The Herbal Bitter Drug Gentiana Lutea Modulates Lipid Synthesis in Human Keratinocytes in Vitro and in Vivo

International Journal of Molecular Sciences

Article The Herbal Bitter Drug Gentiana lutea Modulates Lipid Synthesis in Human Keratinocytes In Vitro and In Vivo

Ute Wölﬂe 1,* ID , Birgit Haarhaus 1, Jasmin Seiwerth 1, Anja Cawelius 2, Kay Schwabe 3, Karl-Werner Quirin 2 and Christoph M. Schempp 1

1 Department of Dermatology, Medical Center, Faculty of Medicine, University of Freiburg, 79085 Breisgau, Germany; [email protected] (B.H.); [email protected] (J.S.); [email protected] (C.M.S.) 2 Flavex Naturextrakte GmbH, 66780 Rehlingen, Germany; ac@flavex.com (A.C.); wq@flavex.com (K.-W.Q.) 3 BSI-Beauty Science Intelligence GmbH, 30855 Langenhagen, Germany; [email protected] * Correspondence: ute.woelfl[email protected]; Tel.: +49-761-2706-8250; Fax: +49-761-2706-7831

Received: 28 June 2017; Accepted: 15 August 2017; Published: 22 August 2017

Abstract: Gentiana lutea is a herbal bitter drug that is used to enhance gastrointestinal motility and secretion. Recently we have shown that amarogentin, a characteristic bitter compound of Gentiana lutea extract (GE), binds to the bitter taste receptors TAS2R1 and TAS2R38 in human keratinocytes, and stimulates the synthesis of epidermal barrier proteins. Here, we wondered if GE also modulates lipid synthesis in human keratinocytes. To address this issue, human primary keratinocytes were incubated for 6 days with GE. Nile Red labeling revealed that GE significantly increased lipid synthesis in keratinocytes. Similarly, gas chromatography with flame ionization detector indicated that GE increases the amount of triglycerides in keratinocytes. GE induced the expression of epidermal ceramide synthase 3, but not sphingomyelinase. Lipid synthesis, as well as ceramide synthase 3 expression, could be specifically blocked by inhibitors of the p38 MAPK and PPARγ signaling pathway. To assess if GE also modulates lipid synthesis in vivo, we performed a proof of concept half side comparison on the volar forearms of 33 volunteers. In comparison to placebo, GE significantly increased the lipid content of the treated skin areas, as measured with a sebumeter. Thus, GE enhances lipid synthesis in human keratinocytes that is essential for building an intact epidermal barrier. Therefore, GE might be used to improve skin disorders with an impaired epidermal barrier, e.g., very dry skin and atopic eczema.

Keywords: keratinocytes; Gentiana lutea; lipid synthesis; p38 MAPK; PPARγ; ceramide synthase 3

1. Introduction The stratum corneum consists of terminally differentiated keratinocytes (corneocytes) that are surrounded by an extracellular lipid bilayer and bound water. Lipids for this bilayer are secreted from lipid droplets, which are intracellular organelles that are specialized in assembling, storing and then supplying lipids [1]. Although sebocytes are specialized in lipid synthesis and storage, almost every cell has the capacity to store and accumulate lipids in lipid droplets [2]. In keratinocytes, lipid droplets are generated when keratinocytes differentiate and reach their maximal number in the stratum granulosum, where also, the highest level of extracellular calcium exists [3]. There are various lipids that contribute to an optimal epidermal barrier function, of which ceramides, the backbone structure of all sphingolipids, are particular important. They represent nearly 50% of the lipid mass in the stratum corneum [4,5]. A decrease in the amount of ceramides occurs in multiple skin disorders with barrier abnormalities, for example, atopic dermatitis and psoriasis [6,7].

Int. J. Mol. Sci. 2017, 18, 1814; doi:10.3390/ijms18081814 www.mdpi.com/journal/ijms Int. J. Mol. Sci. 2017, 18, 1814 2 of 16

Altered expression of key enzymes that are involved in the ceramide metabolism in keratinocytes might account for this, e.g., ceramide synthase 3 (CerS3) and acidic sphingomyelinase. CerS3 is a crucial enzyme for the synthesis of very long-chain ceramides in the epidermis. These ceramides are important for maintaining epidermal lipid homeostasis and terminal differentiation [8]. Acidic sphingomyelinases catalyze the hydrolysis of sphingomyelin (ceramide phosphorylcholine) into ceramide and phosphorylcholine, and have also been related to cellular processes, such as apoptosis, cell proliferation, cell differentiation, and membrane fusion/fission [9]. Furthermore, during aging, stratum corneum ceramides decrease by approximately 10–15% per decade after the age of 20 years [10]. Aging also influences the composition and amount of free fatty acids [11]. A deficiency in fatty acids (e.g., by delayed fatty acid synthesis) results in abnormalities in the stratum corneum structure and delayed recovery of the permeability barrier. A complete skin barrier requires not only an intact lipid layer but also terminal differentiation proteins (see Figure1).

Figure 1. Illustration of the differentiation process in keratinocytes. The triangles indicate increasing or decreasing concentrations of the respective molecules within the epidermis. Bcl-2—B-cell lymphoma 2; BAK—Bcl-2 antagonist/killer1; MDM2—mouse double minute 2 homolog; Ca2+—calcium. The graphic was adapted from [12].

Generally, differentiation of keratinocytes is induced by an increase of the intracellular calcium concentration. Recently, we could show that amarogentin, a characteristic bitter compound of Gentiana lutea, induces calcium influx in keratinocytes and promotes keratinocyte differentiation by inducing keratin 10, involucrin, and transglutaminase expression [13]. Additionally, we demonstrated that human keratinocytes express the bitter taste receptors TAS2R1 and TAS2R38 that can be activated by amarogentin [14]. Meanwhile, more effects of bitter compounds become apparent, as the expression of TAS2Rs could be detected in various extra-gustatory organs. The signaling pathway of bitter compounds is dependent on the particular cell type. Deshpande and colleagues showed that activated TAS2Rs expressed on bronchial smooth muscle cells eventually lead to bronchial relaxation [15]. In airway epithelial cells, bitter compounds induce the synthesis of NO, and increase the ciliary beat frequency [16]. In gustatory cells, intracellular calcium influx induces transmitter release [17], and in keratinocytes, ligand binding induces calcium influx and the expression of differentiation proteins [12,13]. In this work, we addressed the question if the TAS2R agonist Gentiana lutea extract (GE) may improve the skin barrier by modulating the lipid metabolism in keratinocytes apart from inducing differentiation.

Int. J. Mol. Sci. 2017, 18, 1814 3 of 16

2. ResultsInt. J. Mol. Sci. 2017, 18, 1814 3 of 16

2.1. GE2. Results Enhances Lipid Synthesis in Keratinocytes

To2.1. analyze GE Enhances if GE Lipid has anSynthesis impact in on Keratinocytes lipid metabolism in keratinocytes, we performed a quantitative fluorescenceTo assayanalyze with if GE Nile has Red. an impact GE dose on dependentlylipid metabolism enhanced in keratinocytes, neutral lipid we accumulationperformed a in HaCaTquantitative keratinocytes. fluorescence At 200 µassayg/mL with there Nile was Red. a statistically GE dose dependently significant increase enhanced in lipidneutral production lipid of GE-treatedaccumulation compared in HaCaT to keratinocytes. sham treated At cells200 µg/mL (p ≤ there0.05; was Figure a statistically2A). Next, significant 6 samples increase of human in primarylipid keratinocytes production of (hPKs) GE-treated were compared incubated to sham with GEtreated (200 cellsµg/mL) (p ≤ 0.05; for Figure 6 days. 2A). In Next, GE treated 6 samples hPKs, a significantlyof human enhanced primary neutral keratinocytes lipid accumulation (hPKs) were incu wasbated measured with GE (p ≤(2000.001; µg/mL) Figure for2 B).6 days. No toxic In GE effect of GEtreated could behPKs, detected a significantly in hPKs enhanced as examined neutral with lipid an ATPaccumulation assay (Figure was measured2C). The increased(p ≤ 0.001; numberFigure of lipid droplets2B). No toxic in hPKs effect afterof GE GEcould treatment be detected can in also hPKs be as seen examined in the with fluorescence an ATP assay microscope (Figure 2C). after The Nile increased number of lipid droplets in hPKs after GE treatment can also be seen in the fluorescence Red labeling (Figure2D). microscope after Nile Red labeling (Figure 2D).

