A new eco‐friendly and water‐resistant sunscreen agent: Lecithin‐based multilamellar liposomes

UV skin exposure is an important matter of public health, as the worldwide rising prevalence of skin cancers indicates. However, a wide majority of commercially available sunscreens are responsible for ocean ecosystem damages such as coral reef degradation and phytoplankton mortality.


| INTRODUC TI ON
UV skin exposure is well known to cause several damages on both epidermis and dermis, such as sunburn cell features, cell vacuolization, cell dysplasia, cell nuclei amorphism, or Langherans cell population diminution. 1 UV exposure also alters intercellular cohesion between corneocytes and even keratinocytes resulting in an impaired skin barrier, which leads to skin dehydration and may favor pathogens entry into the tissue. 2Chronic UV exposure is responsible for pigmentary disorders (solar lentigo, melasma) and dermis fiber fragmentation and elastosis, which lead to premature skin aging. 3ronic UV exposure is also the main risk factor for skin cancer development.UV light favors direct DNA lesions such as cyclobutane pyrimidine dimer (CPD) formation.It also favors reactive oxygen species accumulation, which promotes not only collagen and elastin breakdown but also DNA, protein, and lipid oxidation. 3All these deleterious effects affect melanocytes and keratinocytes giving rise to melanoma and carcinoma (basocellular or squamous), respectively.
The rising prevalence of these three skin cancer types makes solar protection a major health issue. 4nce the mid-20th century, a large variety of chemicals (octocrylene, avobenzone, oxybenzone, etc.) and minerals (TiO 2 nanoparticles, ZnO nanoparticles, etc.) UV filters have been developed and commercialized to answer the need for solar protection.
However, ocean and lake pollution by both types of filters is a major environmental issue.Active molecules and/or nanoparticles pollute waters either directly through swimmers wearing sunscreens or through the wastewater pathway.The total amount of sunscreen released in oceans is estimated to be 25 000 tons per year despite the development of water-resistant formula. 5Notable active ingredients consequences for the environment are coral reef degradation 6 and local phytoplankton population diminution. 7Reefs represent the major sea life habitat and phytoplankton is the very base of the aquatic food chain.Hence, there is a growing urge for new technologies development associating effective sun protection and environment preservation.
Liposomes are known to be biocompatible and harmless to the environment as they are easily biodegraded or even metabolized. 8posomal sunscreens appeared on the market during the 1990s.
Liposome incorporation into sunscreen formula was believed to improve filters stability, cutaneous penetration, cutaneous moisturizing content, and overall performance.Some groups did demonstrate those advantages even though few literatures exist on the matter.

Golmohammadzadeh et al. have shown that octyl methoxycinnante
encapsulation into multilamellar liposomes slightly improved its sun-screening efficacy compare to classic formula. 9The same group has also demonstrated that the use of safranal loaded liposome increased the hydration content of the stratum corneum (SC) until 5 h after application. 10Korting et al. showed a better resistance of several liposomal sunscreen to water and perspiration exposure. 11 et al. demonstrated that encapsulation of a chemical filter into small uni-lamellar liposomes made from phosphatidylcholine and cholesterol protected it against degradation leading to a longer UV protection. 12Nevertheless, the different mechanisms by which filter quality is improved depends on liposomes properties that differ from a study to another (lamellarity, size, elasticity, charge, drug loading, composition, etc.).Those mechanisms are not fully understood and further investigation is necessary, specifically regarding liposomes use alone.
Multilamellar liposomes (MLLs) of controlled physicochemical parameters made from soybean lecithin were developed to target skin layers, as demonstrated in previous articles. 13,14Those MLLs display distribution sizes suited for low wavelength light scattering such as UV light.Studied lead on other nanoparticles, and notably solid lipid nanoparticles, have demonstrated their ability to synergistically improve the photoprotective effect of sun-screening agent; the latter upgrade being explained by the light scattering induced by nanosized particles. 15Furthermore, lecithin is known to absorb light in both the UVA and UVB sections of the light spectrum. 16Ls could therefore be efficient as both a chemical and a physical sun-screening agent.MLLs prepared by shearing of a lamellar phase (Section 2.2) are nontoxic, more stable than classical liposomes 17 and exhibit a cream-like texture allowing direct on skin spreading.
Lecithins are considered by the European Department of Health and Services as non-toxic for aquatic organisms and as non-persistent in the eventually of an environmental release as they are readily biodegradated. 18MLLs can be easily incorporated in a formula due to their good water affinity. 17In light of the aforementioned information, we decided to investigate the photoprotective potential of two types of elastic MLLs designed to target either the stratum corneum (SC-MLLs) or the living epidermis (Epi-MLLs). 13,14Presumably, MLLs skin penetration will also offer a better adhesiveness of the product compared to classical creams and therefore a more persistent protection during and after bathing.Comparison between both MLLs type will allow investigation of skin layers targeting importance when it comes to sun-screening efficacy.
Effectiveness of Epi-MLLs, SC-MLLs and a SPF50+ water-resistant commercial cream (CC) containing liposomes were evaluated ex vivo.All preparations were deposited onto living skin explants prior to light irradiation.In some cases, preparations were rinsed off before irradiation to assess their adhesiveness to the skin.
Hematoxylin and eosin staining (H&E) and CPD immunostaining were performed to evaluate skin UV damages.Trans-epithelial electrical resistance (TEER) was also performed as a common indicator of skin barrier quality.

