Dissolving Undercut Microneedle Arrays for Multicomponent Cutaneous Vaccination

Abstract

The skin is an attractive tissue target for vaccination, as it is readily accessible and contains a dense population of antigen-presenting and immune-accessory cells. Microneedle arrays (MNAs) are emerging as an effective tool for in situ engineering of the cutaneous microenvironment to enable diverse immunization strategies. Here, we present novel dissolving undercut MNAs and demonstrate their application for effective multicomponent cutaneous vaccination. The MNAs are composed of micron-scale needles featuring pyramidal heads supported by undercut stem regions with filleted bases to ensure successful skin penetration and retention during application. Prior efforts to fabricate dissolving undercut microstructures were limited and required complex and lengthy processing and assembly steps. In the current study, we strategically combine three-dimensional (3D) laser lithography, an emerging micro-additive manufacturing method with unique geometric capabilities and nanoscale resolution, and micromolding with favorable materials. This approach enables reproducible production of dissolving MNAs with undercut microneedles that can be tip-loaded with multiple biocargos, such as antigen (ovalbumin) and adjuvant (Poly(I:C)). The resulting MNAs fulfill the geometric (i.e., sharp tips and smooth edges) and mechanical-strength requirements for failure-free penetration of human and murine skin to simultaneously deliver multicomponent (antigen plus adjuvant) vaccines to the same cutaneous microenvironment. Cutaneous vaccination of mice using these MNAs induces more potent antigen-specific cellular and humoral immune responses than those elicited by traditional intramuscular injection. Together, the unique geometric features of these undercut MNAs and the associated manufacturing strategy, which is compatible with diverse drugs and biologics, could enable a broad range of non-cutaneous and cutaneous drug delivery applications, including multicomponent vaccination.

1. Introduction

Skin is the most accessible human tissue with diverse functions as both a passive and active barrier, sensory organ, and thermoregulator [1, 2]. Importantly, the skin is also an active immune organ with high densities of both antigen-presenting cells (APCs) and immunologically active accessory cells that tune innate immune responses [3–5]. Indeed, there is a growing body of evidence supporting the notion that the skin is an ideal anatomic target for immunization [6–8]. However, the true potential of skin-targeted immunization has yet to be realized due to the lack of safe, efficient, patient-friendly, and broadly-applicable cutaneous vaccine delivery platforms.

Currently, most vaccines are administered using hypodermic needle-based injections [9, 10], which have several disadvantages, including a need for administration by trained healthcare personnel, excessive costs of cold chain storage and transport, risk of disease transmission and needle-stick injuries, trypanophobia (fear of needles), and poor patient compliance [11–13]. Furthermore, parenteral injections deliver vaccines into muscle or subcutaneous tissues, which contain fewer immune cells than viable skin [10, 14]. Thus, vaccines administered by conventional injections may result in sub-optimal immunogenicity by essentially bypassing the cutaneous immune system [15]. These factors could limit vaccine efficacy and lead to a lower rate of global immunization coverage.

Cutaneous vaccination is a viable and attractive alternative for immunization due to the aforementioned advantages of the skin as a target site [16, 17]. Indeed, pre-clinical and clinical studies have demonstrated that skin contains a high density of APCs (e.g., Langerhans cells and dermal dendritic cells), and intradermal vaccination induces more potent and durable immune responses than those obtained by intramuscular injections [18, 19]. Since the stratum corneum (outermost layer of the skin) is an effective physical barrier, with limited permeability to hydrophilic and higher molecular weight (> 500 Da) molecules, including most vaccine components [12], intradermal (ID) injections are needed to deliver vaccines through the stratum corneum into the cutaneous microenvironment [6, 16]. However, ID injections are challenging to perform successfully and reproducibly, which renders ID vaccination inconsistent, inefficient, and limited [20]. Accordingly, alternative skin-targeted vaccine delivery platforms are needed to enable safe, effective, reproducible, and convenient immunization strategies.

Microneedle arrays (MNAs) have emerged as an appealing technology to bridge the gap between topical applications and traumatic hypodermic needle injections for patient-friendly vaccine delivery to cutaneous microenvironments [17, 21–23]. Unlike topical delivery approaches, MNAs physically penetrate the stratum corneum, thereby eliminating formulation complexities and resulting in localized deposition of vaccine components in the skin microenvironment [11, 12]. In contrast to injections with traditional needles, microneedles enable pain-free immunization [24]. Initial studies with MNAs have shown that coated-microneedles created from metals or silicon can effectively deliver vaccines to the skin [25–27]. Dissolving MNAs have gained much recent attention as potential vaccine delivery technologies due to their higher antigen loading capacity, simple manufacturing process, and long-term stability [17, 28]. These MNAs are fabricated from water-soluble polymers that, in their dry-state, are strong enough to penetrate the stratum corneum, and then rapidly dissolve and release their cargo upon insertion into hydrated skin tissue [29–31]. We and others have demonstrated the capacity of dissolving MNAs to precisely deliver vaccines to the cutaneous microenvironments with improved efficiencies, thereby requiring relatively lower vaccine doses compared to traditional injections [13, 32–36]. These promising results have inspired exponentially growing interest in development of more effective dissolving MNA designs and scalable fabrication methods [37].

Several manufacturing strategies that involve multiple processing steps have been used to fabricate dissolving MNAs for skin-targeted vaccine and drug delivery [37]. In general, these strategies involve three steps: (1) fabrication of MNA master molds, (2) creation of MNA production molds, and (3) production of dissolving MNAs loaded with biocargos. Fabrication of master molds is a pivotal step for reproducible manufacturing of dissolving MNAs [19, 29]. Prevailing methods for creating MNA master molds include subtractive techniques, such as photolithography [29], laser cutting [38], and micromachining [39]. Micro-additive manufacturing (μAM) has recently emerged as an attractive alternative that offers a high level of simplicity and broad geometric capabilities, thereby enabling fabrication of more innovative microneedle and array designs for application-oriented optimization. Importantly, μAM allows individuals without microfabrication expertise to directly produce their MNA designs from computer-aided design (CAD) drawings without the complex requirements of subtractive fabrication processes [40, 41]. However, the unique advantages of μAM techniques have yet to be exploited for high-throughput fabrication of dissolving MNAs designed specifically for cutaneous drug delivery.

