- Email: [email protected]
A novel mechanism for improvement of dry skin by dietary milk phospholipids: Effect on epidermal covalently bound ceramides and skin inflammation in hairless mice
Accepted Manuscript Title: A novel mechanism for improvement of dry skin by dietary milk phospholipids: Effect on epidermal covalently-bound ceramides and skin inflammation in hairless mice Author: Masashi Morifuji Chisato Oba Satomi Ichikawa Kyoko Ito Keiko Kawahata Yukio Asami Shuji Ikegami Hiroyuki Itou Tatsuya Sugawara PII: DOI: Reference:
S0923-1811(15)00074-2 http://dx.doi.org/doi:10.1016/j.jdermsci.2015.02.017 DESC 2800
To appear in:
Journal of Dermatological Science
Received date: Revised date: Accepted date:
2-9-2014 23-1-2015 24-2-2015
Please cite this article as: Morifuji M, Oba C, Ichikawa S, Ito K, Kawahata K, Asami Y, Ikegami S, Itou H, Sugawara T, A novel mechanism for improvement of dry skin by dietary milk phospholipids: effect on epidermal covalently-bound ceramides and skin inflammation in hairless mice, Journal of Dermatological Science (2015), http://dx.doi.org/10.1016/j.jdermsci.2015.02.017 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Title Page
2
Title of article
3
A novel mechanism for improvement of dry skin by dietary milk phospholipids: effect
4
on epidermal covalently-bound ceramides and skin inflammation in hairless mice.
cr
ip t
1
us
5
Authors
7
Masashi Morifuji, Ph.D. 1, Chisato Oba 1, Satomi Ichikawa 1, Kyoko Ito 1, Keiko
8
Kawahata 1, Yukio Asami, Ph.D. 1, Shuji Ikegami, Ph.D. 1, Hiroyuki Itou, Ph.D. 1,
9
Tatsuya Sugawara, Ph.D. 2
te
d
M
an
6
Ac ce p
10
11
Affiliations
12
1
13
250-0862, Japan
14
2
15
Kitashirakawaoiwakecho, Sakyo-ku, Kyoto, Kyoto 606-8502, Japan
Food Science Research Labs, Meiji Co., Ltd., 540 Naruda, Odawara-shi, Kanagawa
Division of Applied Biosciences, Graduate School of Agriculture, Kyoto University,
16 1
Page 1 of 41
Corresponding author
18
Masashi Morifuji, Ph.D.
19
Food Science Research Labs, Meiji Co., Ltd., 540 Naruda, Odawara-shi, Kanagawa
20
250-0862, Japan.
21
Tel: +81-465-37-3652. Fax: +81-465-37-3638.
22
E-mail: [email protected]
us
an M
23
cr
ip t
17
Conflict of interest
25
The authors declare no conflict of interests
Ac ce p
26
te
d
24
27
Text word count: 3770
28
Number of references: 40
29
Figures | tables: 5 | 1
2
Page 2 of 41
Abstract
31
Background: Dietary milk phospholipids (MPLs) increase hydration of the stratum
32
corneum and reduced transepidermal water loss (TEWL) in hairless mice fed a standard
33
diet. However, the mechanism by which MPLs improve skin barrier functions has yet to
34
be established.
35
Objective: This study was designed to examine the mechanism by which MPLs may
36
affect covalently-bound ceramides and markers of skin inflammation and improve the
37
skin barrier defect in hairless mice fed a magnesium-deficient (HR-AD) diet.
38
Methods: Four-week-old female hairless mice were randomized into four groups (n =
39
10/group), and fed a standard (control) diet, the HR-AD diet, the HR-AD diet
40
supplemented with either 7.0 g/kg MPLs (low [L]-MPL) or 41.0 g/kg MPLs (high
41
[H]-MPL).
42
Results: Dietary MPLs improved the dry skin condition of hairless mice fed the HR-AD
43
diet. MPLs significantly increased the percentage of covalently-bound -hydroxy
44
ceramides in the epidermis, and significantly decreased both thymus and
45
activation-regulated chemokine (TARC) mRNA and thymic stromal lymphopoietin
46
(TSLP) mRNA levels in skin, compared with the HR-AD diet. Furthermore, the MPL
Ac ce p
te
d
M
an
us
cr
ip t
30
3
Page 3 of 41
diets significantly decreased serum concentrations of immunoglobulin-E, TARC, TSLP,
48
and soluble P-selectin versus the HR-AD diet.
49
Conclusion: Our study showed for the first time that dietary MPLs may modulate
50
epidermal covalently-bound ceramides associated with formation of lamellar structures
51
and suppress skin inflammation, resulting in improved skin barrier function.
us
cr
ip t
47
an
52
Keywords
54
Milk phospholipids, skin barrier function, covalently-bound ceramides, skin
55
inflammation, hairless mice
Ac ce p
te
d
M
53
4
Page 4 of 41
56
1. Introduction
Skin provides an effective barrier between the organism and their environment,
58
helping to reduce the risk of physical, chemical, and microbial damage. Oral intake of
59
dietary components is known to play a beneficial role in improving skin barrier
60
functions [1-6]. Bovine milk lipids contain approximately 0.5% to 1% of phospholipids.
61
Milk phospholipid concentrates cause increased hydration of the stratum corneum (SC)
62
and reduce transepidermal water loss (TEWL) in hairless mice fed a standard diet [3, 4].
63
However, the mechanisms by which milk phospholipids (MPLs) improve skin barrier
64
functions have yet to be established.
d
M
an
us
cr
ip t
57
Reduced barrier function seems to be a consequence of inadequate structural and
66
metabolic conditions in the epidermis. Skin barrier properties are primarily localized in
67
the SC, the outermost layer of the epidermis. The SC consists of corneocytes
68
surrounded by a neutral lipid-enriched intercellular matrix. Ceramides, which comprise
69
approximately 50% of the intercellular lipids, play an important role in retaining
70
epidermal water and, in combination with cholesterol and free fatty acids, they influence
71
permeability of the epidermal barrier [7-9]. Recent research in NC/Nga mice
72
demonstrated that dietary sericin [1] or gromwell [2] improved epidermal skin dryness
73
due to increased levels of non-protein bound glucosylceramides and ceramides and
Ac ce p
te
65
5
Page 5 of 41
up-regulation of glucosylceramide synthase, β-glucocerebrosidase, and acidic
75
sphingomyelinase. The epidermis also contains covalently-bound -hydroxy ceramides
76
that are most frequently bound to a structural protein in the epidermal cornified
77
envelope by an ester linkage. Levels of covalently-bound -hydroxy ceramides were
78
shown to be significantly decreased after ultraviolet-B (UV-B) irradiation,
79
tape-stripping, or treatment with sodium dodecyl sulfate, whereas the levels of
80
non-bound ceramides remained unchanged [10]. Lipid species of this type are therefore
81
thought to play a crucial role in the formation of lamellar structures, and are involved in
82
the maintenance of the skin barrier function [11, 12].
M
an
us
cr
ip t
74
Dry skin conditions, such as atopic dermatitis (AD) and psoriasis vulgaris, cause
84
chronic skin inflammation. Although the pathogenesis of dry skin conditions is not
85
completely understood, it is thought to involve a Th2 cell-mediated allergic
86
inflammatory cascade. Skin injury, caused by environmental allergens, scratching, or
87
microbial toxins, activates keratinocytes to release proinflammatory cytokines and
88
chemokines that induce the expression of adhesion molecules on the vascular
89
endothelium and facilitate the extravasation of inflammatory cells into the skin [13].
90
Serum levels of thymus and activation-regulated chemokine (TARC) [14], thymic
91
stromal lymphopoietin (TSLP) [15], and soluble P-selectin (sP-selectin) [16] appear to
92
be significantly higher in patients with AD than in people without this type of skin
Ac ce p
te
d
83
6
Page 6 of 41
condition. Thus, serum markers of chemokines and platelet activator appear to be useful
94
for assessing the severity of dry skin conditions. However, few studies have been able to
95
demonstrate that the oral intake of dietary components can modulate both
96
covalently-bound -hydroxy ceramides and skin inflammation associated with skin
97
dryness.
us
cr
ip t
93
It is well known that feeding hairless mice a magnesium-deficient diet (HR-AD) for
99
an extended period of time causes a skin barrier defect characterized by an increase in
100
TEWL and a decrease in skin hydration [6, 17-19]. Fujii et al. have previously reported
101
that hairless mice fed the HR-AD diet develop skin inflammation accompanied by a
102
skin barrier defect and itch-related scratching [17, 19]. This study was designed to
103
examine the effect of dietary MPLs on covalently-bound -hydroxy ceramides and skin
104
inflammation markers in hairless mice fed the HR-AD diet with the aim of elucidating a
105
novel mechanism by which MPLs may improve skin barrier function.
Ac ce p
te
d
M
an
98
7
Page 7 of 41
2. Materials and Methods
107
2.1 Animals
ip t
106
Forty four-week-old female hairless mice (Hos:HR-1, Nippon SLC Inc., Shizuoka,
109
Japan) were used in this study. All mice were housed individually in plastic cages in a
110
temperature- and humidity-controlled room (22 ± 1 °C and 50 ± 10% relative humidity,
111
respectively) and maintained on a 12 hr light-dark cycle. All of the animal experiments
112
in this study were approved by Meiji Co., Ltd. Institutional Animal Care and Use
113
Committee, and performed in accordance with the Guiding Principles for the Care and
114
Use of Laboratory Animals approved by Meiji Co., Ltd.
d
M
an
us
cr
108
After acclimatization for seven days, mice were randomized into one of four groups
116
(n = 10/group), and fed either a standard (control) diet (F-2; Funabashi Farm, Chiba,
117
Japan), the HR-AD diet (Norsan Corp., Kanagawa, Japan), or the HR-AD diet
118
supplemented with either 7.0 g/kg (low [L]-MPL) or 41.0 g/kg (high [H]-MPL) of
119
MPLs (Phospholipid concentrate 700, Fonterra Co-operative Group Ltd., New Zealand).
120
Mice were allowed free access to food and water. The MPLs consisted of 16.0%
121
sphingomyelin, 31.0% phosphatidylcholine, 3.0% phosphatidylserine, and 8.7%
122
phosphatidylethanolamine. The remaining components included 25.8% other lipids,
123
2.5% moisture, 6.5% lactose, and 6.0% minerals. The magnesium content in the
Ac ce p
te
115
8
Page 8 of 41
124
experimental diets was 200 mg/g (control), 17.6 mg/g (HR-AD), 17.5 mg/g (L-MPL),
125
and 17.2 mg/g (H-MPL).
TEWL and the water content of the SC was measured at baseline (time 0) and after 2,
127
4, 6, and 8 weeks of feeding. After eight weeks of feeding, all mice were euthanized
128
under isoflurane anesthesia. Blood samples were collected and subsequently centrifuged
129
at 1000 g at 4oC; serum samples were stored at -80oC until analysis. Skin was excised
130
quickly and immediately frozen at -80oC until the time of assay.
an
us
cr
ip t
126
2.2 Measurement of TEWL and water content of the SC
d
132
M
131
The TEWL and water content of the SC were assessed under standardized
134
conditions. Both parameters were measured with a Tewemeter MPA580 (Courage and
135
Khazaka Electronic GmbH, Cologne, Germany) and a Corneometer® (Courage and
136
Khazaka Electronic GmbH, Cologne, Germany), respectively.
138
Ac ce p
137
te
133
2.3 Histological analysis and measurement of epidermal thickness
139
Dorsal skin sections were stained with hematoxylin and eosin (H&E). The thickness
140
of the epidermis (the distance from the bottom of the basal layer to the top of the 9
Page 9 of 41
granular layer) was measured with a biomicroscope BX-2 (Olympus Corporation,
142
Tokyo, Japan) and a CCD camera DP-72 (Olympus Corporation, Tokyo, Japan).
143
Sections were digitally assessed by image measurement and analysis with WinROOF
144
software (Mitani Corporation, Tokyo, Japan).
cr
ip t
141
2.4 Analysis of covalently-bound-hydroxy ceramides
an
146
us
145
Extraction of covalently-bound ceramides was carried out using a modification of the
148
method reported by Macheleidt et al. [20]. Epidermal sheets were obtained from skin
149
samples by treating them with Dispase®II (Roche, IND, USA) at 4oC overnight. Tissues
150
were homogenized in chloroform/methanol (2:1, v/v) using a glass homogenizer. After
151
removal of the supernatant by centrifugation, a chloroform/methanol solution was added
152
to wash the protein residue twice. After drying, protein pellets were incubated in 1 M
153
KOH in 95% methanol at room temperature overnight to release the lipids
154
covalently-bound to the stratum corneum by ester-like bonds. The methanolic layer was
155
removed after centrifugation and neutralized with 1 N HCl. The protein pellets were
156
washed using chloroform/methanol (2:1, v/v). The organic phases were combined, dried,
157
and redissolved in methanol. The protein pellet was immersed into 0.1 M sodium
158
hydroxide solution containing 1% sodium dodecyl sulfate and incubated at 60oC for 2 hr
Ac ce p
te
d
M
147
10
Page 10 of 41
to solubilize the protein. After incubation, the solution was neutralized with 1 N HCl.
