Exogenous Ceramide Serves as a Precursor to Endogenous Ceramide Synthesis and as a Modulator of Keratinocyte Differentiation

Since ceramide is a key epidermal barrier constituent and its deficiency causes barrier-compromised skin, several molecular types of ceramides are formulated in commercial topical agents to improve barrier function. Topical ceramide localizes on the skin surface and in the stratum corneum, but certain amounts of ceramide penetrate the stratum granulosum, becoming precursors to endogenous ceramide synthesis following molecular modification. Moreover, exogenous ceramide as a lipid mediator could modulate keratinocyte proliferation/differentiation. We here investigated the biological roles of exogenous NP (non-hydroxy ceramide containing 4-hydroxy dihydrosphingosine) and NDS (non-hydroxy ceramide containing dihydrosphingosine), both widely used as topical ceramide agents, in differentiated-cultured human keratinocytes. NDS, but not NP, becomes a precursor for diverse ceramide species that are required for a vital permeability barrier. Loricrin (late differentiation marker) production is increased in keratinocytes treated with both NDS and NP vs. control, while bigger increases in involucrin (an early differentiation marker) synthesis were observed in keratinocytes treated with NDS vs. NP and control. NDS increases levels of a key antimicrobial peptide (an innate immune component), cathelicidin antimicrobial peptide (CAMP/LL-37), that is upregulated by a ceramide metabolite, sphingosine-1-phosphate. Our studies demonstrate that NDS could be a multi-potent ceramide species, forming heterogenous ceramide molecules and a lipid mediator to enhance differentiation and innate immunity.

Stimulation of endogenous Cer production is a therapeutic strategy used to improve epidermal permeability barrier integrity. Yet, compounds to specifically promote Cer production have not yet been developed. Since increases in precursor levels increase Cer production [26], the supplementation of Cer and its precursor is an achievable approach to elevate epidermal Cer levels. In theory, dihydroCer (NDS) is converted into diverse types of Cer, while NS can only be a precursor of Cer species containing S ( Figure 1B). Similarly, NP can only be a precursor of Cer species containing P ( Figure 1B). However, the fate of exogenous Cer in the epidermis has not yet been elucidated, nor has the role(s) of exogenous Cer in the modulation of KC proliferation/differentiation.
The amount of exogenous Cer absorption into the epidermis is dependent upon the barrier integrity, and the penetration of exogenous Cer is relatively low in barrier-competent (normal) skin [27]. Recent studies have aimed to increase Cer absorption to increase the efficacy of Cer to further improve barrier function [28,29], suggesting that exogenous Cer can be absorbed into nucleated layers of the normal epidermis. Hence, elucidating the fate of exogenous Cer in the epidermis is important to understand whether exogenous Cer modulates KC differentiation, as well as contributes to barrier formation.
Here, we aim to define the fates of exogenous Cer, the contribution of exogenous Cer to increase Cer levels in differentiated KC, and the effect of exogenous Cer on KC differentiation. As above, several types of Cer species are used in topical agents. Because NDS and NP are widely used as skincare ingredients, we employed these Cer in this study.

Cell Culture
Human primary keratinocytes (KCs) from Life Technologies (Carlsbad, CA, USA) were maintained in serum-free KC growth medium containing 0.07 mM Ca 2+ . Cells at 60-70% confluence were further cultured in a differentiation-inducing medium in two consecutive steps as described ( Figure S1). First, KC were incubated in DMEM and Ham's F-12 (2:1, v/v) containing 1.2 mM calcium, 10% FBS, and 10 µg/mL insulin, along with vitamin C (50 µg/mL) for 8 days with Cer, as described previously [30,31]. Cells were then incubated in DMEM containing 1.2 mM calcium, 10% FBS, and 10 µg/mL insulin, along with vitamin C (50 µg/mL) for 2 days incubated with Cer. Synthesized 17NDS or 17NP were dissolved in 95% aqueous ethanol in a sonication bath for 30 min at 40 • C at a concentration of 20 mM prior to each treatment.

Cell Viability Assay
Cell viability or cytotoxicity was determined by the water-soluble tetrazolium salt (WST) method using the Cell Counting Kit-8 (CCK-8, Dojindo, Japan) in accordance with the manufacturer's instructions.
Protein content was determined using the BCA protein assay method (Pierce, Rockford, IL, USA), using bovine serum albumin as the standard.

