed which not only could provide synergistic effect to PTX but also could inhibit the drug efflux, thus increasing the therapeutic effect of PTX in hepatocellular cancer treatment.⁴⁹

RNA nanoparticles constructed with incorporation of different siRNA using RCT approaches, including RNAi microsponges and siRNA nanosheets, have been applied for the treatment of different diseases. In 2012, the RNAi-microsponge was reported as a new siRNA carrier for the first time.⁷⁵ More recently, bubbled RNA-based cargo (BRC) has been constructed with siRNA targeting UBA6-specific E2 conjugating enzyme 1 (USE1) for lung cancer treatment.²⁹ The in vivo study showed effectively suppressed transcription of target genes, which leads to tumor growth suppression in A549 tumor xenograft mice treated with BRC.

Similarly, RNA origami molecules have also been constructed with different therapeutic RNA fragments such as siRNA and RNA aptamer. A 3D scaffold RNA origami wireframe structure, RNA octahedron, with intrinsic functioning siRNA embedded, demonstrated efficient silencing of a target reporter (eGFP) induced by the release of siRNA (siEGFP) after dicer recognition and process in vitro.⁶² Self-folding, 2-helix RNA origami molecules bearing two thrombin binding RNA aptamers have also been reported (Figure 32B). These functional RNA origami molecules demonstrated strong binding affinity to thrombin, and their anticoagulant activity is more than 7-fold higher than free RNA aptamer.³⁵

In addition, RNA/DNA hybrid nanoparticles with the ability to controllable release RNA functionalities after triggering have been developed.^525–528 The controlled release requires two cognate nonfunctional units that contains one functional RNA strand and one complementary DNA strand with an overhanging toehold sequence. When two nonfunctional units 1 and 2 meets each other, their toehold sequence will form a zipped toehold, leading to the release and formation of RNA functionalities such as siRNA and RNA aptamer. These split functionalities were then incorporated into RNA/DNA hybrid nanoparticles with the cube structure to achieve the conditional activation of RNA interference in human cells.

11.3. RNA Nanotechnology and Immunotherapy

The desirable immune response can be used to treat diseases such as cancer. RNA nanoparticles have been used as the delivery platform for immunomodulators such as DNA and RNA oligos with specific sequences.⁴¹² The specific sequences can be rationally designed to be incorporated in RNA nanoparticles and synthesized together with the structural RNA oligos. RNA nanoparticles were first constructed with CpG sequences to trigger immune response.¹ CpG oligodeoxynucleotides are potent TLR9 agonists. RNA nanoparticles with different size, shape, and loading cargos showed different immune response in macrophage like cell models and in mouse models (Figure 33C). Besides the number and sequence of loaded sequences, the structure of the RNA nanoparticles could also be tuned to trigger different levels of immune stimulation. It can be divided into three categories: globular, planar, and fibrous nanoparticles. Compared to planar nanoparticles, 3D nanoparticles are more easily to trigger immune response. The immune response is evaluated mainly by the in vitro and in vivo measurement of cytokines. Later on, RNA nanoparticles with different size and shape with specific RNA sequences that can trigger immune response have been constructed.³⁶ Their immune stimulation effect has also been verified in vitro and in vivo.

Besides being used as the carrier for immune reagent delivery, the RNA nanoparticle itself can also be used as immunomodulator with specific internalization approach. Without internalization approaches such as ligand conjugation, specific internalization sequence incorporation, or external transfection method, RNA nanoparticles themselves do not trigger much immune response. The main reason that RNA nanoparticles will not trigger immune response is that they could not be internalized into cells due to their negative surface charge. Systematic studies have been done to investigate the immune response of different RNA nanoparticles using transfection agents.⁵²⁹ The immune response of RNA nanoparticles is mainly done by inducing the cytokine production. The results showed that the extent of immune responses triggered by different RNA nanoparticles is largely dependent on their physicochemical properties such as structure and composition. RNA nanoparticles with large size and 3D structure more easily trigger immune response. A corresponding program, QSAR, has been built to predict the immune response of a new nanoparticle based on its parameters.⁵²⁹ Mechanism studies have also been done in human primary blood cells to demonstrate that TLR is involved in the immune response triggered by RNA nanoparticles.³⁷

Toll-like receptors (TLRs) are single-pass membrane-spanning receptors that are involved in maintaining the innate immune system. TLRs are subgrouped into two subfamilies:⁵³⁰ TLRs expressed on the cell surface (TLR1, 2, 4, 5, 6), recognizing molecules from bacterial, fungal, and protozoan components, and TLRs localized to specialized endosomal compartments (TLR3, 7, 8, and 9), detecting the presence of nucleic acids.⁵³¹ TLR-induced activation of the transcription of mRNAs coding for several cytokines and chemokines that continuously stimulate autoreactive B cells.⁵³² Previous studies have identified the potential role of TLR7 and 9 in the initiation of the interferon response to RNA nanoparticles such as RNA cubes and rings.

