Content uploaded by Mitsuhiro Denda
Author content
All content in this area was uploaded by Mitsuhiro Denda on May 06, 2019
Content may be subject to copyright.
ORIGINAL ARTICLE
See related Commentary on page vi
NMDA-Type Glutamate Receptor Is Associated
with Cutaneous Barrier Homeostasis
Shigeyoshi Fuziwara, Kaori Inoue, Mitsuhiro Denda
Shisedo Research Center, Fukuura Kanazawa-Ku, Yokohoma, Japan
Glutamate receptors play an important role in the exci-
tatory synaptic action of the central nervous system. In
this study, e¡ects of glutamate receptor agonists and an-
tagonists on skin barrier homeostasis were studied using
hairless mouse. Topical application of L-glutamic acid,
L-aspartic acid (non-speci¢c glutamate receptor ago-
nists) and N-methyl-D-aspartate (NMDA, NMDA type
receptor agonist) delayed the barrier recovery rate after
barrier disruption with tape stripping. On the other
hand, topical application of D-glutamic acid (non-spe-
ci¢c antagonist of glutamate receptor), MK 801 and D-
AP5, (NMDA-type receptor antagonists) accelerated the
barrier repair. The non-NMDA type receptor agonist,
a-amino-3-hydroxy-5-methyl-4 -isoxazole propionate
(AMPA), did not a¡ect the barrier recovery. Topical ap-
plication of MK-801 also promoted the healing of epi-
dermal hyperplasia induced by acetone treatment
under low environmental humidity. Immediately after
barrier disruption on skin organ culture, secretion of
glutamic acid from skin was signi¢cantly increased.
Immunohistochemistry, reverse transcription poly-
mearse chain reaction (RT-PCR) and in situ hybridiza-
tion showed an expression of NMDA-type receptor-like
protein on hairless mouse epidermis. NMDA increased
intercellular calcium in cultured human keratinocytes
and the increase was blocked by MK 801. These results
suggest that glutamate plays an important role as a sig-
nal of cutaneous barrier homeostasis and epidermal
hyperplasia induced by barrier disruption. Key words:
keratinocyte/epidermis/glutamate receptor/calcium channel.
J Invest Dermatol 120:1023 ^1029, 2003
The uppermost thin layer of the skin, stratum cor-
neum, plays a crucial role as a water impermeable
barrier. Acute disruption of the barrier results in an
increase in epidermal DNA synthesis (Proksch et al,
1991) and cytokine production (Wood et al,1992).
Even when the damage of the barrier is relatively small, when it
is repeated (Denda et al, 1996) or under low environmental hu-
midity (Denda et al, 1998), the damage induces an obvious epider-
mal hyperplasia and in£ammation. Moreover, various kinds of
dermatoses, such as atopic dermatitis, psoriasis and contact der-
matitis, are associated with barrier dysfunction (Grice, 1980).
The stratum corneum is composed of two components,
i.e., protein-rich nonviable cells and intercellular lipid domains
(Feingold and Elias, 2000). When the barrier function of the stra-
tum corneum is damaged by a surfactant, organic solvent or tape
stripping, the homeostatic system is accelerated and the barrier
function recovers its original level (Feingold et al,2000).
First, exocytosis of lipid-containing granules, lamellar bodies,
is accelerated and the inside lipids are secreted into the intercellu-
lar domain between the stratum granulosum and stratum
corneum and forms a water impermeable membrane (Feingold
et al,2000).
Previous studies suggested that ion gradation in the epidermis
is strongly associated with the barrier repair system (Mauro et al,
1998; Denda et al, 2000; Menon et al, 1994; Lee et al, 1994). When
the barrier was damaged, gradation of calcium and magnesium in
the epidermis immediately disappeared (Mauro et al, 1998; Denda
et al, 2000).When the calcium concentration in the upper epider-
mis increases by sonophoresis, the lamellar body secretion is pre-
vented (Menon et al,1994). Lee et al demonstrated that Ca
2þ
channel blocker, verapamil, prevents the delay of the barrier re-
pair induced by increased extracellular calcium concentration
(Lee et al, 1994). These results suggest that a calcium £ux into the
keratinocyte perturbs the lamellar body secretion and conse-
quently delays the barrier repair. However, the type of the chan-
nels responsible for the skin barrier homeostasis has not yet been
clari¢ed.