160
The protein concentration was assayed using a commercial kit (Micro BCA assay kit,
161
Pierce Biotechnology, Inc., IL, USA).
ip t
159
Covalently-bound ceramides in the mice epidermis were identified using a high
163
performance liquid chromatography system coupled to a tandem mass spectrometer
164
(HPLC-MS/MS) (Quattro premier XE, Waters Corporation, Milford, MA, USA). All the
165
analyses were performed on a 2 X 100 mm column with a particle size of 1.7 µm
166
(ACQUITY UPLC® BEH C18, Waters Corporation). Mobile phase A consisted of 5 mM
167
ammonium acetate in 95% methanol, whereas mobile phase B consisted of 5 mM
168
ammonium acetate in acetonitrile. The initial eluent composition was 100% A, followed
169
by an increase to 100% B for 30 min, 100% B for 2 min, and then a reduction to 0% A
170
for 3 min. Total running time was 35 min. The eluent flow was 0.4 mL/min and the
171
column temperature was set at 40°C. Analytes were detected using electrospray
172
ionization in the positive mode. Multiple-reaction-monitoring (MRM) was performed
173
using characteristic fragmentation ions (m/z 750.7/264.3 for d18:1-C30:0, m/z
174
778.8/264.3 for d18:1-C32:0, m/z 776.8/264.3 for d18:1-C32:1, m/z 806.8/264.3 for
175
d18:1-C34:0, m/z 804.8/264.3 for d18:1-C34:1, m/z 832.8/264.3 for d18:1-C36:1, m/z
176
764.8/250.2 for d17:1-C32:0, m/z 762.8/250.2 for d17:1-C32:1, m/z 790.8/250.3 for
177
d17:1-C34:1, m/z 818.8/250.2 for d17:1-C36:1). The parameters for HPLC-MS/MS
Ac ce p
te
d
M
an
us
cr
162
11
Page 11 of 41
analysis were as follows: capillary voltage 3000 V, source temperature 120°C,
179
desolvation temperature 400°C, desolvation gas flow 850 L/hr, cone gas flow 50 L/hr,
180
cone voltage 40 V, and collision energy 30 eV.
ip t
178
2.4. Skin mRNA analysis
us
182
cr
181
The skin samples were snap frozen in liquid nitrogen and ground to a powder using a
184
mortar and pestle. Total RNA was isolated from skin with TRIzol® reagent (Life
185
Technologies Corporation, Carlsbad, CA, U.S.A.), and purified using the RNeasy Mini
186
Kit (Qiagen, Qiagen, Hilden, Germany). Extracted RNA was then dissolved in
187
diethylpyrocarbonate-treated water. Reverse transcription was used to produce cDNA
188
synthesized with a RivertAid First Strand cDNA Synthesis Kit (Thermo Fisher
189
Scientific, Waltham, MA, USA). The cDNA was stored at -80oC for subsequent
190
analysis.
M
d
te
Ac ce p
191
an
183
Real-time PCR was performed using the ABI 7500 Fast Realtime PCR systems
192
(Applied Biosystems, Foster City, CA, USA). Primers and probes (TaqMan® Gene
193
Expression Assays) were designed at Applied Biosystems from gene sequences obtained
194
from GenBank (TSLP: Mm01157588_m1, TARC: Mm01244826_g1, β-actin [internal
195
control]: Mm00607939_s1). Data were analyzed by 7500 software using 2Ct 12
Page 12 of 41
196
methods [21] and the results expressed as arbitrary units.
2.5. Serum analysis
cr
198
ip t
197
Serum total immunoglobulin E (IgE) concentrations were analyzed using an
200
enzyme-linked immunosorbent assay (ELISA) kit (Shibayagi, Gunma, Japan). Serum
201
TRAC, TSLP, sP-selectin were determined using an ELISA kit (R&D systems, MN,
202
USA).
M
an
us
199
2.6. Statistical analysis
te
204
d
203
All data are presented as means +/- standard error (SEM). Data were subjected to
206
one-way ANOVA with post hoc analyses being carried out using Tukey’s honestly
207
significant difference test (SPSS ver.22.0, SPSS, IL, USA). The correlation between
208
-hydroxy ceramide content and both TEWL and water content of the SC was
209
calculated. Associations between the variables were examined using Pearson’s
210
correlation coefficient. Differences among groups were considered to be significant at P
211
< 0.05.
Ac ce p
205
212
13
Page 13 of 41
3. Results
213
3.1. TEWL and the water content of the SC
ip t
212
Figure 1 shows longitudinal changes in TEWL and the water content of the SC.
215
TEWL was significantly increased in the HR-AD group compared with the control
216
group at 4, 6, and 8 weeks. The addition of MPLs to the HR-AD diet significantly
217
suppressed an increase in TEWL at 4, 6, and 8 weeks. TEWL were significantly lower
218
in the H-MPL group than in the L-MPL group at 6 and 8 weeks. TEWL was not
219
different between the control group and H-MPL group at every time point assessed
220
(Figure 1a).
d
M
an
us
cr
214
A significant decrease in the water content of the SC was observed in the HR-AD
222
group, as compared with the control group at 2, 4, 6, and 8 weeks. Supplementing the
223
HR-AD diet with MPLs significantly increased the water content of the SC at 4, 6, and
224
8 weeks. Skin hydration was significantly higher in the H-MPL group than in the
225
L-MPL group at 6 and 8 weeks. The water content of the SC was not different between
226
the control group and the H-MPL group at 2 and 8 weeks (Figure 1b).
Ac ce p
te
221
227
228
3.2. Appearance of mice, histology and epidermal thickness
14
Page 14 of 41
The overall appearance and histological images of the skin of mice, as well as the
230
epidermal thickness data, are shown in Figure 2. Supplementation with MPLs improved
231
the skin condition in a dose-dependent manner (Figure 2a). In mice fed the HR-AD diet,
232
layered parakeratosis, hyperplasia, keratinocyte apoptosis, and inflammatory cell
233
infiltration were observed (Figure 2b). Epidermal thickness was low in mice fed the
234
control and the H-MPL diet compared with those fed the L-MPL and the HR-AD diet
235
(Figure 2c).
237
3.3. Covalently-bound-hydroxy ceramides in the epidermis
d
M
236
an
us
cr
ip t
229
A typical chromatogram of covalently-bound -hydroxy ceramides in the epidermis
239
of hairless mice obtained using HPLC-MS/MS is shown in Figure 3a. Based on recent
240
research that identified sphingolipids containing three molecular species of sphingoid
241
bases, which were located in the epidermis of hairless mice (specifically d17:1, d18:1,
242
and d18:0) [22], we performed a qualitative analysis of covalently-bound -hydroxy
243
ceramides targeting these molecular species of sphingoid bases using HPLC-MS/MS.
244
From the epidermis of hairless mice, we identified eleven molecular species of
245
protein-bound -hydroxy ceramides which consisted of d18:1-C30:0, d18:1-C32:0,
246
d18:1-C32:1, d18:1-C34:0, d18:1-C34:1, d18:1-C36:1, d17:1-C32:0, d17:1-C32:1,
Ac ce p
te
238
15
Page 15 of 41
247
d17:1-C34:1,
and
d17:1-C36:1.
The
peak
intensity
of
248
tetratriacontenoic-sphingosine (d18:1-C34:1) was highest in the epidermis.
-hydroxy
The data for covalently-bound -hydroxy ceramide percentages in the epidermis are
250
summarized in Table 1. The HR-AD diet significantly suppressed all -hydroxy
251
ceramide molecular species compared with the control diet. The levels of -hydroxy
252
ceramides in d18:1-C32:1 and d18:1-C34:1 were significantly increased in both the
253
H-MPL and the L-MPL groups compared with the HR-AD group, while the levels of
254
other -hydroxy ceramides were significantly increased in the H-MPL group only
255
compared with the HR-AD group. The levels of -hydroxy ceramides linked to very
256
long-chain fatty acids in d18:1-C36:1, d17:1-C36:1 were not different between the
257
control and the H-MPL group.
cr
us
an
M
d
te
Ac ce p
258
ip t
249
Figure 3b shows the correlation between -hydroxy ceramide contents and both
259
TEWL and the water content of the SC. A significant and strong correlation was
260
observed between the covalently-bound -hydroxy ceramide d18:1-C34:1 and skin
261
hydration (r = 0.87, P < 0.001) and the covalently-bound-hydroxy ceramide
262
d18:1-C34:1 and TEWL (r = -0.91, P < 0.001).
263
16
Page 16 of 41
264
3.4. Skin chemokine mRNA levels
Levels of chemokine mRNA in the skin of mice are shown in Figure 4. Significantly
266
higher levels of mRNA for both TSLP and TARC were observed in the HR-AD group
267
compared with the control group. The level of TARC mRNA was significantly lower in
268
both the L-MPL and the H-MPL groups compared with the HR-AD group, whereas the
269
level of TSLP mRNA was significantly lower in the H-MPL group only. Furthermore,
270
the levels of both TSLP and TARC mRNA were low in mice fed the control and the
271
H-MPL diet compared with those fed the L-MPL and HR-AD diet.
M
an
us
cr
ip t
265
te
274
3.5. Serum parameters
Figure 5 shows serum concentrations of IgE, TARC, TSLP, and sP-selectin. Serum
Ac ce p
273
d
272
275
IgE, TARC, TSLP, and sP-selectin were significantly higher in the HR-AD group than
276
the control group. Compared with the HR-AD group, mice fed the L-MPL or H-MPL
277
diet had significantly lower serum concentrations of IgE, TSLP, and sP-selectin; in
278
addition, the H-MPL-fed mice also had significantly lower serum concentrations of
279
TARC. Moreover, the H-MPL diet resulted in lower serum IgE, TARC, TSLP, and
280
sP-selectin concentrations that were not different from the control mice.
281 17
Page 17 of 41
281
4. Discussion
This study showed that dietary MPLs improved dry skin conditions in hairless mice
283
fed the HR-AD diet. Interestingly, MPLs attenuated a decrease in epidermal
284
covalently-bound -hydroxy ceramide levels and an increase in inflammation marker
285
levels in both serum and skin. Thus, MPLs attenuated dry skin conditions by
286
modulating epidermal covalently-bound -hydroxy ceramides that are associated with
287
formation of lamellar structures and skin inflammation in hairless mice fed the HR-AD
288
diet.
M
an
us
cr
ip t
282
It is well known that long-term feeding of the HR-AD diet to hairless mice causes a
290
skin barrier defect characterized by an increase in TEWL and a decrease in skin
291
hydration [6, 17-19]. However, it remains unclear whether reduced barrier function
292
causes a consequential inadequacy of the structural condition of the epidermis. We
293
focused on epidermal covalently-bound -hydroxy ceramides, which are thought to
294
play a most crucial role in the formation of lamellar structures, resulting in retention of
295
epidermal water and in the epidermal barrier [11, 12]. A previous report showed that the
296
amount of protein-bound -hydroxy ceramides in healthy epidermises comprised 46%
297
to 53% of the total protein-bound lipids, whereas this percentage was 10% to 25% lower
298
in human subjects with affected atopic skin areas [20]. We identified covalently-bound
Ac ce p
te
d
289
18
Page 18 of 41
-hydroxy ceramides from murine epidermises using HPLC-MS/MS. Eleven species of
300
covalently-bound long-chain (C30-36) -hydroxy ceramides containing either C18:1 or
301
C17:1 sphingoid base were identified in mice epidermises. Our studies showed the
302
HR-AD diet was significantly associated with a lower level of all molecular species of
303
covalently-bound -hydroxy ceramides in the epidermis of mice. A significant and
304
strong correlation was observed between covalently-bound-hydroxy ceramide
305
(d18:1-C34:1) and skin hydration, as well as covalently-bound-hydroxy ceramide
306
(d18:1-C34:1) and TEWL (Figure 3b). Thus, one possible explanation for the observed
307
dysfunction in the skin barrier may be associated with the reduction in covalently-bound
308
ceramides in the epidermis.
