Dissolving Microneedles for Intradermal Vaccination against Shigellosis

Intradermal (ID) immunization is of increasing interest due to the easy accessibility and excellent immunogenic properties of the skin. Among ID immunization methods, dissolving microneedles (MNs) have appeared as an alternative to traditional hypodermic immunization, offering many advantages, such as being an easily administered method, with no need for health personnel, painless, and avoiding the use of needles and sharp wastage. In this study, an affordable and easy-to-produce MNs method was developed based on aqueous blends of 30% w/w poly (methyl vinyl ether-co-maleic anhydride). As an antigen model, a subunit vaccine candidate based on outer membrane vesicles from Shigella flexneri was used. Both unloaded and antigen-loaded MNs were synthetized and characterized. The MNs were successfully validated in an in vitro Parafilm M® skin model and in a pig skin ex vivo model. Biodistribution studies were performed in BALB/c mice using 99mTcO4− radiolabeled samples. Results indicated that the vesicle vaccine was successfully released from the MNs and targeted gastrointestinal tract after 6 h post-administration. In vivo immunization and protection studies were performed in BALB/c mice. Mice were intradermally immunized through ear skin with one single dose of 200 μg antigenic complex, eliciting the production of specific systemic IgG and mucosal IgA. Moreover, MNs were able to protect mice from an experimental infection with 1×106 CFU/mouse of S. flexneri four weeks after immunization. This work demonstrates for the first time the potential of outer membrane vesicle-loaded dissolving MNs for ID vaccination against enteropathogens like Shigella.


Introduction
Vaccination is one of the main strategies to prevent and control infectious diseases [1]. However, during recent decades, many efforts have been being made to obtain alternatives to conventional intramuscular (IM) or subcutaneous (SC) vaccine administration routes, which are mainly administered via hypodermic needles and need trained healthcare personnel [2]. The intradermal (ID) immunization route is of great interest, due to the easy accessibility to the large surface area of the skin and because it has unique immunological properties. The skin contains a rich skin-associated lymphoid tissue

Outer Membrane Vesicle Obtention
An antigenic complex based on outer membrane vesicles was obtained from the S. flexneri ∆tolR mutant strain, as previously described [18]. The parental strain S. flexneri 2a is a clinical isolate ("Clínica Universidad de Navarra", Spain). Both the wildtype and the mutant strains were cultured on tryptone soy agar (TSA, Biomerieux, France) or in tryptic soy broth (TSB, Biomerieux, Spain). Incubations were performed at 37 • C with shaking (140 rpm) to log phase (OD 600 ∼0.3). Bacterial cultures were heat inactivated (HT) by flowing steam (100 • C) for 15 min. After centrifugation at 6 000 × g for 20 min, the HT-∆tolR containing supernatants were filtered (0.22 µm) and purified using 100 kDa-tangential filtration (Millipore). The retenant was then harvested and centrifugated (51,000 × g, 1 h), and the pellets resuspended in deionized water, lyophilized, and stored at room temperature (RT) until use. The HT-∆tolR antigenic complex was incorporated to MN, as described in Section 2.2.

Polymeric MN Arrays Formulation
Although the polymer used for this work was Gantrez ® AN119, other Gantrez ® polymers were also studied to compare efficacy and toxicity. Briefly, each microneedle (MN) was prepared from an aqueous blend of a copolymer of methyl-vinyl-ether-co-maleic anhydride (Gantrez ® ). Briefly, a stock of each polymer was prepared and then diluted with an aqueous suspension containing the HT-∆tolR-based antigenic complex, to reach a determined final concentration of the polymers (Table 1). Once the formulations were homogeneous, they were centrifuged (3500 rpm, 15 min) in order to remove air bubbles. To make the microneedle (MN) arrays, one hundred mg of the formulation containing 200 µg of the antigens was poured into each silicon micromold, avoiding bubbles, and centrifuged at 3500 rpm during 15 min to lower the formulation into the mold [19,20]. Finally, MNs were dried at room temperature (RT) for 48 h and removed from the molds. Each mold contained 361 (19 × 19) pyramidal shaped micro-needles, 500 µm tall and 300 µm wide.