Here, we describe the μAM enabled development and application of novel dissolving MNAs with undercut geometries for multicomponent cutaneous vaccination. The MNA designs include uniquely-shaped micron-scale needles composed of sharp pyramid heads and undercut stems with filleted bases to ensure both successful skin penetration and retention. High-quality master MNAs were fabricated by an emerging three-dimensional (3D) μAM approach, 3D direct laser writing, which offers transformative potential for the MNA field with its unparalleled design capabilities and nanoscale resolution [42]. Conversely, traditional 3D-printing methods, such as stereolithography and fused deposition modeling, which recently have been used to produce non-dissolvable MNAs [43–47], can only achieve resolution several orders of magnitude less [48]. For scalability, high fidelity replicas of master MNAs were obtained from a mechanically-strong, moldable resin using a two-step micromolding approach. These replica master molds were then used to prepare several flexible polydimethylsiloxane (PDMS) production molds, which enabled fabrication of dissolving undercut MNAs tip-loaded with a model antigen (ovalbumin, OVA), with or without an adjuvant (Poly(I:C)) using a spin-casting method. To our knowledge, this is the first study introducing fully dissolving MNAs with true undercut features for effective cutaneous drug delivery. Moreover, the novel MNA designs, along with favorable mold materials, strategically enabled direct fabrication steps without interfering with the molding processes (i.e., easy removal of undercut MNAs from flexible production molds). Importantly, the fabricated MNAs effectively penetrated the stratum corneum of human skin to deliver their cargos to the cutaneous microenvironment, and transcutaneous vaccination of mice with OVA ± Poly(I:C) MNAs induced potent antigen-specific cellular and humoral immune responses. Together, these results demonstrate that synergistic combination of 3D μAM with micromolding provides an effective means for flexible fabrication of high-fidelity dissolving undercut MNAs compatible with a broad range of small molecule and biological cargos. These MNAs fulfilled the geometric and mechanical-strength requirements for effective skin penetration for multicomponent cutaneous vaccination, thereby representing a promising platform for novel skin-targeted immunization.

2. Materials and Methods

2.1. Materials

Ovalbumin (OVA; #A5503), polyinosinic-polycytidylic acid sodium salt (Poly(I:C); #P1530), carboxymethylcellulose (CMC, 90 kDa MW), D-(+)-trehalose dihydrate, polyvinylpyrrolidone (PVP, 40 kDa MW), polyvinyl alcohol (PVA, 87–90% hydrolyzed, 30–70 kDa MW), Allura Red AC (R40 dye), doxorubicin, 4,4’,5,5’-tetramethylbenzidine (TMB) peroxidase substrate, carbonate-bicarbonate buffer (pH 9.6), and Tween20 were purchased from Sigma-Aldrich (St. Louis, MO). Polydimethylsiloxane (PDMS) SYLGARD® 184 and VeroWhiteplus-RGD835 UV-curable resin were obtained from Dow Corning (Midland, MI) and Stratasys (Eden Prairie, MN), respectively. Green fluorescent Degradex PLGA microspheres (10 μm diameter) were acquired from Phosphorex (Hopkinton, MA). Alexa 555-labeled OVA (Invitrogen), Alexa 680-labeled OVA (Invitrogen), Texas Red-labeled dextran (40 kDa MW; Invitrogen), Pierce Micro BCA Protein Assay Kit, SYBR Green EMSA nucleic acid stain, endotoxin-free HyClone Cell Culture Grade Water, RNase-free Ambion TE Buffer (pH 8.0), carboxyfluorescein succinimidyl ester (CFSE; Invitrogen), and DAPI were purchased from Thermo Fisher Scientific (Waltham, MA). Anti-OVA IgG1 (Cayman Chemical, Ann Arbor, MI), anti-OVA IgG2c (Chondrex, Redmond, WA), normal goat serum and biotinylated goat anti-mouse IgG1 and IgG2c secondary antibodies (Jackson ImmunoResearch, West Grove, PA), streptavidin-HRP (BD Biosciences, San Jose, CA), and OVA_257-264 (SIINFEKL) peptide (Anaspec, Fremont, CA) were used for immune assays.

2.2. Fabrication of dissolving microneedle arrays

2.2.1. Microneedle and array designs

The unique microneedle array (MNA) design utilized in this study is shown in Fig. 1. This particular microneedle design consisted of a sharp-tipped pyramid head and an undercut stem portion with a filleted base. The microneedle was 750 μm in height with a 30° apex angle. The stem portion of the microneedle was 150 μm in width and extended from the bottom of a square pyramid head (250 μm x 250 μm base area) to the backing layer of MNA with a 35 μm radius filleted connection. The fillet was specifically designed at the microneedle base to avoid sharp corners and associated mechanical stress concentration, considerably increasing microneedle strength performance during manufacturing processes and skin insertion [19, 41]. The apex angle, width, and height of the microneedles were chosen based on skin anatomy and skin insertion mechanics to ensure failure-free penetration [19, 49]. Notably, this design introduces a novel undercut, or anchor feature, which improves skin retention during application, but still allows direct removal of MNAs from flexible production molds throughout the manufacturing process. The tip-to-tip distance between microneedles in the 5x5 arrays was 650 μm, and the size of MNA was 4.75 mm x 4.75 mm. The array design (microneedle spacing) was based on solid mechanics considerations and skin insertion mechanics to avoid a “bed of nails” effect during skin penetration [19, 31, 50]. The three-dimensional micro-additive manufacturing (3D-μAM) approach provides a simple, reproducible, and revolutionary means to produce the proposed unique MNA design from a 3D-CAD drawing, and allows individuals with no microfabrication expertise to easily create a broad range of MNA designs.

Fig. 1.

Computer-aided design (CAD) drawing of MNAs that include sharp microneedles with undercut features. (A) Geometric parameters of unique microneedle design are depicted on a 2D-CAD drawing. (B) 3D-CAD drawing of a 5x5 MNA.

2.2.2. Manufacturing strategy

The manufacturing strategy used to fabricate dissolving MNAs with novel microneedle designs is graphically summarized in Fig. 2. This strategic six-step approach exploits μAM and micromolding to create dissolving undercut MNAs, while simultaneously achieving high-throughput fabrication: (1) 3D-CAD drawing of the MNA design; (2) direct production of a master MNA from the CAD drawing by 3D direct laser writing using a non-dissolvable resin (IP-S); (3) high-fidelity replication of master MNA with UV-curable resin (VeroWhite) by micromolding; (4) creation of MNA master molds that consist of multiple master MNA replicas on 3D-printed MNA holders; (5) manufacturing of elastomer (PDMS) MNA production molds by micromolding; and (6) fabrication of tip-loaded, dissolving MNAs with undercut microneedles incorporating a vaccine or other biocargo in a water-soluble biocompatible material (e.g., carboxymethylcellulose (CMC) and trehalose) through a spin-casting method. The last step of the process can be modified depending on the biocargo of interest, and typically involves spin-casting cargo (e.g., vaccine) into the tip of the PDMS production molds, followed by spin-casting a dissolvable hydrogel (e.g., CMC/trehalose) into the production molds to serve as the structural material. Notably, the master MNA, master molds consisting of multiple master MNA replicas, and elastomer production molds are reusable, reducing the fabrication costs for dissolving undercut MNAs. At each stage of the fabrication process, optical stereomicroscopy (ZEISS Stemi 2000-C microscope with Olympus OM-D E-M511 camera) was used to assess geometric integrity of the microneedles.