te
d
M
an
us
cr
ip t
299
Considering that ceramides in the stratum corneum comprise more than 350
310
molecular species, it is important to confirm that those specific molecular species of
311
-hydroxy ceramides are implicated in skin barrier function [23]. A recent clinical study
312
showed that the smaller species of ceramides (<40 total carbons) were expressed at
313
significantly higher levels and the larger species (>50 total carbons) were expressed at
314
significantly lower levels in patients with AD sites versus healthy subjects [24]. In
315
addition, it is well known that the ratio of unsaturated and saturated fatty acids in
316
phospholipids influences membrane fluidity. Bouwstra et al. demonstrated that the
317
degree of saturation of the fatty acid chain of ceramide-1 had marked effects on lamellar
Ac ce p
309
19
Page 19 of 41
and lateral lipid organization in vitro [25]. The long periodicity phase was present
319
predominantly in mixtures prepared with ceramide-1 linoleate, and absent in mixtures
320
prepared with ceramide-1 stearate. Our study demonstrated that mice fed the HR-AD
321
diet had notably reduced mean percentages of covalently-bound-hydroxy ceramides
322
with saturated fatty acids (C30:0, C32:0, and C34:0) ranging from 3.5% to 11.4%,
323
whereas ceramides with unsaturated fatty acids (C32:1, C34:1, and C36:1) ranged from
324
14.3% to 34.9% (Table 1). We consider that it is likely that the difference in carbon
325
chain length and the ratio of unsaturated fatty acids to saturated fatty acids of
326
-hydroxy ceramides is also associated with skin barrier structure, although the manner
327
by which such effects may be exerted is not fully understood.
d
M
an
us
cr
ip t
318
This study provides evidence that the oral intake of MPLs improved the skin barrier
329
function in mice fed the HR-AD diet. MPLs contains 16% of sphingomyelin, which
330
consists of three main components; a phosphocholine head group, a sphingoid base, and
331
a fatty acid. Several studies have reported that dietary sphingolipids, such as
332
sphingomyelin
333
Orally-administered milk sphingomyelin was implicated in the water holding capacity
334
of skin in hairless mice [4]. Sphingomyelin from porcine brain was used to accelerate
335
the recoveries of damaged skin caused by skin barrier function impairments in both
336
HR-AD-fed and tape-stripped injured mouse models [6]. Dietary glucosylceramide was
Ac ce p
te
328
and
glucosylceramide,
improve
skin
barrier
function.
20
Page 20 of 41
also shown to improve the recovery of SC flexibility and reduce TEWL in acutely
338
barrier-perturbed mice [6, 26]. Dietary sphingolipids are hydrolyzed by intestinal
339
enzymes to their components, sphingoid bases, fatty acids, and the polar head group and
340
then taken up into mucosal cells [27-30]. Therefore, ingestion of sphingolipids which
341
structurally consist of sphingoid bases may contribute to skin barrier function.
us
cr
ip t
337
MPLs caused significant increases in the percentage of covalently-bound -hydroxy
343
ceramides in the epidermis. Results of a recent in vitro study using cultured
344
keratinocytes showed that sphingadienine (d18:2) enhanced cornified envelope
345
formation via expression of transglutaminase-1 (TGase-1), and played a potential role in
346
the covalent bonding of -hydroxy ceramides to peptide moieties in the outer cornified
347
envelope [31]. Our study showed that the HR-AD diet led to significant increases in
348
mRNA level of TGase-1, while MPLs attenuated the increase in this gene expression in
349
a dose-dependent manner (data not shown). The decrease in TGase-1 mRNA level
350
despite an increase in covalently-bound -hydroxy ceramides is presumably due to
351
negative feedback regulation. Furthermore, milk sphingomyelin consists mainly of
352
sphingosine (d18:1), but not sphingadienine (d18:2) [32]. As a consequence, it is
353
possible that factors other than TGase-1 may regulate the production of
354
covalently-bound -hydroxy ceramides in mice fed MPLs. However, further studies are
355
needed to clarify this mechanism.
Ac ce p
te
d
M
an
342
21
Page 21 of 41
Dry skin conditions, such as AD and psoriasis vulgaris cause chronic skin
357
inflammation. This study provided evidence that levels of TSLP and TARC mRNA, and
358
serum concentrations of IgE, TARC, TSLP, and sP-selectin were significantly higher in
359
the HR-AD group compared with the control group. The focus of this study was on
360
changes in the levels of skin TSLP mRNA and serum concentration changes in TSLP,
361
and these were 20-fold and 110-fold higher, respectively, in mice fed the HR-AD diet,
362
representing a dramatic difference in comparison with mice fed the standard diet.
363
Recent compelling evidence indicates that TSLP may have a determinant role in the
364
initiation and maintenance of allergic immune responses [33, 34]. TSLP modulates
365
polarization of dendritic cells by increasing secretion of Th2 cell-attracting chemokines,
366
such as TARC. TSLP was highly expressed by keratinocytes in AD lesions and its
367
expression was associated with the migration and activation of Langerhans cells [35].
368
Thus, TSLP might be a sensitive marker of dry skin conditions, and play a crucial role
369
in the pathogenesis of such in hairless mice.
cr
us
an
M
d
te
Ac ce p
370
ip t
356
This study showed for the first time that MPLs added to the HR-AD diet fed to mice,
371
attenuated inflammation parameters in both skin and serum. In particular, a high dose of
372
MPLs versus a low dose is more effective at attenuating levels of TSLP in the skin and
373
serum. Several potential mechanisms can be proposed to explain the beneficial effect
374
MPLs can have on skin inflammation. One explanation may be related to the
22
Page 22 of 41
improvement in skin barrier function. Skin barrier dysfunction allows entry of allergens,
376
antigens, and chemicals from the environment, which can activate keratinocytes to
377
release proinflammatory cytokines and chemokines [13]. It is therefore suggested that
378
dietary MPLs help to mitigate skin barrier dysfunction and protect against the entry of
379
environmental agents, thereby attenuating skin inflammation in mice. Another
380
possibility is that sphingolipids may be a source of anti-inflammatory properties.
381
Sphingosine is a potent inhibitor of protein kinase C activity in vitro [36]. Treatment
382
with sphingosines inhibited phorbol ester-induced skin inflammation via inactivation of
383
protein kinase C in mice [37]. In addition, the oral administration of sphingolipids
384
significantly downregulated the activation of TNF- at an inflammatory site [38].
385
Although it remains unclear whether oral intake of sphingolipids directly inhibits skin
386
inflammation, future studies are needed to clarify the potential underlying mechanism.
cr
us
an
M
d
te
Ac ce p
387
ip t
375
In conclusion, dietary MPLs improved dry skin conditions in hairless mice fed the
388
HR-AD diet. Furthermore, adding MPLs attenuated the decrease in epidermal
389
protein-bound -hydroxy ceramide levels associated with the intake of the HR-AD diet,
390
and was associated with a lower level of skin and serum markers of inflammation.
391
MPLs, in particular sphingomyelin, might increase the strength of epidermal lamellar
392
structures and suppress skin inflammation, to thereby improve skin barrier functions. A
393
limitation of this study was that the pathogenic mechanism of dry skin induced by a 23
Page 23 of 41
nutritionally-deficient diet (HR-AD) is different from that induced by epicutaneous
395
sensitization or genetic modification [39, 40]. Further studies are therefore needed to
396
elucidate the beneficial effect of MPLs on dry skin condition using other animal models
397
that have similarities to this condition in humans.
ip t
394
Ac ce p
te
d
M
an
us
cr
398
24
Page 24 of 41
399
Funding sources
400
None
ip t
Acknowledgements
Ac ce p
te
d
M
an
us
cr
398
25
Page 25 of 41
401
References
402
[1] Kim H, Lee J, Cho Y. Dietary sericin enhances epidermal levels of
403
glucosylceramides
404
glucosylceramide synthase, beta-glucocerebrosidase and acidic sphingomyelinase in
405
NC/Nga mice. Nutr Res 2012;32:956-64.
406
[2] Kim J, Cho Y. Gromwell (Lithospermum erythrorhizon) supplementation enhances
407
epidermal levels of ceramides, glucosylceramides, beta-glucocerebrosidase, and acidic
408
sphingomyelinase in NC/Nga mice. J Med Food 2013;16:927-33.
409
[3] Haruta Y, Kato K, Yoshioka T. Dietary phospholipid concentrate from bovine milk
410
improves epidermal function in hairless mice. Biosci Biotechnol Biochem
411
2008;72:2151-7.
412
[4] Haruta-Ono Y, Setoguchi S, Ueno HM, Higurashi S, Ueda N, Kato K, et al. Orally
413
administered sphingomyelin in bovine milk is incorporated into skin sphingolipids and
414
is involved in the water-holding capacity of hairless mice. J Dermatol Sci
415
2012;68:56-62.
416
[5] Oba C, Ohara H, Morifuji M, Ito K, Ichikawa S, Kawahata K, et al. Collagen
417
hydrolysate intake improves the loss of epidermal barrier function and skin elasticity
418
induced by UVB irradiation in hairless mice. Photodermatol Photoimmunol Photomed
419
2013;29:204-11.
420
[6] Duan J, Sugawara T, Hirose M, Aida K, Sakai S, Fujii A, et al. Dietary sphingolipids
expressions
of
cr
ip t
ceramides with up-regulating protein
Ac ce p
te
d
M
an
us
and
26
Page 26 of 41
improve skin barrier functions via the upregulation of ceramide synthases in the
422
epidermis. Exp Dermatol 2012;21:448-52.
423
[7] Schurer NY, Elias PM. The biochemistry and function of stratum corneum lipids.
424
Adv Lipid Res 1991;24:27-56.
425
[8] Elias PM, Menon GK. Structural and lipid biochemical correlates of the epidermal
426
permeability barrier. Adv Lipid Res 1991;24:1-26.
427
[9] Brod J. Characterization and physiological role of epidermal lipids. Int J Dermatol
428
1991;30:84-90.
429
[10] Takagi Y, Nakagawa H, Kondo H, Takema Y, Imokawa G. Decreased levels of
430
covalently bound ceramide are associated with ultraviolet B-induced perturbation of the
431
skin barrier. J Invest Dermatol 2004;123:1102-9.
432
[11] Meguro S, Arai Y, Masukawa Y, Uie K, Tokimitsu I. Relationship between
433
covalently bound ceramides and transepidermal water loss (TEWL). Arch Dermatol Res
434
2000;292:463-8.
435
[12] Behne M, Uchida Y, Seki T, de Montellano PO, Elias PM, Holleran WM.
436
Omega-hydroxyceramides are required for corneocyte lipid envelope (CLE) formation
437
and
438
2000;114:185-92.
439
[13] Leung DY, Boguniewicz M, Howell MD, Nomura I, Hamid QA. New insights into
440
atopic dermatitis. J Clin Invest 2004;113:651-7.
Ac ce p
te
d
M
an
us
cr
ip t
421
normal
epidermal
permeability
barrier
function.
J
Invest
Dermatol
27
Page 27 of 41
[14] Kakinuma T, Nakamura K, Wakugawa M, Mitsui H, Tada Y, Saeki H, et al.
442
Thymus and activation-regulated chemokine in atopic dermatitis: Serum thymus and
443
activation-regulated chemokine level is closely related with disease activity. J Allergy
444
Clin Immunol 2001;107:535-41.
445
[15] Alysandratos KD, Angelidou A, Vasiadi M, Zhang B, Kalogeromitros D,
446
Katsarou-Katsari A, et al. Increased affected skin gene expression and serum levels of
447
thymic stromal lymphopoietin in atopic dermatitis. Ann Allergy Asthma Immunol
448
2010;105:403-4.
449
[16] Tamagawa-Mineoka R, Katoh N, Ueda E, Masuda K, Kishimoto S. Platelet-derived
450
microparticles and soluble P-selectin as platelet activation markers in patients with
451
atopic dermatitis. Clin Immunol 2009;131:495-500.
452
[17] Fujii M, Akita K, Mizutani N, Nabe T, Kohno S. Development of numerous nerve
453
fibers in the epidermis of hairless mice with atopic dermatitis-like pruritic skin
454
inflammation. J Pharmacol Sci 2007;104:243-51.
455
[18] Fujii M, Nakashima H, Tomozawa J, Shimazaki Y, Ohyanagi C, Kawaguchi N, et
456
al. Deficiency of n-6 polyunsaturated fatty acids is mainly responsible for atopic
457
dermatitis-like pruritic skin inflammation in special diet-fed hairless mice. Exp
458
Dermatol 2013;22:272-7.
459
[19] Fujii M, Tomozawa J, Mizutani N, Nabe T, Danno K, Kohno S. Atopic
460
dermatitis-like pruritic skin inflammation caused by feeding a special diet to HR-1
Ac ce p
te
d
M
an
us
cr
ip t
441
28
Page 28 of 41
hairless mice. Exp Dermatol 2005;14:460-8.