Western Blot Analysis
Western blot analysis was performed, as described previously [33]. Briefly, cell lysates, prepared in radioimmunoprecipitation assay buffer (RIPA Lysis and Extraction Buffer), were resolved by electrophoresis on 4-12% Bis-Tris protein gel (Invitrogen, Waltham, MA, USA) under denaturing conditions using SDS. All procedures are conducted following the manufacturer's instructions. Resultant bands were blotted onto polyvinylidene difluoride membranes, probed with anti-human keratin 10 (Santa Cruz Biotechnology, Dallas, TX, USA), anti-human involucrin Abcam (Cambridge, UK), or anti-human β-actin (Sigma-Aldrich), and detected using enhanced chemiluminescence (Thermo Fisher Scientific, Waltham, MA, USA). The intensity of bands was measured with a LAS-3000 (Fujifilm, Tokyo, Japan).

Statistical Analysis
Statistical analysis was performed by an unpaired Student's t-test. The p-values were set at < 0.01.

Conversion of 17NDS and 17NP to Other Cer Species
Heterogenous Cer species are synthesized in the late stages of differentiated KCs. An organotypic reconstituted epidermal model (3D KC cultures) can recapture this het-erogenous Cer production. However, we previously established a submerged cultured KC model, which recaptures the epidermis (a quasi-cultured epidermal model), i.e., KC consisting of proliferating and early and late stages of differentiated KC (which are stratified [piled-up]) [30,31,35]. Heterogenous Cer are then produced (a quasi-epidermal model) [30,31,35]. We confirmed the formation of lamellar membrane structures and corneocyte lipid envelopes, both present in the stratum corneum [31]. The advantage of this cultured system is that large amounts of cells are relatively easy to obtain and reproduce. In addition, because insertion into a membrane, used for the conventional 3D KC model, is not used here, cells can be observed under microscopy. We employed this established KC culture system [30,31,35] in this current study. KCs started to incubate with 17NDS and 17NP two days after culturing in the late-stage differentiation induction medium ( Figure S1). Increases in K10 keratin (early KC differentiation marker) [36] expression and involucrin (middle KC differentiation marker) and loricrin (late KC differentiation marker) production [30] became evident in cells following the switch to a differentiation induction medium. These results confirm that KC used in these studies is fully differentiated and is capable of synthesizing heterogenous Cer species.
Note that, following molecular remodeling, i.e., conversion to various sphingoid bases and amide-linked fatty acids, certain amounts of Cer are converted to glycosylated Cer (mainly glucosylCer), omega-O-acylglucosylCers, and sphingomyelin.

Alterations of Cer Levels in KC Incubated with NDS and NP
Cer production is significantly increased in the late stage of differentiated KCs [31,37,38]. Since it is uncertain whether exogenous Cer contributes to further increases in epidermal Cer levels, we next investigated this in KCs using N-stearoyl-D-erythro-dihydrosphingosine (18NDS) and N-stearoyl-D-erythro-4-hydroxydihydrosphingosine (18NP), which are two of the endogenous Cer species in KC.
Total Cer (NDS, NP, and NS) was significantly increased in KC treated with both NDS and NP (250 µM), while statistically significant increases in NS, NP, and NS were observed in KC incubated with 50 µM of NDS ( Figure 3). Only NP was significantly increased in KC treated with incubated 50 µM NP. Moreover, both EOS and EOP are significantly increased in cells treated with NDS, while NP increased EOP, but not EOS ( Figure 3D,E). S, ND, and P were modestly increased in KC treated with 18NS and 18NP, respectively ( Figure 3F). These results suggest that exogenous Cer (>50 µM) is capable of further elevating Cer and its metabolite levels in the late stage of differentiated KCs. Thus, the supplementation of exogenous Cer should be effective in increasing epidermal Cer levels.

Modulation of KC Differentiation by Exogenous Cer
We next investigated whether exogenous NDS and NP modulate KC differentiation. Many protein levels, including structural proteins and enzymes, are changed during KC differentiation. Since the aim of our study is to describe the effect of exogenous Cer on KC differentiation, we measured established KC differentiation markers' protein, i.e., keratin 10, involucrin, and loricrin in early, mid, and late stages of KC differentiation, Moreover, both EOS and EOP are significantly increased in cells treated with NDS, while NP increased EOP, but not EOS ( Figure 3D,E). S, ND, and P were modestly increased in KC treated with 18NS and 18NP, respectively ( Figure 3F). These results suggest that exogenous Cer (>50 µM) is capable of further elevating Cer and its metabolite levels in the late stage of differentiated KCs. Thus, the supplementation of exogenous Cer should be effective in increasing epidermal Cer levels.

