key: cord-1014510-iuiyky8o
authors: Schlich, Michele; Palomba, Roberto; Costabile, Gabriella; Mizrahy, Shoshy; Pannuzzo, Martina; Peer, Dan; Decuzzi, Paolo
title: Cytosolic delivery of nucleic acids: The case of ionizable lipid nanoparticles
date: 2021-03-20
journal: Bioeng Transl Med
DOI: 10.1002/btm2.10213
sha: 8ccf4b5ebece9c3f744b62add1d3f87dd560cf59
doc_id: 1014510
cord_uid: iuiyky8o

Ionizable lipid nanoparticles (LNPs) are the most clinically advanced nano‐delivery system for therapeutic nucleic acids. The great effort put in the development of ionizable lipids with increased in vivo potency brought LNPs from the laboratory benches to the FDA approval of patisiran in 2018 and the ongoing clinical trials for mRNA‐based vaccines against SARS‐CoV‐2. Despite these success stories, several challenges remain in RNA delivery, including what is known as “endosomal escape.” Reaching the cytosol is mandatory for unleashing the therapeutic activity of RNA molecules, as their accumulation in other intracellular compartments would simply result in efficacy loss. In LNPs, the ability of ionizable lipids to form destabilizing non‐bilayer structures at acidic pH is recognized as the key for endosomal escape and RNA cytosolic delivery. This is motivating a surge in studies aiming at designing novel ionizable lipids with improved biodegradation and safety profiles. In this work, we describe the journey of RNA‐loaded LNPs across multiple intracellular barriers, from the extracellular space to the cytosol. In silico molecular dynamics modeling, in vitro high‐resolution microscopy analyses, and in vivo imaging data are systematically reviewed to distill out the regulating mechanisms underlying the endosomal escape of RNA. Finally, a comparison with strategies employed by enveloped viruses to deliver their genetic material into cells is also presented. The combination of a multidisciplinary analytical toolkit for endosomal escape quantification and a nature‐inspired design could foster the development of future LNPs with improved cytosolic delivery of nucleic acids.

1 | INTRODUCTION

The notion of exploiting nucleic acids (NAs) as therapeutic molecules was conceived for the first time in 1966, in a perspective paper that evoked the possible use of viruses in genetic studies and for gene therapy. 1 However, only in the 1990s, this notion was translated into practice by a series of findings setting the stage for its use in biomedical research and, eventually, in clinical settings. First, in 1990, Wolff et al. demonstrated that the direct intramuscular injection of an in vitro transcribed (IVT) messenger RNA (mRNA) could lead to the expression of the encoded protein. 2 Then, in 1993, the first miRNA was identified in Caenorhabditis elegans and, shortly after, the first mammalian miRNA-let-7-was discovered. 3, 4 Finally, in 1998, Fire and Mello discovered a fundamental mechanism in gene regulation based on RNA interference (RNAi), that was eventually acknowledged with the Nobel Prize in Physiology or Medicine in 2006. 5 Twenty years later, in 2018, the US Food and Drug Administration (FDA) approved the clinical use of patisiran (Onpattro, Alnylam Pharmaceuticals), the first-ever small interfering RNA drug approved for the treatment of a rare genetic disease. 6 In about 30 years, NAs were promoted from the rank of sophisticated laboratory constructs to effective therapeutic compounds with a potentially broad spectrum of medical applications. 6 Nowadays, with the ongoing COVID-19 pandemic, great expectations are placed on mRNA-based vaccines for the immunization against the SARS-CoV-2 virus. [7] [8] [9] The extraordinary medical potential of NAs resides in the fact that they can be designed to modulate the expression of any gene, including those encoding for proteins that are "undruggable" by classical small therapeutic molecules. 10 While small molecules and monoclonal antibodies need to interact with a target protein to activate or block its function, relying exclusively on spatial structural affinity;

