Phycocyanin was kindly donated by Danzi Company (China). GMO and P407 were purchased from Wako Pure Chemical Industries, Ltd., (Japan) and BASF, Ltd., (Germany), respectively. Immortalized human keratinocytes (HaCaT) were obtained from China Center for Type Culture Collection (China). Cell medium, fetal bovine serum, penicillin, and streptomycin were purchased from Gibco America, Ltd., (United States). All other chemicals were analytical grade and used as received.

PC-cubosomes were prepared by melt dispersion emulsification technique with some modifications (Makhlouf and Elnawawy, 2023). As showed in Figure 1, the dispersed phase consisting of GMO and P407 was heated at 50 °C in a water bath to form oil phase. Phycocyanin was dissolved in MilliQ water by a vortex mixer for 5 s to form phycocyanin solution at a concentration of 0.59 μg/mL. Afterwards, the oil phase was gently injected into the preheated aqueous phase (phycocyanin solution) through using a syringe at 50 C and emulsified using a rotor-stator homogenizer (Avestin Em-C3, Ottawa, Canada) at 10,000 rpm for 5 min. The final dispersion was cooled and maintained at room temperature for further investigation.

A three-factor, three-level (3³) full factorial design was applied to identify the impact of mass fractions of GMO, P407, and phycocyanin as independent variables on the PC-cubosomes physicochemical properties (Table 1). Five dependent variables include particle size (Y₁), polydispersity index (PDI, Y₂), zeta potential (ZP) (Y₃), encapsulation efficiency (EE%) (Y₄), and cumulative drug release through an artificial membrane within 24 h (Y₅) were evaluated. The experimental design matrix generated by Design Expert^® (version 8.0.3, State-Ease Inc., Minneapolis, Minnesota) and the composition of PC-cubosomes formulations were showed in Table 1 and Table 2.

Particle size, PDI and ZP of PC-cubosomes were determined by dynamic light scattering (DLS) (Zetasizer, Malvern Instruments, United Kingdom). Samples were diluted with deionized water, adjusted to a suitable light scattering intensity (300 Hz), and measured at 25°C in triplicate.

The non-encapsulated phycocyanin was separated from cubosomes by dialysis as follows (Castangia et al., 2016b). 5 mg PC-cubosomes were loaded into dialysis bag (300 kDa Float-A-Lyzer G2, Spectrum Laboratories Inc., DG Breda, Netherlands) and dialyzed for 2 h at 25 C, which allows dissolution and consequent removal of non-encapsulated phycocyanin. Samples were analyzed for phycocyanin concentration using a fluorescence spectrophotometer (LS55, PerkinElmer, United States) at 615 nm. EE% was calculated according to the following equation: EE  % = Determined   drug   content   post − dialysis   Determined   drug   content   pre − dialysis × 100   % (1)

PC-cubosomes (n = 3) containing 2 mg of phycocyanin were sealed in a cellulosic dialysis membrane (Spectra-Por1 dialysis tubing, cut-off 30 kDa, Spectrum laboratories Inc., United States) and immersed in a beaker containing 200 mL of phosphate buffer saline (PBS, pH 7.4) at 32°C ± 0.5 C with gentle agitation at 100 rpm. A volume of 2 mL of release medium was withdrawn at pre-determined time points and replaced with an equal volume of fresh pre-warmed release medium to maintain sink conditions. The concentration of phycocyanin was analysed by fluorescence spectrophotometry and kinetic models (zero-order, first order, Higuchi, and Korsmeyer-Peppas models) on drug release from PC-cubosomes was also performed. All measurements were carried out in triplicate and the values were expressed as mean ± S.D.

The morphology of the optimal PC-cubosomes was examined using a transmission electron microscope (TEM, JEM−100CX, JEOL, Japan) at an accelerating voltage of 120 kV. Briefly, PC-cubosomes were diluted with distilled water and a drop of suspension was placed on a copper-coated grid to form a thin film. Subsequently, samples were stained by phosphotungstic acid solution (2% w/v, pH 6.8) for 1 min, air dried at room temperature, and observed by TEM.

The stability of phycocyanin encapsulated in the optimal PC-cubosomes and solution was determined according to the method of Zhong with some modifications (Yejun Zhong et al., 2024). PC-cubosomes and PC solution were exposure to UV radiation at 25°C for 10 days. At predetermined time points, 2 mL of each sample were withdrawn and analysed by UV-VIS spectrophotometry (TU−1901, General Instruments Co., Ltd, China) (Zhuang et al., 2023). Encapsulated phycocyanin was extracted from PC-cubosomes using ethanol and free phycocyanin in solution was used as a control.