Recently, we demonstrated that ionotropic receptors, origin-
ally found in the nervous system, played an important role in
epidermal barrier homeostasis (Denda et al, 2002b; Denda et al,
2002c). These results suggest that other ligand^gated ion channel
also associates with the skin barrier function. Glutamic acid is the
principal neurotransmitter with a primarily excitatory synaptic
action in the central nervous system, and several di¡erent types
of receptors have been reported in the nerve system (Shepherd,
1994). One group of receptors comprises the G-protein-coupled
metabotropic receptors and another group the glutamate-gated
ion channels, such as NMDA (N-methyl-D-aspartate)-type or
AMPA (a-amino-3-hydroxy-5-methyl- 4 -isoxazole propionate)-
type receptor. Previous studies demonstrated that these gluta-
mate-gated ion channels were expressed in rat epidermal
keratinocyte cell membrane and that an antagonist of the receptor
Reprint requests to: Mitsuhiro Denda, Shiseido Research Center; 2-12-1
Fukuura, Kanazawa-ku, Yokohama, 236 - 8643 Japan; E-mail: mitsuhiro.
denda@to.shiseido.co.jp
Abreviations: TEWL: transepidermal water loss, NMDA: N-methyl-D-
aspartate, AMPA: a-amino-3-hydroxy- 5-methyl-4 -isoxazole propionate
Manuscript received June 10, 2002; revised November 5, 2002; accepted
for publication December 17, 2002
0022- 202X/03/$15.00 .Copyright r2003 by The Society for Investigative Dermatology, Inc.
1023
a¡ected the colony formation of cultured keratinocytes (Genever
et al, 1999; Morhenn et al, 1994). Especially, NMDA receptor
passes calcium ion selectively (Hollemann et al, 1991) and a vol-
tage-dependent magnesium ion block regulates the permeability
of the channel (Paoletti et al, 1995). Moreover, we demonstrated
that topical application of magnesium salts accelerated the skin
barrier repair after its damage (Denda et al, 19 99). We also reported
that an external electric potential altered calcium and magnesium
ion gradation in the epidermis and also a¡ected the skin barrier
homeostasis (Denda et al, 2002a).
From these previous studies, we hypothesized that a glutamate-
gated ion channel is associated with cutaneous barrier homeosta-
sis. In this study, we examined the e¡ects of topical application of
glutamate receptor agonists and antagonists on the skin barrier
recovery rate after barrier disruption and the e¡ects of these re-
agents on epidermal hyperplasia induced by barrier disruption
under low environmental humidity. We also examined the e¡ects
of glutamate receptor agonists and antagonists on the calcium
concentration in cultured human keratinocytes.
MATERIALS AND METHODS
Animals All experiments were performed on 7- to 10 -wk-old male
hairless mice (HR-1, Hoshino, Japan). All procedures of the measurement
of skin barrier function, disruption of the barrier and application of test
sample were carried out under anesthesia. All experiments were approved
by the Animal Research Committee of the Shiseido Research Center in
accordance with the National Research Council Guide (1996).
Reagents (þ)-MK 801, AMPA, NMDA and D-AP5 were purchased
from TOCRIS (Tocris Cookson, Inc., Ballwin, MO, USA). L-glutamic
acid, L-aspartic acid and D-glutamic acid were purchased from WAKO
(Osaka, Japan). Glutamate detection kit was purchased from Molecular
Probes (Amplex TM Red Glutamic Acid / Glutamate Oxidase Assay Kit,
Molecular Probes, Eugine OH). Normal human keratinocytes (neonatal
skin) were purchased from BioWhittaker (Walkersville, MD, USA).