Figure 2. Gentiana lutea extract (GE) enhances lipid synthesis in hPK. (A–D) HPKs were treated with Figure 2. Gentiana lutea extract (GE) enhances lipid synthesis in hPK. (A–D) HPKs were treated with 200 µg/mL GE (unless otherwise stated) for 6 days. The graph in (A,B) reports the quantitative µ 200 g/mLmeasurement GE (unless of intracellular otherwise neutral stated) lipids for as 6 asse days.ssed Theaftergraph Nile Red in labeling. (A,B) reports Graph theC shows quantitative cell measurementviability ofas intracellularassessed with neutral the ViaLight lipids asPlus assessed bioassay after kit; Nile Graph Red ( labeling.D) shows Graph the qualitative C shows cell viabilitymeasurement as assessed of withintracellular the ViaLight neutral Plus lipids bioassay after labeled kit; Graph by Nile (D Red;) shows the nuclei the qualitative were counterstained measurement of intracellularwith DAPI. neutralThe white lipids arrows after point labeled to lipid by droplets. Nile Red; (ns: thenot nucleisignificant, were * p counterstained < 0.05; ** p < 0.01, with *** p DAPI.< The white0.001). arrows The pictures point were to lipid photographed droplets. at (ns: a magnification not signiﬁcant, of 400×. * pThe< 0.05;bar indicates ** p < 0.01,10 µm. *** p < 0.001). The pictures were photographed at a magniﬁcation of 400×. The bar indicates 10 µm. Int. J. Mol. Sci. 2017, 18, 1814 4 of 16

Int. J. Mol. Sci. 2017, 18, 1814 4 of 16 The content of fatty acids (newly synthesized or separated from ceramides) in HaCaT cells and hPKsThe was content determined of fatty byacids gas (newly chromatography synthesized or with separated Flame from Ionization ceramides) Detector in HaCaT 96 (GC-FID) cells and and comparedhPKs was with determined data from by the gas literature. chromatography The percentage with Flame of fattyIonization acids Detector and the 96 fatty (GC-FID) acid proﬁles and of hPKscompared isolated fromwith data biopsies from werethe literature. comparable The topercenta culturedge of keratinocytes fatty acids and (see the Tablefatty 1acid). profiles of hPKs isolated from biopsies were comparable to cultured keratinocytes (see Table 1). Table 1. The composition of fatty acids isolated from hPKs or the keratinocyte cell line HaCaT, as well as Table 1. The composition of fatty acids isolated from hPKs or the keratinocyte cell line HaCaT, as from isolated human epidermis keratinocytes and skin models is indicated as percentage. Own GC-FID well as from isolated human epidermis keratinocytes and skin models is indicated as percentage. data wereOwn GC-FID compared data with were the compared data from with the the literature. data from the literature.

% Fatty% Fatty Acids Acids% Fatty% Fatty Acids Acids % Fatty Fatty Acids Acids in in Freshly Freshly % Fatty% Acids Fatty in Acids in Fatty AcidFatty Acid in hPKsin hPKs in HaCaTin HaCaT Cells Cells Isolated Keratinocytes Keratinocytes of of Skin ModelsSkin Models [18] [18] (This(This Study) Study) (This(This Study) Study) Human Epidermis Epidermis [18] [18 ] C16:0 15.5% 10.8% 15.8% 22.2% C16:0C16:1 w7 15.5% 4.5% 10.8% 0.83% 1.8% 15.8% 13.1% 22.2% C16:1 w7C18:0 4.5%10% 0.83% 7.8% 14.8% 1.8% 10.8% 13.1% C18:0C18:1 w9 10% 11.8% 7.8% 10% 16.6% (only 14.8%marked as18:1) 45.8% 10.8% C18:1 w9C18:2 w6 11.8% 11.8% 10% 17.5% 23.3% 16.6% (o (onlynly marked marked as C18:2) as18:1) 1.1% 45.8% C18:2 w6 11.8% 17.5% 23.3% (only marked as C18:2) 1.1% There were some differences in C16:1 or C18:2 expressions. This might be caused by the cell Thereculture weremedium. some Ponec differences and colleagues in C16:1 described, or C18:2 for expressions. example, that This fetal might calf beserum-containing caused by the cell culture medium contains low levels of C18:2, and high levels of C16:1, apart from individual culture medium. Ponec and colleagues described, for example, that fetal calf serum-containing culture differences within different hPKs [18]. Nevertheless, the results are comparable and provide a medium contains low levels of C18:2, and high levels of C16:1, apart from individual differences suitable model to study epidermal lipid metabolism. Eight samples of hPKs were incubated with GE within(200 different µg/mL), hPKs and showed [18]. Nevertheless, a significantly the increased results amount are comparable of fatty acids and (p provide < 0.01) (Figure a suitable 3). hPKs model to studyeither epidermal responded lipid to metabolism. GE stimulation, Eight or showed samples no of increase hPKs wereof fatty incubated acids (Figure with 3, GE second (200 graph).µg/mL), and showed aThe signiﬁcantly reason why increased some hPKs amount do not respond of fatty to acids GE might (p < 0.01) be the (Figure occurrence3). hPKs of polymorphisms either responded in to GE stimulation,bitter taste receptors or showed [19]. no increase of fatty acids (Figure3, second graph).

FigureFigure 3. GE 3. increasesGE increases the the amount amount of of fatty fatty acidsacids inin vitr vitroo. hPKs. hPKs were were treated treated with with 200 µg/mL 200 µg/mL GE for GE6 for days, and the lipid fraction was isolated and analyzed by GC-FID. The results were shown in a 6 days, and the lipid fraction was isolated and analyzed by GC-FID. The results were shown in a scatter scatter blot (left graph), and in an untreated versus treated graph that shows the individual treatment blot (left graph), and in an untreated versus treated graph that shows the individual treatment effect effect within cells of the same skin biopsy (right graph). The p value is indicated in the figure by an withinasterisk cells of(* p the < 0.05). same skin biopsy (right graph). The p value is indicated in the ﬁgure by an asterisk (* p < 0.05). GE-induced lipid production was seen in both young and old keratinocytes (Supplementary TheTable reason S1A,B; why Supplementary some hPKs Figure do not S1). respond In particular, to GE might the fatty be the acids occurrence palmitic ofacid polymorphisms (C16:0) and in bitterlinoleic taste receptors acid (C18:2) [19 ].were increased after GE treatment (Supplementary Table S1) and showed an average increase of these fatty acid between 2 -and 4-fold. The absolute values vary between 0.02 and GE-induced lipid production was seen in both young and old keratinocytes (Supplementary 1.3 mg/mL of these fatty acids. Table S1A,B; Supplementary Figure S1). In particular, the fatty acids palmitic acid (C16:0) and linoleic acid (C18:2) were increased after GE treatment (Supplementary Table S1) and showed an average increase of these fatty acid between 2 -and 4-fold. The absolute values vary between 0.02 and 1.3 mg/mL of these fatty acids. Int. J. Mol. Sci. 2017, 18, 1814 5 of 16

2.2. PathwaysInt. J. Mol. Sci. Involved 2017, 18, 1814 in GE-Induced Lipid Production 5 of 16

2.2.Peroxisome Pathways Involved proliferator in GE-Induced activated Lipid receptors Production (PPARs) play a crucial role in lipid and glucose homeostasis. Rau and colleagues showed that PPARγ can be activated by GE [20]. Therefore, Peroxisome proliferator activated receptors (PPARs) play a crucial role in lipid and glucose we tested if a PPARγ antagonist (GW9962) can inhibit the lipid synthesis in hPKs. GE-induced homeostasis. Rau and colleagues showed that PPARγ can be activated by GE [20]. Therefore, we lipidtested synthesis if a PPAR in hPKsγ antagonist could be (GW9962) antagonized can inhibit by GW9962 the lipid (p synthesis< 0.05; Figure in hPKs.4A). GE-induced To elucidate lipid further signalingsynthesis pathways in hPKs possibly could be involved antagonized in GE-induced by GW9962 lipid(p < synthesis,0.05; Figure we 4A). used To severalelucidate inhibitors further that are involvedsignaling pathways in intracellular possibly signaling involved in GE-induced human keratinocytes, lipid synthesis, e.g., we phosphatidyl-inositol-3-kinase used several inhibitors that (PI3K)are inhibitorinvolved in (LY294002), intracellular inhibitors signaling in of human the MAP-kinase keratinocytes, pathway e.g., phosphatidyl-inositol-3-kinase (ERK inhibitor (PD98059) and p38(PI3K) mitogen-activated inhibitor (LY294002), protein inhibitors kinase (MAPK) of the MAP-kinase inhibitor (SB203580)),pathway (ERK or inhibitor NFκB inhibitor (PD98059) (GIV and 3727). As seenp38 inmitogen-activated Figure4A, GE-induced protein kinase lipid synthesis(MAPK) inhibitor was not (SB203580)), modiﬁed by or the NF PI3KκB inhibitor inhibitor, (GIV ERK 3727). inhibitor or NFAsκ Bseen inhibitor in Figure (Figure 4A, GE-induced4A). In contrast, lipid synthesi the PPARs wasγ inhibitor,not modified as well by the as thePI3K p38 inhibitor, MAPK ERK inhibitor, inhibitor or NFκB inhibitor (Figure 4A). In contrast, the PPARγ inhibitor, as well as the p38 MAPK signiﬁcantly antagonized the effect of GE-induced lipid production (p ≤ 0.01; Figure4A). These inhibitor, significantly antagonized the effect of GE-induced lipid production (p ≤ 0.01; Figure 4A). inhibitors showed no cytotoxic effects in hPKs (Figure4B). This suggests an important role of the These inhibitors showed no cytotoxic effects in hPKs (Figure 4B). This suggests an important role of MAPKthe MAPK p38 and p38 PPAR and γPPARpathwayγ pathway in GE-induced in GE-induced keratinocyte keratinocyte lipid lipid synthesis. synthesis.

