Determination of wetting angle of quercetin

To assess the hydrophilicity of quercetin concerning Soluplus® and deionised water, the wetting angle was measured, as depicted in Fig. 1. Figure 1C illustrates that the wetting angle of Soluplus® 0.5% (w/v) was significantly lower than that of deionised water (p < 0.05). This suggests that Soluplus® effectively reduces surface tension, resulting in a significant decreased wetting or contact angle (p < 0.05). This finding aligns with our prior research, demonstrating that the surfactant solution lowered contact angles between both hydrophilic and hydrophobic drugs in comparison to the control group (deionised water) [16]. Quercetin, classified as a BCS class II drug [3], exhibits high permeability but low solubility. Improved solubility could enhance the drug’s efficacy and delivery potential. Given that the contact angle of quercetin with Soluplus® is 2.5-fold lower than that in deionised water, it suggests that the Soluplus® solution tends to spread across the surface of quercetin, forming a flatter droplet and thereby reducing the contact angle [28, 29]. This result indicates that surfactant could help quercetin to disperse better in the formulation, which could lead to higher drug content and faster dissolution of tips of MAPs [16]. Therefore, this strategy, by using solubility enhancement, could contribute to better drug delivery outcomes in the skin.

Fig. 1

Representative images for wetting angle measurement for (A) Soluplus® 0.5% (w/v) and (B) deionised water. (C) Wetting angle measurement for quercetin with Soluplus® 0.5% (w/v) and deionised water (means + SD, n = 3)

Fabrication and characterisation of quercetin loaded dissolving MAPs

After measuring the wetting angle of quercetin to Soluplus® and deionised water, dissolving MAP formulations containing surfactant was fabricated using the double casting technique, as previously detailed [17, 25]. Figure 2 displays the resulting MAPs, exhibiting uniform arrays of microneedles on a smooth, flat baseplate. Notably, there were no observed air bubbles or unevenly formed needles. The physical appearance of the quercetin-loaded MAPs showed yellowish tips, indicating successful concentration and localization of quercetin at the tips rather than dispersion in the baseplate. This concentration at the tips holds potential to enhance delivery efficiency and reduce drug wastage.

Fig. 2

Digital and SEM images of dissolving MAP loaded with quercetin. F1: MAP formulation prepared from aqueous solution containing quercetin 15% w/w, PP2 20% w/w, and deionised water 65% w/w. F2: MAP formulation prepared from aqueous solution containing quercetin 15% w/w, PP2 20% w/w, and Soluplus® solution (0.5% w/v) 65% w/w

The properties of quercetin in the solid state, post its loading into MAPs, were assessed using DSC and FTIR analyses. This physiochemical characterisation is crucial to monitor during MAP preparation, ensuring the uniformity of the produced MAPs and complete encapsulation of quercetin into the polymeric matrix in the first layer. Figure 3A and B illustrate the DSC thermograms and FTIR spectra, respectively. DSC analysis (Fig. 3A) aimed to examine the physicochemical interactions between quercetin, surfactant, and polymers. It also sought to discern any changes in crystallinity following the MAP fabrication step. The thermogram of pure quercetin displayed a sharp endothermic peak at 324 °C, indicating its melting point. Additionally, pure PVP, PVA, and Soluplus® exhibited several broad endothermic peaks at 89 °C, 314 °C, and 75 °C, respectively, representing their respective excipient melting points in the MAP formulations. However, the DSC thermograms of F1 and F2 did not exhibit any endothermic peaks post-MAP fabrication. This suggests the successful incorporation of quercetin within the polymers, with the drug being present in a state of low crystallinity in the MAP formulations.

The FTIR analysis (Fig. 3B) depicted distinctive spectra for pure quercetin, PVP, PVA, Soluplus®, physical mixture, and MAP formulations. The spectrum of pure quercetin displayed O-H stretching at 3310 cm− 1 and a C = O group at 1610 cm− 1. Both pure PVP and PVA showed C-H stretching at 2980 cm− 1, and pure PVA exhibited an additional O-H stretching band at 3350 cm− 1. Soluplus® displayed peaks at 1477 cm− 1, 1635 cm− 1, and 1734 cm− 1, indicating C-O-C stretching, C = O groups, and C = O ester groups, respectively. Comparing the spectra of pure quercetin and MAP formulations (F1 and F2) revealed similar patterns, with a minor peak at 2980 cm− 1 in the MAPs. This suggests that MAP fabrication did not induce significant structural changes in the drug. These findings align with the DSC analysis, confirming successful encapsulation of the drug within the PVA and PVP polymer structures.