| Products and reagents
DMEM/F12 culture medium was purchased from ThermoFisher Scientific (Waltham, MA, USA).P100, a lecithin mixture, was obtained from Lipoid GmbH (Ludwigshafen, Germany) and Tween®80 from Sigma-Aldrich (Saint-Louis, MO, USA).SPF50+ Daylong extreme liposomal solar cream was purchased at a local pharmacy.Human skin explants were obtained from surgical residues of breast reduction done at Bordeaux hospital (CHU).Informed consent of all donors was obtained as required according to the Helsinki declaration.
MLLs designed to target the whole epidermis (Epi-MLLs) contain 51 w% of P100, 17 w% of Tween®80, and 32 w% of water.MLLs were prepared according to a shearing protocol previously described in the literature. 13Briefly, P100 and Tween®80 were solubilized together in ethanol.Organic solvent was evaporated under nitrogen flux and the resulting mixture was dispersed in water and freeze dried to ensure ethanol traces withdraw.The homogeneous mixture was accurately weighted and hydrated with half the total amount of water.The mixture was sheared for 1 min and centrifuged (12 000 RCF, 5 min).The remaining water was added, and four shearing-centrifugation cycles were applied.Samples were stored at 4°C overnight and three last shearing-centrifugation cycles were performed.This process leads to the formation of a viscous phase composed of MLLs in close contact without any external water.MLLs were characterized in terms of charge, size, and elasticity after dispersion in water.
For skin deposition purposes, MLLs were dispersed in water at a 10 mg mL −1 concentration and ultra-centrifuged on an OptimaMax-E centrifuge (Beckman Coulter, Brea, CA, USA) for 1 h, 105 000 RCF at 4°C.Excess water was removed and 1 mg of the cream-like pellet was then directly used for each skin experimentation.

| MLLs characterization
MLLs were characterized in terms of size, charge, and elasticity following previously described protocols. 13,14Ls size distribution was obtained using a Mastersizer 2000 Hydro SM (Malvern Instruments, S.A.).MLLs were dispersed in water and introduced into the dispersion chamber.MLLs diameter was expressed as the volume-average diameter, D[4;3], relying on Mie's theory.In this study, 1.33 was used as the water refractive index and 1.45, that is, lecithin's refractive index, was used as the refractive index of the dispersed phase.
MLLs zeta potential was measured with a ZetaSizer Nano Serie (Malvern Instruments S.A.).MLLs were dispersed at a 1 mg mL −1 concentration in water and introduced into a DTS1070 cell.Each measure consists of an average of a 100 runs at 25°C.

MLLs elasticity corresponds to the percentage of MLLs crossing
a PVDF filter of pore diameter smaller than the smallest liposome obtained from the size distribution analysis. 13Quantification of MLLs on both sides of the filter is performed following the Rouser phosphorus assay.Maximum absorbance was found at 813 nm using 1 cm quartz cells on a UH5300 Hitachi spectrophotometer (Tokyo, Japan).