FigureFigure 4. 4. GE-inducedGE-induced lipid lipid synthesis synthesis involves involves the MAPK the and MAPK PPARand pathway. PPAR hPKs pathway. were cultured hPKs in were 96 well plates and were treated with vehicle, GE (200 µg/mL), or with GE and one of the inhibitors cultured in 96 well plates and were treated with vehicle, GE (200 µg/mL), or with GE and one (p38 MAPK inhibitor SB203580, PPARγ antagonist GW9962; NFκB inhibitor GIV 3727, of the inhibitors (p38 MAPK inhibitor SB203580, PPARγ antagonist GW9962; NFκB inhibitor GIV phosphatidyl-inositol-3-kinase inhibitor LY294002 or ERK inhibitor PD98059) for 6 days. (A) 3727,Quantitative phosphatidyl-inositol-3-kinase measurement of intracellular inhibitor neutral LY294002 lipids as assessed or ERK after inhibitor Nile Red PD98059) labeling and for (B 6) days. (A)measurement Quantitative of measurement cell viability as of was intracellular assessed with neutral the ViaLight lipids as Plus assessed bioassay after kit. NileAll Data Red are labeling andexpressed (B) measurement as means ± ofSD cellof 6 viabilityindependent as wasexperiments assessed with with 8 replicates the ViaLight respectively Plus (ns—not bioassay kit. All Datasignificant, are expressed ** p < 0.01). as means ± SD of 6 independent experiments with 8 replicates respectively (ns—not signiﬁcant, ** p < 0.01).

Int. J. Mol. Sci. 2017, 18, 1814 6 of 16

2.3. GE Modulates the Activity of Ceramide Metabolic Enzymes Int. J. Mol. Sci. 2017, 18, 1814 6 of 16 To investigate if GE inﬂuences ceramide synthesis, we measured the expression and activity of 2.3. GE Modulates the Activity of Ceramide Metabolic Enzymes key enzymes that are involved in ceramide metabolism, e.g., CerS3 and acidic sphingomyelinase. ImmunohistochemicalTo investigate if stainingGE influences showed ceramide that CerS3synthesi expressions, we measured was the increased expression in GE-treatedand activity of HaCaT cells.key This enzymes effect wasthat are more involved prominent in ceramide after 6 metabolism, days of GE e.g., treatment CerS3 and compared acidic sphingomyelinase. to 2 days (Figure 5A). To assessImmunohistochemical if this effect is PPAR stainingγ and/or showed p38 that MAPK CerS3 dependent, expression hPKswas increased were pre-incubated in GE-treated with HaCaT SB203580 cells. This effect was more prominent after 6 days of GE treatment compared to 2 days (Figure 5A). (p38 MAPK inhibitor) or GW9962 (PPARγ inhibitor) for 1 h before the GE treatment. It turned out that To assess if this effect is PPARγ and/or p38 MAPK dependent, hPKs were pre-incubated with GE-induced CerS3 expression could nearly be reduced to background level, either by blocking the SB203580 (p38 MAPK inhibitor) or GW9962 (PPARγ inhibitor) for 1 h before the GE treatment. It p38 MAPKturned out or that PPAR GE-inducedγ pathway CerS3 (Figure expression5B). These could resultsnearly be indicate reduced that to background an altered level, expression either by of the ceramideblocking metabolic the p38 enzyme,MAPK or CerS3 PPAR thatγ pathway is nearly (Figure exclusively 5B). These expressed results indicate in hPKs, that might an altered account for the increasedexpression lipid of the content ceramide in GE-treatedmetabolic enzyme, hPKs. InCerS3 contrast, that is acidicnearly sphingomyelinaseexclusively expressed activity in hPKs, was not affectedmight by account GE treatment for the (Figure increased5C). lipid content in GE-treated hPKs. In contrast, acidic sphingomyelinase activity was not affected by GE treatment (Figure 5C).

FigureFigure 5. GE 5. inducesGE induces CerS3 CerS3 expression expression but but not not sphingomyelinase sphingomyelinase activity.activity. ((AA) hPKshPKs were treated with GE (200with µGEg/mL) (200 µg/mL) for 0, for 2, or0, 2, 6 or days. 6 days. Then, Then, the the cells cells werewere stained against against CerS3. CerS3. The Thepictures pictures were were photographed at magnification of 100×, and the bar indicates 50 µm; (B) hPKs were treated with GE photographed at magnification of 100×, and the bar indicates 50 µm; (B) hPKs were treated with GE (200 µg/mL), or with GE and one of the inhibitors (PPARγ antagonist GW9962 or p38 MAPK inhibitor (200 µg/mL), or with GE and one of the inhibitors (PPARγ antagonist GW9962 or p38 MAPK inhibitor SB203580) for 6 days. Then, the cells were stained against CerS3. The pictures were photographed at SB203580)a magnification for 6 days. of 400×, Then, and the the cells bar were indicates stained 10 µm; against (C) hPKs CerS3. were The treated pictures with were GE (200 photographed µg/mL) or at a magnificationwith GE and of 400inhibitors,×, and as the indica bar indicatested, for 6 days. 10 µm; Then, (C) hPKsthe sphing wereomyelinase treated with activity GE (200 wasµ measuredg/mL) or with GE andin the inhibitors, cell lysates. as indicated,Data are expressed for 6 days. as mean Then,s the± SD sphingomyelinase of three independent activity experiments was measured (ns—not in the cell lysates.significant). Data are expressed as means ± SD of three independent experiments (ns—not significant).

Int. J. Mol. Sci. 2017, 18, 1814 7 of 16

2.4. GE EnhancesInt. J. Mol. Sci. Lipid 2017 Production, 18, 1814 In Vivo 7 of 16 To analyze2.4. GE Enhances if GE also Lipid increases Production lipid In Vivo synthesis in vivo, the effect of a 5% GE cream was assessed in a proof of conceptTo analyze placebo-controlled if GE also increases halflipid side synthesis comparison in vivo, the in 33effect healthy of a 5% volunteers. GE cream was One assessed of their volar forearmsin a proof was of treated concept with placebo-controlled a deﬁned amount half side of GE comparison cream for in 433 weeks, healthy twice volunteers. a day, One whereas of their the other forearmvolar forearms received was the treated placebo with cream. a defined We choseamount the of GE volar cream forearm, for 4 weeks, because, twice in a thisday, bodywhereas region, the skin lipidsother are forearm nearly received exclusively the placebo produced cream. by We keratinocytes. chose the volar In forearm, other skinbecause, regions, in this the body lipids region, are mainly producedskin lipids by are sebaceous nearly exclusively glands. After produced 2, 3, and by keratinocytes. 4 weeks, the lipidIn other content skin wasregions, measured the lipids using are a sebumeter.mainly Already produced after by 2 weeks sebaceous of treatment, glands. After a signiﬁcant 2, 3, and 4 weeks, increase the in lipid the content lipid content was measured could be using seen a sebumeter. Already after 2 weeks of treatment, a significant increase in the lipid content could be (Figure6). Thirteen volunteers (39%) were non-responders, because they showed no change in the seen (Figure 6). Thirteen volunteers (39%) were non-responders, because they showed no change in lipid content.the lipid Twenty content. volunteers Twenty (60%)volunteers showed (60%) an showed increase an in increase the lipid in contentthe lipid ofcontent at least of 25%,at least and 25%, 8 of them showedand 8 anof them increase showed of 50% an increase or more. of The 50% colored or more. dots The incolored Figure dots6 show in Figure 3 responders. 6 show 3 responders. This effect is sustainedThis throughout effect is sustained the whole throughout study the (Supplementary whole study (Supplementary Figure S2). Figure S2).

