Liposome‐assisted penetration and antiaging effects of collagen in a 3D skin model

Collagen is a major component of the extracellular matrix that supports the epidermal layers of the skin; thus, many strategies have been made to enhance the topical delivery of collagen for antiaging purposes. In addition, our previous study indicated that liposome can help the penetration of active ingredients into the skin.


| INTRODUC TI ON
Liposomes have a hydrophilic core and a hydrophobic lipid bilayer, where they can encapsulate active ingredients, and allow them to permeate the skin.Liposome formulations have high encapsulation efficiency, uniform size distribution, and long-term colloidal stability; therefore, liposomes can protect and transport sensitive bioactive ingredients, including proteins and nucleic acids. 1 Liposomes have been reported to positively affect hydration and skin barrier function in humans. 2 In particular, after 30 min of treatment with liposome formulation, the water content in the skin was 1.3 times than in the untreated group, while the skin's barrier function was similar.Liposomes also have a positive effect on drug delivery through the skin. 3Owing to the similarity between the compositions of liposome and stratum corneum lipids, drugs encapsulated in liposomes are more easily absorbed into the skin.Furthermore, liposomes can be manufactured using various methods.Lasic et al. used thin-film hydration method in which hydrated bilayered phospholipid flakes were manufactured, then water was added to produce liposomes. 4uld-Fogerite et al. produced uniform and small liposomes using calcium-EDTA chelation, rotary dialysis techniques, and agarose plug diffusion from large liposomes fabricated by phase change. 5These methods can obtain liposomes with high efficiencies, up to 40% of the weight of the starting materials.Liposomes produced through calcium-EDTA chelation showed jelly roll-like structures.In other study, liposomes were fabricated through bubble method, in which a water/gas interface was created by aggregating lipid monolayers on its surface. 6This method yielded liposomes with a uniform size range from 0.2 to 0.5 μm.It also maintained a very stable state in Tris buffer (pH 7.4) for more than 60 h without a change in the size.Highpressure homogenization is widely employed to satisfy the commercial requirements of cosmetic and pharmaceutical liposomes, such as a monodisperse distribution of particle size and a stable emulsion. 7,8aller and uniformly sized liposomes can be easily obtained by repeatedly forcing liposomes through the valves under high pressure and shear force.In addition, the size of phospholipid liposomes produced through ultrasonication and high-pressure homogenization methods were reported to be uniformly similar, ranging from 0.05 to 0.25 μm. 8 This size can also be controlled by changing homogenization cycles, pressure, and temperature.As the number of cycles increased, the size of the liposomes decreased from 120 to 25 nm, and as the temperature and pressure increased, smaller liposomes were obtained. 9posomes possess various surface charges depending on the type of lipid that constitutes the bilayers.Surface-charged liposomes can improve the skin permeation and penetration of active ingredients.Negatively charged liposomes can penetrate the stratum corneum through hair follicles and sweat glands. 10When the surface-charged liposomes are applied to the skin, the liposomes are dissolved in the lipid membrane located around hair follicle and hair gland, facilitating the delivery of the encapsulated active ingredients.Furthermore, positively charged liposomes facilitate the skin penetration of active ingredients.Since the pI value of the keratin chain constituting the epidermal layer is approximately 5.3, the surface of the skin has a negative charge and forms an electrostatic interaction with positively charged liposomes to facilitate the absorption and permeation of liposomes and active ingredients. 11We have previously evaluated the absorption of drugs into artificial skin according to the surface charge of liposomes.Groups treated with positively and negatively charged liposomes had a 7.65-and 1.76-fold higher drug penetration than the neutrally charged liposomes group. 12 this study, we fabricated surface-charged liposomes containing collagen using a high-pressure homogenization method.To evaluate the effect of liposomes on the skin penetration of collagen, we compared the adherence of collagen between collagen and liposomes containing collagen against artificial membranes.We also determined the degree of skin penetration of liposomes in 3D skin and compared the gene expression levels in the epidermis and dermis between collagen and collagen-containing liposomes.Thus, we showed that percutaneous absorption of collagen can be largely affected by liposome formulation.
Secondly, phase B was gradually added to phase A, which contained 1% collagen in distilled water (DW) at 25°C, the mixture subsequently underwent homogenization for 5 min at 2000-3000 rpm.
Finally, liposomes were obtained through three cycles of 1000-bar homogenization using a high-pressure homogenizer (MN400BF, Micronox, Sangdaewon 1 Il Dong, Republic of Korea).The size and zeta potential of the synthesized liposomes were determined by dynamic light scattering (SZ-100; Horiba) at 25°C every week for 21 days.All analyses were conducted in triplicate, and the data were shown as the mean ± SD.
In the wash-off test and 3D skin experiments, FITC-labeled collagen was used instead of pure collagen.FITC-collagen was synthesized using ethylenediamine (EDA).Briefly, 1000.0 mg of collagen was dissolved in 80 mL of DW, 390.9 mg of 1-eth yl-3-(3-dimethylaminopropyl)carbodiimide and 275.7 mg of 1-hydroxybenzotriazole were dissolved in 4 mL of EtOH/DW (1/1, v/v) before adding into collagen solution for reaction at 24°C for 30 min.After that, 122.6 mg of EDA was added, and the reaction was continued at 24°C for 24 h (EDA@collagen).Finally, FITC was added to EDA@collagen at a molar ratio of 2:1 to the carboxylate group in collagen, and the mixture was reacted at 24°C in the dark for 20 h, then the unbound FITC was extracted using a dialysis tube (MWCO: 1 kDa).
The encapsulation efficiency of collagen was analyzed using high-performance liquid chromatography (HPLC, 1260 Infinity; Agilent).Liposome solution (1.0%, 5 mL) was purified by overnight dialysis (MWCO = 10 kDa) to remove unloaded collagen.After freeze-drying the purified liposomes, 5 mL of acetonitrile was added to destroy the liposome structure.Subsequently, the amount of collagen was quantified using HPLC, and the encapsulation efficiency (EE) was determined using the following formula: where C E is the amount of collagen encapsulated in liposomes, and C 0 is the initial amount of collagen.The incorporation efficiency of collagen was 5.77%.
Liposome size and morphology were observed under cryogenictransmission electron microscopy (cryo-TEM).5 μL of highly diluted liposome solution was dropped onto a carbon-coated copper grid and soaked in liquid ethane for rapid freezing.Frozen liposome samples were visualized using cryo-TEM at 200 kV (Tecnai F20; FEI).The fluorescence intensity of FITC (n = 4) was calculated using ImageJ software v6.0 (Bio Techniques).