Cutaneous barrier function Permeability barrier function was
evaluated by measurement of transepidermal water loss (TEWL) with an
electric water analyzer, as described previously (Denda et al,1996).Four
mice were used for the evaluation of each treatment. In total, 13^14 points
were measured for each treatment. The barrier recovery results are
expressed as percent of recovery because of variations from day to day in
the extent of barrier disruption. In each animal, the percentage of recovery
was calculated by the following formula: (TEWL immediately after barrier
disruption - TEWL at indicated time point)/(TEWL immediately after
barrier disruption - baseline TEWL) 100%. Topical application of each
reagent was carried out immediately after barrier disruption. In each case,
10 0 ml of 1 mM aqueous solution was applied o n 2 4cm
2
of tape stripped
£ank skin and occluded with plastic membrane for 20 min. Then the
membrane was removed and the treated area left to dry spontaneously.
Epidermal hyperplasia induced by barrier disruption under low
humidity Animals were kept separately in 7.2 liter cages in which the
relative humidity was maintained at less than 10% with dry air as
described previously (Denda et al, 1998). Four mice were used for the
evaluation of each treatment. Animals were ¢rst kept in the dry
conditions for 48 h and then bilateral £ank skin was treated with acetone-
soaked cotton balls, as described previously (Denda et al, 1998). The
procedure was terminated when TEWL reached 2.5^3.5 mg per cm
2
per
h. Immediately after treatment, 200 ml of 1 mM test sample 8 cm
2
of back
skin of mouse, and water was applied on the other side. Then the animals
were again kept in the dry conditions for 48 h. After the experiments,
animals were euthanaized with diethyl ether inhalation, and skin samples
were taken from the treated area. One hour before the euthanization, 20 ml
per g body weight bromodeoxyuridine (BrdU) 10mM solution was
injected intraperitoneally. Untreated control mice were also treated with
BrdU at the same time. After ¢xation with 4% paraformaldehyde, full
thickness skin samples were embedded in para⁄n, sectioned (4 mm), and
processed for hematoxylin and eosin staining. For the assessment of DNA
synthesis, the sections were immunostained with anti-BrdU antibodies. On
each section, ¢ve areas were selected at random; the thickness of the
epidermis was measured with an optical micrometer, and the mean value
was calculated. In total, 16^20 sections were used for each evaluation.
Measurements were carried out in an observer-blinded fashion.
Quanti¢cation of glutamate release from the skin We quanti¢ed
glutamate by the method arranged from previous reports (Wood et al,
1996; Ashida et al, 2001a). Four mice were used for the experiment. After
the animals were euthanized, skin samples were immediately taken from
both £anks. Subcutaneous fat was removed with a scalpel and the skin
samples were cut into 2 3cm
2
exactly. The two pieces of skin from
both £anks were placed, epidermis side upwards, in separate culture dishes
kept in an ice water bath, and one of them was tape stripped four times.
The other piece of skin was not treated. Three pieces of circle-shaped
section (exactly 5 mm diameter) was cut from each skin. Each section was
put in a well (12 mm diameter) separately. Then, 200 ml of chilled bu¡ered
saline solution containing 150 mM NaCl,10 mM glucose, 25 mM HEPES,
5 mM KCl, 1.2 mM NaH
2
PO
4
, 1.2 mM MgCl
2
, and 1.8 mM CaCl
2
,
adjusted to pH 7.4 with NaOH, was added to each well and incubated for
30 m at 371C. After the incubation, Sample solution (25 ml) and 0.1 M Tris-
HCl pH 7.5 (25 ml) were mixed to make 50 ml of reaction mixture supplied
in the AmplextRed Glutamic Acid/Glutamate Oxidase Assay Kit
(Molecular Probes, Inc. Eugene, OR, USA) according to the
manufucturer’s instruction. Brie£y, L-glutamate was oxidized with
glutamate oxidase to produce a-ketoglutarate, NH
3
and H
2
O
2
.L-
glutamate was regenerated by transamination of a-ketoglutarate, resulting
in multiple cycles of the initial reaction and a signi¢cant ampli¢cation of
the H
2
O
2
produced. The hydrogen peroxide reacted with 10-acetyl-3,7-
dihydroxyphenoxazine in a 1:1 stoichiometry in the reaction catalyzed by
horseradish peroxidase to generate the high £uorescent product, resoru¢n
which has absorption and £uorescence emission maxima of 563 nm and
587 nm, respectively. After incubation for 5 min at 371C the £uorescence
was detected using a Fluoroskan Ascent FL Plate Reader (Thermo
Labsystems, Inc., Helsinki, Finland). Excitation and emission wavelength
were 548 nm and 587 nm. Standard calibrations performed with L-
Glutamic acid ranged from pmol to 2 nmol under same conditions.