Figure 6. FigureGE increases 6. GE increases lipid synthesis lipid synthesisin vivo in. vivo. In a placeboIn a plac controlledebo controlled half half side side comparison, comparison, 33 adult volunteersvolunteers were treated were treated on their on their volar volar forearms forearms with with 5% 5% GEGE cream, or or placebo placebo cream. cream. The box The blots box show blots the lipid content of the skin after 2 weeks. The red, green and blue dots and squares correspond to one show the lipid content of the skin after 2 weeks. The red, green and blue dots and squares correspond volunteer respectively. The arrows show the difference in the lipid content between the placebo and to one volunteer respectively. The arrows show the difference in the lipid content between the placebo GE-treated forearm. The p value is indicated in the figures by asterisks (** p < 0.01). and GE-treated forearm. The p value is indicated in the figures by asterisks (** p < 0.01). 3. Discussion 3. Discussion Skin surface lipids, and bound water in the stratum corneum, contribute to an intact skin Skinbarrier. surface An lipids, increase and in fatty bound acids water and their in themetabo stratumlites can corneum, activate the contribute nuclear transcription to an intact factors skin barrier.PPARs, An increase and induce in fatty the gene acids expression and their of proteins metabolites that are can required activate for thecorneocyte nuclear formation transcription (e.g., factors PPARs,involucrin) and induce[1]. PPARs the genealso expressionstimulate the of formation proteinsthat and are packaging required of for lipids, corneocyte and can formation induce ceramide synthesis [21]. Rau and colleagues described that a GE extract activates PPARγ [20]. These (e.g., involucrin) [1]. PPARs also stimulate the formation and packaging of lipids, and can induce findings fit to our results that GE induces lipid synthesis in hPKs, and that this effect can be blocked ceramide synthesis [21]. Rau and colleagues described that a GE extract activates PPARγ [20]. with a PPARγ inhibitor. PPARγ is a key regulator of lipid metabolism in different cell types (e.g., in These findingsmacrophages fit to our[22], resultssebocytes that [23], GE and induces dendritic lipid cells synthesis [24]). Another in hPKs, pathway and involved that this in effectthe lipid can be blockedmetabolism with a PPAR of keratinocytesγ inhibitor. is the PPAR MAPKγ is pathway, a key regulator because ofGE-induced lipid metabolism lipid production in different could cellbe types (e.g.,diminished in macrophages by a p38 [MAPK22], sebocytes inhibitor. [ 23Inappropriate], and dendritic MAPK cells signalin [24]).g Anothercontributes pathway to the development involved in the lipidof metabolism several diseases, of keratinocytes for example, the is the metabolic MAPK synd pathway,rome [25]. because p38 MAPK GE-induced also influences lipid productionlipogenesis could bein diminished adipocytes [26]. by a This p38 is MAPK in contrast inhibitor. to the effe Inappropriatect of p38 MAPK MAPK in the signalinghepatic metabolism, contributes where tothe it developmentstimulates of several gluconeogenesis, diseases, for while example, simultaneously the metabolic inhibiting syndrome lipogenesis [25]. [27]. p38 MAPKThe MAPK also pathway influences is lipogenesisalso in involved adipocytes in endocannabinoid-induced [26]. This is in contrast tolipid the sy effectnthesis of p38in sebocytes, MAPK inas the shown hepatic by Dobrosi metabolism, and colleagues [23]. However, epidermal lipids are composed of a mixture of almost equal proportions where it stimulates gluconeogenesis, while simultaneously inhibiting lipogenesis [27]. The MAPK of ceramides, free fatty acids and cholesterol, whereas sebaceous lipids are non-polar lipids, such as pathway is also involved in endocannabinoid-induced lipid synthesis in sebocytes, as shown by triglycerides, wax esters, and squalene [28,29]. In the present study, we could show that GE induces Dobrosi andlipogenesis colleagues in hPKs. [23]. Bonté However, and epidermalcolleagues descri lipidsbed are already composed in 1996 of a that mixture the bitter of almost extract equal of proportionsSimarouba of ceramides, amara, a tree free that fatty grows acids inand in the cholesterol, rainforest of whereas Brazil, shows sebaceous a marked lipids increase are non-polar in total lipids, suchlipids as in triglycerides, air-exposed keratinocyte wax esters, cultures and squalene [30]. This [28 bitter,29]. compound In the present can possibly study, induce we could the lipid show that GE inducessynthesis lipogenesis in keratinocytes in hPKs. comparable Bonté andto GE, colleagues although this described publication already attached in 1996 no importance that the bitter to extract of Simarouba amara, a tree that grows in in the rainforest of Brazil, shows a marked increase in total lipids in air-exposed keratinocyte cultures [30]. This bitter compound can possibly induce the lipid synthesis in keratinocytes comparable to GE, although this publication attached no importance to Int. J. Mol. Sci. 2017, 18, 1814 8 of 16 the fact that Simbarouba amara extract tastes bitter and might activate bitter taste receptors. Another substance that modulates lipogenesis, and especially, ceramide production in keratinocytes, is vitamin C. However, the activity of sphingomyelinase, a hydrolase enzyme that catalyzes the conversion of sphingomyelin to ceramide, remained unaltered after vitamin C treatment [5]. These observations are in line with our finding that GE increases CerS3 expression in keratinocytes without affecting acidic sphingomyelinase activity. CerS3 is particularly important for skin cornification [31]. Its function cannot be replaced by any of the other five CerS, and ceramide 3 is exclusively expressed in the skin and testis [32]. In general, ceramides are sphingosines (fatty amino alcohols) that are N-acylated with various fatty acids to form many ceramide species [33] that are involved in keratinocyte maturation. In atopic dermatitis and psoriasis, the ceramide concentration in the stratum corneum is diminished [6,34,35]. In particular, the levels of ceramide 1 and 3 are significantly lower compared to healthy people [36]. Vitamin D or vitamin D analogues, such as calcipotriol, are used for the treatment of psoriasis. They increase the ceramide level by inducing sphingomyelin hydrolysis in human keratinocytes [37]. In this context, it is interesting that ceramides also have an anti-proliferative effect [38]. Certain amphiphilic lipids, in particular sphingosines and free fatty acids (e.g., lauric acid), decrease the growth of microbes, and thereby reduce infections [29]. A decreased level of sphingosine is associated with increased Staphylococcus aureus colonization in atopic dermatitis patients [39]. Moreover, free fatty acids that are formed by phospholipid breakdown contribute to the acidification of the stratum corneum [40]. This acidic environment is very important as it regulates many enzymes in the stratum corneum, e.g., acidic sphingomyelinase [41]. In a placebo-controlled double-blind half side comparison, we tested the effect of GE on lipid synthesis on the volar forearm. We could demonstrate that GE also increases the lipid content of the stratum corneum in vivo. This could be of therapeutic value in atopic dermatitis that is characterized by an impaired epidermal barrier with reduced lipid content. Similarly, body regions in which lipids are exclusively produced by epidermal cells, for example around the lips, in the hollow of the knee and the bend of the elbow, might benefit from topical GE treatment. Interestingly, these body regions are also particularly affected by atopic dermatitis. It is noteworthy that a ceramide, cholesterol, and fatty acid mixture, corresponding to the epidermal lipid composition, can accelerate the recovery of the skin barrier [42,43]. It is commonly accepted that ceramides are important for skin health. However, studies performed with human fibroblasts that were incubated with a ceramide analogue in an interleukin-6 (IL-6) rich environment, have shown that ceramides also modulate the secretion of prostaglandin E2 (PGE2)[44]. This indicates that ceramides are not only beneficial for skin barrier recovery, but may also enhance the inflammatory responses to inflammatory stimuli. [36]. Also, it has been shown that ceramides can activate the NLRP3 inflammasome in obesity-induced inflammation [45]. To confirm that GE does not induce inflammatory responses in keratinocytes, we measured PGE2 and IL-6 release of hPKs after GE treatment. No expression of these inflammatory mediators could be detected in our study (data not shown). On the other hand, Danso and colleagues showed in skin equivalents that cytokine addition modulates lipid synthesis. As a consequence, inflammation might cause lower ceramide levels that are also observed in atopic dermatitis and psoriasis [46]. In summary, the lipid content of the stratum corneum is essential for an optimal epidermal barrier function. GE enhances epidermal lipid synthesis and might therefore be useful to improve skin disorders with barrier abnormalities, e.g., very dry skin and atopic dermatitis. Int. J. Mol. Sci. 2017, 18, 1814 9 of 16