Compression Test
In order to study the mechanical properties of the MNs after application, the compression of the needles after applying a determined force [21] was studied. For this purpose, MNs were attached to the probe of a TA.XTPlus Texture Analyser (Stable MicroSystems, Surrey, UK) and the probe was lowered vertically at 1.19 mm s −1 until the required force of 32 N/array was applied during 30 s. MNs were visually compared before and after the compression test using a light microscope (GXMGE-5 digital microscope, Laboratory Analysis Ltd, Devon, UK), and the heights of individual needles were measured using ImageJ ® software. Finally, the percentage (%) of change in MN height was calculated using Equation (1), where HBC is the needle height before the compression and HAC is the needle height after the compression:

Insertion Test
The insertion test was performed using the previously described Parafilm M ® method [21]. Briefly, eight layers of 1 cm 2 Parafilm M ® were put on top of each other, forming a 1 mm thick layer. The Texture Analyser (TA) insertion was then performed in compression mode, as previously described. MNs were then removed from the Parafilm and each layer was unfolded. The number of holes created in each layer were counted using a Leica EZ4-D digital microscope (Leica, Wetzlar, Germany) as a measure of insertion depth.

Skin Insertion Studies
Ex vivo studies in skin were performed in order to confirm MN insertion and to determine the time of dissolution to establish the optimal application time for further in vivo studies. Full thickness neonatal porcine skin can be considered a good model for human skin in terms of hair sparseness and physical properties [22]. It was obtained from stillborn piglets and excised <24.0 h after birth. MNs were inserted directly in the skin for 30 s, applying a 32 N force, and the insertion was analyzed at 0, 5, 15, 30, and 60 min by optical coherence tomography (OCT), using an EX1301 VivoSight ® OCT Microscope (Michelson Diagnostics Ltd., Kent, UK) with a hand-held probe. Once insertion studies were performed in pig skin, studies in murine ears were carried out in order to characterize the insertion at the place of administration. Briefly, mice (n = 3) were sacrificed, and ears were removed and incubated at 37 • C till use. MNs were then inserted, applying pressure for 30 s. MNs were observed at time 0, 5, 15, 30, and 60 min, using OCT.

Immunization and Protection In Vivo Studies
All mice used in these studies were treated in accordance with institutional guidelines for treatment of animals (Ethical Committee for Animal Experimentation of the University of Navarra, Spain; Protocol ref. CEEA 164/14).

Comparative Study between Gantrez ® Polymers
In order to determine the suitability of three different Gantrez ® polymers, nine-week-old female BALB/c mice (20 ± 1 g) were separated into seven randomized groups of six animals and were immunized with the empty or the antigen-loaded MNs formulated with different Gantrez ® polymers (S97, AN139 and AN119) containing 200 µg of HT-∆tolR antigens. Non-immunized mice were used as control.
MNs were manually applied on the dorsal surface of the ear of the mouse under mild anesthesia, applying pressure for 5 min, and were removed after 24 h. Blood was taken before immunization (time 0) and at week 1, 2, 3, and 4 post-immunization. Total IgG or specific IgG1, IgG2a, IgG3 antibodies in sera, and IgA antibody in feces against HT-∆tolR were determined by ELISA. Briefly, 96-well plates (MaxiSorb; Nunc, Germany) were coated with HT-∆tolR (10 µg/well) in coating buffer (60 mM carbonate buffer, pH 9.6). Unspecific binding sites were blocked with 3% bovine serum albumin (BSA) in phosphate buffered saline (PBS) for 1 h at RT. Sera from the immunized mice were diluted 1:100 in PBS with 1% BSA and incubated for 4 h at RT. After washing with PBS Tween20 (PBS-T) buffer, class-specific goat anti-mouse conjugated antibodies were added and incubated for 1 h at RT. Absorbance was measured with an ELISA plate reader (Sunrise; Tecan-Austria, O.D. 405 nm) after incubation with H2O2-ABTS substrate-chromogen for 15 min at RT.
Four weeks after immunization, all groups were challenged with 2 × 10 6 colony forming units (CFU) per mice of Shigella flexneri intranasally administered. The doses were confirmed by viable count in TSB. Two days after the infection, mice were sacrificed, and lungs were harvested for colony counting.