Localization of glutamate in mouse epidermis Glutamate detection
was performed in situ on a gel mounted with AmplextRed Glutamic
Acid/Glutamate Oxidase Assay Kit reaction mixture. The method was a
modi¢cation of our previous method (Denda et al, 2000a). Brie£y, a gel
on a slide glass was made of 2% agarose. The reaction mixture of the
assay kit was diluted 10 times with 0.1 M Tris bu¡er (pH 7.5), and 50 mlof
the solution was spread over the gel. One side of the mouse £ank skin
surface was stripped with adhesive tape three times, and the other side was
not treated as a control. After the animals were killed under anesthesia,
skin sections were taken from each side and frozen immediately. The
tissues were subjected to frozen sections with its 6 mm thickness. The
section was put on the gel and incubated at 371C for 5 min. The detection
of £uorescence was immediately executed when the skin slice was mounted
onto the reaction gel. To con¢rm the reproducibility, we observed ¢ve
di¡erent sections from three di¡erent animals each.
Analysis of expression of NMDA receptorf1 mRNA in epidermis
and brain tissue Epidermis of the skin tissue was removed by
incubation with 10 mM EDTA PBS solution at 371C for 30 min. Mouse
total RNA was obtained from epidermis using TRIzol reagent (Gibco,
BRL, Grand Island, NY, USA). cDNA was produced from the RNA by
random hexamer priming and reverse transcription was undergo using
SuperScript II (Life Technologies), according to the manufacturer’s
instructions. RT-PCR was carried out using TaKaRa ExTaq (Takara
shuzo, Shiga, Japan) with 2 mM sense and antisense oligonucleotide
primers designed for mouse NMDA receptor subunit z1 (NMDARz1)
and GAPDH, respectively (Yamazaki et al, 1992; Meguro et al, 1992; Ikeda
et al, 1992). The primer sequences were as follows: NMDAz1sense:
CTGTTATGGCTTCTGCGTTGA, antisense: GTTCACCTTAAATCGG-
CCAAA, GAPDH sense: CCCATCACCATCTTCCAG, antisense: CCT-
GCTTCACCACCTTCT. The denaturing temperature and the poly-
merase working temperature were 941Cand721C for 1 min, respectively.
The annealing was carried out at 57.51C for 1 min, respectively. All
reactions were carried on 30 cycles.
Imunohistochemistry for NMDARf1All mouse tissues were
subjected to frozen sections with 5 mm thickness. The detection for
mouse NMDARs was carried on using anti NMDARz1(C-20) poly-
clonal antibody (derived from goat serum) supplied by Santa Cruz
Biotechnology, Inc. (Santa Cruz, CA, USA). The goat polyclonal
antibody raised against a peptide mapping at the carboxy terminus of
glutamate (NMDA) receptorz1 of human origin and the sequence is
identical to corresponding mouse and rat sequence. The binding of the
¢rst antibody was detected by the £uorescent anti-goat secondary
antibody (Alexa Fluors488 rabbit anti-goat IgG (H þL), Molecular
Probes, Inc.). Nuclei were stained with DAPI using Vectashield with DAPI
Vector Laboratories, Inc., Burlingame, CA, USA). To con¢rm the
1024 FUZIWARA ET AL THE JOURNAL OF IN VESTIGATIVE DERMATOLOGY
reproducibility, we observed ¢ve sections from three di¡erent animals
each.