Ex-vivo transdermal permeability of phycocyanin solution and PC-cubosomes was evaluated using modified Franz diffusion cells with a contact area of 3.14 cm² and a receptor volume of 8.5 mL (Chen et al., 2023). After removal of hairs and subcutaneous fat, the abdominal skin isolated from SD rats was placed in between the donor and receptor compartments with the stratum corneum (SC) side facing the donor compartment. The receptor compartment was filled with 8.5 mL of PBS (pH 7.4) to maintain sink conditions (El-Enin and Al-Shanbari, 2018). The diffusion cells were kept in a water bath at 32°C ± 0.5 C with shaking at 100 rpm in order to mimic the in vivo environment (Tu et al., 2014). After equilibration of 15 min, formulations containing 6.0 mg of phycocyanin were added to the donor compartment. The medium in the receptor compartment were withdrawn and replaced by equal volume of fresh pre-warmed PBS. Each sample was filtered through a 0.22 μm syringe filter before quantified by fluorescence spectrophotometry. Cumulative penetration amounts (Q) were plotted against time (t) by Equation (2). Js = dQ dt (2)

At the end of the ex-vivo transdermal permeation study, the skin was carefully removed from Franz diffusion cells and cleaned by gauze soaked with fresh PBS (El-Enin and Al-Shanbari, 2018). To separate the SC from the remaining epidermis (E) and dermis (D), the skin sections were stripped with 15 pieces of adhesive tape. All tapes, except the first one, were immersed in 5 mL of water and sonicated for 10 min after vortex stirring. The remaining E and D were cut into small pieces and immersed in 10 mL of water and sonicated for 10 min. All samples were centrifuged for 10 min at 15,000 rpm, the supernatant was collected and analysed by fluorescence spectrophotometry.

The protective activity of PC-cubosomes and phycocyanin solution against hydrogen peroxide (H₂O₂) induced oxidative damage was evaluated on HaCaT cells (Rachitha et al., 2023). The cells were co-treated with H₂O₂ and PC-cubosomes or phycocyanin solution (final drug concentration of 100 μg/mL) for 3 or 6 h, followed by repeated washing with PBS. Then the absorbance of all groups were determined by a standard MTT assay (Murgia et al., 2015). All measurements were performed in sextuplicate. Untreated cells and the cells treated with H₂O₂ only (1:30,000 dilution) were used as negative and positive controls, respectively.

Full factorial design (3³) is a commonly used method to optimize experiments with multiple experimental designs involved to generate polynomial mathematical relationships and to map the response over the experimental domain. In this work, full factorial design was applied to optimize PC-cubosome formulations.

Twenty-seven formulations were prepared according to the optimization parameters in full factorial design (3³) (Table 2). All the formulations were valued by the factors mentioned above (Y₁, Y₂, Y₃, Y₄, and Y₅). F2, F11, and F20 were set to evaluate the influence of GMO. The influence of P407 mass fraction was investigated through F1, F2, and F3. Furthermore, F2, F5 and F8 were used to explore the impacts of phycocyanin mass fraction. Thirteen formulations were successfully prepared to form cubosomes, while viscous bulk cubic gels were observed in the remaining formulations after cooling down. Only the formulations with cubosome properties were further analysed. Responses of PC-cubosome parameters generated by (3³) full factorial experimental design were showed in Table 3.

For formulation optimization, particle size and PDI of different formulations were firstly compared. Nanoformulations have many advantages for transdermal drug delivery system such as high skin penetration ability, greater skin deposition and improved physicochemical stability of the loaded drug (Ghasemiyeh and Mohammadi-Samani, 2020). The desirability constrains were minimizing particle size, polydispersity index and maximizing zeta potential as absolute value. According to the results, F1-F6, F12, and F15 were excluded due to their broad particle size distribution with PDI values of higher than 0.3. Formulation 18 was also excluded from analysis due to its large particle size of 338.2 ± 0.8 nm. In vitro release kinetics of F8-F11 was subsequently evaluated in order to choose the optimal formulation, as we discuss later. F8 was eventually identified as the optimal cubosome formulation with further characterize and study.