| Preparations deposition on skin explants
Experiments were conducted on five different white skins aged 17-73.Skin samples were placed on grid set in six wells plates.They were maintained at 37°C in a 5% CO 2 atmosphere and under sterile conditions throughout all experiment.Six milliliter of DMEM/F12 supplemented with 10% SVF medium was set under each skin piece to insure its viability.When needed, 1 mg of the cream-like MLLs or the commercial sunscreen (CC) were deposited on a 0.5 cm 2 surface.
Deposition was performed using a positive piston micropipette based on preparations densities.Preparations were deposited either 2 h or 10 min before skin irradiation.For adhesiveness tests, deposition was made 2 h prior to irradiation and skins were rinsed off twice with 1 mL of phosphate buffered saline (PBS) right before irradiation.An Atlas Sunset CPS+ (Berwyn, PA, USA) equipped with a Xenon lamp set to 750 W m −2 was used for skin irradiation.Xenon lamp mimics sunlight exposure.Exposition lasted 10 min.DMEM/F12-SVF was removed from wells before irradiation and replaced with fresh medium afterwards.TEER measurements were performed 24 h after irradiation and skin samples were thereafter fixed in formol for histology purposes.

| TEER measurements
TEER was measured using an EVOM 2 device (WPI, Sarasota, FL, USA) equipped with a STX2 electrode.Skin samples were cut and placed in a 0.5 cm 2 culture insert matching irradiated surface area.
Eight milliliter of DMEM/F12-SVF was placed under the insert and 300 μL were deposited above each skin sample.Electrode probes were set over and under the skin following manufacturer guidelines.

| Statistical analysis
Statistical analysis was performed on a commercially available software (GraphPad Prism, GraphPad Software, Inc., Boston, MA).One way analysis of variance with a Tukey's multiple comparisons test was used to compare all TEER values.Experiments were repeated three times.Results were considered significant for p-value inferior to 0.05.

| Hematoxylin and Eosin staining
Skin samples remained at 24 h maximum in formol and were then dehydrated in successive ethanol baths of increasing purity followed by three xylene baths.Samples were then imbedded in paraffin and cut off into 4 μm thick sections on a Leica RM2255 microtome (Wetzlar, Germany).Sections were classically hematoxylin/eosin stained and observed on a Nikon Eclipse Ni-U microscope (Tokyo, Japan) equipped with a 40×/0.75objective.

| CPD immunostaining
Four micrometer thick sections of skin samples were cut off and deposited on microscope slides.Sections were deparaffined and hydrated.CPD accessibility was performed for 20 min in a PT link station (Agilent Technologies, Santa Clara, CA, USA).pH 6 citrate buffer was used.Maximum temperature was set to 95°C.TDM-2 mouse anti-CPD antibody bought from Cosmo Bio (Carlsbad, CA, USA) diluted to the 1/100 was used as primary antibody.Donkey to mouse Alexa Fluor™ 568 antibody (abcam, Cambridge, UK) diluted to the 1/500 was used as secondary antibody.Nuclei were counterstained using DAPI (Biolegen, Amsterdam, The Netherlands).Sections were observed on a Nikon Eclipse Ni-U microscope equipped with 568 nm laser beam.

| MLLs characterization
MLLs size, elasticity, and zeta potential were measured.Epi-MLLs and SC-MLLs sizes were respectively measured to be 655 ± 22 nm and 180 ± 21 nm.Zeta potentials were equal to −14 ± 4 for both MLLs type.Elasticity was measured for Epi-MLLs and estimated based on previous work for SC-MLLs as the size distribution of this preparation exhibits particles inferior to the filter diameter.The estimation is based on the Tween®80/P100 ratio of the preparation. 13Epi-MLLs elasticity was found to be 30 ± 3%.SC-MLLs elasticity was estimated to be similar: 32%.
Before testing MLLs ex vivo on explants, we verified in vitro their non toxicity on HaCat cell lines (keratinocytes cells lines; Supporting Information 1).
The pellet obtained after ultracentrifugation (see Section 2.2) was directly deposited on skin explants since its texture and viscosity are close to those of a cream (Supporting Information 2).MLLs concentration in the pellet has been estimated to be 13 g mL −1 , resulting in MLLs being in a near close-contact state.In previous papers, we explained that elastic MLLs could penetrate into the skin by a fusion of their outer membranes with endogenous lipids. 13,14gger elastic MLLs penetrate deeper into the skin because they exhibit more fusiogenic available material (Supporting Information 3).