In situ hybridization We subcloned the PCR product of NMDA
receptor subunitz1 into pCR II TOPO (Invitrogen, San Diego, CA,
USA) vector and generated the pCR II TOPO-NMDAR 1-1 vector. The
NMDAR 1 digoxigenin labeled antisense and sense cRNA probe was
generated by in vitro transcriptions by usi ng T7 and SP6 RNA polymerase
(Roche Diagnostics, Mannheim, Germany) with pCR II TOPO-
NMDARz1-1 respectively. All the mouse tissues were subjected to frozen
sections with 5 mm thickness on the poly L-Lysine coated slide.We carried
out the in situ hybridization using digoxigenin labeled cRNA probes by
the method arranged from a previous report (Schaeren-Wiemers and
Ger¢n-Moser, 1993). The sections were ¢xed in 4% paraformaldehyde/
PBS for 10 min at room temperature and acetylated with aqueous solution
containing 0.25% acetic anhydride, 0.1 mM triethanolamine, and 0.175%
HCl. The acetylated sections were washed twice with PBS. Then the
sections were pre-hybridized with hybridization solution containing 50%
formamide, 250 ml/ml herring sperm DNA (Roche Diagnostics,
Mannheim, Germany), 500 mg/ml yeast derived total RNA (Sigma-
Aldrich Japan, Tokyo, Japan) and 1 Denhart’s solution (Sigma-Aldrich
Japan,Tokyo, Japan) in 4 SSC (1 SSC: 150 mM NaCl, 15 mM sodium
citrate, pH 7.4) at room temperature for 2 h. Then the sections were
hybridized with the hybridization solution containing the heat denatured
200 ng/ml sense or antisense digoxigenin labeled cRNA probe in a
humidi¢ed chamber at 601C overnight. Hybridized sections were washed
¢rst with 5xSSC solution at 501C and then washed with 0.2xSSC solution
twice at 501C for 30 min. Again the sections were washed with 0.2xSSC
solution at room temperature for 5 min. The s ections were washed with B1
bu¡er (0.1 M TrisHCl, 0.15 M NaCl, pH 7.5) and then incubated with B1
bu¡er including 10% heat inactivated goat serum (HINGS) for 1h at room
temperature. After the removal of the solution, the sections were incubated
with B1 solution including 0.02% anti-digoxigenin-Ap, Fab fragments
(Roche Diagnostics, Mannheim, Germany) and 1% HINGS at 41C
overnight. The sections were washed with B1 bu¡er three times, bu¡ered
with a solution (0.1 M TrisHCl pH 9.5, 0.1 M NaCl, 50 mM MgCl
2
)and
the staining reaction was started with a staining solution (75 mg/ml
nitroblautetrazoliumchloride, 50 mg/ml, 5 -bromo- 4 -chloro-3 -indolyl-
phosphate p-toluidine salt and 0.24 mg/ml levamisole in 0.1 M TrisHCl
pH 9.5, 0.1 M NaCl, 50 mM MgCl
2
). After 6 h incubation at room
temperature, the reaction was stopped with 1 mM EDTA in TrisHCl
10 mM, pH 7.5 and the sections were rinsed, air-dried, dehydrated
with ethanol, treated with xylene and mounted with Mount-Quick
(Cosmobio, Tokyo, Japan).
Calcium dynamics in keratinocyte culture system Evaluation of
intracellular calcium concentration ([Ca2
þ
]i) in cultured human
keratinocytes was measured with fura-2 as described by Grynkiewicz et al
(1985) with minor modi¢cations (Koizumi and Inoue,1997). Details of the
method using keratinocyte was the same as previously reported (Denda
et al, 2002a). The coverslip was mounted on a £uorescence microscope
(IX70,TS Olympus,Tokyo, Japan) equipped with a 75 W xenon-lamp and
band-pass ¢lters of 340 nm and 380 nm wavelengths. Measurements were
carried out at room temperature. Imaging data, recorded by a high-
sensitive silicon intensifer target camera (C4742, Hamamatsu Photonics,
Hamamatsu, Japan), were regulated by a Ca
2þ
analyzing system (AQUA/
RATI01, Hamamatsu Photonics, Hamamatsu, Japan).
Statistics The results are expressed as the mean 7SD. Statistical
di¡erences between two groups were determined by a two-tailed
Student’s t-test. In the case of more than 2 groups, di¡erences were
determined by ANOVA test (Fisher’s protected least signi ¢cant di¡erence).