| Wash-off test on artificial membrane
To evaluate the adsorption of collagen in liposomes and solution on an artificial membrane, 8 μL of 1.0% (w/v) collagen liposomes and collagen solution (collagen) were dropped and rubbed with a cotton swab onto the membrane (1.766cm 2 ).After 1 h of air-drying, collagen on the membrane was repeatedly washed with 1 mL of DW up to four times.The concentration and the amount of collagen in the washing solution were quantified using Micro BCA™ protein assay kit (Thermo Scientific) following the manufacturer's protocol.Briefly, 50 μL of the washing solution and 50 μL of bicinchoninic acid (BCA) reagent were mixed and incubated at 60°C for 1 h.After that, the absorbance of the mixture was measured at 562 nm using a microplate reader (SpectraMax; Molecular Devices).The remaining collagen (%) was calculated using the following formula: where C 0 is the initial amount of attached collagen, and C n is the amount of attached collagen after washing.The retained FITC-labeled collagen on the membrane was completely dried after the washing step and was subsequently visualized by in vivo fluorescence imaging (VISQUE®InVivo Smart-LF; Vieworks, Gyeonggi-do, Republic of Korea) at the Korea Basic Science Institute Chuncheon Center (λ ex = 495 nm; λ em = 515 nm), and the remaining fluorescence was calculated using the region of interest value.