The immune stimulation of RNA nanoparticles can be controlled to achieve different aims. The immune inert property of RNA nanoparticles is beneficial for them to be used as safe drug delivery system with minimal immune toxicity. However, RNA nanoparticles could also be used to trigger immune response with the incorporation of specific DNA/RNA sequences and with specific physiochemical properties. RNA nanoparticles could be constructed with designed immune stimulation effect by modulating their structure, sequence, and composition, which could be further used in clinical application for immunotherapy. Also, RNA aptamers that can trigger immune responses have been reported, they can also easily be incorporated into RNA nanoparticles to achieve desirable immune stimulation effect.

12. TARGETING RNA FOR DISEASE TREATMENT

12.1. Therapeutic Oligonucleotides

12.1.1. Antisense Oligonucleotides.

Antisense oligonucleotides (ASOs) are usually small (~18–30 nucleotide), single-stranded, highly modified, and stabilized nucleic acid analogues. ASOs can be subdivided into two major categories in terms of their mechanism of action: occupancy-mediated degradation and occupancy-only (Figure 34A). The mostly studied and used mechanism of occupancy-mediated degradation includes the recruitment of RNase H to degrade the target mRNA after the binding of ASO. Gapmer ASOs (or chimeric ASOs), which contain a central DNA sequence and flanked by 2′-modified nucleotides, can form DNA/RNA hydrides with the target RNA and cause degradation of the RNA strand by recruiting RNase H. The occupancy-only mechanism includes three types of downregulation mechanisms. ASOs can also bind to the transcript as a steric block, leading to direct blockage of mature mRNA translation. They can also inhibit the translation of target mRNA by guiding the cleavage of cap structures on mRNA or altering polyadenylation on pre-mRNA. The steric block mechanism could also cause the upregulation of desirable proteins by regulating the splicing of precursor mRNA without inducing degradation. The first ASOs have been granted FDA approval is fomivirsen, which is used for the treatment of cytomegalovirus retinitis in immunocompromised patients. Later on, two other RNase H-mediated ASOs, including mipomersen and inotersen, as well as three splice-switching ASOs, including eteplirsen, golodirsen, and nusinersen, have received regulatory approval.

Figure 34.

Drugs that target RNA for disease treatment. (A) Regulation mechanism of ASOs including alternative splicing, degradation by RNase H (pink), alternation of polyadenylation, and inhibition or enhancement of translation (left). Regulation mechanism of siRNAs for mRNA degradation after forming RISC complex (proteins in blue, yellow, and green) (right).⁴⁵ (B) 3D module of neomycin B (green) binding to RNA (gray) within bacterial ribosome (protein in blue). (C) Design principles of small molecules targeting RNA from human genome sequence by the Inforna approach. Reproduced with permission from ref 68. Copyright 2019 American Chemical Society. Reproduced with permission from ref 71. Copyright 2020 Elsevier Ltd.

12.1.2. siRNA-based Drugs.

Another type of therapeutic oligonucleotides besides ASO are RNA interference drugs such as siRNAs. siRNAs are double-stranded RNA fragments that are normally 20–27 base pairs long, which contain one guide and passenger strand. After interacting with a multiprotein RNA-induced silencing complex (RISC), the passenger strand of siRNA will be degraded. The guide strand will then complement to the target transcript, which leads to the regulation of gene silencing through cleavage of target mRNA (Figure 34A). There are currently three FDA approved siRNA drugs: patisirna for hereditary transtheyretin-mediated amyloidosis (hATTR), givosiran for acute hepatic porphyria (AHP), and lumasirna for primary hyperoxaluria type 1. Although this area has been rising rapidly, there are still issues with the current therapeutic oligos including the stability, effective delivery, and off-target effect.⁵³³ The stability issue has been largely solved by introducing different chemically modified nucleotides into the RNA oligonucleotides. The effective delivery issue has been solved to some extent by using targeting ligand such as GalNAc and nanoparticles such as liposomes.⁵²⁴ As for the off-target effect, different approaches such as local injection and strand length change have been tried.

12.2. Small Molecule Drugs Targeting RNA

Besides oligonucleotides, researchers have also been trying to develop small molecules that can target RNAs for therapeutic effects (Figure 34B). To date, most work has focused on therapeutically well-validated targets and there has been a notable recent focus on screening against RNA targets with drug-like lead compounds.⁵³⁴ Despite the successes in the antibacterial and antiviral field, the development of small molecules targeting mammalian RNA is still at an early stage.⁵³⁵ However, a growing number of small-molecule ligands targeting different classes of RNA have been identified, leading to burgeoning interest in the field.⁶⁸ The overarching challenge is to identify disease-causing RNAs that have known structure and appropriate ligand-binding site as well as the selection of RNA-binding ligands with high affinity and specificity.