Fig. 3

(A) DSC and (B) FTIR profile of pure quercetin, pure PVP, pure PVA, pure Soluplus® physical mixture F1, Physical mixture F2, MAP F1 and MAP F2

Evaluation of mechanical properties

The mechanical properties of the MAPs were assessed post-fabrication. Figure 4A illustrates that both formulations, F1 and F2, experienced a reduction in needle height of less than 10% during a compression test employing a force of 32 N, equivalent to the thumb pressure typically applied during MAP insertion [30]. Notably, the results displayed no significant differences in needle height before and after compression for either formulation (p > 0.05). However, these findings contradict our earlier research involving surfactant incorporation into MAP formulations [16]. Previous studies indicated that surfactants could enhance needle mechanical resistance due to molecular interactions with the polymer matrix within the MAPs. Yet, it is essential to recognise that Soluplus®, employed in the current formulation, differs in properties from other surfactants like Pluronic® F88, Lutrol® F108, and Tween® 80 utilised previously. Soluplus® initially exists in a glassy and brittle state, potentially causing fragility and challenging removal from the mould after the casting process [31, 32]. However, in the current setup, the initial layer of MAPs comprises Soluplus®, PVA, PVP, and quercetin, possibly counteracting the Soluplus® brittleness, resulting in comparable properties with the other formulation lacking Soluplus®. Moreover, there was no significant difference in the percentage reduction of needle height between F1 and F2 (p > 0.05). This outcome was anticipated since the MAPs prepared had similar needle dimensions, with no significant variation in height and diameter, as the moulds used were identical. In this study, a compression force of 32 N was applied, equivalent to the pressure exerted by a human thumb during MAP application [30]. However, it is important to consider the potential consequences of excessive force during insertion, as reported by Ando et al., where needle deformation leading to buckling and unbuckling can be predicted based on the needle aspect ratio (needle height/base diameter) [33]. With a needle aspect ratio of 2.8, buckling may occur with excessive force application, whereas with a ratio of 1.8, unbuckling may occur. Given that the dimensions of the MAPs in this study are 850 μm and 300 μm for the needle height and base diameter, respectively, buckling may occur as the mechanical failure mechanism if excessive force is applied during application. Since this ratio is crucial for predicting the effect of needle aspect ratio on needle fracture force, it is imperative to consider these parameters seriously during formulation development. Furthermore, the observed reduction in height for both fabricated MAP formulations remained below 15%, closely aligning with values reported for dissolving MAPs in prior studies [34,35,36,37]. Thus, these results indicate that the MAPs in this study exhibit the mechanical robustness necessary to withstand the compression forces likely exerted during MAP application.

Figure 4B demonstrates the penetration profile of MAPs into eight layers of Parafilm®, a validated skin simulant utilized in insertion studies to assess needle penetration capabilities [30, 38]. Both formulations (F1 and F2) successfully breached the initial layer of Parafilm® (~ 126 μm thickness) at a force of 32 N. Subsequently, approximately 60% of F2 needles penetrated the second layer, whereas F1 achieved 100% penetration in the same layer. This difference might relate to the mechanical properties of Soluplus® in F2, which exhibits some brittleness compared to the control formulation (F1). However, there were no significant differences between F1 and F2 in the penetration profiles upon reaching the third and fourth layers (p > 0.05). Notably, no observable penetration was achieved beyond the fifth layer.

Fig. 4

(A) MAPs height reduction for needles loaded with quercetin following the application of a 32 N compressive force (means + SD, n = 20). (B) Percentage of channels formed per Parafilm® M layer upon the application of quercetin loaded dissolving MAPs (means ± SD, n = 3). Insertion depth of quercetin loaded dissolving MAPs into (C) Parafilm® M and (D) full thickness ex vivo neonatal porcine skin monitored in situ using optical coherence tomography (OCT) (means + SD, n = 20). Scale bar = 1 mm