4. Materials and Methods

4.1. Antibodies and Reagents The polyclonal rabbit-anti-human CerS3 antibody was from Antikörper-online (Aachen, Germany; 1:50). The secondary antibodies were the swine anti-rabbit-FITC antibody (Dako, Hamburg, Germany; 1:30) and the donkey anti-rabbit-Alexa-Fluor 555 antibody (Thermo Fisher Scientific, Darmstadt, Germany). DAPI was from Sigma-Aldrich GmbH (Taufkirchen, Germany). The following specific inhibitors were used: PPARγ antagonist (GW9962, Enzo, Lörrach, Germany), PI3K inhibitor (LY294002; Sigma-Aldrich GmbH), ERK inhibitor (PD98059, New England Biolabs GmbH, Frankfurt, Germany), p38 MAPK inhibitor (SB203580, Sigma-Aldrich GmbH) and NFκB inhibitor (GIV 3727, Merck Millipore, Darmstadt, Germany). Methanolic sodium hydroxide solution, hexane standards, loganic acid, and loganin were from Carl Roth GmbH (Karlsruhe, Germany). Boron trifluoride–methanol complex was from Merck KGaA (Darmstadt, Germany), and standard methyl heptadecanoate was from Sigma-Aldrich GmbH. Orto-phosphoric acid, 85%, Ph. Eur. p.a. grade was supplied by VWR. Amarogentin and gentiopicroside were purchased from Chromadex Inc. (Santa Ana, CA, USA).

4.2. Preparation of Yellow Gentian Root High Pressure Ethanol (HPE) Extract Whole gentian root was purchased from a Bavarian farm specialized in root drugs (Berghof Kräuter GmbH, Heilsbronn, Germany). The root was passed through a cutting mill with 3 mm sieve (Pallmann Maschinenfabrik GmbH, Zweibrücken, Germany). The powder was ﬁlled into a high pressure extractor and percolated with 8 kg of 96% (v/v) ethanol at 110 bar/60 ◦C per kg feedstock to give the crude HPE (high pressure ethanol) extract. The major ethanol amount was removed by vacuum distillation, in order to adjust the ﬁnal spissum extract to 50% dry matter, according to a drug/extract ratio of about 2.5/1.

4.3. HPLC-DAD Analysis of Yellow Gentian Root HPE Extract All solvents were of HPLC grade and supplied by VWR International GmbH (Darmstadt, Germany). Standards were dissolved in methanol at 0.3–0.7 mg/mL concentration. The yellow gentian root HPE extract was dissolved under sonication in methanol at 8–10 mg/mL concentration and filtered through a PTFE (polytetrafluoroethylene) filter with 0.45 µm pore diameter before injection. A Merck Hitachi LaChrom Elite HPLC system was used, consisting of auto sampler L-2200, pump L-2130, DAD detector L-2420, column oven L-2350, operated with software EZ Chrom Elite Version 3.3.2 Build 1037 (SP2); column Lichrospher 100 RP-18e (5 µm), 250 × 4 (mm length × internal diameter), Merck KGaA (Darmstadt, Germany). Eluent A was water acidified with o-phosphoric acid 85% to pH 2.3–2.6, eluent B acetonitrile, eluent C 1-propanol; gradient profile: ramp 1: 98% A, 1% B, 1% C to 70% A, 15% B, 15% C within 20 min; ramp 2: to 60% A, 20% B, 20% C within 2 min., isocratic for 10 min. Eluent flow was 1 mL/min, oven temperature 30 ◦C, detector 232 nm, injection volume 10 µL. The concentration of Swertiamarin is calculated as Gentiopicroside, because this compound was not available as reference standard. Peak identification was based on literature. A HPLC-UV chromatogram of yellow gentian root CO2-extract is shown in Figure7. Int. J. J. Mol. Mol. Sci. Sci. 20172017,, 1818,, 1814 1814 10 of 16

Figure 7. HPLC-UV chromatogram of yellow gentian root CO -extract. The detection was at 232 nm: Figure 7. HPLC-UV chromatogram of yellow gentian root CO2-extract. The detection was at 232 nm: peak identiﬁcationsidentifications (concentrations) are: loganic acid (3.1%), swertiamarin (0.79%), gentiopicroside (12.3%), loganin (0.41%), and amarogentin (0.05%).

4.4. Cell Culture The humanhuman keratinocyte keratinocyte cell cell line line HaCaT HaCaT was fromwas CLSfrom (Cell CLS Lines (Cell Service; Lines Heidelberg,Service; Heidelberg, Germany) Germany)and was cultured and was in cultured Dulbecco’s in Dulbecco’s modiﬁed essential modified medium essential (DMEM; medium Biochrom, (DMEM; Biochrom, Berlin, Germany) Berlin, Germany)containing containing 10% fetal calf10% serum fetal calf (FCS; serum PAA, (FCS; Pasching, PAA, Pasching, Austria). Austria). hPKs were hPKs prepared were prepared from juvenile from juvenileforeskin foreskin or adult skinor adult obtained skin obtained from dermatological from dermatological surgery, and surgery, cultured and according cultured toaccording the method to the of methodRheinwald of Rheinwald and Green and [47] Green in DermaLife [47] in DermaLife medium (Cell medium System, (Cell Goisdorf, System, Germany).Goisdorf, Germany). All cells were All ◦ cellscultured were at cultured 37 C in at a humidiﬁed37 °C in a humidified atmosphere atmosphere with 5% CO with2. 5% CO2.

4.5. Immunofluorescence Immunofluorescence µ hPKs were were incubated incubated for for 6 6days days with with 200 200 µg/mLg/mL GE GE and and centrifuged centrifuged on slides. on slides. Then, Then, the cells the werecells werestained stained with with the thepolyclonal polyclonal rabbit rabbit anti-human anti-human CerS3 CerS3 antibody. antibody. The The required required secondary antibodies (donkey(donkey anti-rabbit-Alexa-Fluoranti-rabbit-Alexa-Fluor 555 555 or or swine swine anti-rabbit-FITC) anti-rabbit-FITC) were were applied applied for 1 for h at 1 room h at roomtemperature, temperature, according according to the manufacturer’s to the manufacturer’s instructions. instructions. Then, the Then, cells werethe cells stained were with stained DAPI with and DAPImounted and in mounted fluorescent in fluorescent mounting mediummounting (Dako). medium Images (Dako). were Images taken were with ataken fluorescence with a fluorescence microscope microscope(Carl Zeiss, Oberkochen,(Carl Zeiss, Oberkochen, Germany) equipped Germany) with equipped Axiovision with softwareAxiovision version software 4.1. version 4.1. 4.6. Cytotoxicity Assay 4.6. Cytotoxicity Assay hPKs were incubated in a 96 well plate (2000 cells per well) and incubated for 6 days with GE hPKs were incubated in a 96 well plate (2000 cells per well) and incubated for 6 days with GE (200 µg/mL) before the cytotoxicity was assessed with the ViaLight Plus ATP assay (Lonza, Basel, (200 µg/mL) before the cytotoxicity was assessed with the ViaLight Plus ATP assay (Lonza, Basel, Switzerland), according to the manufacturer’s instruction. The method is based on the bioluminescent Switzerland), according to the manufacturer’s instruction. The method is based on the measurement of ATP that is present in metabolically active cells. Luciferase catalyzes the formation bioluminescent measurement of ATP that is present in metabolically active cells. Luciferase of light from ATP and luciferin. The emitted light intensity is directly proportional to the ATP catalyzes the formation of light from ATP and luciferin. The emitted light intensity is directly concentration, and is measured with a luminometer (Sirius HT, BioTek, Bad Friedrichshall, Germany). proportional to the ATP concentration, and is measured with a luminometer (Sirius HT, BioTek, Bad Friedrichshall,4.7. Enzyme Activity Germany). Assay for Sphingomyelinase

4.7. EnzymehPKs were Activity incubated Assay for for Sphingomyelinase 6 days with 200 µg/mL GE. Subsequently, the cells were lysed and processed according to the manufacturer’s protocol (Acid Sphingomyelinase Activity Colorimetric AssayhPKs Kit, BioVision,were incubated San Francisco, for 6 days CA, with USA). 200 The µg/m absorptionL GE. Subsequently, was measured the at cells 570 nmwere on lysed an ELISA and processedreader (Sirius according HT). to the manufacturer’s protocol (Acid Sphingomyelinase Activity Colorimetric