12.2.1. Design and Identification of Small Molecules Targeting RNA.

Successful targeting of RNA with small molecules for therapy requires three distinct components: (1) identification of target RNA with disease-related function, (2) screening of RNA binding drug-like molecules with high binding affinity and specificity, and (3) validation of the biological function change of target RNA after ligand binding.⁵³⁴ The design of RNA ligands including two categories: sequence-based and structure-based design. Two-dimensional combinatorial screening (2DCS) approach is to identify small molecules binding RNA secondary structure in a high-throughput manner. RNAs that bind to certain small molecules in the microarray are identified by methods such as RNA-seq and the RNA-small molecule interaction will then be scored through statistical methods such as structure–activity relationships through sequencing (StARTS). Structure-based approaches are emerging other essential tools that can accelerate drug screening process for RNA targets (Figure 34C). Such strategies utilize the information on known RNA secondary and tertiary structures and the currently existing RNA ligand. Various small molecules have been designed to target HIV RNA, miRNA precursors, RNA repeat expansions, and riboswitches.⁵³⁶

12.2.2. Structural and Functional Validation Strategies.

To provide methods for the validation of the direct interaction between RNAs and small molecules, various approaches have been developed, ranging from cross-linking and cleavage to altering RNA sequences and nuclease recruitment.⁶⁸ Not only have these approaches established which RNA targets are directly engaged by small molecules, but they have also helped to define factors that influence bioactivity. In chemical cross-linking and isolation by pull down (Chem-CLIP), a small-molecule RNA binder is appended with a cross-linking module and a tag for purification.⁵³⁷ The RNA targets that are engaged with the binding of small molecule will be purified and then analyzed by RNA-seq or RT-qPCR. Binding sites can be deduced by similar approaches that include digestion of the bound RNA fragments followed by RT-qPCR analysis with a set of gene-specific primers. In addition to cross-linking, cleavage approaches have been used to define the RNA targets of small molecules in cells. In RiboSNAP-Map, RNA binding modules are appended with bleomycin A5, which facilitates the cleave of target RNA both in vitro and in cells.⁵³⁸ RiboSNAP-Map was applied to oncogenic pri-miR-96 using a dimeric small molecule, TGP-96. The RiboSNAP-Map approach provides an approach to map the binding sites of small molecules in cellular RNAs via RT-PCR amplification of total RNA, which is complementary to Chem-CLIP. Besides the identification of small molecules with high binding affinity to the target site, additional experiments are required to validate the on and off target effect of selected small molecules. A first assessment in cells includes measuring the changes in the transcript level caused by small molecule binding to target RNA. Further validation of changes in the proteome via LC-MS/MS as well as reversing phenotype should also be performed.⁶⁸

13. CONSTRUCTION OF STABLE AND TRANSLATABLE MRNA FOR MRNA VACCINES

The mRNA vaccines⁵³⁹ have been found to be efficient in inducing antibody or cellular immunity. With the aid of ligand-assisted targeting, the mRNA can be delivered to specific antigen processing cells, including macrophages, B-cells, and dendritic cells that can present foreign antigens to helper T-cells for the production of antigens efficiently. Currently, mRNA vaccines include viral vaccines^540,541 and cancer vaccines.⁵⁴²

13.1. The mRNA Vaccine for SARS-CoV-2

The recent outbreak of a novel coronavirus disease (known as COVID-19) shocked the world.^543–545 The causative agent is named severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). On March 11th of 2020, the WHO declared COVID-19 as a pandemic. As of April 16, 2021, the COVID-19 viral disease has swept into over 200 countries, causing 140 063 831 illnesses and 3 004 146 deaths. Currently, there are more than 160 SARS-CoV-2 vaccine candidates in preclinical trials.^412,546 Among them, Moderna and Pfizer’s mRNA vaccine candidates, China’s inactivated vaccine candidates, and UK’s chimpanzee adenovirus-vectored vaccine are now in phase III clinical trials.⁵⁴⁰ Coronavirus SARS-CoV-2 is causing a life-threatening pandemic pneumonia.^{544,547–550} Safe vaccines that rapidly induce potent and long-lasting virus-specific immune responses are urgently needed. The state-of-art mRNA (mRNA) vaccine strategy has several advantages compare to conventional vaccines, including stronger immune potency, faster R&D development, lower cost, NS fewer barriers in large-scale manufacturing and safer administration.⁵⁵¹ However, mRNA instability and endosome entrapment are two major challenges for vaccine delivery.⁵⁵²

In the U.S., the first clinical trial of a lipid nanoparticle (LNP)-based mRNA vaccine candidate, mRNA-1273, encoding a stabilized prefusion of S protein with two prolines (S-2Pro), developed by Moderna was initiated on March 16, 2020.^553,554 The mRNA-1273 vaccine induced SARS-CoV-2 specific immune responses in all participants, and no trial-limiting safety concerns were identified.⁵⁴¹ Separately, Pfizer developed two LNP-based mRNA vaccine candidates, BNT162b2 and BNT162b1, which encodes an optimized SARS-CoV-2 full-length S and receptor binding domain (RBD) trimers, respectively. Phase I/II clinical trials showed that humans immunized with these mRNA vaccine candidates developed SARS-CoV-2 neutralizing antibodies.⁵⁵⁵ The mRNA-based vaccine is one of the most promising vaccines against SARS-CoV-2. The encapsulation of the mRNA into a liposome allows for the protection of mRNA during delivery. However, a concern with liposome delivery vehicles is the need for endosome escape pathways and a significant portion of the mRNA cargo being degraded rather than reaching the cytosol of cells.

