Hyaluronan molecular weight: effects on dissolution time of dissolving microneedles in the skin and on immunogenicity of antigen

Biomaterials used as matrix for dissolving microneedles (dMNs) may affect the manufacturing process as well as the potency of the active pharmaceutical ingredient, e.g. the immunogenicity of incorporated vaccine antigens. The aim of this study was to investigate the effect of the molecular weight of hyaluronan, a polymer widely used in the fabrication of dMNs, ranging in molecular weight from 4.8 kDa to 1.8 MDa, on the dissolution of microneedles in the skin in time as well as the antibody response in mice and T-cell activation in vitro. Hyaluronan molecular weight (HA-MWs) did not affect antibody responses (when lower than 150 kDa) nor CD4+ T-cell responses against model antigen ovalbumin. However, the HA-MWs had an effect on the fabrication of dMNs. The 1.8 MDa HA was not suitable for the fabrication of dMNs. Similarly, the 4.8 kDa HA generated dMN arrays less robust compared to the other HA-MWs requiring optimization of the drying conditions. Finally, higher HA-MWs led to longer application time of dMN arrays for a complete dissolution of microneedles into the skin. Specifically, we identified 20 kDa HA as the optimal HA-MW for the fabrication of dMNs as with this MW the dMNs are robust and dissolve fast in the skin without affecting immunogenicity.


Introduction
Microneedles are intensively studied for dermal and transdermal drug and vaccine delivery (1,2). Microneedles are needle-like structures up to 1 mm in length, capable of piercing the upper layer of the skin, the stratum corneum, enabling drug and vaccine delivery in a minimally invasive and painless way (3,4). Several microneedle concept are in development: i) hollow microneedles to inject liquid formulations through the bore of the microneedle, ii) microneedles for skin pretreatment to create microchannels prior application of a patch containing the drug or vaccine formulation, iii) porous, coated, hydrogel-forming microneedles and dissolving microneedles that release the drug or vaccine upon insertion into the skin (5)(6)(7)(8). Dissolving microneedles (dMNs), completely dissolve after insertion into the skin, thereby releasing the encapsulated drug or vaccine (3). As an additional advantage, the dry state of dMNs combined with the presence of excipients can increase protein thermostability (3). dMNs are produced from different materials, ranging from low molecular weight carbohydrates to biodegradable polymers (3). The selection of the matrix material is based on different factors such as i) safety, ii) compatibility with the active compound and the manufacturing procedure, iii) capability of the manufactured dMNs to efficiently pierce the skin and subsequently dissolve in a short time period and iv) potential adjuvant properties.
One of the most frequently used matrix material is the hyaluronan (HA) (3,(9)(10)(11)(12)(13). High molecular weight HA is a non-toxic, biodegradable, biocompatible and non-inflammatory linear polysaccharide (14) naturally present in the skin and approved by the FDA as inactive material.
In a previous study from our group, the dissolving microneedle properties of the only 150 kDa HA, such as penetration and dissolution in the skin, have been reported (30).
The aim of this study was to determine whether i) HA with different MW had an effect on the immune response for both antibody (in vivo injections by hMN) and cellular (in vitro T-cell exposure) responses and ii) it was possible to fabricate dMNs by micromolding from HA with different molecular weight. To this end, first the potential adjuvant effect of HA was assessed by i) performing immunization studies investigating the antibody response in mice injected with the model antigen ovalbumin (OVA) in presence of a range of different HA-MWs and ii) investigating T-cell activation in vitro upon exposure to several HA-MWs with OVA.
Subsequently, different HA-MWs were used to fabricate dMNs assessing the compatibility of the matrix material of each MW with the manufacturing procedure. These dMNs were used to investigate their capability to pierce the skin and to dissolve in reasonable short time period.
Subsequently, the capillaries were etched during 4 h in ≥ 48% hydrofluoric acid and the polyimide coating was removed from the ends of the capillaries by diving them into heated (250 °C) sulfuric acid for 5 min.