Fig. 2.

Microneedle array manufacturing strategy involves six distinct steps. (1) 3D-CAD drawing of the target MNA design. (2) Direct production of the master MNA from the 3D-CAD drawing by 3D direct laser printing. (3) High-fidelity replication of the master MNA using a two-step micromolding approach. (4) Creation of the MNA master molds consisting of multiple master MNA replicas (e.g., six MNA replicas) on 3D-printed MNA holders. (5) Manufacturing of the MNA production molds from PDMS using micromolding. (6) Fabrication of tip-loaded, dissolving undercut MNAs incorporating the target vaccine using the spin-casting method.

2.2.2.1. Fabrication of master MNA

The unique MNA geometry was designed in SolidWorks 2018 CAD software and directly created from the 3D-CAD drawing (Fig. 1) using 3D laser printing (Nanoscribe Photonic Professional, GT; Nanoscribe Struensee, Germany) with the photopolymeric resist IP-S. The Nanoscribe printing system was equipped with a laser generator, an optical cabinet, a Zeiss optical microscope attached to a lens to focus the laser beam, a Galvo mirror system to direct the laser-beam scanning, a piezoelectric stage for precise motion control, and software (Nanowrite) to execute 3D printing. The whole system was placed on an optical table to eliminate vibrations during the printing process.

To fabricate the master MNA, the CAD design was converted into ‘STL’ (Stereolithography) format. The STL file was loaded into the specialized software (DeScribe, Germany) for the Nanoscribe system to select the processing conditions (distance of slicing, hatching, and splitting). Finally, the STL file was converted into ‘GWL’ (General Writing Lithography) format and exported to the Nanowrite software to print the master MNA. The master MNA was fabricated using Galvo-scan mode in XY plane and piezo-scan mode in Z direction. The master MNA was split into 220 μm × 220 μm × 200 μm blocks within the working range and then stitched together. Laser power and writing speed were set to 100 mW and 6 cm/s, respectively. Minimum and maximum slicing distances of 0.3 μm and 0.5 μm, respectively, were used. The master MNA was then printed through two-photon polymerization of the IP-S photoresist by a femtosecond pulsed laser at a wavelength of 750 nm using a unique deep-in-liquid mode with a 25x NA0.8 objective in Shell and Scaffold mode. After printing, the master MNA was developed in the photoresist solvent propylene glycol monomethyl ether acetate (PGMEA) for 30 min, followed by a 5 min isopropyl alcohol rinse. The master MNA was then air-dried and placed under UV light (365 nm, 16 mW/cm² intensity) for 30 min to further crosslink the body to make the master MNA structure strong.

2.2.2.2. Replication of master MNA

A two-stage micromolding method was used to replicate the master MNA with high-fidelity using a UV-curable resin. First, an elastomer mold, which is a negative mold of the master MNA, was manufactured from polydimethylsiloxane (PDMS) by soft-lithography. Elastomer molding with PDMS is a well-established technique for rapid, accurate, and reproducible replication of high-fidelity micron-scale structures [51, 52]. Briefly, the master MNA was mounted in a petri-dish with a diameter of 5 cm, and PDMS was prepared using a two-component curable silicone elastomer, SYLGARD® 184 (10:1 base-to-curing agent). The PDMS was poured over the master MNA mounted in the petri-dish and degassed for 15 min. Next, the master MNA with degassed PDMS was cured at 70 °C for 1 h. The cured PDMS was cooled to room temperature for 5 min and then separated from the master MNA to obtain the negative PDMS mold.

The second processing step used the negative PDMS mold to fabricate positive master MNA replicas from a UV-curable resin (VeroWhiteplus-RGD835). For each PDMS mold, 20 μL of liquid resin was poured onto the molds, and then the molds were centrifuged (4500 RPM at 20 °C for 1 min; Thermo Fisher Scientific Sorvall Legend XTR centrifuge with Swinging Bucket Rotor TX-750) to fill the microneedle-shaped wells with resin. The resin was then treated under UV light (365 nm) with 21.7 mW/cm² intensity for 5 min from both the top and bottom to cure the base and the microneedle tips. To ensure the backing layers of the master MNA replicas were flat, an additional 50 μL of UV-curable resin, which exceeded the remaining volume available, was deposited onto the PDMS mold. A glass slide was placed on top of the mold to get rid of the excess resin, thereby creating a uniform flat surface at the base. The liquid resin was then cured from the top side for 5 min and demolded to obtain a replica of the master MNA.

2.2.2.3. Creation of MNA master molds and production molds

To improve productivity of the manufacturing process for dissolving MNAs, the MNA master molds were created by assembling six master MNA replicas onto MNA holders fabricated by Stratasys® from a non-dissolvable photo-polymer (VeroWhite) using a high-resolution Polyjet 3D printing system (Objet Connex 500 multi-material). A 3D model of the MNA holder was created using SolidWorks 2018 CAD software and then converted into the ‘STL’ (StereoLithography) file format. Subsequently, the specialized software (Objet Studio) sliced this 3D model into 2D cross-sectional layers, creating a computer file that was sent to the 3D printer system at Stratasys. Channels in the 3D printed MNA holder were designed to serve as pockets in the MNA production molds to assist as reservoirs for both the bioactive cargo (e.g., vaccine) and the structural hydrogel material of dissolving MNAs during the spin-casting process. The MNA master molds were baked at 80 °C overnight in a vacuum oven to facilitate effective molding of elastomer MNA production molds. Subsequently, MNA production molds that included microneedle-shaped wells for six MNAs were fabricated from PDMS as described for replication of the master MNA. Notably, a single MNA master mold can be used repeatedly to fabricate multiple PDMS production molds.

2.2.2.4. Production of dissolving MNAs

Dissolving MNAs with novel undercut microneedles tip-loaded with multicomponent vaccines (OVA antigen ± Poly(I:C) adjuvant) were manufactured through a spin-casting technique with centrifugation at room temperature. First, 5 μL of an aqueous solution of OVA (25 mg/mL) was dispensed to each MNA reservoir on the PDMS production molds, and production molds were centrifuged (1 min at 4500 rpm) to fill the microneedle-shaped cavities. Excess OVA solution within the reservoir was then recovered, and production molds were centrifuged (30 min at 4500 rpm) to ensure that dry OVA cargo was located at the tip portion of the microneedle-shaped cavities in the production molds. For MNAs integrating OVA and Poly(I:C), the aforementioned process was repeated with 5 μL of an aqueous solution of Poly(I:C) (62.5 mg/mL). The final MNAs used for cutaneous vaccination experiments included 10 μg OVA and 25 μg Poly(I:C) per MNA. The tip-loading process and recovery of excess biocargo is depicted in Fig. S1.