462
[20] Macheleidt O, Kaiser HW, Sandhoff K. Deficiency of epidermal protein-bound
463
omega-hydroxyceramides in atopic dermatitis. J Invest Dermatol 2002;119:166-73.
464
[21] Schmittgen TD, Livak KJ. Analyzing real-time PCR data by the comparative C(T)
465
method. Nat Protoc 2008;3:1101-8.
466
[22] Shimada E, Aida K, Sugawara T, Hirata T. Inhibitory effect of topical maize
467
glucosylceramide on skin photoaging in UVA-irradiated hairless mice. J Oleo Sci
468
2011;60:321-5.
469
[23] Masukawa Y, Narita H, Sato H, Naoe A, Kondo N, Sugai Y, et al. Comprehensive
470
quantification of ceramide species in human stratum corneum. J Lipid Res
471
2009;50:1708-19.
472
[24] Ishikawa J, Narita H, Kondo N, Hotta M, Takagi Y, Masukawa Y, et al. Changes in
473
the ceramide profile of atopic dermatitis patients. J Invest Dermatol 2010;130:2511-4.
474
[25] Bouwstra JA, Gooris GS, Dubbelaar FE, Ponec M. Phase behavior of stratum
475
corneum lipid mixtures based on human ceramides: the role of natural and synthetic
476
ceramide 1. J Invest Dermatol 2002;118:606-17.
477
[26] Tsuji K, Mitsutake S, Ishikawa J, Takagi Y, Akiyama M, Shimizu H, et al. Dietary
478
glucosylceramide improves skin barrier function in hairless mice. J Dermatol Sci
479
2006;44:101-7.
480
[27] Sugawara T, Tsuduki T, Yano S, Hirose M, Duan J, Aida K, et al. Intestinal
Ac ce p
te
d
M
an
us
cr
ip t
461
29
Page 29 of 41
absorption of dietary maize glucosylceramide in lymphatic duct cannulated rats. J Lipid
482
Res 2010;51:1761-9.
483
[28] Nilsson A. Metabolism of cerebroside in the intestinal tract of the rat. Biochim
484
Biophys Acta 1969;187:113-21.
485
[29] Nilsson A. Metabolism of sphingomyelin in the intestinal tract of the rat. Biochim
486
Biophys Acta 1968;164:575-84.
487
[30] Schmelz EM, Crall KJ, Larocque R, Dillehay DL, Merrill AH, Jr. Uptake and
488
metabolism of sphingolipids in isolated intestinal loops of mice. J Nutr
489
1994;124:702-12.
490
[31] Hasegawa T, Shimada H, Uchiyama T, Ueda O, Nakashima M, Matsuoka Y.
491
Dietary glucosylceramide enhances cornified envelope formation via transglutaminase
492
expression and involucrin production. Lipids 2011;46:529-35.
493
[32] Byrdwell WC, Perry RH. Liquid chromatography with dual parallel mass
494
spectrometry and 31P nuclear magnetic resonance spectroscopy for analysis of
495
sphingomyelin and dihydrosphingomyelin. II. Bovine milk sphingolipids. J Chromatogr
496
A 2007;1146:164-85.
497
[33] Liu YJ. Thymic stromal lymphopoietin: master switch for allergic inflammation. J
498
Exp Med 2006;203:269-73.
499
[34] Liu YJ, Soumelis V, Watanabe N, Ito T, Wang YH, Malefyt Rde W, et al. TSLP: an
500
epithelial cell cytokine that regulates T cell differentiation by conditioning dendritic cell
Ac ce p
te
d
M
an
us
cr
ip t
481
30
Page 30 of 41
maturation. Annu Rev Immunol 2007;25:193-219.
502
[35] Soumelis V, Reche PA, Kanzler H, Yuan W, Edward G, Homey B, et al. Human
503
epithelial cells trigger dendritic cell mediated allergic inflammation by producing TSLP.
504
Nat Immunol 2002;3:673-80.
505
[36] Hannun YA, Bell RM. Regulation of protein kinase C by sphingosine and
506
lysosphingolipids. Clin Chim Acta 1989;185:333-45.
507
[37] Gupta AK, Fisher GJ, Elder JT, Nickoloff BJ, Voorhees JJ. Sphingosine inhibits
508
phorbol ester-induced inflammation, ornithine decarboxylase activity, and activation of
509
protein kinase C in mouse skin. J Invest Dermatol 1988;91:486-91.
510
[38] Duan J, Sugawara T, Sakai S, Aida K, Hirata T. Oral glucosylceramide reduces
511
2,4-dinitrofluorobenzene induced inflammatory response in mice by reducing
512
TNF-alpha levels and leukocyte infiltration. Lipids 2011;46:505-12.
513
[39] Jin H, He R, Oyoshi M, Geha RS. Animal models of atopic dermatitis. J Invest
514
Dermatol 2009;129:31-40.
515
[40] Tanaka A, Amagai Y, Oida K, Matsuda H. Recent findings in mouse models for
516
human atopic dermatitis. Exp Anim 2012;61:77-84.
Ac ce p
te
d
M
an
us
cr
ip t
501
517 518
31
Page 31 of 41
Figure legends
520
Figure 1
521
Effect of dietary milk phospholipids (MPLs) on changes in transepidermal water loss
522
(TEWL) (a) and the water content of the stratum corneum (b). Values are means +/-
523
SEM, n = 10/group. * Significantly different from the HR-AD group (P < 0.05). †
524
Significantly different from the L-MPL group (P < 0.05). § Significantly different from
525
the H-MPL group (P < 0.05).
M
an
us
cr
ip t
519
d
526
Figure 2
528
Effect of dietary milk phospholipids (MPLs) on the overall skin appearance of mice (a),
529
and skin histological images of mice (b). Arrow head shows layered parakeratosis (A),
530
hyperplasia (B), keratinocyte apoptosis (C), and inflammatory cell infiltration (D).
531
Effect of dietary MPLs on epidermal thickness (c). Values are means +/- SEM, n =
532
10/group. * Significant difference between groups (P < 0.05).
Ac ce p
te
527
533
534
Figure 3
32
Page 32 of 41
Typical chromatogram of covalently-bound -hydroxy ceramides in the epidermis of
536
hairless mice fed the normal diet using HPLC-MS/MS (a). Correlation between
537
covalently-bound -hydroxy ceramide levels and both transepidermal water loss
538
(TEWL) and the water content of the stratum corneum (b).
cr
ip t
535
us
539
Figure 4
541
Effect of dietary milk phospholipids (MPLs) on chemokines, thymic stromal
542
lymphopoietin (TSLP) (a), and thymus and activation-regulated chemokine (TARC) (b),
543
and levels of mRNA in the skin. Values are means +/- SEM, n = 10/group. * Significant
544
difference between groups (P < 0.05).
M
d
te
Ac ce p
545
an
540
546
Figure 5
547
Effect of dietary milk phospholipids (MPLs) on serum immunoglobulin E (IgE) (a),
548
thymus and activation-regulated chemokine (TARC) (b), thymic stromal lymphopoietin
549
(TSLP) (c), and soluble P –selectin (sP-selectin) levels (d). Values are means +/- SEM, n
550
= 10/group. * Significant difference between groups (P < 0.05).
551
33
Page 33 of 41
551
Tables
552
Table 1
553
covalently-bound ω-hydroxy ceramides in the epidermis of mice.
ip t
Effect of dietary milk phospholipids (MPLs) on the percentages of
HR-AD
d18:1 – C30:0
100.0 ± 14.4 * † §
8.3 ± 0.8
d18:1 – C32:0
100.0 ± 7.9 * † §
11.4 ± 1.0
d18:1 – C32:1
100.0 ± 4.7 * † §
28.3 ± 2.0
45.0 ± 1.5 *
63.7 ± 2.9 * †
d18:1 – C34:0
100.0 ± 11.5 * † §
9.1 ± 0.9
11.8 ± 1.0
66.3 ± 6.5 * †
d18:1 – C34:1
100.0 ± 4.2 * † §
M
L-MPL
34.9 ± 2.2
49.8 ± 1.7 *
84.8 ± 3.3 * †
d18:1 – C36:1
100.0 ± 6.8 * †
24.6 ± 1.8
33.1 ± 1.8
97.3 ± 5.4 * †
d17:1 – C32:0
100.0 ± 7.1 * † §
3.9 ± 0.4
5.6 ± 0.5
50.0 ± 4.4 * †
d17:1 – C32:1
100.0 ± 5.2 * † §
14.3 ± 1.1
22.0 ± 1.7
69.1 ± 3.2 * †
us
Control
H-MPL
14.2 ± 1.2
43.4 ± 3.6 * †
16.6 ± 1.1
59.1 ± 4.7 * †
an
d
te
Ac ce p
554
cr
Diet
d17:1 – C34:0
100.0 ± 8.5 * † §
3.5 ± 0.4
4.2 ± 0.5
59.6 ± 5.9 * †
d17:1 – C34:1
100.0 ± 4.0 * † §
17.7 ± 1.2
24.1 ± 1.6
83.7 ± 3.9 * †
d17:1 – C36:1
100.0 ± 6.2 * †
12.2 ± 1.1
15.9 ± 1.4
95.7 ± 6.6 * †
555
Means +/- SEM (n = 10/group)
556
Data expressed as relative peak area per epidermal protein content.
34
Page 34 of 41
* Significantly different from the HR-AD group (P < 0.05).
558
† Significantly different from the L-MPL group (P < 0.05).
559
§ Significantly different from the H-MPL group (P < 0.05).
Ac ce p
te
d
M
an
us
cr
560
ip t
557
35
Page 35 of 41
560
A novel mechanism for improvement of dry skin by dietary milk phospholipids: effect
561
on epidermal covalently-bound ceramides and skin inflammation in hairless mice.
562
Masashi Morifuji, Chisato Oba, Satomi Ichikawa, Kyoko Ito, Keiko Kawahata, Yukio
564
Asami, Shuji Ikegam, Hiroyuki Itou, Tatsuya Sugawara
ip t
563
cr
565
Highlights
567
1.
Dietary milk phospholipids improved dry skin conditions in hairless mice.
568
2.
Milk phospholipids attenuated a decrease in epidermal protein-bound -hydroxy
569
ceramide levels.
570
3.
an
us
566
M
Milk phospholipids lowered levels of skin and serum markers of inflammation.
Ac ce p
te
d
571
36
Page 36 of 41
Figure 1
b
ip t
a 25
60
* †§
cr
*†
*†
*†
*†
4
8
* 5
*
0
0
2
6
us
Stratum corneum water content (Arbitrary unit)
*
pt
10
ed
M
15
Ac ce
TEWL (g/m2/hr)
*
*
an
20
Fig.1
Control HR-AD L-MPL H-MPL
50
* 40
*
*†§
* *
30
*†
*†
*†
*
*
6
8
20
10 0
(weeks)
0
2
4
(weeks)
Page 37 of 41
Figure 2
A
100 mm
Ac ce pt e
D
C
Control
HR-AD
L-MPL
H-MPL
c B
60
Epidermal thickness (μm)
b
M
HR-AD
d
Control
an
us
cr
ip
t
a
*
*
*
*
40
20
0 Control HR-AD
L-MPL
Fig.2
L-MPL
H-MPL
H-MPL
Page 38 of 41
Figure 3
a
←d17:1-C34:1 ←d17:1-C36:1 ←d17:1-C34:0
←d17:1-C32:1
ip t
←d17:1-C32:0
←d18:1-C34:1
cr
←d18:1-C36:1
us
←d18:1-C32:1
←d18:1-C34:0
an
←d18:1-C32:0
11
12
15
16
17
pt
120
18
100
80
80
60
60
40
40
20
20
0
0 10
20
TEWL (g/m2/hr)
20
21
22
120
100
0
19
140
Ac ce
Covalently bound ceramide d18:1-C34:1 (Arbitrary Unit)
140
Fig.3
14
min
ed
b
13
M
←d18:1-C30:0
30
0
20
40
Stratum corneum water content (Arbitrary Unit)
60 Page 39 of 41
Figure 4
*
ip t
*
*
*
* *
4
cr
25
5
us
20 15
an
(Arbitrary Unit)
TSLP mRNA
30
*
(Arbitrary Unit)
*
35
b
TARC mRNA
a
10
M
5 0
2
1 0
Control HR-AD
L-MPL H-MPL
Ac ce
pt
ed
Control HR-AD L-MPL H-MPL
*
3
Fig.4
Page 40 of 41
Figure 5
a 16000
*
*
*
b 500
*
*
*
400
0
1500
1000
ip t Control
M d
500
*
HR-AD
*
L-MPL
H-MPL
*
*
400
*
500
* (ng/mL)
*
Ac ce
(pg/mL)
TSLP
2000
*
100
H-MPL
sP-Selectin
*
*
L-MPL
ed
2500
HR-AD
200
0
pt
Control
300
200
100
0
0
Control
Fig.5
300
cr
an
4000
c
(pg/mL)
TARC
8000
us
(ng/mL)
IgE
12000
HR-AD
L-MPL
H-MPL
Control
HR-AD L-MPL
H-MPL Page 41 of 41
S0923-1811(15)00074-2 http://dx.doi.org/doi:10.1016/j.jdermsci.2015.02.017 DESC 2800
To appear in:
Journal of Dermatological Science
Received date: Revised date: Accepted date:
2-9-2014 23-1-2015 24-2-2015
Please cite this article as: Morifuji M, Oba C, Ichikawa S, Ito K, Kawahata K, Asami Y, Ikegami S, Itou H, Sugawara T, A novel mechanism for improvement of dry skin by dietary milk phospholipids: effect on epidermal covalently-bound ceramides and skin inflammation in hairless mice, Journal of Dermatological Science (2015), http://dx.doi.org/10.1016/j.jdermsci.2015.02.017 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Title Page
2
Title of article
3
A novel mechanism for improvement of dry skin by dietary milk phospholipids: effect
4
on epidermal covalently-bound ceramides and skin inflammation in hairless mice.