Int. J. Mol. Sci. 2017, 18, 1814 11 of 16

4.8. Determination of Intracellular Lipids in Keratinocytes For semi-quantitative detection of lipids in keratinocytes, hPKs were treated for 6 days with GE (200 µg/mL) and stained with Nile Red (9-diethylamino-5H-benzo[a]phenoxazine-5-one, Sigma-Aldrich GmbH). This dye is only strongly fluorescent in a hydrophobic environment, and can also be applied to cells in an aqueous medium, because it does not dissolve lipids [2]. Staining was performed in formalin-fixed cells with 100 ng/mL Nile Red in phosphate buffered saline PBS for 5 min. Then, the cells were counterstained with 40,6-Diamino-2-phenylindol Dihydrochlorid (DAPI, Sigma-Aldrich GmbH) [2]. For quantitative measurement, hPKs (4000 cells/well) were cultured in 96 white well clear bottom plates in eight replicates, and treated with GE for 6 days according to the literature [10]. For all samples, the average of six experimental settings was used. The supernatants were then discarded, and 100 µL Nile Red solution (1 µg/mL) was added to each well. The emitted fluorescence was measured on an ELISA reader, using 485 nm excitation and 590 nm emission wavelength for neutral lipids [23]. The relative fluorescence units from the untreated control were shown as 100%, and the GE-treated samples related to this value.

4.9. Isolation of Lipids from Keratinocytes hPKs were incubated with 200 µg/mL GE. The lipid content was measured after 6 days of GE incubation. The age of the hPK donors varied from 2 to 92 years. Four samples were classiﬁed as young (<10 years), and 4 as old (>50 years), as shown in the supplementary table to have a representative group. The average age of the young group was 7.3 ± 3.6 years, and of the older group, 71.0 ± 16.8 years. The lipid extraction was performed as described by Bligh and Dyer [48]. Brieﬂy, the cells were pelleted and re-suspended in a water-chloroform-methanol mixture (4:5:10), sonicated, and centrifuged. Then, the upper layer was mixed with chloroform-water (1:1) and centrifuged. The lower chloroform layer with the lipids was dried in a heating block, and stored at −80 ◦C until it was used for analysis of fatty acids by GC-FID.

4.10. Gas Chromatography with Flame Ionization Detector (GC-FID) Analysis of Fatty Acids All solvents and reagents were of analytical grade. Fatty acids were analyzed according to the DGF standard method C-VI 10a (00) by transmethylation with boron trifluoride (BF3). Cell samples were saponified with 200 µL methanolic sodium hydroxide solution, 0.5 mol/L. Resulting soaps were converted to methyl esters by addition of 100 µL of a boron trifluoride–methanol complex, 20% solution. Reaction time of each step was 10 min, and carried out in a water bath at 80 ◦C. The fatty acid methyl esters were diluted in 400 µL hexane and centrifuged. The clear hexane layer was transferred into a vial and used for GC analysis. A Shimadzu GC 17 A with class VP 4.3 software (Shimadzu Europe GmbH, Duisburg, Germany) equipped with AOC20i autosampler and a Roticap FFAP column (Carl Roth GmbH, 30 m × 0.25 mm i.d., 0.25 µm film thickness) was used. Carrier gas was nitrogen, with constant flow of 1.6 mL/min; initial GC oven temperature was 110 ◦C, increased to 220 ◦C at a rate of 5 ◦C/min and held at 220 ◦C for 20 min; injector temperature was 240 ◦C, injection volume was 8 µL with split mode 1:5; FID temperature was 280 ◦C. The absolute content of each fatty acid (newly synthesized or separated from ceramides) in hPKs, and the relative fatty acid composition, were identified by comparing the retention time of samples with the lipid standard “Fatty Acid Methyl Ester Mix” (Sigma-Aldrich GmbH). Quantification of fatty acids was done with external standard methyl heptadecanoate, purity >99% (Sigma-Aldrich) at a concentration of 1 mg/mL. The accuracy was determined by the recovery method to be 98%; repeatability tests showed relative standard deviation (RDS) of less than 4% for all fatty acid methyl esters. A representative GC-FID chromatogram of fatty acids extracted from the cell lipids is shown in Figure8. Int. J. Mol. Sci. 2017, 18, 1814 12 of 16 Int. J. Mol. Sci. 2017, 18, 1814 12 of 16

FigureFigure 8. TheThe graph graph shows shows a arepresentative representative GC-FID GC-FID chromatogram chromatogram of of fatty fatty acids acids GC-FID GC-FID chromatogramchromatogram of of fatty fatty acids acids extracted extracted from from the thecell celllipids. lipids. The first The two ﬁrst peaks two between peaks between 13 and 14 13 min and retention14 min retention time are time C16:0 are C16:0and C16:1n7 and C16:1n7 acids. acids. The Thesecond second group group between between 16 16and and 17 17 min min are are C18:0, C18:0, C18:1n9,C18:1n9, and and C18:1n7 C18:1n7 (shoulder) (shoulder) ac acids,ids, and and the the smaller smaller peak peak at at 17.5 17.5 min min is is C18:2n6 C18:2n6 acid. acid. The The red red line line indicatesindicates the the base base for for peak peak integration. integration.

4.11.4.11. Half Half Side Side Comparison Comparison InIn a a half half side side comparison, comparison, 33 33 adult adult volunteers volunteers with with normal normal to to dry dry skin skin we werere treated treated with with 5% 5% GE GE creamcream onon oneone of of the the volar volar forearms forearms for 4 for weeks. 4 week Thes. other The arm other was arm treated was with treated the vehicle with the unguentum vehicle unguentumemulsificans emulsificans aquosum (INCI: aquosum nonionic (INCI: emulsifying nonionic alcohols, emulsifying 2-ethylhexyllaurate, alcohols, 2-ethylhexyllaurate, glycerol, potassium glycerol,sorbate, aquapotassium dest.). sorbate, The volar aqua forearms dest.). The were volar chosen, forearms because, were atch thisosen, area, because, the lipids at this are area, mainly the lipidsproduced are mainly by keratinocytes, produced by and keratinocytes, not by sebaceous and not glands, by sebaceous as in other glands, regions as in of other the skin. regions The of study the skin.protocol The (codestudy2013-VE01) protocol (code was 2013-VE01) approved by was the appr ethicsoved committee by the ethics of the committee University of of the Freiburg, University and ofwritten Freiburg, informed and written consent informed was obtained consent from was all obta subjects.ined from The all approval subjects. number The approval is 11/12. number Included is 11/12.were subjectsIncluded of were both subjects sexes (4 menof both and sexes 29 women), (4 men and with 29 an women), average agewith of an 43.6 average± 13.2 age years. of 43.6 Exclusion ± 13.2 years.criteria Exclusion were the usagecriteria of cosmeticswere the orusage topical of cosmetics therapeutics or intopical the test therapeutics area. The verum in the or test placebo area. cream The verumwas applied or placebo in a double cream blind was mannerapplied toin the a testdouble area blind by the manner study participants,to the test area two timesby the daily study for participants,4 weeks. Study two visits times were daily at for day 4 1 weeks. (U1), after Study 2 weeks visits (U2),were afterat day 3 weeks1 (U1), (U3) after and 2 weeks after 4(U2), weeks after (U4). 3 weeksAt each (U3) visit, and the after total 4 skin weeks surface (U4). lipids At each (sebum visit, and the cornealtotal skin lipids) surface were lipids measured (sebum in and the testcorneal area lipids)using awere sebumeter measured SM 810 in (Couragethe test area + khazaka using electronica sebumeter GmbH, SM Köln,810 (Courage Germany). + khazaka The mat tapeelectronic of the GmbH,sebumeter Köln, is brought Germany). into contactThe mat with tape skin of and the becomes sebumeter transparent is brought in relation into contact to the sebum with onskin the and test becomesarea. Then, transparent the transparency in relation of the to the tape sebum is measured on the bytest a photocell.area. Then, The the light transparency transmission of the represents tape is measuredthe lipid content. by a photocell. Data are shownThe light in plots. transmission The readings represents were tested the lipid for significant content. Data differences are shown between in plots.U1 and The U2, readings U1 and were U3, tested as well for as significant U1 and U4, diffe respectively,rences between using U1 the and Wilcoxon U2, U1 and test U3, for as pairwise well as U1comparisons. and U4, respectively, The p value using is indicated the Wilcoxon in the figure test for by pairwise asterisks (comparisons.** p ≤ 0.01). All The measured p value is values indicated were inexpressed the figure as theby medianasterisks of (** 3 adjacentp ≤ 0.01). recordings, All measured to avoid values measuring were expressed inaccuracies. as the median of 3 adjacent recordings, to avoid measuring inaccuracies. 4.12. Statistical Analysis 4.12. StatisticalThe in vitro Analysisdata were analyzed using the unpaired Student t-test (two-tailed). The in vivo study was evaluatedThe in vitro by Wilcoxon’sdata were analyzed signed rank using test, the and unpaired statistical Student significance t-test was (two-tailed). established The at * pin≤ vivo0.05 studyand ** wasp ≤ evaluated0.01. The statistical by Wilcoxon’s analysis signed was performedrank test, and with statistical the GraphPad significance Prism software was established (GraphPad at *Software, p ≤ 0.05 and version ** p 5.03,≤ 0.01. Inc., The San statistical Diego, CA,analysis USA). was Data performed are expressed with the as meanGraphPad± SD Prism of at least software three (GraphPad Software, version 5.03, Inc., San Diego, CA, USA). Data are expressed as mean ± SD of at