NA-based therapeutic agents exploit the natural cell machinery to promote gene silencing (RNAi) or protein production (mRNA). 10 The ability of NAs to specifically knockdown or induce gene expression makes them the sole therapeutic approach capable to cope with multifactorial genetic diseases, cancer mutations as well as pandemic viral infections. 11 The RNAi pathway can be exploited in different ways. For example, a gene encoding for a short hairpin RNA (shRNA) could be F I G U R E 1 RNA interference: a miRNA gene is transcribed into primary miRNA (pri-miRNA) that is further processed by Drosha to form pre-miRNA. Exportin-5 translocates the pre-miRNA into the cytoplasm were it is processed by Dicer into mature miRNA. siRNAs can be obtained directly by chemical synthesis and -with the help of a carrier or chemical modifications-can reach the cytoplasm through endocytosis. In the cytosol, the guide (antisense) strand of mature miRNA or siRNA will be assembled into the RNA-induced silencing complex (RISC). The passenger (sense) strand will be discarded. The mature RISC will find the target mRNA sequences through complementary base pairing with the guide strand. As few as 7 complementary bases (seed region) are sufficient for miRNA-mediated RNAi, while full complementarity is usually required for siRNA-induced silencing. Depending on the triggering molecule (siRNA or miRNA), the translation of the target gene could be repressed due to mRNA degradation or translocation to the P bodies. mRNA therapy: once introduced in the cytosol through an appropriate delivery method, a modified, exogenous mRNA could hijack the cell's ribosomes to be translated into a functional protein employed to achieve a sustained production of silencing molecules. In this case, nuclear delivery would be required, and a competition with the endogenous RNAi processing enzymes might occur. Differently, the site of action of a synthetic short interfering RNA (siRNA) is the cytosol, where the guide strand of the siRNA is loaded into the RNAinduced silencing complex (RISC) that then binds to mRNA molecules to modulate their expression ( Figure 1 ). 12 Similarly, in the case of protein expression via mRNAs, the exogenous nucleic acid has to reach the cytosol where the cellular translation machinery resides. 13 Crossing the cell membrane and localizing into the appropriate subcellular compartment have been always recognized as major obstacles to the clinical translation of NA-based therapies. Indeed, only small, neutral, and slightly hydrophobic molecules can passively diffuse across cell membranes, while large and negatively charged molecules, such as RNAs, can only rely on active transport mechanisms, as endocytosis. 14 This results in the confinement of NAs in intracellular organelles, as endosomes, from which NAs should rapidly escape into the cytosol to avoid progressive and fatal degradation. 15 Viral and nonviral vectors are used for the intracellular delivery of nucleic acids.

Viral vectors refer to the use of modified viruses in which the pathogenic part of their genome has been removed, while the nonpathogenic part, which allows them to infect the cell, is retained. 16 These vectors are extremely attractive and have helped to substantially advance the field of gene therapy because of their natural ability of inducing high transfection. Moreover, depending on the type of virus, such vectors can produce long-term gene expression, which is currently difficult to accomplish with non-viral methods. The properties and engineering principles of viral vectors, and of the most clinically advanced type (recombinant adeno-associated virus) were excellently reviewed in References 17 and 18, respectively. Despite the intriguing properties and some clinical successes, the use of viral vectors is still characterized by several limitations and challenges, such as the intrinsic risk for immunogenicity, broad tropism, limited payload packaging capacity, and difficult production. Nonviral vectors represent a valuable alternative as they are generally less immunogenic, easier to design and synthesize, and able to deliver large payloads. 19 Nevertheless, nonviral vectors need to match some other requirements such as the biocompatibility of their constituents. Furthermore, considering that intravenous injection is the preferred route of administration, the ideal vector should maintain long circulation times and guarantee an efficient release of NAs upon reaching the target site. A vast body of literature exists on nonviral vectors for NA delivery. These vectors have been synthesized using different compositions, surface functionalities, and properties and can be broadly classified as conjugates or supramolecular assemblies. The first category comprises all those systems in which the NA is directly linked to another molecule by a covalent bond. Depending on the desired properties of the conjugate, the NA can be bound to a targeting agent, a polymer, or a hydrophobic moiety such as a lipid. [20] [21] [22] A notable example of NA-conjugate is givosiran, an Nacetylgalactosamine-conjugated siRNA clinically approved for the treatment of acute hepatic porphyria. On the other hand, structures formed by noncovalent interactions between NAs and other components belong to the second category. These delivery systems are generally classified based on the materials employed, which can be polymers, lipids, inorganic chemicals or a combination thereof. [23] [24] [25] [26] [27] [28] [29] [30] [31] The recently approved vaccines against SARS-CoV-2, composed of mRNA encapsulated in lipid nanoparticles, belong to this class. The specific features of the different nonviral delivery systems will not be discussed in further detail since they have been reviewed elsewhere. 19, 32, 33 It is important to highlight that nonviral vectors do not have the natural ability of viruses to efficiently overcome cellular barriers. As such, early in their engineering process, they must be optimized to favor endocytosis and endosomal escape with consequent cytosolic release of NAs. Endosomal escape can be accomplished by exploiting different strategies build-into the vector, including the proton-sponge effect of cationic polymers 34 ; the tendency of certain lipids to form non-lamellar phases 35 ; the decoration of the vector surface with specific molecules. 36 In the latter case, great inspiration could come from nature, as viruses and other pathogens have evolved to efficiently transfect their genetic material into host cells. 37 Herein, we have systematically organized the literature describing the endosomal escape triggered by ionizable lipid nanoparticles (LNPs), the most clinically advanced nonviral vector for NAs delivery. 38 After a brief overview of LNPs development from the laboratory bench to the bedside, we will critically review the work that contributed to the description of the LNPs-mediated endosomal escape mechanism through physico-chemical analyses, cell-based studies, and computational simulations. Finally, we will touch on the strategies employed by viruses to deliver their nucleic acids to the cytosol, commenting on their possible exploitation to further improve the LNP technology.