cr
ip t
1
us
5
Authors
7
Masashi Morifuji, Ph.D. 1, Chisato Oba 1, Satomi Ichikawa 1, Kyoko Ito 1, Keiko
8
Kawahata 1, Yukio Asami, Ph.D. 1, Shuji Ikegami, Ph.D. 1, Hiroyuki Itou, Ph.D. 1,
9
Tatsuya Sugawara, Ph.D. 2
te
d
M
an
6
Ac ce p
10
11
Affiliations
12
1
13
250-0862, Japan
14
2
15
Kitashirakawaoiwakecho, Sakyo-ku, Kyoto, Kyoto 606-8502, Japan
Food Science Research Labs, Meiji Co., Ltd., 540 Naruda, Odawara-shi, Kanagawa
Division of Applied Biosciences, Graduate School of Agriculture, Kyoto University,
16 1
Page 1 of 41
Corresponding author
18
Masashi Morifuji, Ph.D.
19
Food Science Research Labs, Meiji Co., Ltd., 540 Naruda, Odawara-shi, Kanagawa
20
250-0862, Japan.
21
Tel: +81-465-37-3652. Fax: +81-465-37-3638.
22
E-mail: [email protected]
us
an M
23
cr
ip t
17
Conflict of interest
25
The authors declare no conflict of interests
Ac ce p
26
te
d
24
27
Text word count: 3770
28
Number of references: 40
29
Figures | tables: 5 | 1
2
Page 2 of 41
Abstract
31
Background: Dietary milk phospholipids (MPLs) increase hydration of the stratum
32
corneum and reduced transepidermal water loss (TEWL) in hairless mice fed a standard
33
diet. However, the mechanism by which MPLs improve skin barrier functions has yet to
34
be established.
35
Objective: This study was designed to examine the mechanism by which MPLs may
36
affect covalently-bound ceramides and markers of skin inflammation and improve the
37
skin barrier defect in hairless mice fed a magnesium-deficient (HR-AD) diet.
38
Methods: Four-week-old female hairless mice were randomized into four groups (n =
39
10/group), and fed a standard (control) diet, the HR-AD diet, the HR-AD diet
40
supplemented with either 7.0 g/kg MPLs (low [L]-MPL) or 41.0 g/kg MPLs (high
41
[H]-MPL).
42
Results: Dietary MPLs improved the dry skin condition of hairless mice fed the HR-AD
43
diet. MPLs significantly increased the percentage of covalently-bound -hydroxy
44
ceramides in the epidermis, and significantly decreased both thymus and
45
activation-regulated chemokine (TARC) mRNA and thymic stromal lymphopoietin
46
(TSLP) mRNA levels in skin, compared with the HR-AD diet. Furthermore, the MPL
Ac ce p
te
d
M
an
us
cr
ip t
30
3
Page 3 of 41
diets significantly decreased serum concentrations of immunoglobulin-E, TARC, TSLP,
48
and soluble P-selectin versus the HR-AD diet.
49
Conclusion: Our study showed for the first time that dietary MPLs may modulate
50
epidermal covalently-bound ceramides associated with formation of lamellar structures
51
and suppress skin inflammation, resulting in improved skin barrier function.
us
cr
ip t
47
an
52
Keywords
54
Milk phospholipids, skin barrier function, covalently-bound ceramides, skin
55
inflammation, hairless mice
Ac ce p
te
d
M
53
4
Page 4 of 41
56
1. Introduction
Skin provides an effective barrier between the organism and their environment,
58
helping to reduce the risk of physical, chemical, and microbial damage. Oral intake of
59
dietary components is known to play a beneficial role in improving skin barrier
60
functions [1-6]. Bovine milk lipids contain approximately 0.5% to 1% of phospholipids.
61
Milk phospholipid concentrates cause increased hydration of the stratum corneum (SC)
62
and reduce transepidermal water loss (TEWL) in hairless mice fed a standard diet [3, 4].
63
However, the mechanisms by which milk phospholipids (MPLs) improve skin barrier
64
functions have yet to be established.
d
M
an
us
cr
ip t
57
Reduced barrier function seems to be a consequence of inadequate structural and
66
metabolic conditions in the epidermis. Skin barrier properties are primarily localized in
67
the SC, the outermost layer of the epidermis. The SC consists of corneocytes
68
surrounded by a neutral lipid-enriched intercellular matrix. Ceramides, which comprise
69
approximately 50% of the intercellular lipids, play an important role in retaining
70
epidermal water and, in combination with cholesterol and free fatty acids, they influence
71
permeability of the epidermal barrier [7-9]. Recent research in NC/Nga mice
72
demonstrated that dietary sericin [1] or gromwell [2] improved epidermal skin dryness
73
due to increased levels of non-protein bound glucosylceramides and ceramides and
Ac ce p
te
65
5
Page 5 of 41
up-regulation of glucosylceramide synthase, β-glucocerebrosidase, and acidic
75
sphingomyelinase. The epidermis also contains covalently-bound -hydroxy ceramides
76
that are most frequently bound to a structural protein in the epidermal cornified
77
envelope by an ester linkage. Levels of covalently-bound -hydroxy ceramides were
78
shown to be significantly decreased after ultraviolet-B (UV-B) irradiation,
79
tape-stripping, or treatment with sodium dodecyl sulfate, whereas the levels of
80
non-bound ceramides remained unchanged [10]. Lipid species of this type are therefore
81
thought to play a crucial role in the formation of lamellar structures, and are involved in
82
the maintenance of the skin barrier function [11, 12].
M
an
us
cr
ip t
74
Dry skin conditions, such as atopic dermatitis (AD) and psoriasis vulgaris, cause
84
chronic skin inflammation. Although the pathogenesis of dry skin conditions is not
85
completely understood, it is thought to involve a Th2 cell-mediated allergic
86
inflammatory cascade. Skin injury, caused by environmental allergens, scratching, or
87
microbial toxins, activates keratinocytes to release proinflammatory cytokines and
88
chemokines that induce the expression of adhesion molecules on the vascular
89
endothelium and facilitate the extravasation of inflammatory cells into the skin [13].
90
Serum levels of thymus and activation-regulated chemokine (TARC) [14], thymic
91
stromal lymphopoietin (TSLP) [15], and soluble P-selectin (sP-selectin) [16] appear to
92
be significantly higher in patients with AD than in people without this type of skin
Ac ce p
te
d
83
6
Page 6 of 41
condition. Thus, serum markers of chemokines and platelet activator appear to be useful
94
for assessing the severity of dry skin conditions. However, few studies have been able to
95
demonstrate that the oral intake of dietary components can modulate both
96
covalently-bound -hydroxy ceramides and skin inflammation associated with skin
97
dryness.
us
cr
ip t
93
It is well known that feeding hairless mice a magnesium-deficient diet (HR-AD) for
99
an extended period of time causes a skin barrier defect characterized by an increase in
100
TEWL and a decrease in skin hydration [6, 17-19]. Fujii et al. have previously reported
101
that hairless mice fed the HR-AD diet develop skin inflammation accompanied by a
102
skin barrier defect and itch-related scratching [17, 19]. This study was designed to
103
examine the effect of dietary MPLs on covalently-bound -hydroxy ceramides and skin
104
inflammation markers in hairless mice fed the HR-AD diet with the aim of elucidating a
105
novel mechanism by which MPLs may improve skin barrier function.
Ac ce p
te
d
M
an
98
7
Page 7 of 41
2. Materials and Methods
107
2.1 Animals
ip t
106
Forty four-week-old female hairless mice (Hos:HR-1, Nippon SLC Inc., Shizuoka,
109
Japan) were used in this study. All mice were housed individually in plastic cages in a
110
temperature- and humidity-controlled room (22 ± 1 °C and 50 ± 10% relative humidity,
111
respectively) and maintained on a 12 hr light-dark cycle. All of the animal experiments
112
in this study were approved by Meiji Co., Ltd. Institutional Animal Care and Use
113
Committee, and performed in accordance with the Guiding Principles for the Care and
114
Use of Laboratory Animals approved by Meiji Co., Ltd.
d
M
an
us
cr
108
After acclimatization for seven days, mice were randomized into one of four groups
116
(n = 10/group), and fed either a standard (control) diet (F-2; Funabashi Farm, Chiba,
117
Japan), the HR-AD diet (Norsan Corp., Kanagawa, Japan), or the HR-AD diet
118
supplemented with either 7.0 g/kg (low [L]-MPL) or 41.0 g/kg (high [H]-MPL) of
119
MPLs (Phospholipid concentrate 700, Fonterra Co-operative Group Ltd., New Zealand).
120
Mice were allowed free access to food and water. The MPLs consisted of 16.0%
121
sphingomyelin, 31.0% phosphatidylcholine, 3.0% phosphatidylserine, and 8.7%
122
phosphatidylethanolamine. The remaining components included 25.8% other lipids,
123
2.5% moisture, 6.5% lactose, and 6.0% minerals. The magnesium content in the
Ac ce p
te
115
8
Page 8 of 41
124
experimental diets was 200 mg/g (control), 17.6 mg/g (HR-AD), 17.5 mg/g (L-MPL),
125
and 17.2 mg/g (H-MPL).
TEWL and the water content of the SC was measured at baseline (time 0) and after 2,
127
4, 6, and 8 weeks of feeding. After eight weeks of feeding, all mice were euthanized
128
under isoflurane anesthesia. Blood samples were collected and subsequently centrifuged
129
at 1000 g at 4oC; serum samples were stored at -80oC until analysis. Skin was excised
130
quickly and immediately frozen at -80oC until the time of assay.
an
us
cr
ip t
126
2.2 Measurement of TEWL and water content of the SC
d
132
M
131
The TEWL and water content of the SC were assessed under standardized
134
conditions. Both parameters were measured with a Tewemeter MPA580 (Courage and
135
Khazaka Electronic GmbH, Cologne, Germany) and a Corneometer® (Courage and
136
Khazaka Electronic GmbH, Cologne, Germany), respectively.
138
Ac ce p
137
te
133
2.3 Histological analysis and measurement of epidermal thickness
139
Dorsal skin sections were stained with hematoxylin and eosin (H&E). The thickness
140
of the epidermis (the distance from the bottom of the basal layer to the top of the 9
Page 9 of 41
granular layer) was measured with a biomicroscope BX-2 (Olympus Corporation,
142
Tokyo, Japan) and a CCD camera DP-72 (Olympus Corporation, Tokyo, Japan).
143
Sections were digitally assessed by image measurement and analysis with WinROOF
144
software (Mitani Corporation, Tokyo, Japan).
cr
ip t
141
2.4 Analysis of covalently-bound-hydroxy ceramides
an
146
us
145
Extraction of covalently-bound ceramides was carried out using a modification of the
148
method reported by Macheleidt et al. [20]. Epidermal sheets were obtained from skin
149
samples by treating them with Dispase®II (Roche, IND, USA) at 4oC overnight. Tissues
150
were homogenized in chloroform/methanol (2:1, v/v) using a glass homogenizer. After
151
removal of the supernatant by centrifugation, a chloroform/methanol solution was added
152
to wash the protein residue twice. After drying, protein pellets were incubated in 1 M
153
KOH in 95% methanol at room temperature overnight to release the lipids
154
covalently-bound to the stratum corneum by ester-like bonds. The methanolic layer was
155
removed after centrifugation and neutralized with 1 N HCl. The protein pellets were
156
washed using chloroform/methanol (2:1, v/v). The organic phases were combined, dried,
157
and redissolved in methanol. The protein pellet was immersed into 0.1 M sodium
158
hydroxide solution containing 1% sodium dodecyl sulfate and incubated at 60oC for 2 hr
Ac ce p
te
d
M
147
10
Page 10 of 41
to solubilize the protein. After incubation, the solution was neutralized with 1 N HCl.