Int. J. Mol. Sci. 2017, 18, 1814 13 of 16 independent experiments. In the in vivo half side comparison, three adjacent body regions at the volar forearm were measured.

Supplementary Materials: Supplementary materials can be found at www.mdpi.com/1422-0067/18/8/1814/s1. Acknowledgments: The research center skinitial receives funding from the Software AG Stiftung and DAMUS-DONATA e.V. Award number: SE-P10042 Author Contributions: Ute Wölfle, Kay Schwabe, Karl-Werner Quirin and Christoph M. Schempp conceived and designed the experiments; Birgit Haarhaus performed the staining experiments, ELISAs and cell culture experiments. Jasmin Seiwerth performed the half side comparison and Anja Cawelius performed the gas chromatography analyses; Ute Wölfle, Birgit Haarhaus, Jasmin Seiwerth and Anja Cawelius analyzed the data; Kay Schwabe prepared the medication for the half side comparison and Karl-Werner Quirin provided and analyzed the GL-extract; Ute Wölfle wrote the paper and Jasmin Seiwerth, Karl-Werner Quirin and Christoph M. Schempp co-wrote the paper. Conflicts of Interest: The authors declare no conflict of interest. The founding sponsors had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the decision to publish the results.

Abbreviations

ASMase Shingomyelinase activity BAK Bcl-2 antagonist/killer1 Ca2+ Calcium Bcl-2 B-cell lymphoma 2 CerS3 Ceramid synthase 3 GC-FID Gaschromatography with ﬂame ionization detector INCI International nomenclature of cosmetic ingredients IL-6 Interleukin-6 HPE High pressure ethanol HPLC High performance liquid chromatography hPKs Human primary keratinocytes MDM2 Mouse double minute 2 homolog p38 MAPK p38 mitogen-activated protein kinases PGE2 Prostaglandin E2 PPARγ Peroxisome proliferator activated receptor γ

References

1. Feingold, K.R.; Schmuth, M.; Elias, P.M. The regulation of permeability barrier homeostasis. J. Invest. Dermatol. 2007, 127, 1574–1576. [CrossRef][PubMed] 2. Greenspan, P.; Mayer, E.P.; Fowler, S.D. Nile red: A selective ﬂuorescent stain for intracellular lipid droplets. J. Cell Biol. 1985, 100, 965–973. [CrossRef][PubMed] 3. Feingold, K.R. The outer frontier: The importance of lipid metabolism in the skin. J. Lipid Res. 2009, 50, S417–S422. [CrossRef][PubMed] 4. Uchida, Y. Ceramide signaling in mammalian epidermis. Biochim. Biophys. Acta 2014, 1841, 453–462. [CrossRef][PubMed] 5. Kim, K.P.; Shin, K.-O.; Park, K.; Yun, H.J.; Mann, S.; Lee, Y.M.; Cho, Y. Vitamin C Stimulates Epidermal Ceramide Production by Regulating Its Metabolic Enzymes. Biomol. Ther. 2015, 23, 525–530. [CrossRef] [PubMed] 6. Motta, S.; Monti, M.; Sesana, S.; Mellesi, L.; Ghidoni, R.; Caputo, R. Abnormality of water barrier function in psoriasis. Role of ceramide fractions. Arch. Dermatol. 1994, 130, 452–456. [CrossRef][PubMed] 7. Imokawa, G.; Abe, A.; Jin, K.; Higaki, Y.; Kawashima, M.; Hidano, A. Decreased level of ceramides in stratum corneum of atopic dermatitis: An etiologic factor in atopic dry skin? J. Investig. Dermatol. 1991, 96, 523–526. [CrossRef][PubMed] Int. J. Mol. Sci. 2017, 18, 1814 14 of 16

8. Mullen, T.D.; Hannun, Y.A.; Obeid, L.M. Ceramide synthases at the centre of sphingolipid metabolism and biology. Biochem. J. 2012, 441, 789–802. [CrossRef][PubMed] 9. Goñi, F.M.; Alonso, A. Sphingomyelinases: Enzymology and membrane activity. FEBS Lett. 2002, 531, 38–46. [CrossRef] 10. Yarosh, D.B.; Both, D.; Brown, D. Liposomal ursolic acid (merotaine) increases ceramides and collagen in human skin. Horm. Res. 2000, 54, 318–321. [CrossRef][PubMed] 11. Kim, E.J.; Kim, M.-K.; Jin, X.-J.; Oh, J.-H.; Kim, J.E.; Chung, J.H. Skin Aging and Photoaging Alter Fatty Acids Composition, Including 11,14,17-eicosatrienoic Acid, in the Epidermis of Human Skin. J. Korean Med. Sci. 2010, 25, 980–983. [CrossRef][PubMed] 12. Raj, D.; Brash, D.E.; Grossman, D. Keratinocyte apoptosis in epidermal development and disease. J. Investig. Dermatol. 2006, 126, 243–257. [CrossRef][PubMed] 13. Wölﬂe, U.; Elsholz, F.A.; Kersten, A.; Haarhaus, B.; Müller, W.E.; Schempp, C.M. Expression and functional activity of the bitter taste receptors TAS2R1 and TAS2R38 in human keratinocytes. Skin Pharmacol. Physiol. 2015, 28, 137–146. [CrossRef][PubMed] 14. Meyerhof, W.; Batram, C.; Kuhn, C.; Brockhoff, A.; Chudoba, E.; Bufe, B.; Appendino, G.; Behrens, M. The molecular receptive ranges of human TAS2R bitter taste receptors. Chem. Sens. 2010, 35, 157–170. [CrossRef][PubMed] 15. Deshpande, D.A.; Wang, W.C.; McIlmoyle, E.L.; Robinett, K.S.; Schillinger, R.M.; An, S.S.; Sham, J.S.; Liggett, S.B. Bitter taste receptors on airway smooth muscle bronchodilate by localized calcium signaling and reverse obstruction. Nat. Med. 2010, 16, 1299–1304. [CrossRef][PubMed] 16. Lee, R.J.; Xiong, G.; Kofonow, J.M.; Chen, B.; Lysenko, A.; Jiang, P.; Abraham, V.; Doghramji, L.; Adappa, N.D.; Palmer, J.N.; et al. T2R38 taste receptor polymorphisms underlie susceptibility to upper respiratory infection. J. Clin. Investig. 2012, 122, 4145–4159. [CrossRef][PubMed] 17. Rozengurt, E.; Sternini, C. Taste receptor signaling in the mammalian gut. Curr. Opin. Pharmacol. 2007, 7, 557–562. [CrossRef][PubMed] 18. Ponec, M.; Weerheim, A.; Kempenaar, J.; Mommaas, A.M.; Nugteren, D.H. Lipid composition of cultured human keratinocytes in relation to their differentiation. J. Lipid Res. 1988, 29, 949–961. [PubMed] 19. Roudnitzky, N.; Behrens, M.; Engel, A.; Kohl, S.; Thalmann, S.; Hübner, S.; Lossow, K.; Wooding, S.P.; Meyerhof, W. Receptor Polymorphism and Genomic Structure Interact to Shape Bitter Taste Perception. PLoS Genet. 2015, 11, e1005530. [CrossRef][PubMed] 20. Rau, O.; Wurglics, M.; Dingermann, T.; Abdel-Tawab, M.; Schubert-Zsilavecz, M. Screening of herbal extracts for activation of the human peroxisome proliferator-activated receptor. Pharmazie 2006, 61, 952–956. [PubMed] 21. Chon, S.-H.; Tannahill, R.; Yao, X.; Southall, M.D.; Pappas, A. Keratinocyte differentiation and upregulation of ceramide synthesis induced by an oat lipid extract via the activation of PPAR pathways. Exp. Dermatol. 2015, 24, 290–295. [CrossRef][PubMed] 22. Chawla, A.; Barak, Y.; Nagy, L.; Liao, D.; Tontonoz, P.; Evans, R.M. PPAR-gamma dependent and independent effects on macrophage-gene expression in lipid metabolism and inﬂammation. Nat. Med. 2001, 7, 48–52. [CrossRef][PubMed] 23. Dobrosi, N.; Toth, B.I.; Nagy, G.; Dozsa, A.; Geczy, T.; Nagy, L.; Zouboulis, C.C.; Paus, R.; Kovacs, L.; Biro, T. Endocannabinoids enhance lipid synthesis and apoptosis of human sebocytes via cannabinoid receptor-2-mediated signaling. FASEB J. 2008, 22, 3685–3695. [CrossRef][PubMed] 24. Szatmari, I.; Töröcsik, D.; Agostini, M.; Nagy, T.; Gurnell, M.; Barta, E.; Chatterjee, K.; Nagy, L. PPARgamma regulates the function of human dendritic cells primarily by altering lipid metabolism. Blood 2007, 110, 3271–3280. [CrossRef][PubMed] 25. Gehart, H.; Kumpf, S.; Ittner, A.; Ricci, R. MAPK signalling in cellular metabolism: Stress or wellness? EMBO Rep. 2010, 11, 834–840. [CrossRef][PubMed] 26. Patel, N.; Holder, J.; Smith, S.; Kumar, S.; Eggo, M. Essential role of p38 mitogen-activated protein kinase in adipogenesis. In Proceedings of the 21st Joint Meeting of the British Endocrine Societies, Harrogate, UK, 8–11 April 2002. Int. J. Mol. Sci. 2017, 18, 1814 15 of 16