| Preparations sun-screening efficacy
H&E staining revealed that the irradiated control shows typical sunburn features including nuclei amorphism, keratinocyte vacuolization, and epidermis/dermis detachment (Figure 1A).CC application 2 h prior to exposition induced a very good protection and prevented damage apparition.Both types of MLLs appeared to offer a similar protection as the commercial SPF50+ sunscreen since no feature of damaged cells and epidermis detachment were noticed.These results were confirmed by CPD assessment.CPD, that is, UV induced direct DNA damages, were evaluated by immunofluorescent staining.As shown Figure 1B, the non-irradiated control showed no sign of any CPD whereas many cell nuclei were damaged in the irradiated control.Consistently with H&E staining, all three preparations appeared to protect the skin from DNA lesions.However, observations suggested that Epi-MLLs offered a slightly less effective protection as few CPD were noticeable when barely any were observed after SC-MLLs or CC application.
Preparations were also applied only 10 min prior to exposition.H&E staining revealed a diminished efficacy of all preparations compared to 2 h prior to exposition depositions.However, SPF50+ sunscreen appeared to have a better performance than MLLs formulations under such conditions (Figure 2A).Indeed, 10 min MLLs protected skin showed very similar features as the irradiated control.
CPD staining brought complementary information.MLLs appeared to have a DNA protective effect that was not suggested by H&E imaging.Indeed, fewer CPD were observed when compared to the irradiated control (Figure 2B).CC CPD results were quite consistent with H&E staining results.

| Preparations adhesiveness on skin
To evaluate the different preparations adhesiveness, MLLs or waterresistant CC were applied 2 h prior to irradiation but rinsed off twice with 1 mL of PBS right before.Comparison of Figures 2 and 3 indicates that the filter quality of all preparations was diminished as expected.H&E staining reveals the presence of few keratinocytes vacuolization on skin explants protected with both types of MLLs.
Not rinsed off skin explants did not exhibit such features.CC filter quality appeared to be more affected than the MLLs' ones.Indeed, anomalous nuclei and more cell vacuolization were noticeable on CC H&E sections.Some local epidermis detachment could also be observed on the rinsed off CC protected skin as shown in Figure 3A.
Altogether, these results suggest a better adhesiveness of both MLLs compositions.Results from CPD staining confirmed these observations.MLLs better protected nuclei from DNA lesions as fewer CPD were noticed after rinsing off compared to the CC.This was particularly true for the SC-MLLs preparation where very few CPD damaged nuclei were noticeable despite rinsing off (Figure 3B).CPD staining reveals differences of protection between the two MLLs types after rinsing off that were not observable through H&E staining.

| Exposition and preparations impact on skin barrier
TEER was used as a skin barrier quality indicator.Measures were taken 24 h after irradiation.As expected, solar irradiation of skin samples resulted in a decrease of TEER (≈15%).Epi-MLLs application 2 h prior to exposition was sufficient to prevent this diminution (Figure 4).SC-MLLs application did not only prevent the diminution but even reinforced the skin barrier quality.A statistically significant 8.5% upswing in TEER value was measured on irradiated SC-MLLs protected skin explant compared to the non-irradiated control.Application of SC-MLLs on non-irradiated skin resulted in an equivalent TEER increased (Supporting Information 4).CC on the other hand diminished TEER value by comparison to the non-irradiated control (8.8%).TEER reduction was, however, less important than on the irradiated control, suggesting a protective effect of the sunscreen.
TEER measurements obtained for preparations deposited 10 min prior to irradiation followed the same tendency as 2 h prior to irradiation, but no statistically significant results could be drawn from our study (data not shown).
Rinsing off formula prior to irradiation did not imply any significant difference compared to not rinsed off formula in terms of TEER.
A slight overall TEER value diminution was nonetheless noticed under such conditions.

