¹Department of Pathology and Laboratory Medicine, University of Pennsylvania Perelman School of Medicine, Philadelphia, Pennsylvania, USA

Correspondence: David B. Weiner, Department of Pathology and Laboratory Medicine, University of Pennsylvania Perelman School of Medicine, 505 SCL-422 Curie Blvd., Philadelphia, Pennsylvania 19104, USA. E-mail: dbweiner@mail.med.upenn.edu

Nucleic acid–based vaccines (NAVs) for induction of antigen-specific immunity have reemerged as important tools in our vaccine/immune therapeutic arsenal. DNA technology appears to more than hold its own alongside viral vectored systems, particularly when the focus is on driving robust T-cell immunity in the clinic.¹ By contrast, studies of vaccines based on RNA in small animals have been disappointing. Instability of the mRNA and manufacturing problems further weakened interest in this platform. However, just as good wine improves with age, cutting-edge technologies can take time to mature, and direct injection of RNA for immunization has recently been shown to be sufficiently improved. Support for RNA delivery as a more effective means of vaccination comes from several very recent reports.^2,3,4

The delivery of DNA or RNA for the in vivo production of antigen was first reported more than two decades ago,^5,6 for both prophylaxis and immune therapy of disease. The attraction of these platforms originated from a paradigm shift engendered by NAV technology.⁷ NAVs enable indefinite repeat boosting because they are not subject to neutralization by the host immune response, thus facilitating boosting even in seropositive individuals. NAVs also drive antigen expression in vivo by the host, and therefore generate antigens that are highly relevant to protection and/or immune treatment of the specific pathogen or cancer. As such, they express host-specific modifications of the relevant antigens that are difficult or impossible to achieve when ex vivo antigen-production systems are used. Although these immunogens are not live, in theory they should induce immunity similar to that following live infection, challenging the vaccine paradigm that only live- or abortive infection–antigen approaches induce diverse CD8 T-cell immunity. NAVs appear to have considerable manufacturing advantages and can easily be combined to form multi-antigen products; this has propelled the platform into clinical studies.

Still, the initial clinical performance of the DNA vaccine platform was surprisingly weak. There were no significant safety issues, but the platform did not induce consistent CD8 T-cell immunity or seroconversion. DNA was subsequently used as a prime-boost modality, a function it served for more than a decade. Within the past few years, however, advances in genetic engineering, delivery, and formulation have resulted in much more robust DNA-based vaccines.^5,8 By contrast, direct RNA vaccination has lagged behind. Reports advancing RNA immunization strategies have relied largely on an ex vivo route^9,10—RNA for antigen expression is transfected into patient-derived dendritic cells (DC) ex vivo, followed by reinfusion of the cells as a way to tailor patient-specific immune responses. This approach is being clinically investigated for immunotherapy against colorectal, lung, prostate, and pancreatic cancer; neuroblastoma; and melanoma, among others.

Improvements in methods for mRNA synthesis and stabilization and development of improved self-amplifying RNA have yielded promising results in animals over the past year. In the study by Petsch et al., protection was observed in both young and old mice, which was encouraging, and combinations of mRNA were effective in as few as one immunization in mice.² Through optimization of the GC content of the synthetic mRNA and by optimizing the untranslated region, as well as complexing the mRNA with protamine, to protect the mRNA from RNase, the in vivo performance of this platform was improved. Furthermore, the vaccines induced immune responses in ferrets and pigs that resulted in disease attenuation upon challenge with virus. The results imply that such an mRNA approach could allow for more rapid generation of new influenza vaccines.

A recent report by Geall et al.³ further extends the direct RNA immunization field using self-amplifying RNA technology. This strategy builds off of early work using alphavirus replicons for in vivo immunization and gene delivery.¹¹ Geall et al. report that, by delivering the alphavirus genes encoding the RNA replication machinery along with the recombinant viral target antigens, which in this case was the respiratory syncytial virus fusion glycoprotein, very low doses of RNA can result in the generation of strong antigen-specific immunity in mice. The responses induced compared favorably with those induced by older, unoptimized DNA vaccines delivered by electroporation and were superior to mRNA vaccines encoding the same F antigen. Further studies in larger animal systems are clearly warranted; these should include evaluation of immune performance in nonhuman primate models such as macaques.

RNA approaches have conceptual advantages as well as disadvantages in relation to other vaccine technologies. They are simple, can be delivered directly into the cytoplasm, and do not require nuclear localization to generate expression. Self-amplifying mRNA vectors expand replication by providing the replication machinery for the RNA to increase copy number, they also provide a unique target for host immunity to home in on in the replication-machinery proteins themselves. It will be important to understand which, if any, of these conceptual issues may limit or help these technologies. In addition, stabilized mRNA approaches have conceptual advantages such as longer half-life, which generates longer expression time, and formulations that are simpler than those of viral vector systems. Conversely, RNA approaches strongly stimulate the host’s innate defense system. They do this, in part, through activation of the TLR3 and 7/8 pathways, which recognize double- and single-stranded RNA, respectively, resulting in inflammatory activation and the generation of type 1 interferon (IFN).¹² Type 1 IFN can either promote or inhibit generation of an adaptive immune response and can have unwanted effects on the host, such as induction of fever or flu-like symptoms and increased expression of autoantigens.

In this regard, a study by Pollard et al.⁴ published recently in Molecular Therapy is enlightening. They reported on the feasibility of using mRNA encoding the HIV gag antigen complexed with DOTAP/DOPE as an immunization strategy. They observed a strong effect of interferon (IFN) on the resulting immunity to this vaccine when delivered into DCs for immunization. Whereas low levels of type 1 IFN may help to drive host immunity, higher levels can shut down the cellular transcriptional machinery, thus negatively impacting mRNA vaccine potency. Pollard et al. investigated this hypothesis by studying their mRNA vaccine in normal DCs as well as DCs derived from IFN-α receptor–deficient mice. They observed that IFN-αR-deficient cells were superior for driving gag-specific T-cell responses in their immune-therapy protocol. These results support the idea that IFN-αR signaling is not required for productive innate immune activation and recruitment of DCs in these mRNA/DOTAP immunized mice, suggesting that new synthetic RNA approaches, modified to reduce innate immunity,¹³ may have important advantages for in vivoimmunization and ex vivo DC applications. It is encouraging that new approaches to building this next generation of RNA immunogens are helping to send their immune message loudly and clearly.