Microneedle‐based intradermal delivery of stabilized dengue virus

Abstract Current live‐attenuated dengue vaccines require strict cold chain storage. Methods to preserve dengue virus (DENV) viability, which enable vaccines to be transported and administered at ambient temperatures, will be decisive towards the implementation of affordable global vaccination schemes with broad immunization coverage in resource‐limited areas. We have developed a microneedle (MN)‐based vaccine platform for the stabilization and intradermal delivery of live DENV from minimally invasive skin patches. Dengue virus‐stabilized microneedle arrays (VSMN) were fabricated using saccharide‐based formulation of virus and could be stored dry at ambient temperature up to 3 weeks with maintained virus viability. Following intradermal vaccination, VSMN‐delivered DENV was shown to elicit strong neutralizing antibody responses and protection from viral challenge, comparable to that of the conventional liquid vaccine administered subcutaneously. This work supports the potential for MN‐based dengue vaccine technology and the progression towards cold chain‐independence. Dengue virus can be stabilized using saccharide‐based formulations and coated on microneedle array vaccine patches for storage in dry state with preserved viability at ambient temperature (VSMN; virus‐stabilized microneedle arrays).


| INTRODUCTION
There are two main criteria that denote a successful vaccine-it must elicit life-long protective immunity against disease from wild-type microbial infection and it must be accessible by those who live in pathogen endemic areas. Lessening the dependence on cold chain infrastructure reduces vaccine cost and improves the capacity to distribute doses to geographically remote communities. [1][2][3] The controlled temperature chain (CTC) involves final sections of vaccine transport at ambient temperature through infrastructure-weak areas and has the potential to reduce vaccination cost by up to 50%. 1,3,4 Virus vaccines represent a significant challenge in the transition towards CTC implementation. Most virus vaccines are live-attenuated vaccines (LAVs), which comprise viable virus particles that infect host cells in order to trigger an immune response akin to the natural infection. LAVs induce both humoral and cellular adaptive immunity with robust memory responses; however, despite the immunological advantages, LAVs currently require stringent cold chain for shipment. This cold chain is especially critical for enveloped viruses, such as flaviviruses, that are more unstable than nonenveloped viruses such as poliovirus.
Our efforts are directed towards the development of a vaccine against dengue virus (DENV). DENV infection is the leading cause of mosquito-borne illness throughout the tropical world, where an estimated 2.5 billion people live at risk of infection each year. 5,6 Travelers to dengue endemic countries are also at risk of disease and facilitating viral transmission. 6,7 Since the isolation of DENV in 1943, the development of an effective vaccine that provides balanced, long-lasting immunity has been challenging. This is partly due to the complexity of dengue epidemiology, population immunity, and the interplay between viral pathogenesis and host immunological mechanisms. [8][9][10][11][12] Like other successful flavivirus vaccines against yellow fever and Japanese encephalitis, the majority of dengue vaccine candidates either approved for limited license or in Phase III trials are LAVs. 8,[13][14][15] These require shipment in lyophilized form at 4 C and must be reconstituted immediately before use to avoid loss in potency. 16,17 Therefore, systems enabling the preservation of DENV viability in dry formulation and at ambient temperatures will be essential for the development of LAVs that can be delivered and administered under CTC conditions.
Flaviviruses, such as DENV, are transmitted by mosquitos following skin penetration by the mosquito fascicle, nature's ideal polymeric microneedle (MN) composed of chitin. The virus infects and replicates within dermal dendritic cells and Langerhans cells, which traffic to lymph nodes and are responsible for processing and presentation of viral antigens during the induction of antiviral immune responses. 18,19 Intradermal delivery of dengue LAVs would leverage this immune activation in the skin. Hence, another attractive approach for dengue vaccination involves the use of minimally invasive MN-based skin patches. MNs have been successfully used to deliver a broad range of small molecule drugs, biotherapeutics and vaccines to the skin and mucosal surfaces. 20 By mimicking the immunological context of natural dengue infection through the skin, native pathways are activated to elicit robust protective immunity. [21][22][23] Furthermore, MN-based therapeutics offer additional logistical benefits including low cost of materials, and easy pain-free administration, which lessens the dependence on trained medical staff and offers the potential for improved compliance. 20,24,25 We aimed to develop a dengue LAV vaccine platform that combined the advantages of intradermal delivery using MNs with a tailored stabilizing formulation optimized to preserve virus viability. Here we describe the fabrication and efficacy of dengue virus-stabilized microneedle arrays (VSMNs) formulated to enable virus vaccine storage in the dry state at ambient temperature. VSMNs applied to the skin were shown to induce protective immunity comparable to that of freshly prepared liquid virus. These steps represent important advances towards the development of intradermal dengue vaccines designed for the CTC supply chain.  with a base formulation of 0.5% CMC and 7.5% trehalose was utilized for VSMN production in a single layer dried film.

