Glycoprotein E (gE) is one of the most abundant glycoproteins in varicella-zoster virus and plays pivotal roles in virus replication and transmission between ganglia cells. Its extracellular domain has been successfully used as an antigen in subunit zoster vaccines. The intracellular C-terminal domain was reported to be decisive for gE trafficking between the endoplasmic reticulum, trans-Golgi network and endosomes and could influence virus spread and virus titers. Considering that the trafficking and distribution of mRNA vaccine-translated gE may be different from those of gE translated against the background of the viral genome (e.g., most gE in virus-infected cells exists as heterodimers with another glycoprotein, gI,), which may influence the immunogenicity of gE-based mRNA vaccines, we compared the humoral and cellular immunity induced by LNP-encapsulated mRNA sequences encoding the whole length of gE, the extracellular domain of gE and a C-terminal double mutant of gE (mutant Y569A with original motif AYRV, which targets gE to TGN, and mutants S593A, S595A, T596A and T598A with the original motif SSTT) that were reported to enhance virus spread and elevate virus titers. The results showed that while the humoral and cellular immunity induced by all of the mRNA vaccines was comparable to or better than that induced by the AS01B-adjuvanted subunit vaccines, the C-terminal double mutant of gE showed stable advantages in all of the indicators tested, including gE-specific IgG titers and T cell responses, and could be adopted as a candidate for both safer varicella vaccines and effective zoster vaccines.

Cell-mediated immunity (CMI) plays a key role in the effectiveness of varicella zoster virus (VZV) vaccines, and mRNA vaccines have an innate advantage in inducing CMI. Glycoprotein E (gE) has been used widely as an antigen for VZV vaccines, and carboxyl-terminal mutations of gE are associated with VZV titer and infectivity. In addition, the untranslated regions (UTRs) of mRNA affect the stability and half-life of mRNA in the cell and are crucial for protein expression and antigenic translational efficiency. In this study, three UTRs were designed and connected to the nucleic acid sequence of gE-M, which is double mutated in the extracellular region of gE. Then, mRNA with different nucleic acids was encapsulated in lipid nanoparticles (LNPs), forming three LNP-mRNA VZV vaccines, named gE-M-Z, gE-M-M, and gE-M-P. The immune response elicited by these vaccines in mice was evaluated at intervals of 4 weeks, and the mice were sacrificed 2 weeks after the final immunization. In the results, the gE-M-P group, which retains the nucleic acid sequence of gE-M and is connected to Pfizer/BioNTech’s BNT162b2 UTRs, induced the strongest humoral immune response and CMI. Because CMI is crucial for protection against VZV and for the design of VZV vaccines, this study provides a feasible strategy for improving the effectiveness and economy of future VZV vaccines.

Varicella zoster virus (VZV) causes two diseases: varicella upon primary infection and herpes zoster when latent viruses in the sensory ganglia reactivate. While varicella vaccines depend on humoral immunity to prevent VZV infection, cell-mediated immunity (CMI), which plays a therapeutic role in the control or elimination of reactivated VZV in infected cells, is decisive for zoster vaccine efficacy. As one of the most abundant glycoproteins of VZV, conserved glycoprotein E (gE) is essential for viral replication and transmission between ganglion cells, thus making it an ideal target subunit vaccine antigen; gE has been successfully used in the herpes zoster vaccine Shingrix^TM on the market. In this report, we found that ionizable lipid nanoparticles (LNPs) approved by the Food and Drug Administration (FDA) as vectors for coronavirus disease 2019 (COVID-19) mRNA vaccines could enhance the synergistic adjuvant effect of CpG oligodeoxynucleotides (CpG ODNs) and QS21 on VZV-gE, affecting both humoral immunity and CMI. Vaccines made with these LNPs showed promise as varicella vaccines without a potential risk of herpes zoster, which identifies them as a novel type of herpes zoster vaccine similar to Shingrix^TM. All of the components in this LNP-CpG-QS21 adjuvant system were proven to be safe after mass vaccination, and the high proportion of cholesterol contained in the LNPs was helpful for limiting the cytotoxicity induced by QS21, which may lead to the development of a novel herpes zoster subunit vaccine for clinical application.