160
The protein concentration was assayed using a commercial kit (Micro BCA assay kit,
161
Pierce Biotechnology, Inc., IL, USA).
ip t
159
Covalently-bound ceramides in the mice epidermis were identified using a high
163
performance liquid chromatography system coupled to a tandem mass spectrometer
164
(HPLC-MS/MS) (Quattro premier XE, Waters Corporation, Milford, MA, USA). All the
165
analyses were performed on a 2 X 100 mm column with a particle size of 1.7 µm
166
(ACQUITY UPLC® BEH C18, Waters Corporation). Mobile phase A consisted of 5 mM
167
ammonium acetate in 95% methanol, whereas mobile phase B consisted of 5 mM
168
ammonium acetate in acetonitrile. The initial eluent composition was 100% A, followed
169
by an increase to 100% B for 30 min, 100% B for 2 min, and then a reduction to 0% A
170
for 3 min. Total running time was 35 min. The eluent flow was 0.4 mL/min and the
171
column temperature was set at 40°C. Analytes were detected using electrospray
172
ionization in the positive mode. Multiple-reaction-monitoring (MRM) was performed
173
using characteristic fragmentation ions (m/z 750.7/264.3 for d18:1-C30:0, m/z
174
778.8/264.3 for d18:1-C32:0, m/z 776.8/264.3 for d18:1-C32:1, m/z 806.8/264.3 for
175
d18:1-C34:0, m/z 804.8/264.3 for d18:1-C34:1, m/z 832.8/264.3 for d18:1-C36:1, m/z
176
764.8/250.2 for d17:1-C32:0, m/z 762.8/250.2 for d17:1-C32:1, m/z 790.8/250.3 for
177
d17:1-C34:1, m/z 818.8/250.2 for d17:1-C36:1). The parameters for HPLC-MS/MS
Ac ce p
te
d
M
an
us
cr
162
11
Page 11 of 41
analysis were as follows: capillary voltage 3000 V, source temperature 120°C,
179
desolvation temperature 400°C, desolvation gas flow 850 L/hr, cone gas flow 50 L/hr,
180
cone voltage 40 V, and collision energy 30 eV.
ip t
178
2.4. Skin mRNA analysis
us
182
cr
181
The skin samples were snap frozen in liquid nitrogen and ground to a powder using a
184
mortar and pestle. Total RNA was isolated from skin with TRIzol® reagent (Life
185
Technologies Corporation, Carlsbad, CA, U.S.A.), and purified using the RNeasy Mini
186
Kit (Qiagen, Qiagen, Hilden, Germany). Extracted RNA was then dissolved in
187
diethylpyrocarbonate-treated water. Reverse transcription was used to produce cDNA
188
synthesized with a RivertAid First Strand cDNA Synthesis Kit (Thermo Fisher
189
Scientific, Waltham, MA, USA). The cDNA was stored at -80oC for subsequent
190
analysis.
M
d
te
Ac ce p
191
an
183
Real-time PCR was performed using the ABI 7500 Fast Realtime PCR systems
192
(Applied Biosystems, Foster City, CA, USA). Primers and probes (TaqMan® Gene
193
Expression Assays) were designed at Applied Biosystems from gene sequences obtained
194
from GenBank (TSLP: Mm01157588_m1, TARC: Mm01244826_g1, β-actin [internal
195
control]: Mm00607939_s1). Data were analyzed by 7500 software using 2Ct 12
Page 12 of 41
196
methods [21] and the results expressed as arbitrary units.
2.5. Serum analysis
cr
198
ip t
197
Serum total immunoglobulin E (IgE) concentrations were analyzed using an
200
enzyme-linked immunosorbent assay (ELISA) kit (Shibayagi, Gunma, Japan). Serum
201
TRAC, TSLP, sP-selectin were determined using an ELISA kit (R&D systems, MN,
202
USA).
M
an
us
199
2.6. Statistical analysis
te
204
d
203
All data are presented as means +/- standard error (SEM). Data were subjected to
206
one-way ANOVA with post hoc analyses being carried out using Tukey’s honestly
207
significant difference test (SPSS ver.22.0, SPSS, IL, USA). The correlation between
208
-hydroxy ceramide content and both TEWL and water content of the SC was
209
calculated. Associations between the variables were examined using Pearson’s
210
correlation coefficient. Differences among groups were considered to be significant at P
211
< 0.05.
Ac ce p
205
212
13
Page 13 of 41
3. Results
213
3.1. TEWL and the water content of the SC
ip t
212
Figure 1 shows longitudinal changes in TEWL and the water content of the SC.
215
TEWL was significantly increased in the HR-AD group compared with the control
216
group at 4, 6, and 8 weeks. The addition of MPLs to the HR-AD diet significantly
217
suppressed an increase in TEWL at 4, 6, and 8 weeks. TEWL were significantly lower
218
in the H-MPL group than in the L-MPL group at 6 and 8 weeks. TEWL was not
219
different between the control group and H-MPL group at every time point assessed
220
(Figure 1a).
d
M
an
us
cr
214
A significant decrease in the water content of the SC was observed in the HR-AD
222
group, as compared with the control group at 2, 4, 6, and 8 weeks. Supplementing the
223
HR-AD diet with MPLs significantly increased the water content of the SC at 4, 6, and
224
8 weeks. Skin hydration was significantly higher in the H-MPL group than in the
225
L-MPL group at 6 and 8 weeks. The water content of the SC was not different between
226
the control group and the H-MPL group at 2 and 8 weeks (Figure 1b).
Ac ce p
te
221
227
228
3.2. Appearance of mice, histology and epidermal thickness
14
Page 14 of 41
The overall appearance and histological images of the skin of mice, as well as the
230
epidermal thickness data, are shown in Figure 2. Supplementation with MPLs improved
231
the skin condition in a dose-dependent manner (Figure 2a). In mice fed the HR-AD diet,
232
layered parakeratosis, hyperplasia, keratinocyte apoptosis, and inflammatory cell
233
infiltration were observed (Figure 2b). Epidermal thickness was low in mice fed the
234
control and the H-MPL diet compared with those fed the L-MPL and the HR-AD diet
235
(Figure 2c).
237
3.3. Covalently-bound-hydroxy ceramides in the epidermis
d
M
236
an
us
cr
ip t
229
A typical chromatogram of covalently-bound -hydroxy ceramides in the epidermis
239
of hairless mice obtained using HPLC-MS/MS is shown in Figure 3a. Based on recent
240
research that identified sphingolipids containing three molecular species of sphingoid
241
bases, which were located in the epidermis of hairless mice (specifically d17:1, d18:1,
242
and d18:0) [22], we performed a qualitative analysis of covalently-bound -hydroxy
243
ceramides targeting these molecular species of sphingoid bases using HPLC-MS/MS.
244
From the epidermis of hairless mice, we identified eleven molecular species of
245
protein-bound -hydroxy ceramides which consisted of d18:1-C30:0, d18:1-C32:0,
246
d18:1-C32:1, d18:1-C34:0, d18:1-C34:1, d18:1-C36:1, d17:1-C32:0, d17:1-C32:1,
Ac ce p
te
238
15
Page 15 of 41
247
d17:1-C34:1,
and
d17:1-C36:1.
The
peak
intensity
of
248
tetratriacontenoic-sphingosine (d18:1-C34:1) was highest in the epidermis.
-hydroxy
The data for covalently-bound -hydroxy ceramide percentages in the epidermis are
250
summarized in Table 1. The HR-AD diet significantly suppressed all -hydroxy
251
ceramide molecular species compared with the control diet. The levels of -hydroxy
252
ceramides in d18:1-C32:1 and d18:1-C34:1 were significantly increased in both the
253
H-MPL and the L-MPL groups compared with the HR-AD group, while the levels of
254
other -hydroxy ceramides were significantly increased in the H-MPL group only
255
compared with the HR-AD group. The levels of -hydroxy ceramides linked to very
256
long-chain fatty acids in d18:1-C36:1, d17:1-C36:1 were not different between the
257
control and the H-MPL group.
cr
us
an
M
d
te
Ac ce p
258
ip t
249
Figure 3b shows the correlation between -hydroxy ceramide contents and both
259
TEWL and the water content of the SC. A significant and strong correlation was
260
observed between the covalently-bound -hydroxy ceramide d18:1-C34:1 and skin
261
hydration (r = 0.87, P < 0.001) and the covalently-bound-hydroxy ceramide
262
d18:1-C34:1 and TEWL (r = -0.91, P < 0.001).
263
16
Page 16 of 41
264
3.4. Skin chemokine mRNA levels
Levels of chemokine mRNA in the skin of mice are shown in Figure 4. Significantly
266
higher levels of mRNA for both TSLP and TARC were observed in the HR-AD group
267
compared with the control group. The level of TARC mRNA was significantly lower in
268
both the L-MPL and the H-MPL groups compared with the HR-AD group, whereas the
269
level of TSLP mRNA was significantly lower in the H-MPL group only. Furthermore,
270
the levels of both TSLP and TARC mRNA were low in mice fed the control and the
271
H-MPL diet compared with those fed the L-MPL and HR-AD diet.
M
an
us
cr
ip t
265
te
274
3.5. Serum parameters
Figure 5 shows serum concentrations of IgE, TARC, TSLP, and sP-selectin. Serum
Ac ce p
273
d
272
275
IgE, TARC, TSLP, and sP-selectin were significantly higher in the HR-AD group than
276
the control group. Compared with the HR-AD group, mice fed the L-MPL or H-MPL
277
diet had significantly lower serum concentrations of IgE, TSLP, and sP-selectin; in
278
addition, the H-MPL-fed mice also had significantly lower serum concentrations of
279
TARC. Moreover, the H-MPL diet resulted in lower serum IgE, TARC, TSLP, and
280
sP-selectin concentrations that were not different from the control mice.
281 17
Page 17 of 41
281
4. Discussion
This study showed that dietary MPLs improved dry skin conditions in hairless mice
283
fed the HR-AD diet. Interestingly, MPLs attenuated a decrease in epidermal
284
covalently-bound -hydroxy ceramide levels and an increase in inflammation marker
285
levels in both serum and skin. Thus, MPLs attenuated dry skin conditions by
286
modulating epidermal covalently-bound -hydroxy ceramides that are associated with
287
formation of lamellar structures and skin inflammation in hairless mice fed the HR-AD
288
diet.
M
an
us
cr
ip t
282
It is well known that long-term feeding of the HR-AD diet to hairless mice causes a
290
skin barrier defect characterized by an increase in TEWL and a decrease in skin
291
hydration [6, 17-19]. However, it remains unclear whether reduced barrier function
292
causes a consequential inadequacy of the structural condition of the epidermis. We
293
focused on epidermal covalently-bound -hydroxy ceramides, which are thought to
294
play a most crucial role in the formation of lamellar structures, resulting in retention of
295
epidermal water and in the epidermal barrier [11, 12]. A previous report showed that the
296
amount of protein-bound -hydroxy ceramides in healthy epidermises comprised 46%
297
to 53% of the total protein-bound lipids, whereas this percentage was 10% to 25% lower
298
in human subjects with affected atopic skin areas [20]. We identified covalently-bound
Ac ce p
te
d
289
18
Page 18 of 41
-hydroxy ceramides from murine epidermises using HPLC-MS/MS. Eleven species of
300
covalently-bound long-chain (C30-36) -hydroxy ceramides containing either C18:1 or
301
C17:1 sphingoid base were identified in mice epidermises. Our studies showed the
302
HR-AD diet was significantly associated with a lower level of all molecular species of
303
covalently-bound -hydroxy ceramides in the epidermis of mice. A significant and
304
strong correlation was observed between covalently-bound-hydroxy ceramide
305
(d18:1-C34:1) and skin hydration, as well as covalently-bound-hydroxy ceramide
306
(d18:1-C34:1) and TEWL (Figure 3b). Thus, one possible explanation for the observed
307
dysfunction in the skin barrier may be associated with the reduction in covalently-bound
308
ceramides in the epidermis.
te
d
M
an
us
cr
ip t
299
Considering that ceramides in the stratum corneum comprise more than 350
310
molecular species, it is important to confirm that those specific molecular species of
311
-hydroxy ceramides are implicated in skin barrier function [23]. A recent clinical study
312
showed that the smaller species of ceramides (<40 total carbons) were expressed at
313
significantly higher levels and the larger species (>50 total carbons) were expressed at
314
significantly lower levels in patients with AD sites versus healthy subjects [24]. In
315
addition, it is well known that the ratio of unsaturated and saturated fatty acids in
316
phospholipids influences membrane fluidity. Bouwstra et al. demonstrated that the
317
degree of saturation of the fatty acid chain of ceramide-1 had marked effects on lamellar
Ac ce p
309
19
Page 19 of 41
and lateral lipid organization in vitro [25]. The long periodicity phase was present
319
predominantly in mixtures prepared with ceramide-1 linoleate, and absent in mixtures
320
prepared with ceramide-1 stearate. Our study demonstrated that mice fed the HR-AD
321
diet had notably reduced mean percentages of covalently-bound-hydroxy ceramides
322
with saturated fatty acids (C30:0, C32:0, and C34:0) ranging from 3.5% to 11.4%,
323
whereas ceramides with unsaturated fatty acids (C32:1, C34:1, and C36:1) ranged from
324
14.3% to 34.9% (Table 1). We consider that it is likely that the difference in carbon
325
chain length and the ratio of unsaturated fatty acids to saturated fatty acids of
326
-hydroxy ceramides is also associated with skin barrier structure, although the manner
327
by which such effects may be exerted is not fully understood.
d
M
an
us
cr
ip t
318
This study provides evidence that the oral intake of MPLs improved the skin barrier
329
function in mice fed the HR-AD diet. MPLs contains 16% of sphingomyelin, which
330
consists of three main components; a phosphocholine head group, a sphingoid base, and
331
a fatty acid. Several studies have reported that dietary sphingolipids, such as
332
sphingomyelin
333
Orally-administered milk sphingomyelin was implicated in the water holding capacity
334
of skin in hairless mice [4]. Sphingomyelin from porcine brain was used to accelerate
335
the recoveries of damaged skin caused by skin barrier function impairments in both
336
HR-AD-fed and tape-stripped injured mouse models [6]. Dietary glucosylceramide was
Ac ce p
te
328
and
glucosylceramide,
improve
skin
barrier
function.