Immunogenicity and Protection with Gantrez ® AN119
Gantrez ® AN119 was selected to perform further mouse experiments. Briefly, nine-week-old female BALB/c mice (20 ± 1 g) were separated in randomized groups of six animals and immunized with empty or antigen-loaded Gantrez ® AN119 MNs. With the aim of comparing two different doses of HT-∆tolR in MNs, a low concentration (20 µg/mouse) and a high concentration (200 µg/mouse) of antigens were tested. A group immunized with one single intranasal dose of HT-∆tolR (20 µg) and a non-immunized group were used as controls. Blood was taken before immunization (time 0) and at week 1, 2, 3, and 4 post-immunization. Specific serum IgG1, IgG2a, and fecal IgA antibodies against HT-∆tolR were determined by ELISA, as previously described. Four weeks after immunization, all groups were challenged intranasally with 2 × 10 6 CFU of S. flexneri per mouse. Two days after the infection, mice were sacrificed, and lungs were harvested for colony counting.

Biodistribution Studies
Radiolabeling of HT-∆tolR was performed by technetium-99m ( 99m TcO 4 − ) reduction with Sn ++ . Electric) following the manufacturer's instructions. Forty µL of SnCl 2 ·2H 2 O (1 mg/mL) was used and no 99m Tc-tin colloids were produced during the radiolabeling reaction. HT-∆tolR (300 mg) were pre-tinned with a HCl acidified tin chloride solution, 99m TcO 4 − in saline added and reduction carried out in a non-oxidizing atmosphere using He-purged vials and solutions. The radiochemical purity of radiolabeled vesicles was checked by thin layer chromatography (TLC) using Whatmann 3MM strips (GE Healthcare, Chicago, IL, USA) developed with 0.9% NaCl. Radioactivity distribution was measured and quantified using a radioTLC system (miniGITA, Raystest GmBH, Straubenhardt, Germany). Radiolabeling proceeded with >95-97% yield, thus avoiding the need for further purification of the radiolabeled product. After the quality control, 146 L of the radiolabeled vesicles was mixed with 439 mg of Gantrez ® AN 119 (40%) to reach a concentration of 0.44 µg radiolabeled HT-∆tolR/mg formulation in a 30% Gantrez ® solution. Finally, MNs were fabricated as mentioned above (see Section 2.1), obtaining 20 µg of radiolabeled HT-∆tolR/patch. MNs were manually applied on the dorsal surface of the ear of the mouse under mild anesthesia and removed after 6 h. The selection of this period of time was due to the half-life of the 99m Tc selected for this study, which is 6 h, and hence any activity dose would have dropped after that period of time [23].

Statistical Analysis
Statistical analysis was carried out using parametric t-student test, one-way ANOVA, or the non-parametric Mann-Whitney U test, as required. p values of <0.05 were considered statistically significant. All calculations were performed using GraphPad Prism5 ® software.