Small particle size of nanoparticles generally facilitates transdermal permeation of drugs. The regression equations showed the influence of mass fractions of GMO (A), phycocyanin (B), and P407 (C) on particle size (Y₁) and PDI (Y₂) in terms of coded values according to equation 3 and 4, respectively. According to the results, the particle size and PDI values of PC-cubosomes were significantly affected by the mass fractions of GMO, P407, and phycocyanin (p < 0.05). The mass fraction of P407 (C) showed a noticeable negative effect on particle size (Y₁), while both mass fractions of GMO (A) and phycocyanin (B) showed positive effects. In contrast, mass fractions of GMO (A) and phycocyanin (B) displayed negative effects on PDI (Y₂), whereas P407 (C) exhibited a positive effect on it. Y 1 = − 151.522 + 46.627 A + 806.022 B − 19.713 C (3) Y 2 = 0.438 − 0.027 A − 1.053 B + 0.082 C (4)

It was found that the particle size of cubosomes increased with increasing the amount of GMO, which could be attributed to its high viscosity. Also, higher mass fraction of phycocyanin could decrease the surface tension and lead to the formation of cubosomes with uniform particle sizes (Chen et al., 2024). The positive effect of amphiphilic polymer P407 on PDI was mainly due to its ability to overcome intrinsic adhesion of GMO and to promote the formation of well-dispersed cubosomes (Shamma and Aburahma, 2014).

ZP is an important parameter to evaluate the stability and bio-distribution of colloidal dispersions (Zhang and Wu, 2021). Generally, high surface charges lead to electrostatic repulsion between particles, thereby preventing them from aggregation (Makhlouf and Elnawawy, 2023). In this study, the prepared PC-cubosomes showed a high absolute ZP values ranged from −26.4 ± 1.3 mV (F9) to −33.9 ± 1.1 mV (F1) (Table 3), which could effectively avoid particle aggregation.

The ANOVA analysis indicated that the mass fractions of GMO (A), phycocyanin (B), and P407 (C) significantly affected the ZP parameter (p < 0.05) according to Equation 5. Y 3 2 = 1284.490 + 4.590 A − 1808.440 B − 153.473 C (5)

The absolute value of ZP increased with increasing the mass fraction of GMO, which consistent with previous reports (Al-Shoubki et al., 2023). This may be due to the small amount of free oleic acid, which could absorb on the surface of cubosomes and exhibit a negative surface charge based on the presence of carboxyl groups at the terminal of oleic acid chains. Both mass fractions of phycocyanin and P407 showed negative impacts on the ZP value, which was probably due to their adsorption on the outer surface of PC-cubosomes (Sherif et al., 2014). Overall, the high ZP value could allow enhanced carrier penetration due to the electric potential differences across the skin.

The EE% values of preliminary screening 13 formulations were measured with data shown in Table 3. All the EE% values of PC-cubosomes ranged from 82.1% ± 1.7% (F4) to 87.2% ± 2.7% (F8), which might be attributed to the strong attraction between phycocyanin and the hydrophilic domain of cubic phase bilayer (Driever et al., 2013). P407 is a commonly used moisturizer in transdermal drug delivery systems. Since both hydrophilic (ethylene oxide blocks) and hydrophobic (propylene oxide blocks) segments exist, P407 could facilitate phycocyanin to be incorporated into cubosomes through intermolecular hydrogen binding or molecular interactions between the copolymers and drugs (Dumortier et al., 2006).

In vitro drug release kinetics of the formulations with a particle size smaller than 300 nm and uniform particle size distribution (PDI <0.3) (i.e., F8, F9, F10 and F11) were investigated (Figure 2). It was observed that the viscosity of formulations increased with increasing the mass fraction of GMO (F10 and F11), which in turn restricted drug diffusion from the outer monoglyceride bilayer to the aqueous release medium (Atlibatur et al., 2023). Meanwhile, according to the literature, GMO as a basic cubosomes component might lead to decreasing the drug partitioning rate from the oily medium to the aqueous one when compared to the free drug which could diffuse easily to the dissolution media (Nguyen et al., 2017; El-Enin and Al-Shanbari, 2018). Compared with the phycocyanin solution, PC-cubosomes showed decreased drug release rates, particularly for the formulations of F10 and F11, which could be attributed to the neutral stabilizers of P407 (Driever et al., 2013). Moreover, the cubosomes prepared with a higher mass fraction of P407 (F9) exhibited delayed drug release compared to the formulation with a low mass fraction of P407 (F8). The possible explanation is the surface of F9 was covered more by P407, which restricted release of encapsulated phycocyanin (Moura et al., 2020). Therefore, F8 was considered the optimal formulation and its release kinetics were further analysed.

The correlation coefficients (R²) were obtained after fitting the in vitro release data to various release kinetic models. Results indicate that best fit was obtained with Higuchi kinetic model (R² = 0.9901), which was consistent with previously reported findings (Guo et al., 2010). The calculated value of 0.70059 (0.43 < n < 0.85) obtained by fitting the release data to Korsmeyer Peppas model indicated that the drug release mechanism was anomalous (non-Fickian) and coupled with the swelling of glyceryl monooleate (Hu et al., 2023).