RESULTS
Barrier recovery Figure 1 shows the e¡ects of topical
application of L, D-glutamic acid and L-aspartic acid on skin
barrier recovery after tape stripping. L-glutamic acid and L-
aspartic acid, both non-speci¢c agonists of the glutamate
receptor, delayed the barrier recovery and D-glutamic acid,
which is a weak antagonist against the glutamate receptor,
slightly but signi¢cantly accelerated the barrier recovery. Topical
application of NMDA delayed the barrier recovery and the delay
was blocked by MK 801, antagonist of NMDA receptor
(Fig 2A). Topical application of MK 801 or D-AP5 (another
NMDA speci¢c antagonist) alone accelerated the barrier recovery
(Fig 2B). Topical application of AMPA, AMPA receptor-speci¢c
agonist, did not a¡ect the barrier recovery rate (Fig 2A).
Figure 2Cshows the barrier recovery rate 30 min and 24 h after
tape stripping. The e¡ects of NMDA and MK 801were observed
in the earlier stage and also one d after treatment. These results
suggest that the NMDA type glutamate receptor or very similar
receptor is associated with ski n barrier homeostasis. The e¡ects of
NMDA receptor antagonists suggest that the presence of
endogeneous glutamate increase in the epidermis after barrier
disruption.
Epidermal hyperplasia Figure 3 shows the e¡ect of topical
application of NMDA and MK 801 on epidermal hyperplasia
induced by acetone treatment under dry conditions. As
previously demonstrated (Denda et al, 1998), the barrier
disruption induced obvious epidermal hyperplasia under dry
conditions (Fig 3B) compared to untreated skin (Fig 3A). The
epidermal hyperplasia was almost completely prevented by the
topical application of MK 801 (Fig 3C), but topical application
of NMDA even worsened the hyperproliferative response
(Fig 3D). The results of quanti¢cation of the epidermal
thickness are shown in Figure 3E. Evaluation of epidermal
proliferative response using BrdU showed the same tendency
(Fig 3F). These results suggest that the NMDA receptor or
similar receptor also plays an important role on the epidermal
proliferative response induced by barrier insults.
Localization of glutamate in the epidermis Figure 4A
shows the amount of glutamate released from skin organ culture
with or without barrier disruption by tape stripping. The barrier
disruption signi¢cantly increased the secretion of glutamate from
the skin tissue within 30 min after barrier disruption. Visual
Figure1. Topical application of non-speci¢c glutamate receptor
agonists, L-glutamic acid and L-aspartic acid, delayed the barrier
recovery of the hairless mice and the non-speci¢c antagonist of glu-
tamate receptor, D-glutamic acid, slightly but signi¢cantly acceler-
ated the barrier recovery at 2, 4 and 6 hours after the barrier
disruption by tape stripping. Ordinate shows recovery percent of the
barrier to the original level. (mean7SD) n ¼14 each.
n
:Po0.05,
nn
:
Po0.005,
nnn
:Po0.0 0 05
NMDAR RELATES BARRIER HOMEOSTASIS 1025VOL. 120, NO. 6 JUNE 20 03
images of glutamate localization with and without tape stripping
are shown in Figure 4Band C. Glutamate was localized in the
upper layer of the epidermis of untreated normal skin (Fig 4B:
n
red color,blue color shows DAPI stai ning of DNA). The localization
disappeared by the barrier disruption with tape stripping and
some of the glutamate re-localized or re-generated in the basal
layer of the epidermis (arrows,Fig 4C). With this method,
we could not distinguish the stratum corneum and stratum
granulosum. Rawlings et al (1994) previously demonstrated the
low content of glutamate in the stratum corneum (Ikeda et al,
1992) and moreover, the red color appears only by the chemical
reaction with glutamate as we described in the Methods. Thus,
the result of the untreated skin likely shows the localization of
glutamate in the stratum granulosum. These results suggest that
glutamate in the epidermis might be associated with the
signaling of the skin barrier damage.