| Collagen penetration through 3D skin tissue
For the visualization of collagen penetration, FITC-collagen and FITC-collagen liposomes were applied to a 3D skin model and visualized with a super-sensitive confocal laser scanning microscope (LSM 800 BIO; Carl Zeiss) at the Central Laboratory of the Seoul National University of Science and Technology, Republic of Korea.
The degree of green fluorescence intensity was determined for a proportional amount of collagen.A day before collagen treatment, the 3D skin was incubated with a dedicated medium.FITC-collagen liposomes and FITC-collagen (1.0% (w/v), 30 μL) were dropped on the top surface of 3D skin and rubbed with a cotton swab for 30 s.
After 4-h incubation, the skin tissue was rinsed five times with PBS, fixed with 4.0% (w/v) formaldehyde solution for 1 h, and rinsed once with PBS.The 3D skin cell sheets were detached using a knife and trypsin/EDTA, then embedded in O.C.T. compound for sectioning at a thickness of 20 μm using a cryostat microtome (CM1850; Leica).
The sectioned samples were stained with DAPI for nuclei, then visualized using confocal laser scanning microscope (DAPI: 405 nm, FITC: 488 nm).The fluorescence intensity of FITC was analyzed using the ImageJ software (n = 4).
Collagen penetration was determined by comparing the expression levels of several genes (Table 1), including collagens 1 and 3 for collagen synthesis, keratin 5 and 14 for keratinocyte synthesis, aquaporin for the water channel in the epidermis, and involucrin for differentially expressed genes in corneocytes.To damage the 3D skin, 30% EtOH was treated for 5 min, 40 μL of collagen and collagen liposomes (1.0%, w/v) were subsequently dropped and rubbed with cotton swab on each 3D skin, then further co-incubated for 4 h, and finally the treated 3D skins were rinsed twice with PBS.
Subsequently, sample-treated 3D skins were incubated at 37°C for 3 days.

| Gene expression
Total RNA was extracted from each 3D skin sample using GeneAll® Hybrid-R™ (Geneall, Seoul, Republic of Korea), according to the manufacturer's protocol.Briefly, cDNA synthesis was conducted using the reverse transcription premix for 40 cycles of reaction (94°C for 30 s; annealing temperature for 30 s; 72°C for 45 s for each cycle), then quantitative reverse transcription-polymerase chain reaction (qRT-PCR) was conducted to analyze the gene expression levels.The primer sequences and annealing temperatures are listed in Table 1.
Related gene expression levels were normalized to GAPDH expression.The results are presented as target gene expression fold change (2 −ΔΔCt ), in which: and