13.1.1. The mRNA Coding for the Prefusion S Is Ideal for SARS-CoV-2 Vaccine Development.

The coronavirus (CoV) spike (S) protein is the main target for neutralizing antibodies for protection against future infection. The S protein is a “class 1” fusion protein trimer that possesses both receptor binding and fusion activity.^556–558 The ectodomain of S consists of the S1 subunit that contains the receptor-binding domain (RBD) and the S2 subunit that contains the membrane-fusing mechanism.^556–558 Both S and its RBD domain have been shown to be immunogenic for many CoVs.^559–561 The native S in the virion is in its “prefusion” form. Upon triggering, the prefusion S undergoes significant conformational changes to arrive at its postfusion S form. It is believed that antibodies to the “prefusion” form of S protein have significantly higher neutralizing activity than antibodies to their “postfusion” forms.^540,558,562 Thus, mRNA coding for the stabilized version of the prefusion S protein can be used for SARS-CoV-2 vaccine development.

13.1.2. Capping of mRNA to Ensure Stability and Intracellular Expression of Antigen.

Eukaryotic mRNA expression requires the presence of a cap on the 5′ end of the mRNA for the translation into protein. The vaccinia virus capping enzyme is required to produce stable mRNA for high efficiency transcription in eukaryote systems. The 5′-terminal cap structure m7G(5′)pppN- is a typical feature of eukaryotic mRNAs that is required for translation and stability.³³⁰ The cap is formed by posttranscriptional modification of the nascent mRNA by an enzyme with RNA guanylyltransferase and mRNA (guanine-N7-)-methyltransferase activities that was first isolated from vaccinia virus.⁵⁶³ This multifunctional enzyme can remove the terminal y-phosphate from an RNA chain, transfer the GMP moiety of GTP to the now diphosphate-ended RNA, and then transfer a methyl group from S-adenosylmethionine (AdoMet) to position 7 of the added guanosine.^564–566 An enzyme-guanylate intermediate, in which a GMP residue is covalently attached to the large subunit of the enzyme via a phosphoramidate bond to a lysine E amino group, was demonstrated.^567,568 The poxviruses, of which vaccinia virus is the prototype, are unique in their ability to replicate and express their genes in the cytoplasm of infected cells.⁵⁶⁹ All of the enzymes and factors necessary for the formation of capped and polyadenylylated mRNA are contained within the infectious virus particle. The vaccinia virus capping enzyme is a 127 kDa protein composed of two subunits of 95 and 31 kDa⁵⁶³ that are thought to be encoded by the D1 and D12 open reading frames (ORFs), respectively.^329,570 Expression of either D1 or D12 alone or mixing of the purified D1 and D12 in vitro did not lead to the capping activity.³³⁰ Coexpression of both ORFs to form the heterodimer was needed for stability and high activity. However, coexpression of the D1 and D12 in the same cell led to the aggregation and lost the capping activity. Guo et al. developed a novel method by coexpression and ping-ponging the expression to produce an active D1/D2 complex that has capping activity.³³⁰ The large subunit alone was shown to have guanylyltrans. By coexpression of these two ORFs in Escherichia coli, Guo et al. obtained the recombinant functional D1/D2 capping enzyme that can be used for the capping of mRNA. The recombinant protein is a valuable reagent for specifically labeling the 5′ ends of mRNA and enhancing their translatability in mRNA vaccine development.

13.1.3. Chemical Modification of mRNA for Maximized Protein Expression.

Moderna’s current mRNA vaccine (mRNA-1273) is a LNP-based stabilized prefusion S protein.^541,571 This protein, S-2Pro, was stabilized by replacing two amino acids (K986 and V987) with prolines, mutating the furin cleavage site to prevent cleavage between S1 and S2 and replacing the C-terminal transmembrane/cytoplasmic tail (TM/CT) domain with a T4 fibritin self-trimerizing domain. Recently, a more stable soluble prefusion S protein, HexaPro (S-6Pro), expressed 3-fold more than S-2Pro.⁵⁷² HexaPro is S-2Pro with 4 additional strategic amino acids replaced with prolines and comes from the McLellan lab.⁵⁷² Pfizer mRNA vaccine BNT162b1 is based on RBD trimers by addition of a T4 fibritin-derived “foldon” trimerization domain.