RT-PCR assay, immunohistochemistry and in situ
hybridization of NMDA receptor RT-PCR analysis for
mouse NMDARz1 and GAPDH on the total RNA from mouse
epidermis also showed positive bands at the expected sizes
(Fig 5A). The expected bands for z1 were observed on the total
RNA from the mouse brain (data not shown). Immunoreactivity
against anti NMDARz1(Fig 5B) was observed in the living
layer of the epidermis. Blank assay without primary antibody
did not show any reactivity (data not shown). In situ
hybridization analysis of the mouse skin sections with the
Figure 2. Topical application of NMDA receptor speci¢c agonist,
NMDA delayed the barrier recovery after tape stripping and the de-
lay by NMDA was blocked by MK 801, speci¢c antagonist of
NMDA receptor. Topical application of AMPA, AMPA receptor speci¢c
agonist did not a¡ect the barrier recovery rate after tape stripping (Fig 2A).
Topical application of MK 801 or D-AP5, another NMDA receptor spec i¢c
antagonist, alone accelerated the barrier recovery (Fig 2B). The e¡ects of
NMDA and MK 801 were observed in the earlier stage (30 min after tape
stripping) and also one d after tape stripping (Fig 2C). Ordinate shows
recovery percent of the barrier to the original level. (mean7SD) n ¼
13 -14 each.
n
:Po0.0 5,
nn
:Po0.0 05,
nnn
:Po0.0005
Figure 3. The e¡ects of topical application of NMDA and MK 801
on epidermal hyperplasia induced by acetone treatment under dry
conditions. As previously demonstrated (Denda
¤et al,1998), the barrier dis -
ruption induced obvious epidermal hyperplasia under a dry condition (Fig
3B) compared to untreated normal skin (Fig 3A).The epidermal hyperpla-
sia was almost perfectly prevented by the topical application of MK 801
(Fig 3C) and topical application of NMDA worsened the hyperprolifera-
tive response (Fig 3D). The results of quanti¢cation of the epidermal
thickness are shown in Figure 3E. Evaluation of epidermal prolifer-
ative response using BrdU shows same tendency (Fig 3F). Ordinate in
Figure 3Eshows the average thickness of epidermis of each treatment.
(mean7SD) n ¼16- 20 each. Bars in A,B,C,D¼10 mm,
nnn
:Po0.0005
1026 FUZIWARA ET AL THE JOURNAL OF IN VESTIGATIVE DERMATOLOGY
antisense probe showed NMDARz1 mRNA expression in the
epidermal cells broadly (Fig 5C). The analysis with sense probe
did not show any staining (Fig 5D). These results show the
existence of NMDA receptor or a very structurally similar
protein in the epidermis.
Calcium concentration in the cultured keratinocytes The
intracellular calcium concentration in a single cell using
fura-2AM was increased when we added 50 mM NMDA (Fig
6A,Band E). Most of the increase was blocked by pre-
incubation with 50 mMMK801(Fig 6C,Dand E).
Approximately 31% of the keratinocytes showed an increase of
intracellular calcium among 56 observations within 1 min after
we applied NMDA. When we pre-i ncubated with MK 801, only
4.2% among 47 observations showed the calcium response
against NMDA application. These results suggest that the
NMDA receptor or a functionally very similar protein exists in
the epidermis. To con¢rm a contamination of other cell line in
the keratinocyte culture system, we carried out immuno-
histochemical study with keratinocyte speci¢c antibody (against
CK14) on the culture system of this study. Almost all cells (more
than 99%) showed positive immnunoreactivity against the
keratinocyte speci¢c protein antibody (data not shown).
DISCUSSION
An increase of intercellular and intracellular calcium in the kera-
tinocyte has been suggested to perturb the exocytosis of lamellar
bodies and delay skin barrier repair (Menon et al, 1994; Lee et al,
1994). In the healthy epidermis, a high concentration of calcium is
observed in the uppermost epidermis and the gradation disap-
peared immediately after barrier disruption by tape stripping or
organic solvent treatment (Mauro et al, 1998; Denda et al, 2000a).