After loading vaccine biocargo at the tips of microneedles, the MNA structural biomaterial was prepared by dissolving a 70:30 mixture of sodium carboxymethylcellulose (CMC) and D-(+)-trehalose dihydrate in endotoxin-free water at a total solute concentration of 30 % w/w. The resulting CMC/trehalose hydrogel was loaded onto each MNA in the PDMS production molds (40 μL each) to fill the remaining volume of the microneedles and to form the MNA backing layer. Hydrogel-loaded production molds were centrifuged (5h at 4500 rpm) to obtain the final dissolving undercut MNAs for cutaneous vaccination experiments. MNAs were then removed from production molds with tweezers, or forceps, by pulling two diagonal corners of the MNA base away from the mold. To demonstrate the broader material capabilities of our manufacturing strategy for dissolving MNAs with undercut microneedles, MNAs were also fabricated using a 40 % w/w hydrogel with a 40:60 mixture of polyvinylpyrrolidone (PVP) and polyvinyl alcohol (PVA) through the spin-casting method.

2.3. Quantification of antigen and adjuvant loading

Microneedles were dissolved in TE Buffer, and concentrations of OVA and Poly(I:C) measured using a Micro BCA protein assay and SYBR Green nucleic acid assay, respectively. Loading error, defined as the difference between measured and theoretical amounts of biocargo in microneedles as a percentage of the theoretical amount, was calculated. To determine loading efficiency, excess biocargo recovered from the MNA production mold reservoir after loading and prior to drying (Fig. S1C) was quantified. Loading efficiency = [biocargo in microneedles / (biocargo loaded to mold − excess biocargo recovered)] × 100 %. Results are reported as mean ± SD (N = 6).

2.4. Cutaneous vaccine delivery to human skin explants using MNAs

2.4.1. Preparation of ex vivo human skin explants

Human skin explants were prepared as described previously [53]. Briefly, normal human skin from deidentified healthy donors undergoing plastic surgery was acquired through the Pitt Biospecimen Core and used according to University of Pittsburgh Medical Center guidelines. Tissue was rinsed in 70% ethanol and then in phosphate-buffered saline (PBS). Human skin explants (approximately 1 mm thick) were harvested using a Silver’s miniature skin graft knife (Padgett, Integra Miltex, Plainsboro, NJ), and then cut into 20 mm x 20 mm square pieces. The resulting human skin samples comprised epidermis and a thin layer of underlying dermis.

2.4.2. Imaging analysis

To evaluate undercut MNA-directed intradermal biocargo (e.g., vaccine) delivery to living human skin explants, several imaging analyses were performed. Tip-loaded MNAs incorporating a red cargo (Allura Red R40 dye) were fabricated using the manufacturing strategy described above. Prior to application of MNAs to human skin explants, MNAs were imaged using an optical stereomicroscope. Subsequently, MNAs were applied to human skin explants and removed after 10 min. An optical stereomicroscope was then used to image the patterns of colored biocargo deposited from MNAs into the human skin. Remaining MNA materials after application were also imaged. For further qualitative assessment of MNA-directed intradermal vaccine delivery to human skin, MNAs containing both Alexa555-labeled OVA and Alexa488-labeled Poly(I:C) were fabricated, applied to human skin explants for 10 min, and removed. Targeted areas of the human skin explants were fixed in 2% paraformaldehyde, cryopreserved with 30% sucrose solution, flash frozen in optimum cutting temperature (OCT) compound, and cryo-sectioned into 10 μm thick sections. Human skin cross-sections were counter-stained with a fluorescent nuclear dye (DAPI) and imaged using a bright-field and epifluorescence microscope (Nikon Eclipse E800) to detect Alexa555-OVA and Alexa488-Poly(I:C), with bright-field images taken to better visualize the stratum corneum breaching.

2.5. MNA-directed skin immunization in vivo

2.5.1. Mice

Female C57BL/6J mice were purchased from The Jackson Laboratory (Bar Harbor, ME) and used at 8–10 weeks of age. Mice were maintained under specific pathogen-free conditions at the University of Pittsburgh, and all experiments were conducted in accordance with the institutional animal care and use committee (IACUC) guidelines.

2.5.2. Quantification of antigen and adjuvant delivery with MNAs

OVA + Poly(I:C) MNAs were applied to murine abdominal skin for 5, 10, or 20 min, and then remaining MNAs were removed and dissolved in TE Buffer. Concentrations of OVA and Poly(I:C) were measured using a Micro BCA protein assay and SYBR Green nucleic acid assay, respectively. Quantities of OVA and Poly(I:C) delivered to skin were calculated by subtracting the amount remaining from the mean amount loaded, and delivery is reported as percentage of the initial amount loaded (mean ± SD, N = 6).

2.5.3. In vivo IVIS imaging

In vivo intradermal vaccine delivery with dissolving undercut MNAs was demonstrated on a C57BL/6J mouse. Tip-loaded CMC/trehalose MNAs integrating both Alexa555-labeled OVA and Alexa488-labeled Poly(I:C) were created using the manufacturing strategy described above, applied to the abdomen of an anesthetized mouse for 10 min, and then removed. Fluorescent OVA + Poly(TC) MNAs were imaged before and after in vivo application by optical stereomicroscopy and epifluorescence microscopy. To show delivery of the fluorescent multicomponent vaccine, the mouse was imaged with an IVIS 200 in vivo imaging system (PerkinElmer, Waltham, MA), using the corresponding filters to detect Alexa488-Poly(I:C) and Alexa555-OVA at the MNA application site. Images were then post-processed using Living Image software (PerkinElmer).

2.5.4. Cell-mediated and humoral immune responses

Mice were immunized by application of 10 μg OVA ± 25 μg Poly(I:C) MNAs to the right and left sides of abdomen (two MNAs per mouse) or by two intramuscular injections of 10 μg OVA in PBS into the hindlimb gastrocnemius muscles. Control mice were left untreated (i.e., naïve), or treated with blank MNAs (without antigen or adjuvant). Cutaneous or intramuscular immunizations were repeated 7 days later. In vivo OVA-specific cytotoxic T-cell activity and OVA-specific antibody responses were evaluated 5 days after the second immunization (booster dose) using well-established techniques [53, 54].