The morphology of F8 was investigated by TEM (Figure 3) with showing a spherical shape. The particle size was around 200 nm (Figure 3A), which was consistent with the result obtained by dynamic light scattering. Figure 3B shows the cubical structure of PC-cubosomes with light lines and black dots representing water channels and lipid bilayers, respectively. These results therefore indicated that the PC-cubosomes were successfully formed in a highly ordered cubic internal structure (Kwon et al., 2012).

The stability of PC-cubosomes and phycocyanin in solution was evaluated by exposure the formulations to UV radiation for 10 days (25 C). As shown in Figure 4, over 70% of free phycocyanin was degraded, whereas only 3% of the encapsulated phycocyanin in cubosomes was degraded on the day 10, indicating cubosomes protected the encapsulated phycocyanin from degradation. P407 in cubosomes also could contribute to the protective effect by increasing the hydrophilicity of proteins, demonstrating that cubosomes could efficiently protect phycocyanin from rapid degradation due to their high encapsulation capacity of phycocyanin (Guo et al., 2010; Chong et al., 2011).

Figure 5A showed the cumulative drug penetration per unit area across the excised SD rat abdominal skin. The drug permeation rates of F8 and free drug at steady state (Js) were obtained from the slope of the linear portion. PC-cubosomes exhibited a higher permeation rate (39.79 μg⋅cm⁻²⋅hour⁻¹) than the phycocyanin solution (16.33 μg⋅cm⁻²⋅hour⁻¹), suggesting that cubosomes improved the permeability of phycocyanin across the skin tissue. This was possibly due to the presence of penetration enhancer GMO, which modifies the intercellular structure of lipid layer and changes the fluidity of cell membrane. Moreover, the repulsive force between cubosomes and lipids in the skin also could contribute to the improved drug penetration (Jia et al., 2014).

As shown in Figure 5B, the amounts of phycocyanin released from PC-cubosomes retained in SC and (E + D) were 98.47 ± 3.34 μg and 45.38 ± 2.37 μg, respectively. For the free drug group, only 41.03 ± 3.21 μg (SC) and 19.38 ± 2.34 μg (E + D) of phycocyanin were retained in the skin tissue after release studies. This may be attributed to the fact that cubosomes with molecular weight copolymers could restrict drug transport from the skin to the whole body, delay drug clearance by the systemic circulation, and prolong drug retention in the target tissue (Chong et al., 2011). Cubosomes therefore can be considered as a sustained release depot for transdermal drug delivery (Zakaria et al., 2022).

The cytotoxicity of PC-cubosomes (F8) and phycocyanin in solution was evaluated using HaCaT cells. The cells were treated with PC-cubosomes and drug solutions with different phycocyanin concentrations ranged from 20 to 400 μg/mL for 48 h. In the phycocyanin solution group, the rates of cell viability were 85.13% ± 2.94% and 69.87% ± 2.73% at the lowest (20 μg/mL) and highest concentrations (400 μg/mL) (Figure 6A). The cell viability reduced with increasing the drug concentration, because high concentration of phycocyanin may impair the mitochondrial respiratory chain, reduce energy supply, and induce cell death (Chong et al., 2011). In contrast, the relative viability of cells treated with PC-cubosomes at all the concentrations was higher than 95.98% ± 4.54%, as phycocyanin was slowly released from cubosomes, which allows the cells to compensate for the change of phycocyanin concentration in a certain degree with their metabolism. Cell viability results indicated that PC-cubosomes were a safe carrier for transdermal drug delivery.

As Figure 6B showed, the cells treated with H₂O₂ only were used as a positive control with cell viability decreasing to 36.13% ± 3.18% and 21.79% ± 2.87% after 3 h and 6 h of incubation, respectively. In contrast, higher cell viability was achieved in the phycocyanin solution group with 73.69% ± 5.26% and 75.14% ± 3.92% observed after the same treatments. These results indicated that the oxidative stress damage induced by H₂O₂ was rescued by the treatment with phycocyanin due to its antioxidant potentials. Interesting, the PC-cubosomes group showed a further improvement on the cell viability with 101.73% ± 2.74% and 110.04% ± 4.01% (p < 0.05 versus all others) survived after 3 h and 6 h of treatment, respectively. This was possibly due to the lipid structures of GMO and P407 that facilitated the exchange between proteins and intracellular lipids, which are crucial for maintaining cell function (Zeng et al., 2023). Therefore, this study demonstrated that phycocyanin could counteract the ROS oxidative activity towards keratinocyte damages and loading phycocyanin into cubosomes further improved its aptitude due to the enhanced cell permeability (Chen et al., 2024).