Modulation of KC Differentiation by Exogenous Cer
We next investigated whether exogenous NDS and NP modulate KC differentiation. Many protein levels, including structural proteins and enzymes, are changed during KC differentiation. Since the aim of our study is to describe the effect of exogenous Cer on KC differentiation, we measured established KC differentiation markers' protein, i.e., keratin 10, involucrin, and loricrin in early, mid, and late stages of KC differentiation, respectively. Both keratin 10 and involucrin protein levels are increased in vehicle-treated KC on day 3, while loricrin levels are increased on day 5 ( Figure 4A). Increases in loricrin levels in cells treated with both NDS and NP were higher than the control, while involucrin levels were higher in NDS-treated cells at 3 and 5 days vs. NP and control (Day 3: 1.9-fold and 1.4-fold, vs. vehicle and NP, respectively) and Day 5 (2.8-fold and 1.3-fold, respectively) ( Figure 4A).  Figure 4A). These results suggest that both exogenous NDS and NP, and/or their metabolite(s) can promote KC differentiation. In addition, NDS and/or its metabolites likely specifically stimulate involucrin production independent of global stimulation of differentiation.

Discussion
Our study using 17NDS and 17NP characterized the metabolic fate of exogenous Cer in the late stages of differentiated KC as helping to recapture features of healthy epidermis, including synthesizing diverse Cer species required for epidermal permeability barrier formation [3,4]. We demonstrated that structural (compositional) remodeling of exogenous 17NDS and 17NP occurs in KC. Exogenous Cer is hydrolyzed These results suggest that both exogenous NDS and NP, and/or their metabolite(s) can promote KC differentiation. In addition, NDS and/or its metabolites likely specifically stimulate involucrin production independent of global stimulation of differentiation.