13.2. Cancer mRNA Vaccines

Cancer mRNA vaccines are among the latest therapeutic methods in cancer immunotherapy.^101,573 Different from immune checkpoint inhibitors and adoptive cell therapy, cancer vaccines may be more effective and safer in preventing and treating cancer. The vaccine teaches the immune system to recognize cancer cells as foreign and therefore need to be eliminated. There are some special proteins markers known as cancer antigens displayed on surface of cancer cells. By targeting these cancer antigen proteins, the immune system can specifically eliminate cancer cells without harming healthy cells. Moreover, the vaccine can prevent the recurrence of cancer and remove the remaining cancer cells after treatment. In the past 30 years, cancer vaccines have been extensively investigated in animal models involving many different types of cancers. In 2020, cancer vaccines for various types of cancer have entered clinical trials.

One cancer mRNA vaccine is the mRNA-4157 vaccine developed by Moderna in the United States. The Moderna personalized cancer vaccine (PCV) mRNA-4157 is a vaccine candidate that is a combination of several genes into a single mRNA concatemer. On November 11, 2020, Moderna announced the latest clinical trial data code-named KEY-NOTE-603, suggesting that the initial development of a cancer vaccine for solid tumors was successful. It is believed the vaccine can trigger the immune system, which translates into a more sensitive response to PD-1 inhibitors and reduces the risk of cancer recurrence. Their interim phase 1 trial data is the vaccination with mRNA-4157 in combination with the anti-PD-1 drug pembrolizumab. The data demonstrated that mRNA-4157 combining with the drug is well tolerated by patients and produced responses shown by tumor shrinkage while having no adverse effects.

13.3. Endosome Entrapment Limits mRNA Vaccine Development

mRNA instability and endosome entrapment are two major challenges for the realization of an effective mRNA vaccine delivery. The good news is both Moderna and Pfizer used LNP to deliver mRNA vaccine which significantly enhanced the encapsulated mRNA stability.⁵⁷⁴ However, lipofected mRNAs must first travel through endosomes until released into the cytosol. The degradation and sequestration of mRNA in endosomes reduces final expression of the target protein in cytosol.^{102,574–577} It has been reported that LNP-based mRNA vaccine delivery has limited capability to escape endosome trapping due to the fact that a major proportion of LNP (more than 95%) is endocytosed within half an hour.^100,576 It is estimated that <2% of the RNA delivered via LNP escapes the endosomes to reach the cytosol.^100,578 Thus, endosome entrapment remains a challenge for SARS-CoV-2 mRNA vaccine development.

13.3.1. Projected Use of Exosomes for mRNA Delivery to Overcome Endosome Entrapment.

As described above in detail, exosomes or extracellular vesicles are derived from late endosome/multivesicular body (MVB), and the Peixuan Guo lab has proven their use for delivery siRNAs. We believe that such exosomes could encapsulate and deliver mRNAs to offer unique advantages as a mRNA vaccine (Figure 29A), due to their abilities to fuse with plasma membrane or organelle membranes,^8,576,579 they are able to escape the endosome^8,22,576 and efficiently deliver cargos to the cytosol for protein translation.^576,580 Thus, exosome-based delivery can overcome the endosome entrapment problem, the bottleneck problem for RNA-based therapy, drug delivery, and vaccine development.^22,545,580 The exosome-based RNA delivery system has significantly improved RNA-based therapeutic effects against a wide range of human cancers.^22,545 Furthermore, the exosomes have been produced with ligand-displaying RNA nanoparticles on the exosome surface for target-specific delivery of RNA.^8,22,581 We predict to use exosomes to for specific delivery of mRNA coding for spike S (S) or receptor-binding domain (RBD) antigen of SARS-CoV-2 to dendritic cells as vaccine. An exosomal vaccine could increase the delivered mRNA, allowing for higher efficiency in S protein production and stronger immune response. RNA nanotechnology has the ability to serve as both a therapeutic target and therapeutic itself.

14. CONCLUSION

The biological macromolecules DNA, RNA, and proteins are intrinsically polymeric in nature and have self-assembling properties from their components to form into nanostructures and thus serve as powerful building blocks for the fabrication of nanostructures and nanodevices in material science and medicine. RNA is an anionic polymer ubiquitously synthesized in cells.¹³² Earlier, RNA was categorized into mRNA, tRNA, and siRNA; however, the categorization was rewritten with the discovery of ribozymes.¹²¹ As early as 1987, Peixuan Guo predicted that there are a large number of unknown novel small RNA species in cells and dubbed this category as sRNA and published in Science.¹⁶ Sequencing of the human genome revealed that about 98.5% of the genome does not encode for proteins but rather produce short and long noncoding RNAs that regulate cell activity. Thus, RNA plays a very critical role in the human body. In addition to chemical and protein drugs, it was predicted as early as 2014 that RNA drugs or RNA-implanted drugs would become the third milestone in pharmaceutical drug development.⁴⁵²

RNA nanostructures can be synthesized with a simplicity similar to DNA while having elastic nature and versatility in function as proteins. Although RNA nanotechnology holds some similarity to DNA, there are important differences between the two, including the thermal stability and structural diversity. RNA is thermodynamically more stable than DNA and displays a higher T_m than DNA.^41,46 In addition, magnesium supports RNA folding but is not required for highly stable structures, while the assembly of DNA nanoparticles typically requires at least 10 mM magnesium. The fact that the human body only has 1 mM magnesium makes RNA nanoparticles more suitable for in vivo applications.⁴² The noncanonical base parings in RNA, which is not in DNA, leads to structural versatility and the generation of various strong RNA structural motifs.²⁵⁰ An RNA strand 100 nt in length can fold into 1.61 × 10⁶⁰ different structures. The formation of loops, bulges, hairpins, and stem-loops as dovetail enables RNA to become diverse and act as LEGOs for the construction of complicated RNA nanoarchitectures via self-assembly. RNA molecules, such as aptamers,⁴⁰³ ribozymes,¹²¹ and siRNA,¹⁵⁸ can have special functions and can be included in the RNA nanoparticles.