Theoretically, the subunit herpes zoster vaccine Shingrix^TM could be used as a varicella vaccine that avoids the risk of developing shingles from vaccination, but bedside mixing strategies and the limited supply of the adjuvant component QS21 have made its application economically impracticable. With lipid nanoparticles (LNPs) that were approved by the FDA as vectors for severe acute respiratory syndrome coronavirus 2 vaccines, we designed a series of vaccines efficiently encapsulated with varicella-zoster virus glycoprotein E (VZV-gE) and nucleic acids including polyinosinic-polycytidylic acid (Poly I:C) and the natural phosphodiester CpG oligodeoxynucleotide (CpG ODN), which was approved by the FDA as an immunostimulator in a hepatitis B vaccine. Preclinical trial in mice showed that these LNP vaccines could induce VZV-gE IgG titers more than 16 times those induced by an alum adjuvant, and immunized serum could block in vitro infection completely at a dilution of 1:80, which indicated potential as a varicella vaccine. The magnitude of the cell-mediated immunity induced was generally more than 10 times that induced by the alum adjuvant, indicating potential as a zoster vaccine. These results showed that immunostimulatory nucleic acids together with LNPs have promise as safe and economical varicella and zoster vaccine candidates.

Previous studies have shown that the herpes zoster subunit vaccine Shingrix™ performs well in clinical trials. However, the key ingredient in its adjuvant, QS21, is extracted from rare plants in South America, so vaccine production is limited. Compared with subunit vaccines, mRNA vaccines have the advantages of faster production and not requiring adjuvants, but currently, there is no authorized mRNA vaccine for herpes zoster. Therefore, this study focused on herpes zoster subunit and mRNA vaccines. We prepared a herpes zoster mRNA vaccine and compared the effects of vaccine type, immunization route, and adjuvant use on vaccine immunological efficacy. The mRNA vaccine was injected directly into mice via subcutaneous or intramuscular injection. The subunit vaccine was mixed with adjuvants before immunization. The adjuvants include B2Q or alum. B2Q is BW006S + 2395S + QS21. BW006S and 2395S are phosphodiester CpG oligodeoxynucleotides (CpG ODNs). Then, we compared the cell-mediated immunity (CIM) and humoral immunity levels of the different groups of mice. The results showed that the immune responses of mice inoculated with the mRNA vaccine prepared in this study were not significantly different from those of mice inoculated with the protein subunit vaccine supplemented with the B2Q. The mRNA vaccine-induced immune responses following subcutaneous or intramuscular injection, and the different immunization routes did not lead to significant differences in immune response intensity. Similar results were also observed for the protein subunit vaccine adjuvanted with B2Q but not alum. The above results suggest that our experiment can provide a reference for the preparation of mRNA vaccines against herpes zoster and has certain reference significance for the selection of the immunization route; that is, there is no significant difference in the immune response caused by subcutaneous versus an intramuscular injection, so the injection route can be determined according to the actual situation of individuals.

A recently reported parallel preclinical study between a severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) mRNA vaccine and an inactivated SARS-CoV-2 vaccine adjuvanted with alum showed pulmonary immunopathology typical of eosinophil accumulation in a mouse pneumonia model for the latter, which implied a potential role of cellular immunity in the difference in the protection rate between these two forms of vaccines. For those who have been vaccinated with alum-adjuvanted subunit or inactivated SARS-CoV-2 vaccines, whether the Th2 responses that have been established and the absence of induced cellular immunity could be changed is an open question. Using two heterologous boosts with Th1-oriented CpG ODN-adjuvanted S1-based SARS-CoV-2 subunit vaccines for mice that were primed with two doses of Th2-oriented alum-adjuvanted S1-based SARS-CoV-2 subunit vaccines, we demonstrated that established Th2 orientation could not be reversed to Th1 orientation and that no cellular immunity was induced, which should have been induced if the boosting vaccines were used as the prime vaccines. These results remind us that if widely administered alum-adjuvanted SARS-CoV-2 vaccines cannot overcome the challenge of coronavirus disease 2019 (COVID-19) and that if cellular immunity is important for the efficacy of SARS-CoV-2 vaccines in the future, the choice of more powerful heterologous boosting vaccine forms that can induce cellular immunity should be considered very carefully before application.