Polymeric MN Arrays Formulation
The obtained Gantrez ® based MNs were 500 µm high and 300 µm wide at the base, containing 19 × 19 pyramidal shaped needles with sharp tips, ideal for insertion into the skin ( Figure 1). The different Gantrez ® polymers were able to form complete arrays of needles, and the antigens did not seem to affect the structural properties of the MNs. in a non-oxidizing atmosphere using He-purged vials and solutions. The radiochemical purity of radiolabeled vesicles was checked by thin layer chromatography (TLC) using Whatmann 3MM strips (GE Healthcare, Chicago, IL, USA) developed with 0.9% NaCl. Radioactivity distribution was measured and quantified using a radioTLC system (miniGITA, Raystest GmBH, Straubenhardt, Germany). Radiolabeling proceeded with >95-97% yield, thus avoiding the need for further purification of the radiolabeled product. After the quality control, 146 L of the radiolabeled vesicles was mixed with 439 mg of Gantrez ® AN 119 (40%) to reach a concentration of 0.44 μg radiolabeled HT-ΔtolR/mg formulation in a 30% Gantrez ® solution. Finally, MNs were fabricated as mentioned above (see Section 2.1), obtaining 20 μg of radiolabeled HT-ΔtolR/patch. MNs were manually applied on the dorsal surface of the ear of the mouse under mild anesthesia and removed after 6 h. The selection of this period of time was due to the half-life of the 99m Tc selected for this study, which is 6 h, and hence any activity dose would have dropped after that period of time [23].

Statistical Analysis
Statistical analysis was carried out using parametric t-student test, one-way ANOVA, or the nonparametric Mann-Whitney U test, as required. p values of <0.05 were considered statistically significant. All calculations were performed using GraphPad Prism5 ® software.

Polymeric MN Arrays Formulation
The obtained Gantrez ® based MNs were 500 μm high and 300 μm wide at the base, containing 19 × 19 pyramidal shaped needles with sharp tips, ideal for insertion into the skin ( Figure 1). The different Gantrez ® polymers were able to form complete arrays of needles, and the antigens did not

Mechanical Characterization of MNs
In order to characterize the mechanical properties of Gantrez ® AN119 MNs, arrays were submitted to a force of 32 N, previously stablished by Larrañeta et al., as the mean force that is exerted on the skin by humans when applying the patches [21], and their compression and insertion capacities were evaluated. After the application of the force, both empty and loaded-MNs decreased in height (10.4% and 12.5%, respectively), showing no significant differences between them, thus demonstrating that antigens did not affect the integrity of the patch (Figure 2a-c).

Mechanical Characterization of MNs
In order to characterize the mechanical properties of Gantrez ® AN119 MNs, arrays were submitted to a force of 32 N, previously stablished by Larrañeta et al., as the mean force that is exerted on the skin by humans when applying the patches [21], and their compression and insertion capacities were evaluated. After the application of the force, both empty and loaded-MNs decreased in height (10.4% and 12.5%, respectively), showing no significant differences between them, thus demonstrating that antigens did not affect the integrity of the patch (Figure 2a,b,c).
Moreover, both empty and loaded-MNs could successfully insert in Parafilm M ® (Sigma-Aldrich, St. Louis, MO, USA), obtaining a similar percentage of holes created in each layer, with 80% of the needle tips inserted through the first two layers. Moreover, less than 20% of the needle tips were capable of being inserted through the third and fourth layers. Thus, the MN insertion depth was between 300 μm and 400 μm, which represents approximately 60-80% of the needle length (Figure 2d,e,f).

Insertion in Pig Skin
Ex vivo insertion studies were first performed in porcine skin, as a good model for human skin [22], and were visualized using optical coherence tomography (OTC). Both empty and antigenloaded MN arrays could successfully insert into porcine skin. As observed in Figure 3, MN array started to dissolve after the application, and was completely dissolved after 45 min.  Moreover, both empty and loaded-MNs could successfully insert in Parafilm M ® (Sigma-Aldrich, St. Louis, MO, USA), obtaining a similar percentage of holes created in each layer, with 80% of the needle tips inserted through the first two layers. Moreover, less than 20% of the needle tips were capable of being inserted through the third and fourth layers. Thus, the MN insertion depth was between 300 µm and 400 µm, which represents approximately 60-80% of the needle length (Figure 2d-f).

Insertion in Pig Skin
Ex vivo insertion studies were first performed in porcine skin, as a good model for human skin [22], and were visualized using optical coherence tomography (OTC). Both empty and antigen-loaded MN arrays could successfully insert into porcine skin. As observed in Figure 3, MN array started to dissolve after the application, and was completely dissolved after 45 min.