When visualising the insertion of MAPs into Parafilm® using OCT, both MAP formulations (F1 and F2) were observed to reach the fourth layer, as depicted in Fig. 4C. This resulted in insertion depths of approximately 326 μm for F1 and 284 μm for F2. The significant difference in Parafilm® insertion depth between F1 and F2 (p < 0.05) aligns with earlier results from insertion profile studies, indicating that F2, due to its brittle nature, generated fewer holes, particularly in the second Parafilm® layer. However, upon evaluating skin insertion in excised full-thickness neonatal porcine skin, both MAP formulations (F1 and F2) achieved significantly greater insertion depths (p < 0.05) compared to those in Parafilm®, as shown in Fig. 4D. This discrepancy may be attributed to Parafilm®’s nature derived from paraffin wax, which inadequately represents the properties of actual skin [30]. Additionally, as the MAPs consist of hydrophilic polymers and even though the drug is classified as BCS class II, it is completely encapsulated within the polymer matrix. The excised full-thickness neonatal porcine skin containing interstitial fluid, rich in moisture, might facilitate lubrication and promote needle insertion into deeper layers compared to the dry, moisture-lacking Parafilm® [16]. This observation is further supported by the results depicted in Fig. 4D, demonstrating that F2 exhibited significantly deeper insertion into the excised full-thickness porcine skin (p < 0.05). This could be attributed to the incorporation of Soluplus® in the MAP formulation, enhancing hydrophilicity and facilitating insertion into ex vivo skin. In summary, the disparity in insertion depth between Parafilm® and excised full-thickness porcine skin aligns with prior evaluations of MAP insertion profiles [18, 34, 37,38,39,40,41]. However, despite all the differences between these two models, Parafilm® has been validated as an alternative for insertion study of MAP platform, which can be used for comparative formulation studies [42]. Parafilm® has been demonstrated and validated that it can be used as an alternative of excised porcine skin for insertion studies. In the previous study, Larraneta et al. has shown that the force that generated to apply MAPs by human volunteer was in the average of 20–40 N for 30 s [30]. Gantrez® S-97 and PEG based MAPs were used in the study which contain no drugs, therefore there is no discrepancy was observed between the Parafilm and excised skin during the insertion study. However, the result from insertion studies of the current work is aligned with our previous studies [18, 33, 36,37,38,39] where the dissolving MAPs containing the active compounds and the insertion in Parafilm® is significantly lower compared to excised porcine skin. Overall, the Parafilm® can be used as a skin simulant for MAP insertion evaluation as long as the force applied during insertion is in the range of 20–40 N for 30 s. This condition is equal to the force that generated by human thumb and still in the acceptable range of force to avoid buckling or any deformation due to excessive force application.

Evaluation of drug content of dissolving MAPs

In relation to drug content, Fig. 5 illustrates a significant impact (p < 0.05) on the quantity of drugs that can be loaded into the MAPs due to the incorporation of Soluplus® as a surfactant within the formulation. Previously mentioned, quercetin falls under BCS Class II, characterised by low permeability and solubility [43]. Incorporating Soluplus® is known to enhance the solubility of quercetin [32], thereby enabling a greater quantity of the drug to be loaded into the needle tips of the MAPs. These values were equal to 55% and 69% of drug content observed in F1 and F2, respectively, compared to the theoretical drug content in the needle tips (approximately 3.51 mg for both F1 and F2). These discrepancies might be affected by the manual removal of excess formulation during the preparation of dissolving MAPs.

Fig. 5

Drug content of quercetin loaded dissolving MAPs (means + SD., n = 3)

In situ skin dissolution studies

The in situ dissolution studies aimed to determine the timeframe and process involved in the complete dissolution of MAPs upon skin insertion. As depicted in Fig. 6A, after a 1-hour application, the needle layer of F2 MAP underwent full dissolution, transferring all drug content into the skin. However, only 57% of needle tips were dissolved after 1-hour for F1 (without any surfactants) (Fig. 6B). Furthermore, upon leaving F2 MAP on the skin for an hour, complete dissolution of the entire patch, including the polymeric baseplate, was observed, resulting in an occlusive and adhesive polymer gel forming on the skin surface. In contrast, F1 MAP required a longer duration to achieve complete dissolution and deposit its content into the skin. The addition of Soluplus® in the formulation of F2 demonstrated enhanced solubility of quercetin loaded in the MAP, contributing to improved dissolution of the entire patch, as previously reported [16, 32].

Fig. 6

(A) Digital images of needle dissolution at 0, 30 and 60 min, following insertion into and removal from excised neonatal porcine skin ex vivo. (B) Profile of needle height reduction (%) during the in situ skin dissolution studies (means ± SD, n = 10). F1: MAP formulation prepared from aqueous solution containing quercetin 15% w/w, PP2 20% w/w, and deionised water 65% w/w. F2: MAP formulation prepared from aqueous solution containing quercetin 15% w/w, PP2 20% w/w, and Soluplus® solution (0.5% w/v) 65% w/w

Ex vivo deposition studies

The ex vivo skin deposition studies using full-thickness neonatal porcine skin were conducted, and the findings are presented in Fig. 7. The assessment of quercetin delivery across different skin layers revealed a significantly higher amount of drug delivered from formulation F2 compared to F1 within the epidermis layer (p < 0.05). Specifically, approximately 1850 µg (~ 76%) from F2 and approximately 1500 µg (~ 77%) from F1 were observed, suggesting an accumulation of quercetin within this layer, forming a secondary drug reservoir. This reservoir potentially facilitates a slow release of the drug into the deeper dermis layers over time, beneficial for its anti-inflammatory effects, as quercetin’s primary site of action is predominantly localised in the epidermis layer [44,45,46,47]. Moreover, in the dermis layer, F2 exhibited significantly higher drug delivery (p < 0.05) with approximately 106 µg (~ 4.3%) compared to 94 µg (~ 4.9%) for F1. However, no significant difference was found between F1 and F2 in transdermal delivery over 24 h (p > 0.05). Consequently, both MAP formulations showed an overall delivery efficiency of more than 80% within 24 h.