For OVA-specific antibody responses, blood was collected from anesthetized mice at the time of sacrifice by cardiac puncture, and serum was isolated using BD Microtainer serum separator tubes (BD Biosciences, San Jose, CA). OVA-specific IgG1 and IgG2c antibodies in serum were measured by indirect ELISAs. Costar EIA/RIA plates (Corning Inc., Corning, NY) were coated with OVA (100 μg/mL in 0.5 M carbonate-bicarbonate buffer) by overnight incubation at 4 °C. Plates were washed (3x) with 0.05% Tween20 in PBS, and blocked with 1% goat serum in PBS for 1 hour at 37 °C. Serum samples and standards (anti-OVA IgG1 or anti-OVA IgG2c) were diluted with 1% goat serum, added to plates, and incubated 2 hours at 37 °C. After washing (3x), plates were incubated for 1 hour at 37 °C with biotinylated secondary antibodies (goat anti-mouse IgG1 or IgG2c, 1:20,000 in 1% goat serum). Plates were then washed (3x) and incubated for 30 min with streptavidin-HRP (1:1000 in 1% goat serum). Plates were washed (3x) again and incubated at room temperature with TMB peroxidase substrate for 2–3 minutes, and the reaction quenched with 1.0 M H₂SO₄. For all ELISAs, absorbance at 450 nm (OD450) was read with a SpectraMax 340PC plate reader (Molecular Devices, Sunnyvale, CA), and serum concentrations calculated from standard curves.

To assess OVA-specific cytotoxic T-cell (CTL) activity, splenocytes from naïve mice were pulsed with 2 μg/ml OVA_257-264 (SIINFEKL) peptide, or left unpulsed for 1 h. Antigen pulsed splenocytes were washed and stained with high concentration CFSE (10 μM), while unpulsed splenocytes were labeled with low concentration CFSE (1 μM) for 15 min at 37 °C. A 1:1 mixture of pulsed target cells and unpulsed control cells (10⁷ each) was intravenously (IV) injected into immunized and naïve mice. Twenty hours after adoptive transfer, spleens of mice were isolated, and killing of target cells was evaluated by comparison of the antigen pulsed and unpulsed populations by flow cytometry to quantify OVA-specific killing of the high CFSE labeled SIINFEKL-pulsed targets. Specific lysis was calculated and expressed as a percentage of maximum lysis as: % Lysis = {1 − [(mean CFSE^low/CFSE^high ratio from naïve mice) / (CFSE^low/CFSE^high ratio from vaccinated mouse)]} × 100 %.

2.5.5. Statistical Analyses

Statistical analyses were performed using GraphPad Prism v8 (San Diego, CA). Data from vaccination experiments were analyzed by one-way independent ANOVA, followed by Tukey’s or Dunnett’s post-hoc testing. Differences were considered significant if p < 0.05.

3. Results and Discussion

3.1. Fabrication of dissolving undercut microneedle arrays

Penetration, dissolution, and delivery efficiency are key parameters for dissolving MNA-mediated cutaneous immunization. These factors depend highly on MNA design, and microneedle geometry is a major contributor to the success of MNA-based cutaneous drug delivery [19, 31]. Over the past few decades, a number of different microneedle geometries such as circular, obelisk, and pyramid microneedles have been used for MNA-directed intradermal drug delivery [37]. We previously showed that MNAs with obelisk microneedle geometries result in better penetration and cutaneous delivery efficiency than those with prevailing pyramid microneedles [19]. Further, localizing biocargo to the skin-penetrating tip portion of the microneedles enhanced delivery efficiency [31]. Maximizing intradermal delivery efficiency is particularly important for effective cutaneous immunization to enable skin microenvironment conditioning while minimizing the necessary quantities of expensive vaccine components. These critical factors were considered when developing novel MNAs and the associated manufacturing strategy.

Here, we introduce a novel design of dissolving microneedles for cutaneous vaccination, which features undercut geometry and biomolecules localized in the needle apex. This undercut microneedle geometry may enable better retention and other practical advantages in skin and non-cutaneous tissues [40, 55–57]. Despite this potential advantage, undercut microneedles have not been widely adopted, or used for cutaneous vaccination, at least in part due to complex fabrication methods (summarized in Fig. S2). Micromolding is a critical method for high-throughput manufacturing of microstructures [51]; however, fabrication of microstructures with undercut features through micromolding has required complex processing steps and precision assembly of separately molded, or machined, microneedle tips and shafts. For the first time, we present a manufacturing strategy and materials to fabricate dissolving MNAs with undercut features (Fig. 1) through micromolding (Fig. 2), eliminating complicated engineering procedures. Importantly, and counterintuitively, we show that undercut microneedles can be directly removed from the flexible production molds which are reusable for several processing cycles, substantially improving cost and productivity. These results suggest that during the micromolding processes, MNA production molds undergo elastic deflection without permanent deformation, and the mechanical stress distribution (caused by removal forces) is smaller than the strength of the microneedle materials, resulting in failure-free removal of undercut microneedles.

The manufacturing strategy we utilize uniquely enables reproducible fabrication of high-quality, tip-loaded dissolving MNAs with undercut features from different and widely-used dissolving microneedle biomaterials, including CMC/trehalose and PVP/PVA compositions [19, 29, 31]. The manufacturing and processing steps schematically depicted in Fig. 2 result in the final products shown in Fig. 3A. Specifically, the master MNA was fabricated from IP-S photoresist by 3D direct laser writing. IP-S is a specific material designed for 3D laser lithography and provides high resolution and mechanical integrity for micro- and nano-structures. We find that 3D laser lithography based on two-photon polymerization provides an effective means for fabrication of undercut MNA designs with smooth edges and sharp tips (>2 μm tip radius), and without any unwanted residues (e.g., machining chips) (Figs. 3B and 3F). To enable more rapid, parallel fabrication of dissolving MNAs, the master MNA was replicated through a two-step micromolding process (Figs. 3C and 3G). The IP-S master MNA was used to fabricate a flexible PDMS mold through soft-lithography, and the resulting PDMS molds were used to manufacture several VeroWhite MNA replicas through UV-curable micromolding. PDMS is a commonly used elastomer with tunable flexibility and low cost for molding of micro- and nano-structures [29, 51]. VeroWhite resin is a wear-resistant, acrylic-based photo-polymer extensively used for 3D Polyjet printers [58], which renders it an ideal material for MNA master molds. Six MNA replicas were then assembled into one MNA master mold, and this master mold was used to produce several PDMS MNA production molds (Fig. 3D). Collectively, these processing steps, along with high geometric capability of 3D direct laser writing, resulted in an effective MNA manufacturing strategy. Furthermore, rapid replication of the 3D printed master MNA using a wear-resistant moldable material improved productivity. We believe other undercut features could be achieved with this versatile approach, potentially using materials with different Young’s modulus for the elastomer molds or/and polymers with different strength properties for microneedles.

Fig. 3.