When the first attempts to encapsulate nucleic acids were made in 1980, liposomes were already a well-established vector for the delivery of small molecules. Initially, liposomes composed of neutral and zwitterionic lipids were employed as carriers for NA (specifically DNA), but were characterized by very low encapsulation efficiencies.

Subsequently, the strategy of pre-condensing the NAs with polycations, such as poly-L-lysine and protamine, was introduced to neutralize the highly negative charge of the therapeutic cargo and pack more genetic material within the aqueous core of the liposomes. 39 In 40 The process was termed lipofection and the lipid mixture, which is still commercially available as Lipofectin and its more recent derivatives, is still widely used for the in vitro delivery of NAs.

Despite the in vitro efficacy and an extensive research campaign, these permanently charged lipids and the related liposomal formulation never succeeded in reaching the clinic mostly due to their unacceptable toxicity, short circulation half-life, and unspecific association to negatively charged cellular and extracellular components. 41 PEGylation became instrumental in masking the cationic surface charge by introducing an hydrophilic, stealth coating around the lipid particle, thus, improving systemic biodistribution and circulation half-life. 42 However, excessive PEGylation turned out to be detrimental for cell access and optimal subcellular distribution. 43 Eventually, the notion of ionizable lipids was introduced whereby the quaternary ammonium head of cationic lipids was substituted with a titratable moiety. 44 The resulting ionizable lipids present an electrostatic charge depending on the lipid pKa and the environmental pH ( Table 1) (Table 1) . Taking a closer look to the LNP development, the first ionizable lipid was 1,2-dioleoyl-3-dimethylammonium propane (DODAP), whose rapid mixing with other lipids and oligonucleotides in the presence of ethanol allowed encapsulation efficiencies as high as 70%. 45 Then, after realizing that polyunsaturated lipids could lead to more efficient transfections, 46 LNPs were synthesized using Nowadays, the research on new ionizable lipids is thriving, in search for molecules with better tolerability, defined organ tropism and improved endosomal escape. 50 It is important to highlight that since the beginning, the LNP development was supported by a coordinated research effort at the T A B L E 1 The evolution of ionizable lipids from permanently charged DOTMA to FDA-approved DLin-MC3-DMA DLin-DMA 6.8 46 DLin-KC2-DMA 6.7 48 DLin-MC3-DMA 6.4 49 Note: A complete overview on the more recent ionizable lipids synthesized and used in preclinical and clinical studies is presented in Reference [50] .

interface between academia, industry, and the clinic. This effort is still ongoing with the objective of using LNPs to target other organs than the liver and deliver larger nucleic acids, such as mRNA. [51] [52] [53] Also, in parallel to the above described translational research effort, several groups focused on studying the fine biophysical interaction of LNPs with different cells. A great deal of work was dedicated to characterizing the endosomal escape and cytosolic delivery of the genetic materials. Quite a few authors concluded that the efficacy of LNPs in promoting the endosomal escape of nucleic acids was extremely limited with less than 2-3% of the intracellular siRNA being visualized in the cytosol. 54, 55 If this modest percentage can successfully induce gene silencing in the liver, where most of the injected LNPs accumulate, it is questionable whether this approach could work for targeting other organs and diseases. 56 Therefore, increasing the percentage of RNA escaping into the cytosol is recognized as a necessary condition to unleash the full potential of LNPs.

In the following sections, key findings on the biophysical mechanisms regulating the interaction of LNP with the endosomal machinery and the release of the genetic cargo into the cytosol are reviewed based on physico-chemical interpretations, cell-based studies, and computational modeling.

2 | HOW DO IONIZABLE LIPID NANOPARTICLES OVERCOME THE ENDOSOMAL MEMBRANE?

The efficiency of an RNA therapy is influenced by several factors: the specificity of NAs; the ability of the vector to protect the NAs from biodegradation; the tropism of the vector for the diseased tissue; and the ability of the vector to release its cargo into the proper subcellular compartment. Assuming that the NA has been properly selected and designed, the efficiency of an RNA therapy depends essentially on the number of NAs reaching the intracellular target. In other words, the higher is the amount of siRNA that can be loaded on RISC, the amount of antimiR pairing with the target miRNA or the amount of mRNA engaging with the ribosome and the higher will be the therapeutic efficiency ( Figure 1 ). All these molecular targets-RISC, miRNA, and ribosome-share the same subcellular location: the cytosol. In this scenario, the endosomal compartment represents a formidable barrier for the cytosolic accumulation of NAs, and it is recognized as the main limiting factor to their efficacy. 57 Understanding the biophysical mechanisms regulating the cytosolic delivery of nucleic acids is fundamental to expand the realm of applications of the RNA therapies.