20
Page 20 of 41
also shown to improve the recovery of SC flexibility and reduce TEWL in acutely
338
barrier-perturbed mice [6, 26]. Dietary sphingolipids are hydrolyzed by intestinal
339
enzymes to their components, sphingoid bases, fatty acids, and the polar head group and
340
then taken up into mucosal cells [27-30]. Therefore, ingestion of sphingolipids which
341
structurally consist of sphingoid bases may contribute to skin barrier function.
us
cr
ip t
337
MPLs caused significant increases in the percentage of covalently-bound -hydroxy
343
ceramides in the epidermis. Results of a recent in vitro study using cultured
344
keratinocytes showed that sphingadienine (d18:2) enhanced cornified envelope
345
formation via expression of transglutaminase-1 (TGase-1), and played a potential role in
346
the covalent bonding of -hydroxy ceramides to peptide moieties in the outer cornified
347
envelope [31]. Our study showed that the HR-AD diet led to significant increases in
348
mRNA level of TGase-1, while MPLs attenuated the increase in this gene expression in
349
a dose-dependent manner (data not shown). The decrease in TGase-1 mRNA level
350
despite an increase in covalently-bound -hydroxy ceramides is presumably due to
351
negative feedback regulation. Furthermore, milk sphingomyelin consists mainly of
352
sphingosine (d18:1), but not sphingadienine (d18:2) [32]. As a consequence, it is
353
possible that factors other than TGase-1 may regulate the production of
354
covalently-bound -hydroxy ceramides in mice fed MPLs. However, further studies are
355
needed to clarify this mechanism.
Ac ce p
te
d
M
an
342
21
Page 21 of 41
Dry skin conditions, such as AD and psoriasis vulgaris cause chronic skin
357
inflammation. This study provided evidence that levels of TSLP and TARC mRNA, and
358
serum concentrations of IgE, TARC, TSLP, and sP-selectin were significantly higher in
359
the HR-AD group compared with the control group. The focus of this study was on
360
changes in the levels of skin TSLP mRNA and serum concentration changes in TSLP,
361
and these were 20-fold and 110-fold higher, respectively, in mice fed the HR-AD diet,
362
representing a dramatic difference in comparison with mice fed the standard diet.
363
Recent compelling evidence indicates that TSLP may have a determinant role in the
364
initiation and maintenance of allergic immune responses [33, 34]. TSLP modulates
365
polarization of dendritic cells by increasing secretion of Th2 cell-attracting chemokines,
366
such as TARC. TSLP was highly expressed by keratinocytes in AD lesions and its
367
expression was associated with the migration and activation of Langerhans cells [35].
368
Thus, TSLP might be a sensitive marker of dry skin conditions, and play a crucial role
369
in the pathogenesis of such in hairless mice.
cr
us
an
M
d
te
Ac ce p
370
ip t
356
This study showed for the first time that MPLs added to the HR-AD diet fed to mice,
371
attenuated inflammation parameters in both skin and serum. In particular, a high dose of
372
MPLs versus a low dose is more effective at attenuating levels of TSLP in the skin and
373
serum. Several potential mechanisms can be proposed to explain the beneficial effect
374
MPLs can have on skin inflammation. One explanation may be related to the
22
Page 22 of 41
improvement in skin barrier function. Skin barrier dysfunction allows entry of allergens,
376
antigens, and chemicals from the environment, which can activate keratinocytes to
377
release proinflammatory cytokines and chemokines [13]. It is therefore suggested that
378
dietary MPLs help to mitigate skin barrier dysfunction and protect against the entry of
379
environmental agents, thereby attenuating skin inflammation in mice. Another
380
possibility is that sphingolipids may be a source of anti-inflammatory properties.
381
Sphingosine is a potent inhibitor of protein kinase C activity in vitro [36]. Treatment
382
with sphingosines inhibited phorbol ester-induced skin inflammation via inactivation of
383
protein kinase C in mice [37]. In addition, the oral administration of sphingolipids
384
significantly downregulated the activation of TNF- at an inflammatory site [38].
385
Although it remains unclear whether oral intake of sphingolipids directly inhibits skin
386
inflammation, future studies are needed to clarify the potential underlying mechanism.
cr
us
an
M
d
te
Ac ce p
387
ip t
375
In conclusion, dietary MPLs improved dry skin conditions in hairless mice fed the
388
HR-AD diet. Furthermore, adding MPLs attenuated the decrease in epidermal
389
protein-bound -hydroxy ceramide levels associated with the intake of the HR-AD diet,
390
and was associated with a lower level of skin and serum markers of inflammation.
391
MPLs, in particular sphingomyelin, might increase the strength of epidermal lamellar
392
structures and suppress skin inflammation, to thereby improve skin barrier functions. A
393
limitation of this study was that the pathogenic mechanism of dry skin induced by a 23
Page 23 of 41
nutritionally-deficient diet (HR-AD) is different from that induced by epicutaneous
395
sensitization or genetic modification [39, 40]. Further studies are therefore needed to
396
elucidate the beneficial effect of MPLs on dry skin condition using other animal models
397
that have similarities to this condition in humans.
ip t
394
Ac ce p
te
d
M
an
us
cr
398
24
Page 24 of 41
399
Funding sources
400
None
ip t
Acknowledgements
Ac ce p
te
d
M
an
us
cr
398
25
Page 25 of 41
401
References
402
[1] Kim H, Lee J, Cho Y. Dietary sericin enhances epidermal levels of
403
glucosylceramides
404
glucosylceramide synthase, beta-glucocerebrosidase and acidic sphingomyelinase in
405
NC/Nga mice. Nutr Res 2012;32:956-64.
406
[2] Kim J, Cho Y. Gromwell (Lithospermum erythrorhizon) supplementation enhances
407
epidermal levels of ceramides, glucosylceramides, beta-glucocerebrosidase, and acidic
408
sphingomyelinase in NC/Nga mice. J Med Food 2013;16:927-33.
409
[3] Haruta Y, Kato K, Yoshioka T. Dietary phospholipid concentrate from bovine milk
410
improves epidermal function in hairless mice. Biosci Biotechnol Biochem
411
2008;72:2151-7.
412
[4] Haruta-Ono Y, Setoguchi S, Ueno HM, Higurashi S, Ueda N, Kato K, et al. Orally
413
administered sphingomyelin in bovine milk is incorporated into skin sphingolipids and
414
is involved in the water-holding capacity of hairless mice. J Dermatol Sci
415
2012;68:56-62.
416
[5] Oba C, Ohara H, Morifuji M, Ito K, Ichikawa S, Kawahata K, et al. Collagen
417
hydrolysate intake improves the loss of epidermal barrier function and skin elasticity
418
induced by UVB irradiation in hairless mice. Photodermatol Photoimmunol Photomed
419
2013;29:204-11.
420
[6] Duan J, Sugawara T, Hirose M, Aida K, Sakai S, Fujii A, et al. Dietary sphingolipids
expressions
of
cr
ip t
ceramides with up-regulating protein
Ac ce p
te
d
M
an
us
and
26
Page 26 of 41
improve skin barrier functions via the upregulation of ceramide synthases in the
422
epidermis. Exp Dermatol 2012;21:448-52.
423
[7] Schurer NY, Elias PM. The biochemistry and function of stratum corneum lipids.
424
Adv Lipid Res 1991;24:27-56.
425
[8] Elias PM, Menon GK. Structural and lipid biochemical correlates of the epidermal
426
permeability barrier. Adv Lipid Res 1991;24:1-26.
427
[9] Brod J. Characterization and physiological role of epidermal lipids. Int J Dermatol
428
1991;30:84-90.
429
[10] Takagi Y, Nakagawa H, Kondo H, Takema Y, Imokawa G. Decreased levels of
430
covalently bound ceramide are associated with ultraviolet B-induced perturbation of the
431
skin barrier. J Invest Dermatol 2004;123:1102-9.
432
[11] Meguro S, Arai Y, Masukawa Y, Uie K, Tokimitsu I. Relationship between
433
covalently bound ceramides and transepidermal water loss (TEWL). Arch Dermatol Res
434
2000;292:463-8.
435
[12] Behne M, Uchida Y, Seki T, de Montellano PO, Elias PM, Holleran WM.
436
Omega-hydroxyceramides are required for corneocyte lipid envelope (CLE) formation
437
and
438
2000;114:185-92.
439
[13] Leung DY, Boguniewicz M, Howell MD, Nomura I, Hamid QA. New insights into
440
atopic dermatitis. J Clin Invest 2004;113:651-7.
Ac ce p
te
d
M
an
us
cr
ip t
421
normal
epidermal
permeability
barrier
function.
J
Invest
Dermatol
27
Page 27 of 41
[14] Kakinuma T, Nakamura K, Wakugawa M, Mitsui H, Tada Y, Saeki H, et al.
442
Thymus and activation-regulated chemokine in atopic dermatitis: Serum thymus and
443
activation-regulated chemokine level is closely related with disease activity. J Allergy
444
Clin Immunol 2001;107:535-41.
445
[15] Alysandratos KD, Angelidou A, Vasiadi M, Zhang B, Kalogeromitros D,
446
Katsarou-Katsari A, et al. Increased affected skin gene expression and serum levels of
447
thymic stromal lymphopoietin in atopic dermatitis. Ann Allergy Asthma Immunol
448
2010;105:403-4.
449
[16] Tamagawa-Mineoka R, Katoh N, Ueda E, Masuda K, Kishimoto S. Platelet-derived
450
microparticles and soluble P-selectin as platelet activation markers in patients with
451
atopic dermatitis. Clin Immunol 2009;131:495-500.
452
[17] Fujii M, Akita K, Mizutani N, Nabe T, Kohno S. Development of numerous nerve
453
fibers in the epidermis of hairless mice with atopic dermatitis-like pruritic skin
454
inflammation. J Pharmacol Sci 2007;104:243-51.
455
[18] Fujii M, Nakashima H, Tomozawa J, Shimazaki Y, Ohyanagi C, Kawaguchi N, et
456
al. Deficiency of n-6 polyunsaturated fatty acids is mainly responsible for atopic
457
dermatitis-like pruritic skin inflammation in special diet-fed hairless mice. Exp
458
Dermatol 2013;22:272-7.
459
[19] Fujii M, Tomozawa J, Mizutani N, Nabe T, Danno K, Kohno S. Atopic
460
dermatitis-like pruritic skin inflammation caused by feeding a special diet to HR-1
Ac ce p
te
d
M
an
us
cr
ip t
441
28
Page 28 of 41
hairless mice. Exp Dermatol 2005;14:460-8.
462
[20] Macheleidt O, Kaiser HW, Sandhoff K. Deficiency of epidermal protein-bound
463
omega-hydroxyceramides in atopic dermatitis. J Invest Dermatol 2002;119:166-73.
464
[21] Schmittgen TD, Livak KJ. Analyzing real-time PCR data by the comparative C(T)
465
method. Nat Protoc 2008;3:1101-8.
