Programmable Antivirals and Just-in-Time Vaccines: Biosecurity Implications of Viral RNA Secondary Structure Targeting

Introduction

There is a novel approach to antiviral drugs that has great promise as a countermeasure against RNA viruses, many of which have pandemic potential. RNA viruses have genomes that are composed of RNA. While DNA forms the familiar double-helix shape, RNA is a single strand that folds upon itself to form many different shapes. Some of these shapes or structures are not just happenstance—they function as tiny molecular machines and perform important steps in the virus life cycle. We have found that drugs designed to bind and disrupt these essential and conserved RNA structures have a high barrier to resistance. These drugs are based on locked nucleic acids (LNAs), which can be designed to bind to specific regions of viral RNA. LNAs have a longer half-life, and are more shelf-stable, than other nucleic acids like mRNA. They can persist in the body after a single dose, making them a convenient drug to administer. In this commentary we will describe challenges with current antivirals and vaccines, and how targeting conserved structures of viral RNA can overcome these challenges. We will propose the concept of a programmable antiviral, as well as a just-in-time vaccine. Finally, we will touch upon the advances needed to bring this new class of therapeutic to the clinic.

The Classical Antiviral Paradigm

Classical antiviral therapeutics target viral proteins that often mutate, resulting in antiviral resistance, as has been seen for HIV, ¹ influenza, ² and SARS-CoV-2. ³ These classical antivirals (also called direct acting antivirals) are designed to tightly bind their protein target, generally a viral enzyme such as a protease or polymerase. ⁴ This confers a highly targeted, narrow-spectrum therapeutic effect that has been successful in the treatment of HIV and hepatitis C virus infections. However, direct-acting antivirals are often subject to unwanted development of resistance due to point mutations in the region of the antiviral binding site. When point mutations prevent antiviral binding, but have no effect on viral protein function, strong (subtype limited) resistance results. This is with the exception of nucleoside analog antivirals, which function by chain termination mechanism; they are often used in combination to overcome resistance.

Current Viral Vaccine Challenges

The problem of antiviral resistance accentuates the importance of vaccine development for viral pathogens. Considerable steps have been made in rapid vaccine platform technology during the COVID-19 pandemic. It is now possible to design an efficacious mRNA vaccine almost immediately after a viral genome has been sequenced. However, our current vaccines continue to have substantial challenges that make them inadequate, particularly during epidemics and pandemics. Vaccines do not confer immediate immunity after administration; days to weeks are required for immunity to develop. As seen during the COVID-19 pandemic, this delay necessitates the use of other nonpharmaceutical measures to reduce transmission in the interim, such as lockdowns. Furthermore, all types of current vaccine platforms target viral protein antigens. Thus, similar to their antiviral counterparts, vaccines can be rendered less efficacious when these antigenic proteins mutate under selection pressure.

Key Viral RNA Secondary Structures Are Essential and Highly Conserved

The public health and biosecurity communities have called for the development of new broad-spectrum antivirals and vaccines, particularly for viruses of pandemic potential. ⁵ For any vaccine or therapeutic to maintain broad-spectrum activity, its target must be (1) essential and (2) broadly conserved across viral variants, and ideally across an entire viral genus or family. These features reduce the degree of freedom for mutations that are compatible with virus function, translating into a high barrier for resistance.

Given the strong public health need for broad-spectrum activity, one current approach is to combine multiple narrow-spectrum vaccines into a single multivalent vaccine, such as the quadrivalent influenza vaccine. ⁶ Others have recently fabricated potential nanoparticle vaccines composed of dozens of antigenic protein variants, such as the influenza haemagglutinin glycoprotein. ⁷ Neither of these approaches are able to reliably confer broad-spectrum protection against novel, previously undescribed variants.

Traditionally, infectious virions are thought to be composed of functional proteins, a protective protein capsid, perhaps a membrane envelope, and an inner core of RNA or DNA genetic material. This static description of a virion fails to capture the dynamic nature of the viral life cycle. In the case of RNA viruses, we now know that genomic RNA not only encodes viral proteins, but also has various functional roles itself.^8,9 RNA molecules form single-stranded and double-stranded arrangements; these secondary structures play essential roles in the viral life cycle, such as for regulatory pathways, transcription, translation, and packaging. The fact that RNA can function as a molecular machine has been well established, and in humans, the amount of RNA sequence transcribed exceeds the number of protein sequences by possibly tenfold.^10,11

We thus hypothesized the existence of RNA secondary structures that are conserved across viral subtypes, and which could be potential targets for antivirals. RNA secondary structures are involved in RNA–RNA and RNA–protein interactions that are required for the correct completion of the infective cycle. ⁹ To form an infectious virion, the components described earlier must be assembled appropriately. Viral capsids self-assemble from protein oligomers with high apparent cooperativity. ¹² The complete viral genome must be packaged within the capsid, a process that has been demonstrated to rely on packaging signals within viral RNA, such as the 5’ packaging signal of HIV. ¹³ In influenza virus, packaging signals exist in regions of high nucleotide conservancy, and where synonymous codon usage is strongly suppressed.^14-16

Essential and Conserved Regions of Influenza A and SARS-CoV-2

We previously found an influenza A RNA structure that serves as an essential packaging signal, and is highly conserved across about 30,000 influenza A isolates.^17,18 This suggested that it may be an ideal therapeutic target, which we subsequently validated in mouse models. Disruption of this structure provided 100% survival to a lethal dose of influenza A inoculum. Importantly, our method of disruption led to the development of strong immunity to rechallenge, suggesting potential as both a therapeutic as well as a vaccine. This method does not involve small molecule or protein design, but rather, the rapid design of specialized nucleic acid oligomers, “locked” nucleic acid oligomers (LNAs), which are unable to be enzymatically cleaved. Our LNA bound to, and disrupted, the essential and conserved RNA structure we identified. This complete efficacy in preventing mortality was demonstrated in mice, even with a single dose of the therapeutic LNA administered 2 weeks prior to a lethal influenza A inoculum, or as a single dose administered 3 days after lethal inoculum, suggesting a role as both a prophylactic vaccine and a therapeutic drug.