Standing in awe of the sensitivity of RNA to RNase degradation has made many people mistakenly flinch away from RNA nanotechnology. Actually, the inclusion of 2′-modifications or 3′−5′ locking modifications in RNAs not only makes it stable to nuclease degradation but also improves its thermodynamic stability.^20,41,95 This facilitates the construction of RNA nanoparticles with controllable shape, size, and stoichiometry.^60,76,164

The multivalency,^14,451 anionic nature,¹³² passive and active targeted delivery,¹²⁰ rubbery property for rapid cancer accumulation and fast renal clearance,³³ favorable nanoscale size to avoid organ trapping,^27,120 favorable PK/PD and biodistribution,¹¹⁸ controllable immunogenicity,³⁶ and undetectable toxicities^12,118 make RNA a unique material in therapeutics, which recently took giant leaps forward with the recent FDA approval of several RNA-based drugs.⁵⁸² Additionally, due to the active targeting of RNA nanoparticles, they have a strong potential to serve in disease diagnosis by the inclusion of imaging agents and binding to disease markers.⁶⁴ RNA nanoparticle beacons have also been created that can serve as diagnostic tools for monitoring gene expressions related to diseases.^583–586 Outside of biomedical applications, RNA nanotechnology has grown into the field in the development of logic gates,^135,587,588 as well as the construction of resistive memory by conjugation to quantum dots.^134,586 These could potentially translate into computation devices in vitro for in the cell.⁵⁸⁷ The potential expansion of the application of RNA nanotechnology in engineering, electronics, semiconductors, computers, sensing, and NEM are expected.

15. PERSPECTIVE: OVERCOMING CHALLENGES IN RNA NANOTECHNOLOGY

15.1. RNA Serving as a Drug Target

As predicted, RNA is becoming the third milestone in pharmaceutical drug development.⁴⁵² This includes two subjects: (1) RNAs as drugs such as siRNA,¹⁵⁸ miRNA,⁵⁸⁹ aptamer,⁴⁰³ ribozyme,¹²¹ and riboswitch,⁵⁸⁷ as well as (2) chemicals or other drugs targeting RNA. In its current state, the field of RNA as drugs is quickly becoming fully realized.⁵⁸² As mentioned earlier, only 1.5% of the human genome codes for proteins, but the rest of the human genome is mainly responsible for noncoding small or long RNAs, providing vast targets for drugs. The subject of RNA serving as a drug has been extensively explored; however, the subject of drugs targeting RNA has been slowing initiatation and is still elusive due to the uncertainty of the nature of RNA. For example, whether the interactions of RNA with potential drugs rely on induced fit or conformational capture remains to be further investigated. A drug predicted by structural computation or in vitro screening may disassociate from the substrate RNA in vivo due to induced conformational changes in the RNA conformation. This will result in low selectivity due to the landscape of conformations induced by the binding of the chemicals. The studies of RNA thermodynamic stability and the finding of drug targeted to the thermostable RNA motif is critically important to elevate the feasibility of potential drugs to inactivate RNA in vivo.^27,41,42

15.2. Prediction and Computation of RNA 4D Structure

A distinct property of RNA is its unique characteristic in folding into different structures. The traditional studies of RNA focus on intramolecular structuring, folding, and nucleotide interactions However, RNA nanotechnology focuses on the study on intermolecular interactions among RNAs. The inter-RNA interaction requires the knowledge of the 4D structure of the RNA architecture. In 2010, it was predicted that the development of 4D predictive studies on RNA folding would be an emerging task.²⁷ During the last 10 years, intense progress has been shown on RNA 3D predictions, building folding libraries, and structural computation, as discussed in detail above.^220,245,260 However, the 4D study on RNA structure and folding has just emerged and has a long way to go. The modeling of 4D structures is pivotal to RNA nanotechnology, as RNA nanostructures rely on interactions and structuring between multiple oligos. This potential technology will include the use of artificial intelligence, machine learning, big data, and atomic computation for RNA quantum structures. Additionally, the in vivo computation of RNA is an exciting field in RNA nanotechnology. The in vivo tasks include the use of RNA sequences and interactions as parameters to build logic gates.^135,587,588 It also includes the prediction and construction of RNA nanoparticle in vivo for 4D folding and assembly using fermentation industry-scale production.^164,590 The complete understanding of RNA nanoparticle assembly, formation, and structure in vitro and in vivo will allow for the better designs and construction of RNA nanoparticles in the future.