Mechanical Characterization of MNs
In order to characterize the mechanical properties of Gantrez ® AN119 MNs, arrays were submitted to a force of 32 N, previously stablished by Larrañeta et al., as the mean force that is exerted on the skin by humans when applying the patches [21], and their compression and insertion capacities were evaluated. After the application of the force, both empty and loaded-MNs decreased in height (10.4% and 12.5%, respectively), showing no significant differences between them, thus demonstrating that antigens did not affect the integrity of the patch (Figure 2a,b,c).
Moreover, both empty and loaded-MNs could successfully insert in Parafilm M ® (Sigma-Aldrich, St. Louis, MO, USA), obtaining a similar percentage of holes created in each layer, with 80% of the needle tips inserted through the first two layers. Moreover, less than 20% of the needle tips were capable of being inserted through the third and fourth layers. Thus, the MN insertion depth was between 300 μm and 400 μm, which represents approximately 60-80% of the needle length (Figure 2d,e,f).

Insertion in Pig Skin
Ex vivo insertion studies were first performed in porcine skin, as a good model for human skin [22], and were visualized using optical coherence tomography (OTC). Both empty and antigenloaded MN arrays could successfully insert into porcine skin. As observed in Figure 3, MN array started to dissolve after the application, and was completely dissolved after 45 min.

Insertion in Mice Skin
After the ex vivo studies in porcine skin, insertion was evaluated in murine ear skin, since it was the place of administration in further in vivo studies, and images were also analyzed by OTC. As observed in Figure 4, both empty and antigen-loaded MN arrays could also insert into mice skin, and different images showed that 60 min after the application of MNs, the MNs were almost completely dissolved.

Insertion in Mice Skin
After the ex vivo studies in porcine skin, insertion was evaluated in murine ear skin, since it was the place of administration in further in vivo studies, and images were also analyzed by OTC. As observed in Figure 4, both empty and antigen-loaded MN arrays could also insert into mice skin, and different images showed that 60 min after the application of MNs, the MNs were almost completely dissolved.

Immunization and Protection Studies in vivo
In a first animal study, mice were immunized with MNs formulated with different Gantrez ® polymers (S97, AN139 and AN119) containing 200 μg of HT-ΔtolR antigens. Blood was taken once