F I G U R E 2 Schematic representation of the endocytic pathway, showing the different possible fates of an internalized LNP. The endocytosed LNP engulfed in an early endosome (EE) can be sent back towards the cell membrane and excreted either directly (fast recycling) or through other intracellular organelles such as the endocytic recycling compartment (ERC) (slow recycling). Alternatively, the EE matures to late endosome (LE), gradually modifying its receptors and enzymatic pool and decreasing its pH. The endosomal escape events were suggested to occur at an intermediate, hybrid compartment stage between EE and LE (see also the section Cell-based Studies). Eventually, the LE fuses with the lysosome (Ly), whose enzymes can dismantle and degrade the entrapped LNPs and their NA payload. On the surface of endo-lysosomal vesicles, the figure shows the main stage-defining markers employed in the works analyzed in this review: EEA1, early endosome antigen 1; RabX, Ras-related protein RabX (X=4, 5, 7, 9, 11); LAMP1, Lysosome-associated membrane glycoprotein 1

The cellular uptake of LNP mainly relies on the endocytic pathway. More in detail, it has been shown that specific serum proteins adsorbed on the surface of LNPs upon intravenous injection can drive the cell internalization. 58 This mechanism has been carefully elucidated for liver-targeting LNPs, which are taken up by hepatocytes following the interaction between apolipoprotein E -adsorbed on the particles-and low-density lipoprotein receptors on the cell membrane. 59 Other receptors may be involved in the cell uptake of LNPs if a targeting ligand (e.g., an antibody) is used to decorate their surface. 60 In general, all these cell uptake processes require, at first, the formation of early endosomes (EE), which are cellular vesicles engulfing the nanoparticles with a pH ranging between 5.5 and 6.5. (Figure 3 ). This proposed mechanism F I G U R E 3 Top: the molecular structure hypothesis: the geometry of the lipid molecule dictates the structure of its aggregates. Cone lipids (e.g. 1-Stearoyl-sn-glycero-3-phosphocholine) form micelles, cylindrical lipids (e.g. 1,2-distearoyl-sn-glycero-3-phosphocholine, DSPC) form bilayers and inverted-cone lipids (e.g. 1,2-Dioleoyl-sn-glycero-3-phosphoethanolamine, DOPE) form hexagonal phases (H II ). Bottom: the geometry of lipids might change upon mixing and ion pair formation, with consequences on the macrostructure. Protonated ionizable lipids interact with anionic lipids adopting an inverted cone shape, which promotes the formation HII phase. Non-bilayer phases are associated with membrane fusion assumed a crucial relevance in the extensive work of optimization of ionizable lipids carried out in the last decades, and will be discussed in details here below. 62 

Since 1978, Cullis and Hope pioneered the study of the physicochemical properties of the system linking the formation of nonbilayer structures with the "fusogenic" property of lipids. 63 In general, lipids can self-assemble in aqueous media to form characteristic mesoscopic phases, whose morphology is dictated by the geometry of the lipid molecule (size of the polar head, length, and unsaturation of the alkyl tails) and can depend on the temperature and presence of ions ( Figure 3 ). 64 When freely dispersed in water, oppositely charged lipids adopt a lamellar structure but, upon mixing, tend to form nonbilayer phases, following the reduction of the combined polar head size due to electrostatic attraction and enlargement of the hydrophobic section ( Figure 3 ). 68 The formation of an H II structure upon contact of lipoplexes and unilamellar anionic liposomes in aqueous media was observed by 31 P NMR, and it was accompanied by the immediate release of DNA from lipoplexes, providing a strong evidence of a link between these two events. 65 In the same work, the authors foresee the possibility to correlate transfection efficiency with the ability of a given cationic lipid to adopt an H II structure when in contact with anionic lipids. The capacity to transition from one state to another can be conveniently measured via the bilayer-to-H II transition temperature (T BH ). T BH is defined as the temperature at which cationic (ionizable or permanent) lipid assemblies shift from the lamellar to the hexagonal phase, upon equimolar mixing with anionic vesicles.

The lower is the T BH , the stronger is the tendency to adopt nonlamellar phases. The positive charge of permanently cationic lipids does not depend on pH and therefore it is always available for ion pair formation with anionic lipids. Conversely, in the case of ionizable lipids, the hydrophilic head must be in its protonated form to trigger the same process.