15.3. Cost and Yield of Production of RNA Nanoparticles

RNA production by biological approaches and solid-state synthesis each have their advantages and disadvantages. For biological synthesis of RNA by T7 RNA polymerase, to be used in clinical applications with long RNA molecules, the main challenge is the contamination of the biological side products that may cause toxicity, immune responses, and pyrogenic effects. These effects will jeopardize FDA approval and clinical applications at later stages. This problem can be improved by using fermentation for the industrial scale production and purification. RNA nanoparticles can be produced in bacterial cells by using implantation of plasmids incorporating ribozymes for self-cleavage and self-assembly into nanoparticles during transcription in the bacterial cells or by in vitro fermentation.¹⁶⁴ Large-scale RNA complexes produced in bacteria escorted by a tRNA vector have been reported.^591,592 Large-scale RNA production by rolling circle replication has also be reported.^75,521,593 On the other hand, concerning the solid-state chemical synthesis, its major hurdle is the premature termination of the RNA synthesis that creates contamination of the shortened RNAs. Commercial RNA chemical synthesis of RNA has significantly improved over the past decade by increasing the nucleotide incorporation efficiency and the yield of production. This in turn resulted in the synthesis of RNA oligos up to 120 bases in length. However, with a 98.5% incorporation efficiency of each nucleotide for a 120-nucleotide RNA strand, RNA oligos will result in 84% contamination of aborted byproducts [1 - (98.5%)¹²⁰ = 84%]. Thus, RNA nanotechnology is one of the remedies because large RNAs can be assembled from small RNA fragments via bottom-up self-assembly. RNA nanoparticles are typically composed of short building blocks that self-assemble like Lego that allows for the creation of the next generation of RNA nanoparticles. Industrial companies have seen the growing interest of RNA for therapeutics and have now developed facilities to produce RNA on the kilogram scale per batch and under GMP standards that can allow for the clinical application of RNA nanotechnology. On the basis of the rapid reduction of cost over the history of DNA synthesis, it is expected that the cost of RNA synthesis will gradually decrease with the development of industrial-scale RNA production technologies.

Biographies

Daniel W. Binzel is a research scientist at The Ohio State University College of Pharmacy. He received his B.E. degree in Chemical Engineering from Miami University in 2010 and Ph.D. degree in in Pharmaceutical Sciences from the University of Kentucky in 2016. He completed his postdoctoral training at The Ohio State University before becoming a Research Scientist in 2018. His research has focused on developing novel RNA nanoparticles and understanding their assembling mechanisms and stabilities for in vivo applications. Daniel has implemented these branched RNA nanoparticles for the treatment of cancers focusing on the delivery of RNAi-based therapeutics to prostate and breast cancer models.

Xin Li received her B.Sc. in Pharmaceutical Analysis at the China Pharmaceutical University and is currently a Ph.D. candidate in the Pharmaceutical Sciences program at The Ohio State University. Her research is focused on the development of complex RNA nanoparticles with high thermostability and high complexity for the loading of chemical and RNAi-based therapeutics. She has become an expert in RNA nanoparticle design for applications of chemical drug delivery towards cancer treatments.

Nicolas Burns is a Research Associate at The Ohio State University in the College of Pharmacy. He completed his undergraduate degree in Biotechnology and Molecular Bioscience at the Rochester Institute of Technology. His research has focused on novel antibiotic discovery and bioremediation at RIT and more recently has shifted to applying the Phi29 bacteriophage for novel applications like single-molecule pore sensing and RNA nanotechnology.

Eshan Khan received his Ph.D. at the Indian Institute of Technology, Indore, India, and is currently a postdoctoral researcher in the laboratory of Dr. Wayne Miles in the Comprehensive Cancer Center at The Ohio State University. His research is focused on the understanding of RNA–protein interactions during post-transcription during cancer progression.

Wen-Jui Lee is currently a Ph.D. student in the laboratory of Dr. Yuan-Soon Ho in the School of Laboratory Science and Biotechnology at Taipei Medical University.

Li-Cheng Chen is an associate research fellow of the TMU Research Center of Cancer Translational Medicine, Taipei Medical University, Taipei, Taiwan. She received her Ph.D. at the Graduate Institute of Medical Science, College of Medicine, Taipei Medical University. Her research is focused on understanding the underlying mechanism of RNA nanoparticle in breast cancer cells and PDX models.

Satheesh Ellipilli is currently a postdoctoral researcher in the laboratory of Dr. Peixuan Guo in the College of Pharmacy at The Ohio State University. He received his Ph.D. in Chemistry from IISER, Pune, India, followed by postdoctoral research at the University of Utah and Emory University under Jennifer Heemstra. He has dedicated his research to nucleic acid chemistry in the synthesis of nucleotides and nucleic acids. More recently, he has applied his strong chemical knowledge toward the construction of stable RNA nanoparticles for the application toward oncologic therapeutics.