Immunization and Protection Studies In Vivo
In a first animal study, mice were immunized with MNs formulated with different Gantrez ® polymers (S97, AN139 and AN119) containing 200 µg of HT-∆tolR antigens. Blood was taken once four weeks post-immunization and specific serum levels of total IgG, IgG1, IgG2a, IgG3, and fecal IgA against HT-∆tolR antigens were determined by ELISA. Results indicated that the immunization with MN elicited significant levels of serum total IgG (p < 0.01). Specifically, serum IgG2a and IgG3 (p < 0.001), were increased four weeks post-immunization with respect to time 0, whereas IgG1 levels were lower (p < 0.01). Finally, immunization with MNs also elicited high levels of IgA in mice feces (p < 0.001) ( Figure 5). four weeks post-immunization and specific serum levels of total IgG, IgG1, IgG2a, IgG3, and fecal IgA against HT-ΔtolR antigens were determined by ELISA. Results indicated that the immunization with MN elicited significant levels of serum total IgG (p < 0.01). Specifically, serum IgG2a and IgG3 (p < 0.001), were increased four weeks post-immunization with respect to time 0, whereas IgG1 levels were lower (p < 0.01). Finally, immunization with MNs also elicited high levels of IgA in mice feces (p < 0.001) ( Figure 5). Four weeks post-immunization, mice were infected via the intranasal route with a sublethal dose of 1 × 10 6 CFU/mouse. Lungs were harvested and bacteria were counted. The three different groups of immunized mice showed a decrease in CFU/mL with respect to non-immunized group, although no significant differences (p > 0.05) were observed between the non-immunized group and the different Gantrez ® formulations groups ( Figure 6). Four weeks post-immunization, mice were infected via the intranasal route with a sublethal dose of 1 × 10 6 CFU/mouse. Lungs were harvested and bacteria were counted. The three different groups of immunized mice showed a decrease in CFU/mL with respect to non-immunized group, although no significant differences (p > 0.05) were observed between the non-immunized group and the different Gantrez ® formulations groups ( Figure 6). Once it was confirmed that Gantrez ® AN119 showed no differences with respect to the other Gantrez ® polymers, and since this particular polymer has not been previously used before to form MNs, this polymer was selected to perform further mice experiments.
Thus, in order to determine whether the immunization with Gantrez ® AN119 MNs was protective against an infection with S. flexneri, mice were immunized with a low (20 μg) and high (200 μg) dose of the antigens loaded in the MNs and blood was taken once per week for four weeks. Specific serum levels of IgG1 and IgG2a and fecal IgA against HT-ΔtolR antigens were determined by ELISA. The immunization with both MN groups and free antigens elicited significant levels of serum IgG2a after four weeks post-immunization with respect to time 0 (p < 0.001) (Figure 7a), whereas only mice immunized with MNs with 200 μg (p < 0.05) or free antigens (p < 0.01) elicited significant IgG1 levels ( Figure 7b). This imbalanced result in the Th1 and Th2 antibody release was represented with the relative amount of IgG2a and IgG1 at week 4 post-immunization. Results indicate that the groups immunized through MNs showed a marked Th1 profile, whereas the difference between both antibodies was lower in the group immunized IN with the free antigens (Figure 7d). Regarding mucosal immune response, only mice immunized with free antigens showed significant levels of IgA in feces in week 4 with respect to time 0 (p < 0.05), although the levels of IgA elicited by MNs with high dose of antigens were significantly higher than this group until week 3 (p < 0.05) when compared between them (Figure 7c). Once it was confirmed that Gantrez ® AN119 showed no differences with respect to the other Gantrez ® polymers, and since this particular polymer has not been previously used before to form MNs, this polymer was selected to perform further mice experiments.
Thus, in order to determine whether the immunization with Gantrez ® AN119 MNs was protective against an infection with S. flexneri, mice were immunized with a low (20 µg) and high (200 µg) dose of the antigens loaded in the MNs and blood was taken once per week for four weeks. Specific serum levels of IgG1 and IgG2a and fecal IgA against HT-∆tolR antigens were determined by ELISA. The immunization with both MN groups and free antigens elicited significant levels of serum IgG2a after four weeks post-immunization with respect to time 0 (p < 0.001) (Figure 7a), whereas only mice immunized with MNs with 200 µg (p < 0.05) or free antigens (p < 0.01) elicited significant IgG1 levels (Figure 7b). This imbalanced result in the Th1 and Th2 antibody release was represented with the relative amount of IgG2a and IgG1 at week 4 post-immunization. Results indicate that the groups immunized through MNs showed a marked Th1 profile, whereas the difference between both antibodies was lower in the group immunized IN with the free antigens (Figure 7d). Regarding mucosal immune response, only mice immunized with free antigens showed significant levels of IgA in feces in week 4 with respect to time 0 (p < 0.05), although the levels of IgA elicited by MNs with high dose of antigens were significantly higher than this group until week 3 (p < 0.05) when compared between them (Figure 7c).
Four weeks post-immunization, mice were infected intranasally with a sublethal dose of 1 × 10 6 CFU/mouse. Lungs were harvested and CFU were counted. Responder and non-responder mice could be identified in all immunized groups. Results showed a decrease in the number of CFU in vaccinated mice with respect to the non-immunized group. Moreover, the bacteria cargo also decreased in the two groups immunized with MNs when compared to mice immunized intranasally with the free antigens. Finally, no differences were obtained between the two different antigen doses tested in MNs (Figure 8). Four weeks post-immunization, mice were infected intranasally with a sublethal dose of 1 × 10 6 CFU/mouse. Lungs were harvested and CFU were counted. Responder and non-responder mice could be identified in all immunized groups. Results showed a decrease in the number of CFU in vaccinated mice with respect to the non-immunized group. Moreover, the bacteria cargo also decreased in the two groups immunized with MNs when compared to mice immunized intranasally with the free antigens. Finally, no differences were obtained between the two different antigen doses tested in MNs (Figure 8).