466
[22] Shimada E, Aida K, Sugawara T, Hirata T. Inhibitory effect of topical maize
467
glucosylceramide on skin photoaging in UVA-irradiated hairless mice. J Oleo Sci
468
2011;60:321-5.
469
[23] Masukawa Y, Narita H, Sato H, Naoe A, Kondo N, Sugai Y, et al. Comprehensive
470
quantification of ceramide species in human stratum corneum. J Lipid Res
471
2009;50:1708-19.
472
[24] Ishikawa J, Narita H, Kondo N, Hotta M, Takagi Y, Masukawa Y, et al. Changes in
473
the ceramide profile of atopic dermatitis patients. J Invest Dermatol 2010;130:2511-4.
474
[25] Bouwstra JA, Gooris GS, Dubbelaar FE, Ponec M. Phase behavior of stratum
475
corneum lipid mixtures based on human ceramides: the role of natural and synthetic
476
ceramide 1. J Invest Dermatol 2002;118:606-17.
477
[26] Tsuji K, Mitsutake S, Ishikawa J, Takagi Y, Akiyama M, Shimizu H, et al. Dietary
478
glucosylceramide improves skin barrier function in hairless mice. J Dermatol Sci
479
2006;44:101-7.
480
[27] Sugawara T, Tsuduki T, Yano S, Hirose M, Duan J, Aida K, et al. Intestinal
Ac ce p
te
d
M
an
us
cr
ip t
461
29
Page 29 of 41
absorption of dietary maize glucosylceramide in lymphatic duct cannulated rats. J Lipid
482
Res 2010;51:1761-9.
483
[28] Nilsson A. Metabolism of cerebroside in the intestinal tract of the rat. Biochim
484
Biophys Acta 1969;187:113-21.
485
[29] Nilsson A. Metabolism of sphingomyelin in the intestinal tract of the rat. Biochim
486
Biophys Acta 1968;164:575-84.
487
[30] Schmelz EM, Crall KJ, Larocque R, Dillehay DL, Merrill AH, Jr. Uptake and
488
metabolism of sphingolipids in isolated intestinal loops of mice. J Nutr
489
1994;124:702-12.
490
[31] Hasegawa T, Shimada H, Uchiyama T, Ueda O, Nakashima M, Matsuoka Y.
491
Dietary glucosylceramide enhances cornified envelope formation via transglutaminase
492
expression and involucrin production. Lipids 2011;46:529-35.
493
[32] Byrdwell WC, Perry RH. Liquid chromatography with dual parallel mass
494
spectrometry and 31P nuclear magnetic resonance spectroscopy for analysis of
495
sphingomyelin and dihydrosphingomyelin. II. Bovine milk sphingolipids. J Chromatogr
496
A 2007;1146:164-85.
497
[33] Liu YJ. Thymic stromal lymphopoietin: master switch for allergic inflammation. J
498
Exp Med 2006;203:269-73.
499
[34] Liu YJ, Soumelis V, Watanabe N, Ito T, Wang YH, Malefyt Rde W, et al. TSLP: an
500
epithelial cell cytokine that regulates T cell differentiation by conditioning dendritic cell
Ac ce p
te
d
M
an
us
cr
ip t
481
30
Page 30 of 41
maturation. Annu Rev Immunol 2007;25:193-219.
502
[35] Soumelis V, Reche PA, Kanzler H, Yuan W, Edward G, Homey B, et al. Human
503
epithelial cells trigger dendritic cell mediated allergic inflammation by producing TSLP.
504
Nat Immunol 2002;3:673-80.
505
[36] Hannun YA, Bell RM. Regulation of protein kinase C by sphingosine and
506
lysosphingolipids. Clin Chim Acta 1989;185:333-45.
507
[37] Gupta AK, Fisher GJ, Elder JT, Nickoloff BJ, Voorhees JJ. Sphingosine inhibits
508
phorbol ester-induced inflammation, ornithine decarboxylase activity, and activation of
509
protein kinase C in mouse skin. J Invest Dermatol 1988;91:486-91.
510
[38] Duan J, Sugawara T, Sakai S, Aida K, Hirata T. Oral glucosylceramide reduces
511
2,4-dinitrofluorobenzene induced inflammatory response in mice by reducing
512
TNF-alpha levels and leukocyte infiltration. Lipids 2011;46:505-12.
513
[39] Jin H, He R, Oyoshi M, Geha RS. Animal models of atopic dermatitis. J Invest
514
Dermatol 2009;129:31-40.
515
[40] Tanaka A, Amagai Y, Oida K, Matsuda H. Recent findings in mouse models for
516
human atopic dermatitis. Exp Anim 2012;61:77-84.
Ac ce p
te
d
M
an
us
cr
ip t
501
517 518
31
Page 31 of 41
Figure legends
520
Figure 1
521
Effect of dietary milk phospholipids (MPLs) on changes in transepidermal water loss
522
(TEWL) (a) and the water content of the stratum corneum (b). Values are means +/-
523
SEM, n = 10/group. * Significantly different from the HR-AD group (P < 0.05). †
524
Significantly different from the L-MPL group (P < 0.05). § Significantly different from
525
the H-MPL group (P < 0.05).
M
an
us
cr
ip t
519
d
526
Figure 2
528
Effect of dietary milk phospholipids (MPLs) on the overall skin appearance of mice (a),
529
and skin histological images of mice (b). Arrow head shows layered parakeratosis (A),
530
hyperplasia (B), keratinocyte apoptosis (C), and inflammatory cell infiltration (D).
531
Effect of dietary MPLs on epidermal thickness (c). Values are means +/- SEM, n =
532
10/group. * Significant difference between groups (P < 0.05).
Ac ce p
te
527
533
534
Figure 3
32
Page 32 of 41
Typical chromatogram of covalently-bound -hydroxy ceramides in the epidermis of
536
hairless mice fed the normal diet using HPLC-MS/MS (a). Correlation between
537
covalently-bound -hydroxy ceramide levels and both transepidermal water loss
538
(TEWL) and the water content of the stratum corneum (b).
cr
ip t
535
us
539
Figure 4
541
Effect of dietary milk phospholipids (MPLs) on chemokines, thymic stromal
542
lymphopoietin (TSLP) (a), and thymus and activation-regulated chemokine (TARC) (b),
543
and levels of mRNA in the skin. Values are means +/- SEM, n = 10/group. * Significant
544
difference between groups (P < 0.05).
M
d
te
Ac ce p
545
an
540
546
Figure 5
547
Effect of dietary milk phospholipids (MPLs) on serum immunoglobulin E (IgE) (a),
548
thymus and activation-regulated chemokine (TARC) (b), thymic stromal lymphopoietin
549
(TSLP) (c), and soluble P –selectin (sP-selectin) levels (d). Values are means +/- SEM, n
550
= 10/group. * Significant difference between groups (P < 0.05).
551
33
Page 33 of 41
551
Tables
552
Table 1
553
covalently-bound ω-hydroxy ceramides in the epidermis of mice.
ip t
Effect of dietary milk phospholipids (MPLs) on the percentages of
HR-AD
d18:1 – C30:0
100.0 ± 14.4 * † §
8.3 ± 0.8
d18:1 – C32:0
100.0 ± 7.9 * † §
11.4 ± 1.0
d18:1 – C32:1
100.0 ± 4.7 * † §
28.3 ± 2.0
45.0 ± 1.5 *
63.7 ± 2.9 * †
d18:1 – C34:0
100.0 ± 11.5 * † §
9.1 ± 0.9
11.8 ± 1.0
66.3 ± 6.5 * †
d18:1 – C34:1
100.0 ± 4.2 * † §
M
L-MPL
34.9 ± 2.2
49.8 ± 1.7 *
84.8 ± 3.3 * †
d18:1 – C36:1
100.0 ± 6.8 * †
24.6 ± 1.8
33.1 ± 1.8
97.3 ± 5.4 * †
d17:1 – C32:0
100.0 ± 7.1 * † §
3.9 ± 0.4
5.6 ± 0.5
50.0 ± 4.4 * †
d17:1 – C32:1
100.0 ± 5.2 * † §
14.3 ± 1.1
22.0 ± 1.7
69.1 ± 3.2 * †
us
Control
H-MPL
14.2 ± 1.2
43.4 ± 3.6 * †
16.6 ± 1.1
59.1 ± 4.7 * †
an
d
te
Ac ce p
554
cr
Diet
d17:1 – C34:0
100.0 ± 8.5 * † §
3.5 ± 0.4
4.2 ± 0.5
59.6 ± 5.9 * †
d17:1 – C34:1
100.0 ± 4.0 * † §
17.7 ± 1.2
24.1 ± 1.6
83.7 ± 3.9 * †
d17:1 – C36:1
100.0 ± 6.2 * †
12.2 ± 1.1
15.9 ± 1.4
95.7 ± 6.6 * †
555
Means +/- SEM (n = 10/group)
556
Data expressed as relative peak area per epidermal protein content.
34
Page 34 of 41
* Significantly different from the HR-AD group (P < 0.05).
558
† Significantly different from the L-MPL group (P < 0.05).
559
§ Significantly different from the H-MPL group (P < 0.05).
Ac ce p
te
d
M
an
us
cr
560
ip t
557
35
Page 35 of 41
560
A novel mechanism for improvement of dry skin by dietary milk phospholipids: effect
561
on epidermal covalently-bound ceramides and skin inflammation in hairless mice.
562
Masashi Morifuji, Chisato Oba, Satomi Ichikawa, Kyoko Ito, Keiko Kawahata, Yukio
564
Asami, Shuji Ikegam, Hiroyuki Itou, Tatsuya Sugawara
ip t
563
cr
565
Highlights
567
1.
Dietary milk phospholipids improved dry skin conditions in hairless mice.
568
2.
Milk phospholipids attenuated a decrease in epidermal protein-bound -hydroxy
569
ceramide levels.
570
3.
an
us
566
M
Milk phospholipids lowered levels of skin and serum markers of inflammation.
Ac ce p
te
d
571
36
Page 36 of 41
Figure 1
b
ip t
a 25
60
* †§
cr
*†
*†
*†
*†
4
8
* 5
*
0
0
2
6
us
Stratum corneum water content (Arbitrary unit)
*
pt
10
ed
M
15
Ac ce
TEWL (g/m2/hr)
*
*
an
20
Fig.1
Control HR-AD L-MPL H-MPL
50
* 40
*
*†§
* *
30
*†
*†
*†
*
*
6
8
20
10 0
(weeks)
0
2
4
(weeks)
Page 37 of 41
Figure 2
A
100 mm
Ac ce pt e
D
C
Control
HR-AD
L-MPL
H-MPL
c B
60
Epidermal thickness (μm)
b
M
HR-AD
d
Control
an
us
cr
ip
t
a
*
*
*
*
40
20
0 Control HR-AD
L-MPL
Fig.2
L-MPL
H-MPL
H-MPL
Page 38 of 41
Figure 3
a
←d17:1-C34:1 ←d17:1-C36:1 ←d17:1-C34:0
←d17:1-C32:1
ip t
←d17:1-C32:0
←d18:1-C34:1
cr
←d18:1-C36:1
us
←d18:1-C32:1
←d18:1-C34:0
an
←d18:1-C32:0
11
12
15
16
17
pt
120
18
100
80
80
60
60
40
40
20
20
0
0 10
20
TEWL (g/m2/hr)
20
21
22
120
100
0
19
140
Ac ce
Covalently bound ceramide d18:1-C34:1 (Arbitrary Unit)
140
Fig.3
14
min
ed
b
13
M
←d18:1-C30:0
30
0
20
40
Stratum corneum water content (Arbitrary Unit)
60 Page 39 of 41
Figure 4
*
ip t
*
*
*
* *
4
cr
25
5
us
20 15
an
(Arbitrary Unit)
TSLP mRNA
30
*
(Arbitrary Unit)
*
35
b
TARC mRNA
a
10
M
5 0
2
1 0
Control HR-AD
L-MPL H-MPL
Ac ce
pt
ed
Control HR-AD L-MPL H-MPL
*
3
Fig.4
Page 40 of 41
Figure 5
a 16000
*
*
*
b 500
*
*
*
400
0
1500
1000
ip t Control
M d
500
*
HR-AD
*
L-MPL
H-MPL
*
*
400
*
500
* (ng/mL)
*
Ac ce
(pg/mL)
TSLP
2000
*
100
H-MPL
sP-Selectin
*
*
L-MPL
ed
2500
HR-AD
200
0
pt
Control
300
200
100
0
0
Control
Fig.5
300
cr
an
4000
c
(pg/mL)
TARC
8000
us
(ng/mL)
IgE
12000
HR-AD
L-MPL
H-MPL
Control
HR-AD L-MPL
H-MPL Page 41 of 41