Adopting the same approach for SARS-CoV-2 indicated multiple regions of RNA structured elements that are highly conserved. ¹⁹ We developed therapeutic LNAs targeting these regions. When compared with a novel COVID-19 antiviral (EIDD-1931, a nucleoside analog), the LNA therapeutics inhibited virus replication by as much as 3log10, outperforming the EIDD control by nearly 1log10. Further testing in a Syrian hamster model demonstrated intranasal pretreatment with a therapeutic LNA, on 2 consecutive days (Day -1 and Day 0) prior to infectious exposure, reduced viral titers in lung tissue by over 3log10.^17,18

In summary, it appears that these LNA therapeutics benefit from a long half-life and thus reduced dosing schedule. In animal models, our LNA drug was still active at desired efficacy levels 2 weeks after a single dose, although the precise half-life and pharmacokinetics have not yet been determined. It does not appear that repeat dosing is needed within the tested 2-week interval.

Our novel drugs, which target RNA secondary structure, meaningfully differ from prior classes of RNA drugs, such as small-interfering (siRNAs), microRNAs (mrRNAs), and antisense oligonucleotides (ASOs). Nucleic acid-based antisense technology aims to inhibit gene expression, and not highly conserved secondary structures. Double-stranded RNA-mediated interference (RNAi) is a more potent approach that takes advantage of RNA-induced silencing to guide target mRNA cleavage, but the aim is still the inhibition of gene expression. ²⁰ In contrast, targeting essential and conserved RNA secondary structures can lead to complete disruption of the viral life cycle in a way that is independent from protein-encoding genes.

A Just-in-Time Vaccination Platform

The identification, design, and targeting of these essential and highly conserved RNA secondary structures, in both influenza A and SARS-CoV-2, suggests the possibility of rapid and effective broad-spectrum antiviral development of the kind that, thus far, has been a substantial challenge. The promise of pangenotypic prophylactic vaccines and therapies for pandemic viral threats carries important biosecurity implications.

Viral pathogens with pandemic potential, such as influenza, pose broad biosecurity risks because of their ability to cause natural pandemics and their ability to be engineered and possibly weaponized. These concerns have led many to consider defense-dominant biotechnology development, such as the development of vaccine platforms that can quickly and reliably generate vaccines for novel pathogens. While we continue to reduce design-test-deploy cycles for mRNA vaccines, we are unable to reduce the time-to-immunity delay seen with these types of vaccines, and we are as yet unable to produce broad-spectrum vaccines. During a global catastrophic biological risk event, these setbacks could prove pivotal.

LNA vaccine platforms may share many features with mRNA vaccine platforms. As a nucleic acid-based platform, which is not based on small molecules or proteins, LNA platform vaccines are quickly and easily designed and synthesized (ie, “programmable”). This assumption was tested as we pivoted our LNA platform from influenza A to SARS-CoV-2 during the COVID-19 pandemic. Importantly, however, just-in-time LNA vaccines function immediately as a therapeutic, allowing them to be used in cases of prophylaxis and postexposure treatment. This is due to their immediate binding to an essential and highly conserved component of the targeted virus. This immediate effect is likely to reduce viral transmission during the earliest stages of outbreaks, further strengthening ring vaccination efforts.

The immediate therapeutic effect of LNAs rapidly reduces viral replication to attenuated levels, allowing natural immunity to the virus to build in the days to weeks following exposure. LNA is also more shelf-stable and degradation resistant than mRNA, due to the locked bonds between nucleotides, which offers promising opportunities for distribution without cold chain requirements. LNAs do not require lipid nanoparticle encasement, suggesting easy self-administration using inhaled or nebulized delivery.

Dual-Use Considerations

Prior to promoting any new experimental approach, careful consideration of possible dual-use implications should be undertaken. The scientific community is beginning to recognize that vaccine platform technologies may pose dual-use risks. ²¹ Some research on virally vectored vaccines involves techniques to circumvent preexisting viral immunity; these techniques carry the potential for misuse. More broadly, developing and disseminating techniques for viral engineering increases the number of individuals capable of misuse.

Among platform vaccine approaches, LNA vaccines are most similar to mRNA vaccines, which feature relatively little dual-use potential. ²¹ No viral engineering is required during research and development. mRNA vaccine platforms have the risk of possibly increasing access to nucleic acid sequences that encode viral proteins, whereas LNA vaccines do not encode viral proteins and thus may carry lower risk than mRNA vaccine platforms.

Paths Forward

Locked nucleic acid therapeutics and just-in-time vaccines have now been tested in animal models, but will still need to progress through clinical trials to demonstrate safety, tolerability, and efficacy in patients. The ability to deliver LNAs as aerosols by inhalation or nebulization makes them ideal for countering pandemic viruses affecting the respiratory tree. However, safety and efficacy trials of these drugs in humans are still needed. Finally, for LNAs targeting RNA secondary structures to become a biosecurity-relevant therapeutic, we must work toward developing cheaper and more scalable manufacturing of LNAs.