Fabrication of novel dissolving MNAs with undercut microneedles. (A) Final products corresponding to each step of the presented manufacturing strategy. Scale bar is 10 mm. (B-I) Geometric quality control of the fabricated MNAs using optical stereomicroscopy. Scale bars are 250 μm. (B) Master MNA created using 3D direct laser writing. (C) Replica of the master MNA created through a two-stage micromolding strategy (elastomer molding combined with UV-curable micromolding). (D) Microneedle-shaped wells in an MNA production mold. (E) Final dissolving CMC trehalose MNA incorporating a multicomponent vaccine (OVA + Poly(I:C)). (F) Higher magnification of an individual undercut microneedle on the 3D printed master MNA (as in B). (G) Higher magnification of an individual undercut microneedle on master MNA replica (as in C). (H) Final dissolving PVP/PVA undercut microneedle tip-loaded with Alexa680-labeled OVA. (I) Final dissolving CMC trehalose undercut microneedle tip-loaded with doxorubicin, a red-colored chemotherapeutic small molecule drug.

Upon fabrication of the MNA master molds with six MNA replicas, dissolving MNAs that integrate vaccine components in the tip portion of the microneedles were fabricated using the conventional three-stage manufacturing strategy through master mold to production mold to final dissolving MNAs [19, 29]. Dissolving MNAs that incorporated the vaccine components (OVA ± Poly(I:C)) in the tip portion of the undercut microneedles were fabricated through the spin-casting process (Fig. 3E). Total OVA and Poly(I:C) content in microneedles was determined to be 10.15 ± 0.87 μg and 24.29 ± 1.60 μg, respectively. With nominal doses of 10 μg OVA and 25 μg Poly(I:C), average loading errors were 5.8 % and 6.1 %, respectively. During the microneedle tip-loading process, recovery of excess biocargo from MNA production mold reservoirs prior to drying (detailed in Fig. S1) reduces biocargo waste and enables a higher loading efficiency of 77.8 ± 5.8 %, which is especially important when working with more expensive vaccine components. For vaccine experiments, we used MNAs made of CMC and trehalose, two FDA-designated “Generally Recognized as Safe” (GRAS) biomaterials. The water-solubility and mechanical strength of CMC make it a good structural material for MNAs [59], while trehalose is a disaccharide known to enhance stability of proteins [60]. We are currently conducting a phase I clinical trial using CMC MNAs to deliver doxorubicin to skin lesions for the treatment of patients with cutaneous T-cell lymphoma (ClinicalTrials.gov # NCT02192021).

To demonstrate compatibility of our MNA fabrication process and the undercut microneedle geometry with another dissolvable biomaterial composition commonly used in the MNA field, we fabricated some MNAs using a PVP/PVA hydrogel (Fig. 3H). Additionally, MNAs with undercut microneedles tip-loaded with a red colored model drug (doxorubicin) were fabricated to facilitate imaging and demonstrate compatibility of the fabrication process with small molecule agents (Fig. 3I). Demonstrated compatibility of the MNA fabrication process and undercut microneedle geometry with different types of cargos and material compositions makes application-driven optimization possible, as materials can be selected based on compatibility with bioactive cargo, dissolution requirements, and/or necessary mechanical properties for insertion into different types of skin (e.g., normal skin vs. psoriatic plaques).

In addition to cutaneous vaccination, these dissolving undercut MNAs can be used for a broad range of intradermal and non-cutaneous (e.g., liver, ocular, and cardiac tissues) drug delivery applications [61–63]. Through the spin-casting process, biocargo(s) of interest can be located either at the tips of microneedles (Fig. 4A), or throughout the entire pyramid region (Fig. 4B), depending on dose requirements. Furthermore, a number of sequential spin-casting steps can be performed to fabricate high-quality MNAs with undercut microneedles that incorporate multiple cargos in their pyramid regions (Figs. 4C–D). As such, the presented approach and novel MNA designs are compatible with single and combination therapies for several cutaneous and non-cutaneous applications. Importantly, same production molds can be re-used to fabricate dissolving MNAs, suggesting that removal of undercut MNAs from flexible PDMS production molds results in elastic deformation for several process cycles without destroying the production molds. For example, the dissolving MNAs in Fig. 4A and Fig. 4C were obtained using the same PDMS production mold at the first and twelfth cycles, respectively. Furthermore, we are currently capable of fabricating 5000+ MNAs per day in our laboratories, and these fabrication processes can be scaled up using industrial grade manufacturing strategies.

Fig. 4.

Successful fabrication of dissolving MNAs with undercut microneedles integrating single or multiple cargos from different dissolvable biomaterial compositions. Scale bars are 250 μm. (A) PVP/PVA MNAs incorporating Texas Red-labeled dextran (40 kDa MW) at the tips of microneedles. (B) Tip-loaded CMC trehalose MNAs integrating Allura Red R40 dye (~500 Da MW) at the pyramid region of microneedles. (C) Tip-loaded PVP/PVA MNAs incorporating multiple cargos, such as Texas Red-labeled dextran and Allura Red R40 dye. (D) Tip-loaded PVP/PVA MNAs incorporating multiple cargos, such as Texas Red-labeled dextran and green fluorescent PLGA microspheres.

Additive manufacturing (AM), or 3D printing, has been a key enabling technology for the preparation of drug delivery systems [64, 65]. Indeed, there are currently FDA approved, 3D-printed drug delivery systems, such as Spritam® tablets [64, 65]. A unique advantage of AM over traditional subtractive fabrication techniques is the possibility for accurate and reproducible manufacturing of 3D complex geometries without design limitations [40]. As such, AM offers a high degree of design flexibility and control, and thus enables rapid design-to-fabrication turnaround for optimal application-driven drug delivery systems [66]. Indeed, micro-scale AM has been effectively used for accurate and reliable fabrication of intradermal drug delivery systems [40]. However, the unique advantages of AM have yet to be exploited for scalable fabrication of high-quality dissolving MNAs with novel designs for a broad range of drug delivery applications.

In this study, we utilized the micro-additive manufacturing process, 3D direct laser writing, to enable fabrication of complex, high accuracy, 3D microneedle geometries with smooth edges and sharp tips, as well as undercut features. This technology offers an unprecedented level of flexibility for MNA designs. To demonstrate the range of geometric capability of 3D direct laser writing, we fabricated microneedle designs with diverse geometries. This technology enabled fabrication of a wide range of microneedle geometries with high-fidelity, supporting application-driven optimization. Furthermore, as shown in Fig. S3, it allows a wide range of design changes, including height, width, apex angle, and geometry of the microneedles without requiring complex and custom processing steps. From a strength of materials standpoint, removal of MNAs with different undercut microneedle geometries is governed by needle geometry and material strength, as well as by elasticity and strength of production mold materials. Microneedles with larger undercut geometries may require more flexible production molds with lower Young’s moduli for failure-free removal. More flexible PDMS molds can be prepared by adjusting the crosslinker ratio and/or curing temperature [67, 68], and hyper-elastic materials, such as Ecoflex, could serve as more flexible alternatives to PDMS [69], allowing removal of even larger undercut features. Collectively, 3D direct laser writing and use of flexible production molds pave the way for fabrication of a broad range of application-driven MNA designs.