This pH-dependence provides a unique benefit, as an ionizable lipid with optimal pKa (around 6.5) is neutral in the circulation, preserving a bilayer structure but becomes protonated at endo-lysosomal pH assuming an hexagonal phase upon contact with anionic membrane lipids. 49 Thus,

T BH and pKa have been used as guiding parameters for the rational design of novel ionizable lipids. 48 However, as pointed out by the authors, these measures do not fully account for the biological activity of ionizable lipids, whose ability to deliver NAs into the cytosolic compartment also depends on other structural features as well as the biological properties of the target cells and tissue. For instance, a flexible linker between the polar head and the lipid tails is thought to be essential to allow sufficient proximity of the cationic head and endosomal membrane lipids. 48 

Cell-based assays can be used to understand the LNP ability to induce were developed for high throughput screening purposes. 70 Although not specifically tested on ionizable LNPs, these sensors might provide a formidable tool for the screening of new cytosol-targeting nanoparticles.

Overall, in addition to the scarce amount of cytosolic siRNA determined, these studies agree on the narrow time frame in which endosomal escape occurs, and they both exclude that cytosolic release could occur from LE or lysosomes. Despite the different molecular signatures found on the releasing endosomes by the two groups, they both indicate a hybrid, maturing endosome as the optimal condition for siRNA escape. The pH of this compartment, although not being experimentally measured, is described as supportive of the ion-pair mechanism observed in the physico-chemical studies discussed in the previous section.

More recently, the use of LNPs as nucleic acid vectors was extended to mRNA. 71 Compared to siRNAs, mRNAs have larger hydrodynamic volume and greater molecular weight. These parameters could be responsible for differences in the process of endosomal escape, which was thus investigated in the works discussed below. 84 In a recent report, the authors investigated if the adjuvant effect of CADs could be extended to nanocarriers having different structures and compositions. 85 Interestingly, the cytosolic delivery mediated by MC3-LNPs could not be improved by CAD treatment, probably due to the stable interaction between the ionizable lipid and the nucleic acid.

The complexity of LNP's transfection dynamics and intracellular trafficking cannot be fully appreciated and elucidated by experiments only, although a broad variety of experimental techniques have been employed. 86 Computational methods, such as molecular dynamics against small-angle X-ray scattering (SAXS) data available in the literature. Two alternative transfection pathways were documented: the perpendicular pathway, where DNA would align parallel to the endosomal membrane, unzipped from lipids, while a pore opens concomitantly; the parallel pathway, where DNA strands oriented normally to the membrane are quickly ejected after the opening of a pore.

As previously observed experimentally, 46 unsaturated lipid tails promoted fusion with the endosomal membrane compared to saturated lipids. The particle size also played a role, as small lipoplexes tend to be more fusogenic to alleviate the state of stress and higher potential energy associated with the larger surface curvature. 

As previously discussed, the interface between the two approaching membranes has to be dehydrated for fusion to initiate. 95 For instance, a lipid composition of the viral envelope affects the endosomal escape performances. 102 The viral envelope derives from the host cell membrane. However, differences in the lipid composition of the two suggest that some viruses "select" lipids to build a more fusogenic envelope. 114 For instance, the composition of influenza virus envelopes was found to be enriched in phosphatidyl ethanolamines (PE) and phospatidyl serines (PS) as compared to the host membrane, where phosphatidyl cholines (PC) were more represented. 115 Both PE and PS promote fusion, as the former increases membrane fluidity while the latter improves the interaction with the annexins of the host. Therefore, the lipidomic analysis of viruses might provide useful hints for the design of LNPs with improved fusogenicity.

In addition to envelope lipids, membrane fusion proteins, or derived peptides, might be exploited to improve non-viral nucleic acid delivery. The first viral peptide used for enhanced endosomal release was the HIV1-TAT. 116 Ever since, several viral-derived or viralinspired synthetic analogs were reported to enhance the delivery of nucleic acids. 112 The mechanism of action of TAT is related to its positive charge (derived from arginine and lysine residues) that leads to endosomal membrane destabilization via direct interaction with the negatively charged membrane lipids. 112 Similar features were also shown by the human papillomavirus L2 capsid peptide, which is characterized by a positive charge and a hydrophobic domain. 117 In other cases, viral peptides exploiting the cited pH-driven conformational change have been employed. HA2 has been used to enhance the delivery of DNA following its incorporation to transferrin-polylysine-DNA complexes. 118 The influenza HA2 fusion protein is a class I fusion protein; as such the α-helical structure is crucial for the facilitation of pH-dependent membrane fusion. 119 Inspired by HA2, several synthetic pH-sensitive α-helical peptides were designed and their structure-activity correlation was studied. 120 For example, GALA, a 30 amino acid peptides that switch conformation from random coil to α-helix at pH 5, was shown to facilitate endosomal escape of nucleic acid delivered by cationic dendrimers. 121 Newer derivatives were developed, such as KALA, which is positively charged and therefore also enables complexation of nucleic acids, 122 and several others including INF7, EBI, and CADY. [123] [124] [125] Exploiting the same mechanism, the glycoprotein G of the vesicular stomatitis virus was used for the enhanced gene delivery through liposomes. 126 Finally, some viral derived peptides can mediate membrane fusion in a pH-independent manner, such as peptides derived from the HIV-1 gp41 protein. 127 A more detailed understanding of these mechanisms could help to identify nature-inspired strategies to improve the performance of LNPs. Previous studies suggest that stable contact between oppositely charged lipid heads can only occur upon dehydration of the interface between the LNP and the endosomal membrane, and membrane destabilization is required for fusion to proceed ( Figure 4 ). As extensively discussed in this work, ionizable lipids induce the cytosolic release of nucleic acid by forming non-bilayer structures upon ion pairing with endosomal anionic lipids. Among the viral peptides here presented, GALA and its derivatives could provide LNPs with a membrane fusion mechanism complementary to the one of ionizable lipids.