Regarding quercetin delivery via MAP, Paleco et al. developed a silicon-based MAP combined with lipid microparticles [44]. Using a ‘poke and patch’ approach with cream containing quercetin-loaded lipid microparticles followed by silicone-based MAP application, they achieved a 5.5-fold increase in intra-epidermal quercetin delivery compared to the control. Their study revealed quercetin mainly localised in the epidermis layer (approximately 2.23 µg/cm2) [44]. However, this amount was significantly lower than that achieved in our current work. Our formulation, integrating surfactants in the dissolving MAPs, successfully improved quercetin delivery to the epidermis layer, surpassing 1.5 mg of quercetin.

Quercetin is known for its diverse therapeutic activities, making it valuable for managing various diseases such as inflammation, cardiovascular disease, neurodegenerative disorders, cancer, ulcers, bacterial and viral infections, allergies, and respiratory conditions like asthma and hay fever [48]. When applied topically, quercetin exhibits potent antioxidant properties, shielding keratinocytes from external oxidative stressors and neutralizing free radicals [49]. This protective effect helps maintain endogenous antioxidant levels and prevents lipid peroxidation induced by UV exposure. Additionally, quercetin has demonstrated strong anti-inflammatory effects, surpassing other flavonoids in reducing inflammation induced by irritants [49]. The combination of its anti-inflammatory and antioxidant properties positions quercetin as a promising candidate for managing wound healing in particularly caused by diabetes mellitus [50]. To maximise this potential, an appropriate delivery vehicle is essential for successful skin delivery. Given that the oral bioavailability of quercetin is typically less than 10% [38], and there is limited research on its bioavailability through topical or transdermal delivery routes, utilising a MAP platform for quercetin delivery holds promise for enhancing local efficacy.

Fig. 7

Amounts of quercetin extracted from different skin layers: epidermis and dermis from excised neonatal porcine skin as well as quercetin delivered transdermally at 24 h in an in vitro Franz cell diffusion study (means + SD, n = 5)

Regarding the proposed patch size, this can be extrapolated using the ex vivo skin deposition data and considering oral bioavailability, as there is limited available data for the transdermal pharmacokinetics of quercetin. To achieve an anti-inflammatory effect, at least 500 mg of quercetin per day is required to be dosed orally [51], with an oral bioavailability of less than 10% [38]. Accordingly, the dose of quercetin that needs to be effectively delivered is 50 mg. Therefore, correlating our ex vivo deposition data, which is able to deposit approximately 2 mg of quercetin and expected to be released sustainably over a period of time, with oral bioavailability, we can estimate the patch size for daily treatment in human adults to be 12.25 cm2. This patch size might be necessary to achieve systematic anti-inflammatory activity. Although the patch size in this study is relatively large compared to marketed transdermal patches, previous research has shown the successful application of large MAPs onto human skin. Moreover, it is important to note that this size is estimated to achieve systemic effect. Extensive studies are required to evaluate the dose needed to achieve anti-inflammatory activity locally, in which case smaller patches might be required to fulfill this purpose.

Cytocompatibility studies

After evaluating both MAP formulations (F1 and F2) for their mechanical properties, insertion profile, and drug delivery efficiency, we opted for the MAP containing Soluplus® (F2) for biocompatibility studies and in vitro anti-inflammatory assessment. This choice was due to the superior characteristics of the MAP F2 formulation, which boasted the highest drug loading, faster dissolution time, and greater drug delivery into the skin compared to F1. Previous reports have indicated that quercetin treatment significantly enhances the viability of HaCaT keratinocytes [52, 53]. To explore this further, we conducted MTT assays, live-dead staining, and proliferation assays to assess the cell viability, cytotoxicity, and proliferation of HaCaT cells following exposure to quercetin-loaded MAPs (F2) and blank MAPs.

Our findings revealed no significant changes in the group treated with blank MAPs compared to the control, as shown in Fig. 8A. This was further supported by cell proliferation analysis (Fig. 8B) and the live-dead assays (Fig.