| DISCUSS ION
The purpose of this study was to investigate the potential of two MLLs formulations as new non-toxic biodegradable, 18,19  as in this study.They can also be easily added to an hydrogel or a water based cream as other hydrophilic nanoparticles. 20Evidence supports the idea that elastic MLLs penetration mechanism in skin depends on the fusion of their outer membranes with endogenous lipids. 13The two elastic MLLs preparations investigated, namely Epi-MLLs and SC-MLLs, target, respectively, the whole epidermis and the SC.Both have similar elasticities and zeta potentials but Epi-MLLs are about four times bigger than SC-MLLs.This difference of size explains the difference of depth penetration.
Epi-MLLs go deeper into the epidermis as they have more fusionavailable material.Our results demonstrated that both Epi-MLLs and SC-MLLs used alone had a filter efficacy similar to the one of a commercial SPF50+ sunscreen when applied 2 h prior to exposition.To the best of our knowledge, this is the first report of liposomal structures used directly as sun-screening agents and not as a sun-screening efficacy boosters. 9,12MLLs sun-screening action is explained both by their particle size that allows UV-light scattering and by their lecithin content. 21Low wavelength light scattering supposes a local decrease in the UV dose received during exposure and lecithin presents a large absorbance band roughly going from 200 to 380 nm with peaks at 235, 271, and 355 nm. 16It covers both the UVB (280-315 nm) and UVA (315-400 nm) regions of the light spectrum, which implies that lecithin could be considered as a natural sunscreen agent.Hence, lecithin-based MLLs offer both physical and chemical photoprotection and can be seen as ideal natural sun-screening agents.
A slightly better DNA protection was found using SC-MLLs suggesting that MLLs targeting has an impact on performance.SC-MLLs targeting suppose the formation of a more concentrated lecithin layer in the SC.Comparison between Epi-MLLs and SC-MLLs suggest that this higher concentration of exogeneous lipids in the SC allows a better UV protection than the more spread out exogenous lipids resulting from Epi-MLLs fusion in the whole epidermis. 14plication of MLLs 10 min prior to exposition resulted in a very low protection only observable through CPD staining.CC application under such conditions was also conducted to a filter quality diminution but to a lesser extent.This result can be explained by a faster skin entrance and diffusion of the CC compared to MLLs due to the numerous penetration enhancers found in the CC formula (pentylene glycol, alcohol, sorbitol, cetyl alcohol and propylene glycol, etc.). 22MLLs seem to require more time to properly diffuse into the skin and form an efficient shield against UV light.More experiments conducted in the lab suggest that a minimum time of 30 min is necessary to reach an effective protection when MLLs are deposited in vitro.However, we believe that this time would be considerably diminished in an in vivo experiment with a proper application of MLLs, that is, with a consumer-like rubbing of the formula.
Experimentation has indeed shown that rubbing increases significantly skin permeation of hydrophilic compounds. 23This inconvenient could also be worked around by encapsulating a small amount of a chemical UV filters within MLLs as they are efficient for encapsulation of a wide variety of molecules. 14,17Ls' better adhesiveness compared to a water-resistant CC was demonstrated by rinsing off preparations before irradiation.
These results are explained by the ability of MLLs to penetrate and diffuse either in the SC or the whole epidermis as we demonstrated in a previous paper. 13They are consistent with the results of Korting et al. who explained that liposomes promoted chemical filter entry in the uppermost layer of the skin making them difficult to wash out compared to classical formula that tend to form a more superficial film at the surface of the skin. 11Skin sequestration of MLLs makes them difficult to wash out in comparison.Rinsing off nevertheless induced a performance diminution of both MLLs preparations.The latter was less important for SC-MLLs than Epi-MLLs suggesting a faster skin entrance of the smaller size stratum corneum targeting composition.
Finally, TEER results confirmed that UV-light impairs skin barrier function.They also pointed out the tendency of commercial products to weaken consumers' skin barrier, as already shown by Gonzales-Bravo et al. on 51 healthy volunteers' cheek. 24The slight TEER improvement compared to the irradiated control, the CPD immunostaining, and H&E staining confirmed that this result could not be explained by a lack of efficacy from the CC sunscreen but rather by its excipient composition.As said previously, the tested CC contains many penetration enhancers that weaken the skin barrier function to favor products entry.They interact and fluidify endogenous and water-resistant sun-screening agents.MLLs are hydrophilic particles exhibiting a creamlike texture which can be used on their own such F I G U R E 1 Hematoxylin and eosin staining (A) and Cyclobutane Pyrimidine Dimer immunostaining (B) images of controls and irradiated skin explants protected with the different preparations 2 h before irradiation.CC, commercial cream; Epi-MLL, epidermis multi-lamellar liposome; SC-MLL, stratum corneum multi-lamellar liposome.F I G U R E 2 Hematoxylin and eosin staining (A) and Cyclobutane Pyrimidine Dimer immunostaining (B) images of irradiated skin explants protected with the different preparations 10 min before irradiation.CC, commercial cream; Epi-MLL, epidermis multi-lamellar liposome; SC-MLL, stratum corneum multi-lamellar liposome.F I G U R E 3 Hematoxylin and eosin staining (A) and Cyclobutane Pyrimidine Dimer immunostaining (B) images of irradiated skin explants protected with the different preparations.Preparations were incubated for 2 h and rinsed off right before irradiation.CC, commercial cream; Epi-MLL, epidermis multi-lamellar liposome; SC-MLL, stratum corneum multi-lamellar liposome.