LNPs may indeed benefit from the synergic action of such peptides, which help establish the first contact between the facing membranes through the formation of biological anchors. Coherently with this vision, a recent work described the development of a novel LNPs decorated with GALA-cholesterol conjugate. 128 The combined effect of the ionizable lipid YSK05 and GALA improved the endosome escape as compared to a control system (composed of the cationic lipid DOTMA and GALA). In another work, the virus-derived peptide KALA was conjugated to a lipid chain and included in LNPs structure for mRNA delivery. 129 The combination of KALA and ionizable lipids promoted the cell uptake and mRNA translation compared to controls (LNPs equipped with another peptide, or LNPs formulated without ionizable lipids). Although a mechanistic analysis of cytosolic delivery was missing, these reports support the notion of combining ionizable lipids with virus-inspired agents to promote endosomal escape.

Different lights have been used to illuminate the mechanisms regulating the endosomal escape of therapeutic RNA loaded into ionizable lipid nanoparticles. The body of work presented in this review includes physico-chemical and biological evidence as well as in silico modeling data and contributes to our understanding of intracellular RNA delivery.

Overall, it is recognized that promoting the instability of endosomal membrane is a fundamental step. This has been realized by designing LNPs with a faceted morphology; employing lipids with a higher tendency to diffuse, protrude, and flip-flop; or biomimicking viruses using complementary escape mechanisms, such as fusion-inducing peptides.

Today, only minimal amounts of siRNA are released into the cell cytosol.

However, this appears to be sufficient to achieve potent silencing and has dictated the successful advancement of LNP from the laboratory benches to the clinic. However, as our understanding of the complex mechanisms regulating the behavior of cells grows, it becomes clear that siRNA, mRNA, and cocktails of nucleic acids should be more efficiently delivered to positively address a variety of disorders, including cancer, cardiovascular, neurological, and infectious diseases. To specifically address the cytosolic delivery issue, the refinement of high-throughput cellular screening systems for the quantitative detection of intra-cytosol nucleic acids is urgently needed, and should be coupled with the traditional in vivo efficacy assays. The design of novel nanoparticles with higher RNA delivery efficiency would require a coordinated, multidisciplinary effort focusing not only on the design and testing of new ionizable lipids, but also on the integration into nanoparticles of additional components promoting the endosomal destabilization, possibly taking inspiration from virus-inspired escape mechanisms. 

The peer review history for this article is available at https://publons. com/publon/10.1002/btm2.10213.

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

https://orcid.org/0000-0002-7214-9045

Paolo Decuzzi https://orcid.org/0000-0001-6050-4188

Molecular biology, nucleic acids, and the future of medicine

Direct gene transfer into mouse muscle in vivo

The C. elegans Heterochronic gene lin-4 encodes small RNAs with antisense complementarity to lin-14

Conservation of the sequence and temporal expression of let-7 heterochronic regulatory RNA

Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans

The Onpattro story and the clinical translation of nanomedicines containing nucleic acid-based drugs

Phase 1/2 study of COVID-19 RNA vaccine BNT162b1 in adults

An mRNA vaccine against SARS-CoV-2 -preliminary report

CureVac Initiates Phase 2a Clinical Trial of COVID-19 Vaccine Candidate

Current prospects for RNA interferencebased therapies

Overcoming cellular barriers for RNA therapeutics

Small silencing RNAs: state-of-the-art

MRNA-based therapeutics-developing a new class of drugs

Mechanisms of endocytosis

Cellular uptake and intracellular trafficking of oligonucleotides: implications for oligonucleotide pharmacology

Gene therapy and DNA delivery systems

Engineering targeted viral vectors for gene therapy

Adeno-associated virus vector as a platform for gene therapy delivery

Non-viral vectors for gene-based therapy

Oligonucleotide conjugates-candidates for gene silencing therapeutics

A divalent siRNA chemical scaffold for potent and sustained modulation of gene expression throughout the central nervous system