Discussion
Our study using 17NDS and 17NP characterized the metabolic fate of exogenous Cer in the late stages of differentiated KC as helping to recapture features of healthy epidermis, including synthesizing diverse Cer species required for epidermal permeability barrier formation [3,4]. We demonstrated that structural (compositional) remodeling of exogenous 17NDS and 17NP occurs in KC. Exogenous Cer is hydrolyzed to the sphingoid base and FA, and the generated sphingoid base and a pool of FA in cells become precursors to Cer synthesis. Cer species consisting of different N-acyl chain lengths of FA and P and P1P are produced from NP, while, as expected, diverse species of Cer can be produced from NDS, but not NP. Although Cer species containing ß-hydroxy FA were not analyzed, as shown in Figure 1, BS can be produced by the metabolic conversion of N-acyl FA. Yet, since 6-hydroxy sphingosine and 4,14-sphingadiene synthetic enzymes are not completely identified, it remains to be resolved if NH and NSD are generated from NDS.
Since Cer production is significantly elevated in the late stage of differentiated KCs, it is possible that negligible amounts of Cer are produced from exogenous Cer. We confirmed that exogenous Cer further increases Cer levels, suggesting the pharmacological relevance of exogenous Cer to elevate Cer levels in the late stages of KC. We previously found that S1P (P1P is less potent vs. S1P [unpublished data]) stimulates the synthesis of a key epidermal antimicrobial peptide, CAMP [30,33]. We here showed that exogenous NDS increases CAMP production, suggesting the amounts of S1P from NDS (NDS→NS→S→S1P and NDS→DS→NDS→NS→S→S1P) are enough to signal CAMP production. These results further suggest that the exogenous application of Cer suffices to increase biologically relevant levels of Cer and its metabolites in cells. In particular, NDS has the ability to increase Cer production at a lower concentration (50 µM). Thus, suitably formulated Cer, which enhances the absorption of Cer [29], is useful for increased endogenous Cer production, leading to improving epidermal permeability barrier integrity and antimicrobial barrier defenses.
Cer and its metabolites modulate cellular proliferation and differentiation in KC [24,43,44]. Kim et al. reported that P promotes differentiation in KC through the transactivation of PPAR [45]. Because NP is decreased in atopic dermatitis and psoriasis, P was a focus of their study [45]. Sigruener showed that not only P but also ND are potent sphingoid bases that induce KC differentiation [46]. Taken together with our current study, not only P but also other sphingoid bases are able to increase KC differentiation. We showed here that NDS is more effective in increasing involucrin levels (Figure 4), and NDS and its metabolite(s) may specifically upregulate involucrin. Although it remains to be resolved (1) which molecule (either Cer or its metabolites) it is and (2) how this particular molecule (derived from NDS) promotes involucrin synthesis, our results suggest that specific step(s) or protein(s) associated with differentiation are regulated by Cer and/or its metabolites.
Like NP, NS could be a precursor for Cer-containing S synthesis and stimulate KC differentiation. However, relative amounts of NS are increased in atopic dermatitis. In addition, we previously reported that a ratio of S to DS is increased in the SC of an atopic dermatitis mouse model, and changes in the ratio of S to ND alter the packing of lamellar bilayers, i.e., increases in the ratio of S to DS decrease tightly packed orthorhombic structure [47]. Because vital acidic and alkaline ceramidase are present in the SC [48,49], DS produced from NDS may modulate the ratio of S/ND related to the packing of lamellar bilayers, and may result in increasing the integrity of the lamellar membrane in the SC [47]. Conversely, S generated from NS may affect orthorhombic structures in SC to attenuate barrier integrity.
The excessive production of Cer induces apoptosis in cells, including in KCs [43], but we did not see any toxic effects using exogenous NDS and NP (up to 250 µM) in our KC studies. In addition, apoptotic activity through Cer channel formation of NS is stronger than with NDS [50] (note that P was not tested). Thus, NDS might be less toxic in cells. However, since NS is the Cer species used for most studies that investigate apoptotic effects of Cer in cells/tissues, the deleterious effects of NDS have not been completely elucidated [51]. Hence, optimal amounts of any Cer in formulations should be carefully considered to maximize positive outcomes in vivo.
Insights gained from our study propose that topical Cer agents are classified into two groups. Group 1 includes Cer agents that improve the epidermal barrier on the skin surface and in the SC. Group 2 includes Cer agents that work in KCs in nucleated layers, i.e., endogenous Cer synthesis increases via increases in Cer synthesis precursors and by promoting KC differentiation. Cer agents used for barrier-compromised skin and formulated with skin absorption enhancers are categorized in Group 2. Groups 1 and 2 can also be classified as first and second generations of topical Cer agents, respectively ( Figure 5).
Our study could become a basis for developing topical agents containing Cer. Since we now know the fate of exogenous Cer and the effect of exogenous Cer on KC differentiation, we could test the effects of exogenous Cer in an appropriate formulation (that enhances the penetration of Cer into the stratum corneum) on epidermal barrier function. Using in vivo/ex vivo human skin. Our study could become a basis for developing topical agents containing Cer. Since we now know the fate of exogenous Cer and the effect of exogenous Cer on KC differentiation, we could test the effects of exogenous Cer in an appropriate formulation (that enhances the penetration of Cer into the stratum corneum) on epidermal barrier function. Using in vivo/ex vivo human skin.

Conclusions
Increased epidermal Cer is a therapeutic strategy to improve epidermal permeability barrier function. We demonstrated that exogenous NDS has a greater ability as a precursor to diverse Cer species required for the vital permeability barrier production than exogenous NP has. Moreover, although both NDS and NP and their metabolites enhance KC differentiation and increase endogenous Cer, NDS increases S1P levels, leading to upregulated cathelicidin (innate immune component) production. Because NDS is a minor component of the SC, this Cer species has not previously received a great deal of attention. Our current study demonstrates that NDS could be a multi-potent Cer species for epidermal barrier formation and a lipid mediator that stimulates KC differentiation and enhances innate immunity ( Figure S4).

Supplementary Materials:
The following supporting information can be downloaded at: www.mdpi.com/xxx/s1, Figure S1: Culture Scheme. Figure S2: Outline of C17NDS and C17 NP preparation and verification by LC-MS/MS. Figure S3: HPLC choromatograms syntheisized

Conclusions
Increased epidermal Cer is a therapeutic strategy to improve epidermal permeability barrier function. We demonstrated that exogenous NDS has a greater ability as a precursor to diverse Cer species required for the vital permeability barrier production than exogenous NP has. Moreover, although both NDS and NP and their metabolites enhance KC differentiation and increase endogenous Cer, NDS increases S1P levels, leading to upregulated cathelicidin (innate immune component) production. Because NDS is a minor component of the SC, this Cer species has not previously received a great deal of attention. Our current study demonstrates that NDS could be a multi-potent Cer species for epidermal barrier formation and a lipid mediator that stimulates KC differentiation and enhances innate immunity ( Figure S4).