Immunization studies
BALB/c mice were anesthetized by intraperitoneal injection of 150 mg/kg ketamine and 10 mg/kg xylazine and the injection site was shaved (approximately 4 cm 2 ). The same day, mice were immunized by intradermal hMN injection (120 µm injection depth) of 10 µl with 0.31 µg OVA with 3.1 µg HA for each HA-MW dissolved in PBS pH 7.4. Intradermal hMN injection of each HA-MWs without OVA and PBS were included as control. For controlled depth intradermal microinjections, a hollow-microneedle applicator was used as reported previously (32).
Immunizations were performed at day 1 (prime), day 22 (boost) and day 43 (2nd boost). Prior to each immunization, a blood sample was collected from the tail vein. At day 63, the blood sample was collected from the femoral artery and all mice were sacrificed. Serum was isolated from the samples and stored at -80°C.

Determination of OVA-specific IgG antibodies
OVA-specific antibodies were analysed by a sandwich enzyme-linked immunosorbent assay (ELISA) as described earlier (10). Briefly, well-plates were coated with OVA for 1.5 h at 37 °C and then blocked with bovine serum albumin (BSA) (Sigma-Aldrich, Zwijndrecht, the Netherlands). After the blocking, three-fold serial dilutions of serum were applied to the plates and incubated for 1.5 h at 37 °C. Then, the plates were incubated with horseradish peroxidase-conjugated goat antibodies against IgG total, IgG1 and IgG2a (Southern Biotech, Birmingham, AL, USA) for 1 h at 37 °C. Finally, 1-step TM ultra 3,3′,5,5′tetramethylbenzidine (TMB) (Thermo-Fischer Scientific, Waltham, USA) was used as substrate and sulfuric acid (H2SO4) (95-98%) (JT Baker, Deventer, The Netherlands) was added to stop the reaction. The absorbance was measured at 450 nm on a Tecan Infinite M1000 plate reader (Männedorf, Switzerland) and the antibody titers were determined as the log10 value of the mid-point dilution of a complete s-shaped absorbance-log dilution curve of the diluted serum level.

Bone marrow-derived dendritic cells (BMDCs)
To examine T-cell activation in vitro, first bone marrow was isolated from the tibias and femurs of C57BL/6 female mouse. A single-cell suspension of bone marrow cells was obtained by using a 70 μm cell strainer (Greiner Bio-One B.V., Alphen aan den Rijn, NL).
On day 11, the BMDCs were harvested from the petri dish and distributed into 96-well plates (100 µl/well, 10 000 cells/well). Then, 100 µl/well of formulations consisting of: i) 5 µg OVA, ii) 5 µg OVA mixed with 50 µg HA (per each molecular weight) or iii) 50 µg HA (per each molecular weight) were added to the wells. OVA (5 µg/well) + LPS (100 ng/well) or LPS (100 ng-well) were added as positive control; medium was included as negative control.
The BMDCs were exposed to the formulations overnight at 37 °C and 5% CO2 and subsequently OVA-specific CD4 + T-cells were transferred on BMDCs in co-culture experiments (see 2.6).

CD4 + T-cell activation by antigen loaded BMDC
OT-II (OVA-specific CD4 + ) T-cells were obtained from the spleen of OT-II transgenic Finally, the arrays were removed from the PDMS mold and inspected for shape and sharpness by light microscopy (Stemi 2000-C, Carl Zeiss Microscopy GmbH, Göttingen, Germany).

Human skin
Human abdomen skin was obtained from a local hospital within 24 hours after cosmetic surgery according to the declaration of Helsinki. The fat excess was removed with a scalpel and the skin was stored at -80°C. Before use, the skin was thawed at 37°C for 1 h in a humid petri dish and stretched with pins on parafilm-covered styrofoam. Before starting the experiment, the skin was cleaned with Milli-Q and 70% ethanol.

Dissolution of microneedles in ex vivo human skin
A dMN array (n=7) was applied on the skin as described in section 2.

Statistics
IgG titers and T-cell response were analysed using one-way ANOVA with respectively Bonferroni and Tukey's post-test suitable in the software Prism (Graphpad, San Diego, USA).
A p-value less than 0.05 was considered to be significant.

Microneedle penetration efficiency was analysed by Kruskal-Wallis test with Dunn's multiple
comparison test (p < 0.05).
The remaining dMN length after dissolution at different time points was analysed by two-way ANOVA with a Tukey's post-test (p < 0.05).