GalNAc-siRNA conjugates: leading the way for delivery of RNAi therapeutics

Polymeric vehicles for nucleic acid delivery

Dendrimers for siRNA delivery

Next-generation lipids in RNA interference therapeutics

Delivery of multiple siRNAs using lipid-coated PLGA nanoparticles for treatment of prostate cancer

Anionic liposomes for small interfering ribonucleic acid (siRNA) delivery to primary neuronal cells: evaluation of alpha-synuclein knockdown efficacy

Hybrid lipid/polymer nanoparticles for pulmonary delivery of siRNA: development and fate upon in vitro deposition on the human epithelial airway barrier

SiRNA-chitosan complexes in poly(lactic-co-glycolic acid) nanoparticles for the silencing of aquaporin-1 in cancer cells

Systemic leukocyte-directed siRNA delivery revealing Cyclin D1 as an antiinflammatory target

Cell specific delivery of modified mRNA expressing therapeutic proteins to leukocytes

Delivery materials for siRNA therapeutics

Paving the road for RNA therapeutics

The proton sponge hypothesis: fable or fact

Nonbilayer phase of lipoplex-membrane mixture determines endosomal escape of genetic cargo and transfection efficiency

A pH-sensitive fusogenic peptide facilitates endosomal escape and greatly enhances the gene silencing of siRNA-containing nanoparticles in vitro and in vivo

Viral escape from endosomes and host detection at a glance

CRISPR-Cas9 genome editing using targeted lipid nanoparticles for cancer therapy

Potentiation of cationic liposome-mediated gene delivery by Polycations

Lipofection: a highly efficient, lipid-mediated DNA-transfection procedure

Lipid-based nanoparticles for nucleic acid delivery

In vitro and in vivo effects of polyethylene glycol (PEG) -modified lipid in DOTAP / cholesterol-mediated gene transfection

PEGylation as a strategy for improving nanoparticle-based drug and gene delivery

Modulation of membrane fusion by asymmetric Transbilayer distributions of amino lipids

Efficient encapsulation of antisense oligonucleotides in lipid vesicles using ionizable aminolipids: formation of novel small multilamellar vesicle structures

Cationic lipid saturation influences intracellular delivery of encapsulated nucleic acids

RNAi-mediated gene silencing in non-human primates

Rational design of cationic lipids for siRNA delivery

Maximizing the potency of siRNA lipid nanoparticles for hepatic gene silencing in vivo

Lipid nanoparticles for nucleic acid delivery: current perspectives

A combinatorial library of lipid nanoparticles for RNA delivery to leukocytes

The challenge and prospect of mRNA therapeutics landscape

Systemic gene silencing in primary T lymphocytes using targeted lipid nanoparticles

Image-based analysis of lipid nanoparticle-mediated siRNA delivery, intracellular trafficking and endosomal escape

Visualizing lipid-formulated siRNA release from endosomes and target gene knockdown

Lipid nanoparticle delivery of siRNA to osteocytes leads to effective silencing of SOST and inhibition of Sclerostin in vivo

Current challenges in delivery and cytosolic translocation of therapeutic RNAs

Selective organ targeting (SORT) nanoparticles for tissuespecific mRNA delivery and CRISPR-Cas gene editing

Targeted delivery of RNAi therapeutics with endogenous and exogenous ligand-based mechanisms

A modular platform for targeted RNAi therapeutics

Enabling cytoplasmic delivery and organelle targeting by surface modification of nanocarriers

Lipid nanoparticle Systems for Enabling Gene Therapies

Effects of fusogenic agent on membrane structure of erythrocyte ghosts and the mechanism of membrane fusion

Lipid polymorphism: the molecular basis of nonbilayer phases

On the mechanism whereby cationic lipids promote intracellular delivery of polynucleic acids

The role of dioleoyl phosphatidylethanolamine in cationic liposome mediated gene transfer

An inverted hexagonal phase of cationic liposome-DNA complexes related to DNA release and delivery

Roles of lipid polymorphism in intracellular delivery

Gal8 visualization of endosome disruption predicts carrier-mediated biologic drug intracellular bioavailability

Genetically encoded Splitluciferase biosensors to measure endosome disruption rapidly in live cells

Sequence-engineered mRNA without chemical nucleoside modifications enables an effective protein therapy in large animals

Endocytic profiling of cancer cell models reveals critical factors influencing LNP-mediated mRNA delivery and protein expression

Linkage between endosomal escape of LNP-mRNA and loading into EVs for transport to other cells

A novel amino lipid series for mRNA delivery: improved Endosomal escape and sustained pharmacology and safety in non-human primates

Optimization of lipid nanoparticles for intramuscular administration of mRNA vaccines