Biodistribution Studies
In order to compare the effect of MNs on the biodistribution of the vaccine, 99m Tc-HT-∆tolR (20 μg) was administered either intradermally with MNs or intranasally, as control. After nasal administration, the percentage of radioactivity in the large intestine increased (50%) as radioactivity in the nose decreased (23%) 6 h post-administration (Figure 9a). Radiolabeled 99m Tc HT-∆tolR were not detected in the other organs studied.

Biodistribution Studies
In order to compare the effect of MNs on the biodistribution of the vaccine, 99m Tc-HT-∆tolR (20 µg) was administered either intradermally with MNs or intranasally, as control. After nasal administration, the percentage of radioactivity in the large intestine increased (50%) as radioactivity in the nose decreased (23%) 6 h post-administration (Figure 9a). Radiolabeled 99m Tc HT-∆tolR were not detected in the other organs studied.
dose of 20 μg of free HT-∆tolR (intranasally) or 20 or 200 μg of HT-∆tolR embedded in Gantrez ® AN119 MNs (six mice per group). Results are presented as degree of colonization (Log CFU/mL) in lungs. Non-immunized and empty-MN groups were included as controls. Error bars represent SD.

Biodistribution Studies
In order to compare the effect of MNs on the biodistribution of the vaccine, 99m Tc-HT-∆tolR (20 μg) was administered either intradermally with MNs or intranasally, as control. After nasal administration, the percentage of radioactivity in the large intestine increased (50%) as radioactivity in the nose decreased (23%) 6 h post-administration (Figure 9a). Radiolabeled 99m Tc HT-∆tolR were not detected in the other organs studied. After the administration of MNs on the ear skin of the mice, almost 60% of the radioactive signal remained in this location. From the amount (40%) distributed to other organs, more than half (56%) of the formulation was detected in the gastrointestinal tract 6 h post-administration (Figure 9b).