Immunization studies
The liquid formulations of OVA alone or mixed with different HA-MWs was intradermally injected by using hollow microneedles. During injection no problems with clogging or leakage were observed. The formation of a blister on the site of the injection indicated a successful intradermal delivery of the formulation. No adverse effects were observed.
The OVA-specific total IgG titers increased after each immunization ( Figure 1A

Analysis of CD4 + T-cell activation in vitro
To determine whether the presence of different HA-MW, with or without OVA, affect celllar responses, OT-II T-cell (OVA-specific CD4 + cells) activation studies were performed in vitro.
As expected, addition of OVA induces OT-II proliferation, which is enhanced in the presence of LPS (Figure 2). The co-exposure of cells to OVA and HA however, regardless of the molecular weight, did not increase OT-II proliferation compared to OVA alone. In line with the antibody responses, this suggests that none of the tested HA polymer provided a measurable adjuvant effect.

Dissolving microneedle fabrication
The

Penetration and dissolution of microneedles in ex vivo human skin
In order to assess whether the molecular weight of HA had an effect on the capability of microneedles to pierce and dissolve into the skin, first their ability to pierce ex vivo human skin was tested. Bright field analysis showed sharp microneedles regardless of the HA-MW ( Figure 3A).The penetration efficiency of microneedles into the skin was not affected by the molecular weight of the HA. The 150 kDa microneedles showed a penetration efficiency of 96 ± 7%; 20 kDa and 4.8 kDa HA microneedles showed a penetration efficiency of 98 ± 4%, (mean ± SD, n=3).

Discussion
When choosing the HA polymer for the fabrication of dMNs, it is relevant to investigate potential immune modulating effects besides effects related to the dMN manufacturing and physicochemical characteristics of dMNs such as capability to pierce the skin and dissolve quickly.
Although LMW-HA can have inflammatory properties in contrast to HMW-HA (19,20,22,36,37), to our knowledge, the role of HA-MW on the antibody response has not been reported yet. To this end, in the present study HA ranging from 4.8 kDa to 1.8 MDa mixed with OVA was injected intradermally in mice to assess the antibody response evoked. We did not observe an increase in antibody titers by adjuvanting with LMW-HA; however we did observe that HMW-HA (1.8 MDa) reduced the antibody level after the second boost compared to OVA only and LMW-HA 4.8 kDa HA mixed with OVA. This may be explained by reports that HMW-HA exerts an anti-inflammatory role reducing the side effects of vaccines (38,39) and displays immunosuppressive properties (40)(41)(42). To this end, it has been reported that HMW-HA up-regulates the transcription factor FOXP3 on regulatory T-cells (Treg) (e.g. CD4 + CD25 + ) (42) involved in the regulatory mechanism of autoantibody production (43) and likely antibody production. Although we did not investigate Treg activation in this study, they provide a potential mechanism regarding the lack of immunogenicity of HMW-HA.
The role of what in literature is defined as LMW-HA has been extensively investigated suggesting a pro-inflammatory effect in vitro: 4-6 oligosaccharides HA induced cytokine synthesis in dendritic cells (19,20); HA < 250 kDa induced inflammatory cytokines levels (36,(44)(45)(46)(47)(48); HA ≤ 800 kDa led to activation of macrophages (37). In this study, the effect of HA-MW was examined on the activation of CD4 + T-cells. Although this is an in vitro system, it has shown to be predictive in the sense that formulations that induce strong in vitro response of OT-II cells also outperformed other formulations in vivo (49,50). Conversely to what has been reported in literature, changes in HA-MW did not influence the proliferation and activation of CD4 + T-cells in vitro. Considerations justifying the referred controversial effects of HA preparations may be related to i) MW of HA that is not always measured accurately or is not homogeneous in the same HA population, for this reason an effect may be attributed to a minor population, ii) the presence of minor contaminants even in highly purified HA and iii) the conformational diversity of HA highly dependent on pH, temperature, salt concentration and specific cations (48). Furthermore, the effects of HA-MW seems to be cell-specific and depending on the HA medium concentration (26,48). Based on the result described above, the 20 kDa HA should be preferred to the 4.8 kDa and 150 kDa HA for the fabrication of dMNs.

Conclusions
The present study underlines the importance of defining a specific HA-MW for the fabrication of dMNs based on the investigation and assessment of the effects of the HA-MW on the immune response and on the physicochemical characteristics of dMNs such as capability to pierce the skin and dissolve quickly.
We consider, among the HA-MWs investigated, the medium-low HA-MW of 20 kDa the optimal MW as it results in robust dMNs characterized by a fast dissolving process in the skin and has no effect on the immunogenicity of the antigen.

Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declaration of competing interest
The authors declare that they have no conflicts of interest.