| RE SULTS AND D ISCUSS I ON
In this study, collagen liposomes were produced by high-pressure homogenization, as described in Figure 1.The synthesized collagenencapsulated liposomes exhibited a high degree of adhesion to the artificial membrane and good penetration through the artificial 3D skin.In addition, fold-changes in skin-related gene expression (collagen 1 and 3, keratin 5 and 14, involucrin, and aquaporin) were observed between collagen solution-and collagen liposome-treated samples.Similar findings were reported in our previous study, in which the surface of various charged liposomes facilitated the 3D penetration of active molecules. 12When liposomes have a negative charge, they diffuse down the stratum corneum through hair follicles and can penetrate the skin. 13Most drugs are delivered to the outermost root sheath of the three-layered hair follicle.This is because sebum released by the sebaceous gland creates a lipophilic pathway that enables drug delivery. 14 shown in Figure 2A, the cryo-TEM images revealed that the liposomes had a multilamellar or unilamellar structure, demonstrating the unique morphology of liposomes synthesized using the high-pressure homogenization method.The prepared liposomes exhibited a uniform size distribution, with a TEM average diam-  surface-charged liposomes, liposomes were observed to have high colloidal stability for 28 days in DW at 25°C. 12In another study on collagen-encapsulated liposomes, no significant change in size was observed after 0.25% (w/v) collagen was encapsulated, and a negligible change in zeta-potential (−66.48 to −72.95 mV) was observed within 28 days. 15 demonstrate that an active-ingredient-encapsulated liposome could enhance the adhesion efficiency of synthetic membranes, the wash-off test was conducted on collagen liposomes or collagentreated membranes.Both membranes were washed up to four times, and the remaining collagen was analyzed by the BCA assay.Figure 3 demonstrates that the retention levels of collagen liposomes were significantly greater than those of native collagen.The collagen was significantly washed off after the first two washes (W 1 -W 2 ), and similar results were observed for the remaining collagen in the subsequent washes (W 3 -W 4 ) (Figure 3A).Using collagen liposomes increased the amount of collagen remaining on the membrane by 1.56-fold after the final wash.The wash-off test results indicated better adhesion properties for collagen when encapsulated in liposomes, and collagen could not be washed off completely owing to its strong attachment and penetration to the outside of the membrane.The artificial membrane used in the experiment imitated the structure of the natural skin tissue.
The stratum corneum is the outermost layer with a bilayer structure and is organized into clusters or columns of cells.Liposomes are very similar to the structure of the corneum and can enhance deeper penetration through the skin tissue. 16Specifically, negatively charged liposomes can settle more effectively in the skin through the hair follicle lipid film of the stratum corneum. 10In addition, the liposome size was approximately 150 nm, which was smaller than the studied membrane model, the Strat M synthetic membrane, and the pore size (approximately 200 nm), which could be an additional reason for the easier penetration of collagen liposomes into the membrane. 17xt, the expression of genes related to various skin factors was determined after applying collagen liposomes and collagen to the 3D skin (Figure 4).To confirm the differences in gene expression between normal and damaged skin, 3D skin was damaged using a 30% EtOH treatment for 5 min.With or without skin damage, the collagen liposome-treated groups exhibited greater gene expression effects than the collagen-treated groups for all six genes.Collagen 1 and 3 are related to collagen synthesis; keratin 5 and 14 are related to keratinocyte synthesis; aquaporin is associated with keratinocyte synthesis; and involucrin is related to the differential expression of genes in corneocytes. 18The difference between the expression levels of keratinocyte-related genes in the EtOH-non-treated and EtOH-treated groups was significant.Moreover, collagen was applied more effectively to fibroblasts in the dermal layer beneath the epidermal layer following damage, resulting in a noticeable difference in the expression levels between samples of collagen 1 and in the construction of a consolidated epidermis. 19In another study, the expression of keratinocyte markers such as keratin 14, laminin subunit alpha 3, and collagen VII was comparable before and after culturing human keratinocytes with collagen. 20Indicating that keratinocytes retain their proliferative phenotype after collagen treatment.Collagen is rich in amino acid peptides, such as proline, glycine, and hydroxyproline, which might enhance the moisture content of the stratum corneum and the differentiation of collagen existing in the skin. 21,22qRT-PCR results confirmed that collagen treatment itself is helpful for the expression of skin-related genes.
Furthermore, we evaluated the deeper penetration of collagen liposomes into 3D skin by observing FITC-labeled collagen crosssections under IVIS (Figure 5).Encapsulated FITC-collagen exhibited 1.27-fold greater fluorescence intensity than non-encapsulated FITC-collagen.Liposomes have been employed as carriers and penetration enhancers for dermal and transdermal delivery, and liposomes prepared from ceramides have been reported to be more penetration-effective than those prepared from phospholipids. 23uorescence labeled phospholipids have been previously used to visualize the penetration depth of liposomes.The results showed that the fluorescence intensity could be observed in the deep dermis at a depth of approximately 1000 μm within a short time, regardless of the formula of the liposomes. 2,24Therefore, we speculated that the synthesized liposomes could transport a larger amount of collagen and simultaneously penetrate deeper into the skin.The qRT-PCR results also confirmed that using liposomes, a larger amount of collagen could penetrate more deeply into the skin, which further stimulated the expression of multiple genes associated with the skin.
Along with the wash-off test, we hypothesized that a greater amount of collagen encapsulated in liposomes remained in the stratum corneum, allowing for greater collagen attachment to the artificial membrane.However, clinical studies are necessary to evaluate a degree of percutaneous absorption of collagen-encapsulated liposomes on human skin.

| CON CLUS ION
In this study, collagen-encapsulated liposomes and collagen solutions were prepared using high-pressure homogenization.We compared skin adhesion, penetration properties, and the expression of several genes with and without liposome encapsulation.Collagen liposomes showed high colloidal stability over extended time periods.
The results also demonstrated that collagen liposomes in artificial membranes have a high skin adhesion property and a high absorption rate in 3D skin.In addition, a positive effect was observed on the expression of various genes that are essential for skin composition.Henceforth, the application of active ingredients, such as collagen, to the skin after encapsulation in liposomes can increase the absorption and expression of effective ingredients into the deeper skin even after several washes.Therefore, the developed liposome can be a potential carrier for collagen and can be utilized in the cosmetic industry.