Compounds and compositions for intracellular delivery of therapeutic agents

Compounds and compositions for intracellular delivery of therapeutic agents

Safety and Immunogenicity Study of 2019-nCoV Vaccine (mRNA-1273) for Prophylaxis of SARS-CoV-2 Infection (COVID-19)

Boosting intracellular delivery of lipid nanoparticle-encapsulated mRNA

Naturally-occurring cholesterol analogues in lipid nanoparticles induce polymorphic shape and enhance intracellular delivery of mRNA

Efficiency of siRNA delivery by lipid nanoparticles is limited by endocytic recycling

Codelivery of zoledronic acid and double-stranded RNA from core-shell nanoparticles

Simulation of polyethylene glycol and calcium-mediated membrane fusion

Repurposing cationic amphiphilic drugs as adjuvants to induce lysosomal siRNA escape in nanogel transfected cells

Cationic Amphiphilic drugs boost the Lysosomal escape of small nucleic acid therapeutics in a Nanocarrier-dependent manner

Methodologies to investigate intracellular barriers for nucleic acid delivery in non-viral gene therapy

Lipid nanoparticles containing siRNA synthesized by microfluidic mixing exhibit an electrondense nanostructured core

On the formation and morphology of lipid nanoparticles containing Ionizable cationic lipids and siRNA

Update of the CHARMM all-atom additive force field for lipids: validation on six lipid types

Ionizable amino lipid interactions with POPC: implications for lipid nanoparticle function

A molecular view on the escape of lipoplexed DNA from the endosome

Development and testing of the OPLS all-atom force field on conformational energetics and properties of organic liquids

Synergistic lipid compositions for albumin receptor mediated delivery of mRNA to the liver

Bioinspired Alkenyl amino alcohol Ionizable lipid materials for highly potent in vivo mRNA delivery

Membrane fusion: overcoming of the hydration barrier and local restructuring

Plasma membranes are asymmetric in lipid unsaturation, packing and protein shape

Insights into the endosomal escape mechanism via investigation of dendrimer-membrane interactions

Membrane lipids destabilize short interfering ribonucleic acid (siRNA)/polyethylenimine nanoparticles

Uncovering Earth's virome

Endocytosis of viruses and bacteria

Structures of enveloped virions determined by cryogenic electron microscopy and tomography

Virus entry: molecular mechanisms and biomedical applications

Respiratory syncytial virus entry and how to block it

Dissecting virus entry via endocytosis

Involvement of microtubular network and its motors in productive endocytic trafficking of mouse polyomavirus

The cell biology of receptor-mediated virus entry

How non-enveloped viruses hijack host machineries to cause infection

Viral membrane fusion

Structural monitoring of a transient intermediate in the hemagglutinin fusion machinery on influenza virions

Mechanisms of influenza viral membrane fusion

Architecture of a nascent viral fusion pore

Current progress of virus-mimicking nanocarriers for drug delivery

Real-time single-molecule imaging of the infection pathway of anadeno-associated virus

The role of the membrane in the structure and biophysical robustness of the dengue Virion envelope

Lipid composition of the viral envelope of three strains of influenza virus-not all viruses are created equal

TAT peptide on the surface of liposomes

A membrane-destabilizing peptide in capsid protein L2 is required for egress of papillomavirus genomes from endosomes

Influenza virus hemagglutinin HA-2 N-terminal fusogenic peptides augment gene transfer by transferrin-polylysine-DNA complexes: toward a synthetic virus-like gene-transfer vehicle

Cellular uptake of an α-helical amphipathic model peptide with the potential to deliver polar compounds into the cell interior non-endocytically

Membranotropic cell penetrating peptides: the outstanding journey

GALA: a designed synthetic pH-responsive amphipathic peptide with applications in drug and gene delivery

Design, synthesis, and characterization of a cationic peptide that binds to nucleic acids and permeabilizes bilayers

A new potent secondary amphipathic cell-penetrating peptide for siRNA delivery into mammalian cells

Delivery of short interfering RNA using endosomolytic cell-penetrating peptides

Fusogenic peptides enhance endosomal escape improving siRNA-induced silencing of oncogenes

Enhanced gene transfer with Fusogenic liposomes containing vesicular stomatitis virus G glycoprotein

Cellular delivery of impermeable effector molecules in the form of conjugates with peptides capable of mediating membrane translocation

Lung-endotheliumtargeted nanoparticles based on a pH-sensitive lipid and the GALA peptide enable robust gene silencing and the regression of metastatic lung cancer

Development of a lipoplextype mRNA carrier composed of an ionizable lipid with a vitamin E scaffold and the KALA peptide for use as an ex vivo dendritic cellbased cancer vaccine

Cytosolic delivery of nucleic acids: The case of ionizable lipid nanoparticles