3.2. Dissolving undercut MNAs deliver multicomponent vaccines to human skin.

To evaluate cutaneous biocargo delivery characteristics of dissolving MNAs with undercut microneedles, MNAs tip-loaded with Allura Red R40 dye were manufactured using the presented fabrication strategy. Allura Red R40 dye-loaded MNAs were applied to living human skin explants and removed after 10 min. Images of these MNAs before (Fig. 5A) and after (Fig. 5B) application demonstrated high-quality MNAs and complete dissolution of the microneedles, respectively. The corresponding deposits of MNA-embedded Allura Red R40 dye in the targeted skin are shown in Fig. 5C.

Fig. 5.

Intradermal vaccine delivery to freshly-excised human skin explants using tip-loaded, dissolving MNAs with undercut microneedles. (A-B) Optical stereomicroscopy images of PVP/PVA MNAs incorporating Allura Red R40 dye before (A) and after (B) application to human skin explants. Scale bars are 250 μm. (C) Optical stereomicroscopy image of Allura Red R40 dye microneedle traces on living human skin samples. Scale bar is 500 μm. (D-I) Intradermal co-delivery of Alexa488-labeled Poly(I:C) and Alexa555-labeled OVA from tip-loaded CMC/trehalose MNAs. Scale bars are 100 μm. Fluorescence microscope composite images demonstrate delivery cavities penetrating the epidermis and upper dermis, and delivery of both antigen and adjuvant to targeted skin microenvironments. (D) DAPI nuclear stain. (E) Alexa488-labeled Poly(I:C). (F) Alexa555-labelled OVA. (G) Brighfield. (H) Merged fluorescent images from (D-F). (I) Merged fluorescent images from (D-F) overlaid on brightfield image from (G). (J) Low magnification merged fluorescent images showing OVA and Poly(I:C) delivered in two parallel microneedle tracks. Scale bar is 200 μm.

Successful vaccine delivery through the stratum corneum into the immune cell-rich cutaneous microenvironments is critical for effective intradermal immunization. To anatomically evaluate the delivery of antigen (OVA) and adjuvant (Poly(I:C)) into human skin, MNAs incorporating both Alexa555-labeled OVA and Alexa488-labeled Poly(I:C) were applied to human skin explants for 10 min and then removed. The targeted human skin was cryo-sectioned and imaged using epifluorescence microscopy. The resulting images demonstrated microneedle cavities penetrating through the epidermis into the dermis (Fig. 5G), and delivery of fluorescent labeled OVA and Poly(I:C) to targeted human skin microenvironments (Fig. 5D–I). Collectively, these results indicate that the MNAs fulfilled the geometric (sharp tips and smooth edges) and mechanical-strength requirements for failure-free human skin penetration (i.e., breaching through the stratum corneum and epidermis), and material requirements for efficient dissolution in the aqueous environment of the skin, thereby presenting an effective cutaneous drug and vaccine delivery platform.

3.3. Dissolving undercut MNAs deliver multicomponent vaccines to murine skin.

To evaluate cutaneous delivery efficiency and kinetics for undercut MNAs, OVA + Poly(I:C) MNAs were applied to murine abdominal skin for 5, 10, or 20 min, then removed and the remaining biocargo content measured. Within 10 min, MNAs delivered 80.2% ± 12.5 % OVA and 79.6 ± 5.0 % Poly(I:C), with nonsignificant additional delivery for either vaccine component by 20 min (Fig. 6A). To visually confirm in vivo intradermal multicomponent vaccine delivery in mice, dissolving MNAs with high-fidelity undercut microneedles incorporating both Alexa555-OVA and Alexa488-Poly(I:C) were fabricated as described above. Prior to application, Alexa555-OVA + Alexa488-Poly(I:C) MNAs were imaged using optical stereomicroscopy and epifluorescence microscopy (Fig. 6B). MNAs were then applied to mice and removed after 10 min. The remaining MNA material after applications was also imaged using optical stereomicroscopy (Fig. 6C). MNA-treated mice were imaged using the IVIS 200 live animal imaging system with filters for detection of both Alexa488-Poly(I:C) and Alexa555-OVA. MNA-directed co-delivery of OVA (antigen) and Poly(I:C) (adjuvant) are shown in Figs. 6D–E. Together, these images demonstrate successful in vivo application of dissolvable undercut MNAs to mice and efficient delivery of both components of a multicomponent vaccine.

Fig. 6.

MNA directed intradermal vaccine delivery in mice. (A) Delivery kinetics for OVA and Poly(I:C) in murine skin. OVA + Poly(I:C) MNAs were applied for 5, 10, or 20 min, and then delivery efficiency of OVA and Poly(I:C) with respect to time was quantified. Data represent percent of initial MNA content delivered (mean ±SD, N = 6). One-way ANOVA and Tukey’s post-hoc tests were used for each biocargo, and significant differences are indicated by *** p < 0.001. (B) Optical stereomicroscopy image of MNAs integrating both Alexa555-OVA (red) and Alexa488-Poly(I:C) (green). The fluorescence microscopy inset shows the distribution of both vaccine components in a pyramid microneedle tip. (C) Representative optical stereomicroscopy image of the Alexa555-OVA and Alexa488-Poly(I:C)-loaded MNAs after application to murine skin. (D-E) Effective co-delivery of (D) Alexa488-Poly(I:C) and (E) Alexa555-OVA to the mouse skin using MNAs with undercut features.

3.4. Multicomponent vaccine MNAs induce potent cellular and humoral immunity.

Upon demonstration of successful intradermal delivery of the vaccine components to mice, we specifically evaluated immunogenicity of MNA-embedded antigen ± adjuvant and compared MNA immunization to vaccination by the clinically common intramuscular (IM) injection route. To this end, MNAs were fabricated with 10 μg OVA ± 25 μg Poly(I:C) per MNA, as described above. We and others have previously shown that proteins integrated in dissolving MNAs maintain their integrity [19, 31]. Mice were immunized twice with MNAs or by IM injections, as detailed above, and OVA-specific cytotoxic T-cell (CTL) and antibody responses were quantified using standard in vivo lytic assay and ELISAs, respectively (Fig. 7). Notably, dissolving MNAs with different geometries have been used to deliver antigens and other adjuvants [70–72], while the Poly(I:C) adjuvant used in this study has only been incorporated in coated or sustained release MNAs [73–75].

Fig. 7.