Wayne Miles is an Assistant Professor in the Department of Cancer Biology and Genetics at The Ohio State University and part of the Center of RNA Biology. He received his Ph.D. from the University of Manchester, UK, and then completed his postdoctorial training at Harvard Medical School and Massachusetts General Hospital. He began his independent laboratory at The Ohio State University Comprehensive Cancer Center in 2016, studying the role of RNA structure and RNA binding proteins in cancer initiation, progression, and metastasis.

Yuan Soon Ho completed his doctorate in Biochemistry at the National Taiwan University in 1994 and then joined the faculty at Graduate Institute of Biomedical Technology, Taipei Medical University, where he directed research combining genomics and breast cancer research with translational research. Dr. Ho is currently the Associate Director of the TMU Research Center of Cancer Translational Medicine, Taipei Medical University (2014–present). His research team found that α9-nicotinic receptor (α9-AChR) was detected in the breast tumor tissues (J. Natl. Cancer Inst. 2010, 102, 1322–1335⁵⁹⁴). This finding led to the development of powerful antitumor agents, and the cloning of nicotinic receptor gene linked to breast cancer and now being investigated as a potential therapeutic target for smoking-induced breast and other cancers. On the basis of α9-AChR study in breast cancer, Dr. Ho recently (Nature Commun. 2019103131, https://www.nature.com/articles/s41467-019-10920-8⁵⁹⁵) presented a systematically integrated method to generate a resource of cancer membrane protein-regulated networks, containing 63 746 new protein–protein interactions for 1962 membrane proteins, using expression profiles from 5922 tumors with overall survival outcomes across 15 human cancers. Dr. Ho is also invited as an Associate Editor of the Journal of Food and Drug Analysis and editor of the Scientific Report (oncology field) and Plos One journals. He is a member of National Academy of Inventors (USA, 2016–present), the director of academic committee in Taiwan Tea Association (2012–present), and the member of council (director of the academic committee) in the Taiwan Oncology Society (2015–2019).

Peixuan Guo, the pioneer of RNA nanotechnology, has held three endowed chair positions at three prestigious universities and is currently the Sylvan G. Frank Endowed Chair in Pharmaceutics and Drug Delivery at OSU, the director of the OSU Center of RNA Nanobiotech and Nanomedicine, and the president of the International Society of RNA Nanotech and Nanomed (ISRNN). He received his Ph.D. from the University of Minnesota and postdoc training from the NIH. He joined Purdue in 1990, tenured in 1993, became a full professor in 1997, and honored as a Purdue Distinguished Faculty Scholar in 1998. He was Director of the NIH Nanomed Development Center (NDC) 2006–2011 and director of the NCI Cancer Nanotech Platform Partnership Program 2012–2017. He has 230 high-impact publications. He first proposed the concept that a large number of novel but undiscovered small RNA existed in cells and named them small RNA (sRNA) (Guo, P., et al. A small viral RNA is required for in vitro packaging of bacteriophage-phi29 DNA. Science 1987, 236, 690⁵⁹⁶), constructed the first viral DNA packaging motor (PNAS 1986), discovered phi29 motor pRNA (Science 1987), revealed pRNA hexamer (Mol. Cell 1998 pioneered NA nanotechnology (Mol. Cell 1998, featured in Cell 1998, four papers in NatureNanotechnol. 2009, 2010, 2011, 2018; Nature Commun., 2019), invented a method for the assembly and production of themRNA capping enzyme (PNAS, 1990) that is currentlyused in COVID-19 mRNA vaccine manufacturing, developed a TIRF Systemto count single-fluorophores (EMBO J. 2007), incorporated-phi29 motor channel into membrane (Nature Nanotechnol. 2009) for single pore sensing and DNA sequencing, discovered a thirdclass of biomotor using a revolution mechanism, developed a new approach for ultrapotent drug development, and found that RNA is like rubberand amoeba with unusually high efficiency for tumor targeting and quick kidney clearance and thus undetectable toxicity. Dr. Guo hasreceived the Pfizer Distinguished Faculty Award, the Purdue FacultyScholar Award, the Lions Club Cancer Research Award, DistinguishedAlumni of the University of Minnesota, and 100 Years DistinguishedChinese Alumni of the University of Minnesota. Dr. Guo is Editor or an editorial board member of 7 journals, was reported numerous times by TV such as ABC and NBC, was featured by NIH, NSF, MSNBC, NCI, and Science Now, a member of two prominent national nano biotech initiatives by NSF, NIH, the National Council of Nanotechnology, and NIST, was a member of two NIH steering committees in NDC and Extracellular RNA, was on the NCI Intramural Research Site Visit Committee two times, was a member of the Selection Committee of the 2019–2020 LifetimeAchievement Award for the American Association for Cancer Research(AACR), is Chairman of the Board of Directors of ExonanoRNA LLC, and was honored as The Ohio State University top “Innovator of the Year” for 2021.