F I G U R E 1
Preparation of collagen liposomes.(A) Schematic representation of collagen liposomes synthesis via highpressure homogenization method.Phases A (water phase) and B (oil phase) were separately mixed before undergoing highshear mixing, followed by homogenization.(B) Components of phases A and B to prepare collagen liposomes and collagen solution.
eter of 117.1 ± 22.2 nm.The hydrodynamic size and zeta potential of collagen-encapsulated liposomes were further monitored every week for 21 days to confirm the colloidal stability of liposomes in distilled water (Figure 2B,C).The size of the liposomes remained almost constant (151.9 ~ 158.4 nm), while the zeta-potential did not show any significant change (−19.2 to −24.3 mV) during the monitoring period, demonstrating the high colloidal stability of liposomes in water.In previous studies on the skin penetration effect of (3) ΔΔCt = ΔCt (target group) − ΔCt (no treatment group), (4) ΔCt = target gene Ct value − GAPDH Ct value.TA B L E 1 Primer sequences and annealing temperatures used for qRT-PCR analysis.

FITC-labeled collagen
Figure 3C, only 69.0 and 85.8% (w/w) of the collagen remained after

F I G U R E 2 F I G U R E 3
Characterization of collagen liposomes.(A) Cryogenic-transmission electron microscopy (cryo-TEM) images of collagenencapsulated liposomes, in which multilamellar and unilamellar structures of liposomes are clearly observed.(B) Hydrodynamic size and (C) Zeta potential of collagen-encapsulated liposomes.Liposomes were incubated with gentle shaking at 25°C and monitored using dynamic light scattering measurements every week for 21 days.Determination of remaining collagen liposomes (yellow-green) and collagen (green) on the artificial membrane after washing.(A) Percentage of remaining collagen based on the number of washes after bicinchoninic acid treatment of an artificial membrane.The remaining collagen (%) was calculated using the following formula: remaining collagen (%) = ((C 0 -C n )/C 0 ) × 100 (%), where C 0 is the initial amount of attached collagen, and C n is the amount of attached collagen after washing count (n = 3).(B) Fluorescence intensity of collagen liposomes and collagen before washing (W 0 ) and in the supernatant after each wash (W 1 -W 4 ).(C) The percentage of fluorescence intensities of remaining collagen liposomes and collagen after each wash.Initial collagen before washing was normalized.Fluorescein isothiocyanate isomer I-labeled collagen was used instead of collagen for fluorescence visualization studies.The fluorescence intensities were visualized using by in vivo Imaging System with λ ex /λ em = 495/515 nm, and the percentage was calculated using the region of interest value.

3 .
The presence of collagen strengthened the expression of involucrin and aquaporin genes, especially collagen in the encapsulated form.Previous studies have reported that keratinocytes, such as HaCaT cells, proliferate at a rate of 114% in collagen-containing scaffolds.In addition, collagen induced a remarkable upregulation of involucrin, filaggrin, and TGase1 genes, indicating that collagen could induce the differentiation of keratinocyte, which plays a vital role F I G U R E 4 Gene expression properties of 3D skin after collagen liposome and collagen treatment using quantitative reverse transcription-polymerase chain reaction.3D skin was damaged by 30% EtOH (EtOH O) or not damaged (EtOH X) and then treated with collagen liposomes (1.0%, w/v) or collagen.Selected genes were collagen type 1 and 3 for collagen synthesis, keratin 5 and 14 for keratinocyte synthesis, aquaporin for the water channel in the epidermis, and involucrin for the differential expression gene for the corneocyte.Y-axis is presented by target gene expression fold change ( −ΔΔCt ), where ΔΔCt = ΔCt (target group) -ΔCt (no treatment group), and ΔCt = target gene Ct value -GAPDH Ct value.F I G U R E 5 Penetration of encapsulated and non-encapsulated collagen to a 3D skin model composed of human keratinocytes.After treating with encapsulated fluorescein isothiocyanate isomer I (FITC)-labeled collagen and non-encapsulated FITC-labeled collagen for 4 h, 20μm sections of 3D-treated skin were visualized using confocal laser scanning microscope (4′6, diamidino-2phenylindole dihydrochloride stained in blue and FITC stained in green).The fluorescence intensity of FITC was calculated using ImageJ software (n = 4).Scale bar = 50 μm.