27. Cao, W.; Collins, Q.F.; Becker, T.C.; Robidoux, J.; Lupo, E.G.; Xiong, Y.; Daniel, K.W.; Floering, L.; Collins, S. p38 Mitogen-activated protein kinase plays a stimulatory role in hepatic gluconeogenesis. J. Biol. Chem. 2005, 280, 42731–42737. [CrossRef][PubMed] 28. Pappas, A. Epidermal surface lipids. Dermatoendocrinology 2009, 1, 72–76. [CrossRef] 29. Drake, D.R.; Brogden, K.A.; Dawson, D.V.; Wertz, P.W. Thematic review series: Skin lipids. Antimicrobial lipids at the skin surface. J. Lipid Res. 2008, 49, 4–11. [CrossRef][PubMed] 30. Bonté, F.; Barré, P.; Pinguet, P.; Dusser, I.; Dumas, M.; Meybeck, A. Simarouba amara extract increases human skin keratinocyte differentiation. J. Ethnopharmacol. 1996, 53, 65–74. [CrossRef] 31. Jennemann, R.; Rabionet, M.; Gorgas, K.; Epstein, S.; Dalpke, A.; Rothermel, U.; Bayerle, A.; van der Hoeven, F.; Imgrund, S.; Kirsch, J.; et al. Loss of ceramide synthase 3 causes lethal skin barrier disruption. Hum. Mol. Genet. 2012, 21, 586–608. [CrossRef][PubMed] 32. Mizutani, Y.; Mitsutake, S.; Tsuji, K.; Kihara, A.; Igarashi, Y. Ceramide biosynthesis in keratinocyte and its role in skin function. Biochimie 2009, 91, 784–790. [CrossRef][PubMed] 33. Maceyka, M.; Spiegel, S. Sphingolipid metabolites in inﬂammatory disease. Nature 2014, 510, 58–67. [CrossRef][PubMed] 34. Imokawa, G. Lipid abnormalities in atopic dermatitis. J. Am. Acad. Dermatol. 2001, 45, S29–S32. [CrossRef] [PubMed] 35. Betts, J.C.; Agranoff, A.B.; Nabel, G.J.; Shayman, J.A. Dissociation of endogenous cellular ceramide from NF-kappa B activation. J. Biol. Chem. 1994, 269, 8455–8458. [PubMed] 36. Di Nardo, A.; Wertz, P.; Giannetti, A.; Seidenari, S. Ceramide and cholesterol composition of the skin of patients with atopic dermatitis. Acta Derm. Venereol. 1998, 78, 27–30. [CrossRef][PubMed] 37. Geilen, C.C.; Bektas, M.; Wieder, T.; Orfanos, C.E. The vitamin D3 analogue, calcipotriol, induces sphingomyelin hydrolysis in human keratinocytes. FEBS Lett. 1996, 378, 88–92. [CrossRef] 38. Lew, B.-L.; Cho, Y.; Kim, J.; Sim, W.-Y.; Kim, N.-I. Ceramides and cell signaling molecules in psoriatic epidermis: Reduced levels of ceramides, PKC-alpha, and JNK. J. Korean Med. Sci. 2006, 21, 95–99. [CrossRef] [PubMed] 39. Arikawa, J.; Ishibashi, M.; Kawashima, M.; Takagi, Y.; Ichikawa, Y.; Imokawa, G. Decreased levels of sphingosine, a natural antimicrobial agent, may be associated with vulnerability of the stratum corneum from patients with atopic dermatitis to colonization by Staphylococcus aureus. J. Investig. Dermatol. 2002, 119, 433–439. [CrossRef][PubMed] 40. Fluhr, J.W.; Kao, J.; Jain, M.; Ahn, S.K.; Feingold, K.R.; Elias, P.M. Generation of free fatty acids from phospholipids regulates stratum corneum acidiﬁcation and integrity. J. Investig. Dermatol. 2001, 117, 44–51. [CrossRef][PubMed] 41. Feingold, K.R. Thematic review series: Skin lipids. The role of epidermal lipids in cutaneous permeability barrier homeostasis. J. Lipid Res. 2007, 48, 2531–2546. [CrossRef][PubMed] 42. Yang, L.; Mao-Qiang, M.; Taljebini, M.; Elias, P.M.; Feingold, K.R. Topical stratum corneum lipids accelerate barrier repair after tape stripping, solvent treatment and some but not all types of detergent treatment. Br. J. Dermatol. 1995, 133, 679–685. [CrossRef][PubMed] 43. Geilen, C.C.; Wieder, T.; Orfanos, C.E. Ceramide signalling: Regulatory role in cell proliferation, differentiation and apoptosis in human epidermis. Arch. Dermatol. Res. 1997, 289, 559–566. [CrossRef] [PubMed] 44. Ballou, L.R.; Chao, C.P.; Holness, M.A.; Barker, S.C.; Raghow, R. Interleukin-1-mediated PGE2 production and sphingomyelin metabolism. Evidence for the regulation of cyclooxygenase gene expression by sphingosine and ceramide. J. Biol. Chem. 1992, 267, 20044–20050. [PubMed] 45. Vieira, J.; Cardoso, C.S.; Pinto, J.; Patil, K.; Brazdil, P.; Cruz, E.; Mascarenhas, C.; Lacerda, R.; Gartner, A.; Almeida, S.; Alves, H.; Porto, G. A putative gene located at the MHC class I region around the D6S105 marker contributes to the setting of CD8+ T-lymphocyte numbers in humans. Int. J. Immunogenet. 2007, 34, 359–367. [CrossRef][PubMed] 46. Danso, M.O.; van Drongelen, V.; Mulder, A.; van Esch, J.; Scott, H.; van Smeden, J.; El Ghalbzouri, A.; Bouwstra, J.A. TNF-α and Th2 Cytokines Induce Atopic Dermatitis–Like Features on Epidermal Differentiation Proteins and Stratum Corneum Lipids in Human Skin Equivalents. J. Investig. Dermatol. 2014, 134, 1941–1950. [CrossRef][PubMed] Int. J. Mol. Sci. 2017, 18, 1814 16 of 16

47. Rheinwald, J.G.; Green, H. Formation of a keratinizing epithelium in culture by a cloned cell line derived from a teratoma. Cell 1975, 6, 317–330. [CrossRef] 48. Bligh, E.G.; Dyer, W.J. A rapid method of total lipid extraction and puriﬁcation. Can. J. Biochem. Physiol. 1959, 37, 911–917. [CrossRef][PubMed]

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).