Intradermal delivery of antigen (OVA) ± adjuvant (Poly(I:C)) with MNAs induces antigen-specific cellular and humoral immunity. Mice were immunized by intramuscular (IM) injection of OVA (2 x 10μg injections per mouse), or by application of OVA ± Poly(I:C) MNAs (10μg OVA ± 25μg Poly(I:C) per MNA, 2 MNAs per mouse) to abdominal skin, and boosted identically seven days later. To determine activity of OVA-specific cytotoxic T lymphocytes (CTLs), equal numbers of unpulsed splenocytes (CFSE^low “control” cells) and OVA_257-264 peptide-pulsed splenocytes (CFSE^high “target” cells) were transferred to naïve and immunized mice (2x10⁷ total cells per mouse) five days later. Spleens and serum were isolated the next day. (A) Representative flow cytometry histograms showing remaining CFSE-labeled cells in spleens of immunized and unimmunized mice. Specific lysis of peptide-pulsed target cells by OVA-specific CTLs is indicated by a reduction in CFSE^high target cells. (B) Quantification of specific cell lysis, with 100% lysis corresponding to complete elimination of target cells (mean ± SD), N = 3 mice per group). (C) Serum concentrations of OVA-specific IgG1 and IgG2c antibodies (bars represent mean values, 3 mice per group). Groups were compared by one-way ANOVA, followed by Tukey’s post-hoc tests (B), or Dunnett’s comparisons to OVA IM control group (C). Significant differences are indicated by * p < 0.05, ** p < 0.01, or *** p < 0.001.

Cutaneous vaccination with MNAs elicited robust antigen-specific cellular immune responses (Figs. 7A–B). As expected, equivalent numbers of antigen-pulsed (CFSE^high) target cells and unpulsed (CFSE^low) target cells were recovered from spleens of unimmunized mice (Fig. 7A), indicating the absence of antigen-specific cytolytic activity. Mice treated with Blank MNAs (without antigen or adjuvant) also did not exhibit OVA-specific CTL responses (comparable to naïve). In contrast, specific lysis of antigen-pulsed target cells was dramatically enhanced in immunized mice, as shown by reduced survival of OVA-pulsed targets compared to unpulsed targets (Fig. 7A). While immunization by IM injections of OVA elicited a relatively low antigen-specific cellular immune response (i.e., OVA-specific lysis), cutaneous vaccination with OVA MNAs led to significantly greater OVA-specific lysis (Figs. 7A–B). Importantly, addition of Poly(I:C), a Toll-like receptor 3 (TLR3) agonist adjuvant [76], to MNAs further improved vaccine immunogenicity, as indicated by a greater CTL response to multicomponent OVA + Poly(I:C) MNAs (Figs. 7A–B). The enhanced CTL response we observed with Poly(I:C) adjuvant is consistent with previous reports that TLR3 ligands activate keratinocytes, innate immune cells, and professional APCs and induce cross-presentation of antigen to prime CD8⁺ T cells [77–79]. Given these results with a model antigen, multicomponent cutaneous vaccination using dissolving undercut MNAs that incorporate pathogen- or tumor-specific antigens and adjuvants could induce robust cellular immunity essential for prevention and/or treatment of many infectious diseases and cancer [3, 53, 54].

In addition to cellular immunity, cutaneous vaccination with OVA ± Poly(I:C) MNAs elicited robust antigen-specific humoral immune responses (Fig. 7C). While IM immunization resulted in modest OVA-specific serum IgG antibody responses, mice immunized with equivalent doses of OVA antigen in dissolving MNAs had significantly higher IgG levels. Mice not exposed to OVA antigen (i.e., naïve and Blank MNA treated mice) had undetectable levels of OVA-specific antibodies. In particular, we measured serum levels of two subclasses of OVA-specific IgG: IgG1 and IgG2c. Typically, IgG1 antibodies are associated with Th2 type immune responses to extracellular pathogens, while IgG2c antibodies are associated with Th1 type immune responses to viruses and other intracellular pathogens [80]. Although both IgG subclasses can potentially neutralize and/or opsonize pathogens, IgG2c antibodies can also activate the complement pathway and typically evoke more potent cellular responses because of a greater affinity for activating Fc receptors (FcγRI, FcγRIII, and FcγRIV) and lower affinity for the inhibitory Fc receptor (FcγRIIB) [80]. In our immunization experiments, the addition of Poly(I:C) adjuvant had minimal effect on OVA MNA induced IgG1 responses, but promoted a modest increase in IgG2c responses, consistent with the enhanced CTL immunity. Compared to IM immunization, the stronger and more balanced IgG1 / IgG2c responses to cutaneous vaccination with MNAs could translate into enhanced protection against different types of pathogens. Taken together, these results demonstrate that dissolving MNAs with undercut microneedles can efficiently deliver antigens ± adjuvants to APC rich microenvironments within the skin to induce potent cellular and humoral immunity. Ultimately, these undercut MNAs represent a novel modular platform technology for the specific and precise delivery of embedded multicomponent vaccines (antigen ± adjuvant) to defined microenvironments within the skin.

4. Conclusions

We have described a comprehensive approach to fabricate novel dissolving MNAs with undercut microneedles for effective multicomponent cutaneous vaccination. Our manufacturing approach strategically combined 3D laser lithography with nanoscale resolution and micromolding with mechanically flexible molds that allow direct removal of undercut MNAs. Reproducible fabrication of dissolvable MNAs with undercut microneedles incorporating multiple cargos was achieved using different biocompatible and water-soluble polymers, and these MNAs successfully delivered biocargos to murine and human skin microenvironments. Importantly, cutaneous vaccination with antigen-loaded MNAs elicited more potent antigen-specific cellular and humoral immune responses than traditional immunization by intramuscular injection. Simultaneous delivery of adjuvant (Poly(I:C)) to the same skin microenvironment as antigen (OVA) enhanced immune responses and may reduce the amount of antigen and/or adjuvant needed, reducing both the risk of systemic toxicity and cost. Ultimately, our approach to fabrication of dissolving MNAs with diverse geometries, including undercut microneedles, may have a broad range of cutaneous and non-cutaneous vaccination and drug delivery applications.

Highlights.

Combines emerging 3D laser lithography with nanoscale resolution and micromolding to manufacture dissolving microneedle arrays (MNAs).

Microneedles with undercut geometries can be directly removed from flexible molds, simplifying fabrication of dissolving undercut MNAs.

First demonstration of dissolving undercut MNAs incorporating antigen plus adjuvant vaccine components.

MNAs simultaneously deliver antigen and adjuvant to the same cutaneous microenvironment in both human and murine skin.

Vaccination with OVA ± Poly(I:C) MNAs generates more potent antigen-specific cellular and humoral immune responses than traditional immunization by intramuscular injection.