Discussion
In recent years, there has been an increasing requirement for alternative vaccine delivery systems to avoid the use of injectable devices, which need trained personnel for their administration and may require specialized storage and transport conditions [24]. This is especially important in the developing world, where numerous infectious diseases are endemic [25]. Furthermore, from an immunological point of view, although injectable vaccines are capable of eliciting robust systemic responses, they sometimes suffer from a lack of mucosal protection, which is essential to protect against specific infectious agents, such as enteropathogens [24]. The main alternative routes of vaccine delivery without needles capable of eliciting both mucosal and systemic immune responses are mucosal and intradermal ones [26]. These routes stimulate the mucosa-associated lymphoid tissue (MALT) (e.g., the gut-associated lymphoid tissue, GALT, and the nasal-associated lymphoid tissue, NALT), and the skin-associated lymphoid tissue (SALT). Moreover, these routes are attractive due to the rapid and wide biodistribution of the antigens, for being self-administered and painless, avoiding the use of needles, and not needing healthcare personnel [27].
The ID route is particularly interesting to get patients' compliance due to the easy accessibility to the skin [28]. A number of strategies have been proposed for the ID delivery of antigens without the employment of needles, such as high-velocity needle-free injection systems, permeability enhancers to disrupt the stratum corneum, or the use of antigen delivery systems with percutaneous absorption properties or diffusive capacity [29]. Among these, microneedles (MNs) are gaining special interest due their easy and self-vaccination process and the lack of generation of biohazardous waste, avoiding problems related to needle-phobia, needle-associated infections, or waste management [30].
Among them, and based on their self-disabling properties, dissolving MNs have been selected for this work. MNs were prepared by using Gantrez ® , a poly(methyl vinyl ether-co-maleic anhydride) polymer, which has been extensively employed for pharmaceutical applications, such as denture adhesives, thickening, suspending agents, or carriers for oral delivery, due to its mucoadhesive properties and low oral toxicity [31]. More recently, it has been used for the preparation of transdermal patches, showing good properties to form dissolving MNs [14,32]. Among the different types of Gantrez ® polymers used, Gantrez ® AN119 has not been reported to perform MNs. We have previously shown that this particular polymer is non-cytotoxic and has immunoadjuvant properties [15,16]. Indeed, AN119 MNs were successfully produced and showed excellent mechanical and insertion properties when analyzed in both Parafilm M ® and neonatal porcine skin ( Figure 2). Moreover, OCT showed that the MNs could successfully penetrate murine skin and the microprojections of the array were completely dissolved in the first hour after application. In vivo studies in mice also demonstrated the safety of this polymer. Previous studies demonstrated that a repeated application of Gantrez ® MNs does not cause any sign of skin alteration in mice [33]. This may be a critical factor, since dissolving MN deposit the polymer on the skin follows complete dissolution and local irritation and erythema have been described elsewhere [10]. In the current study, mice ears did not show any significant irritation or necrotic signs, even after 24 h of application. Moreover, a single dose of the vaccine was considered, which would decrease the risk of any side effect.
MNs were loaded with HT-∆tolR antigenic complex, an outer membrane vesicle-based vaccine obtained from heat-inactivated S. flexneri ∆tolR strain. The complex was previously characterized and demonstrated to be immunogenic and protective in a murine model of shigellosis [18]. To our knowledge, this is the first time that outer membrane vesicles have been administrated through dissolving MNs. To demonstrate the efficacy of the vaccine after cutaneous administration, immunogenicity and protection were evaluated.
In a first in vivo study, a comparison between Gantrez ® AN119 and the polymers AN139 and S97 already used in other studies [11,14,34] was performed. Comparative immunization and protection results did not show significant differences between polymers. The three groups elicited high levels of specific antibodies and were able to protect mice against an experimental infection with S. flexneri ( Figures 5 and 6). These data support the use of the AN119 polymer for transcutaneous immunization, and further in vivo studies were performed. Immunization results confirmed that the antigenic complex embedded in AN119 MNs elicited a predominant IgG2a response (Figure 7), indicating a marked Th1 immunity, pro-inflammatory response which is necessary to detain S. flexneri infection [35]. On the contrary, when the antigenic complex was delivered in solution intranasally, the Th1/Th2 balance was significantly lower. These data suggest a possible role of the Gantrez ® polymer as Th1-adjuvant in the MN delivery. In fact, the role of SALT in inducing a pro-inflammatory response has been previously shown [36]. Furthermore, high levels of IgA were observed in feces from week 1 after immunization with MNs with 200 µg of the antigenic complex, and protection was achieved against an experimental infection. Mucosal secretory IgA is a critical factor to face enteropathogens in the sites of colonization [37,38]. Thus, elevated levels of specific antibodies have been associated with reduced incidence of shigellosis in naturally exposed individuals. Moreover, LPS and IpaB-specific IgA have also been proposed as indicators of protective immunity [39].
All these results indicated the penetration of the HT-∆tolR from the MN patch through the skin layers into the circulation, which was corroborated by the biodistribution studies performed with radiolabeled vaccine. From the vaccine formulation distributed among the different organs, almost 60% was located in the gastrointestinal tract. The connection between the skin and gut is possible through the interconnections of MALT. The great network of immune cells of the SALT recognize the antigens and migrate to peripheral lymph nodes or circulate in the blood to other tissues, including mucosae, reaching the GALT [40].

Conclusions
In summary, the HT-∆tolR antigenic complex was successfully loaded into dissolving MNs for intradermal administration, eliciting specific systemic and mucosal immune responses, which were protective against an experimental infection in mice. This delivery system represents a simple, affordable, and easy-to-use method of vaccine administration against enterobacterial pathogens, facilitating immunization processes in low-and middle-income countries.