| DENV formulation for enhanced stability
Next, the stability of DENV VSMNs was monitored over a 5-day period with different storage conditions, as measured by DENV-2 viability. Storage of dry VSMN at −80 C did not result in loss of As a second method to assess the extent of virus delivery from VSMN, plaque assays were performed to quantify the load of viable virus coated on VSMN prior to and remaining after application. This indicated that approximately 50% of the total coated virus was delivered to the skin (Figure 2d). This included virus transferred from needles that penetrate into the skin, in addition to virus transferred from the array base to the skin surface. It is difficult to determine the precise dose of virus that infects host cells in the skin. However, the replication of delivered virus enables the amplification of antigenic material and inflammatory signals that promote the immune response following vaccination. While a simple coated MN design was favored for this study, future vaccines must permit dose standardization to meet quality control and manufacturing standards. Recent work by

| VSMN-mediated immunity protects from viral challenge
An effective vaccine must provide protection from disease challenge.
To demonstrate the quality of immunity generated by VSMN, we immunized animals with the DENV-2 attenuated vaccine strain PDK53 by either VSMN or subcutaneous administration. Subcutaneous injection of PBS served as a negative unimmunized control. After 2 weeks, all groups were challenged with a high dose (10 7 pfu) of the wild-type DENV-2 16681 strain by intraperitoneal inoculation and viremia was monitored in the serum for 5 days post-challenge, as depicted ( Figure 5a). Following challenge, control animals that received PBS during priming succumbed to rapid acute viremia exceeding 10 10 vRNA copies/mL serum within 4 days (Figure 5b). In contrast, animals immunized with PDK53 by VSMN or by subcutaneous route were protected from disease, with complete viral clearance to below detectable levels by Day 4 postchallenge (Figure 5b). This confirms that immunization with stabilized live-attenuated DENV delivered from our VSMN platform can achieve strong protective immunity that is noninferior to subcutaneous vaccination using fresh virus.

| Conclusions
The stability of virus particles to retain infectivity is critical for LAV efficacy and the protection of immunized patients. Stabilizing

| Mice
Animal studies were conducted at the National University of Singapore in accordance with Institutional Animal Care and Use Committee regulations and approvals. AG129 (Interferon type I/II receptor-deficient) mice breeders were purchased from B&K Universal and bred at InVivos (Singapore) under specific pathogen-free conditions; 7-to 10-week-old female AG129 mice were used for experiments.

| Confocal microscopy
Microscopy was performed using a Zeiss Axio Observer Z1 LSM 700 (488 nm laser) at 10× objective. Normalized signal intensities were determined by integrating the total confocal fluorescence signal from z-stacks collected through the length of individual MNs and normalizing to the total fluorescence of MNs prior to skin application.
Image processing and data analysis was performed using Image J.  BHK-21 cell monolayers were infected with 200 μL incubated virus for 1 hr at 37 C. Cells were overlaid with RPMI/2% CMC/1% pen/strep, incubated for 5 days, then plaques were stained and counted as described for the plaque assay. The serum dilution resulting in 50% inhibition of plaque formation relative to virus only control (PRNT 50 ) was determined in Prism using a nonlinear regression with variable slope (4 parameters) constrained to 100 and 0 for top and bottom, respectively.

| Viral challenge
Mice were immunized with 5 × 10 5 pfu DENV-2 PDK-53 (attenuated vaccine strain) via VSMN intradermal or subcutaneous administration, as described above. Control animals received subcutaneous administration of 200 μL endotoxin-free PBS. All groups were challenged on Day 14 with 1 × 10 7 pfu DENV-2 16681 (parental wild-type strain) by intraperitoneal injection. Serum viremia was analyzed as a measure of protection up to 5 days post-challenge by quantitative PCR.

| Statistical analysis
Data sets were analyzed using t tests, Mann-Whitney tests or Kruskal-Wallis tests using Prism (GraphPad Software, San Diego, CA). Unless otherwise indicated, data is presented as mean ± SEM. with statistical significance defined by p-values; *p < 0.05, **p < 0.01, ***p < 0.001.