The drastic change of calcium gradation in the epidermis might
be a crucial signal for the barrier repair signaling.The phase tran-
sition and fusion of intercellular granular membrane and cell
Figure 4. Glutamate was released from skin immediately after bar-
rier disruption. The barrier disruption signi¢cantly increased the secre-
tion of glutamate from the skin tissue within 30 min after the barrier
disruption (Fig 4A)(mean7SD). n ¼12 each. Visual images of glutamate
localization with and without tape stripping were shown in Fig 5Band C.
In normal skin, glutamate was localized in the uppermost epidermis (Fig
4B,Asterisk). Barrier disruption by tape stripping released glutamate from
epidermal granular layer and some of them re-localized in the epidermal
basal layer (Fig 4C,arrows). Bars in Band C¼5mm.
Figure 5. RT-PCR assay, immunohistochemistry and in situ hybri-
dization analysis of NMDA receptor subtype, NMDARf1. RT- P C R
analysis for mouse NMDARz1 and G3PDH on the total RNA from
mouse epidermis also showed positive bands of the expected sizes respec-
tively (Fig 5A). Immunoreactivity against NMDARz1 was observed in the
living layer of epidermis (Fig 5B,gre en color). In situ hybridization analysis
for NMDARz1 also s howed similar localization (Fig 5C, antisense) and no
signal was detected in an application of the corresponding sense probe (Fig
5D). SC; stratum corneum, epi; epidermis. Bars in B,C,D¼5mm.
Figure 6. E¡ects of NMDA and MK 801 on intracellular calcium
ion in cultured human keratinocytes. Application of 50 mMofNMDA
increased [Ca2 þ]i in the keratinocytes (Fig 6A,Band E)and the increase
was blocked by pre-incubation with MK 801 (Fig 6C,Dand E). Among
56 keratinocytes, 17 cells showed obvious increase of intracellular calcium
by NMDA treatment. After pre-incubation with MK 801, only two cells
among 47 keratinocytes showed the increase of intracellular calcium by
NMDA treatment. Ordinate shows the ratio of relative intensity of 340
nm and 380 nm emission.
NMDAR RELATES BARRIER HOMEOSTASIS 1027VOL. 120, NO. 6 JUNE 20 03
membrane is a crucial stage of the exocytosis of the lamellar
body, and ions such as calcium or magnesium in£uence the phase
of lipid bilayer structure (Denda et al, 1999; Denda et al, 2002a).
Although the mechanism has not been clari¢ed, the relative con-
centration of ions outside of and inside the keratinocyte cell
membrane, i.e., polarity of the membrane, might be crucial for
the skin barrier repair process. Zhang et al (2001) demonstrated
that depolarization of cell membrane causes an outward
movement. Increase of intracellular calcium ion might induce de-
polarization of the keratinocyte cell membrane, perturb inward
movement of cell membrane, the exocytosis of lamellar body
and delay the barrier repair.
Recently, we demonstrated that P2X purinergic receptor an-
tagonist and GABA(A) receptor agonists accelerated the barrier
repair after tape stripping (Denda et al, 2002b; Denda et al,
2002c). The former is a calcium channel and the latter is a chlor-
ide channel. These results suggest that calcium ion in£ux i nto the
keratinocytes delays the barrier repair and on the other hand,
chloride ion in£ux accelerates the barrier recovery. NMDA re-
ceptor is also a calcium channel. In neurons, chloride channels
like GABA(A) receptor or glycine receptor play a role of inhi-
biting depolarization induced by calcium £ux through other
excitatory receptors such as P2X or NMDA receptor. Similar to
the nerve-system, ‘‘excitation’’ induced by calcium in£ux and ‘‘in-
hibition’’ induced by chloride in£ux through the ligand-gated ion
channel might be important for the epidermal homeostasis. The
results of the present study support this hypothesis.
In this study, we ¢rst demonstrated the speci¢c localization and
dynamic of glutamate in the epidermis. In normal skin, gluta-
mate localized in the uppermost epidermis and the localization
disappeared by the barrier disruption. Interestingly, some of the
glutamate re-localized in epidermal basal layer (Fig 4C). Barrier
disruption induces epidermal DNA synthesis (Proksch et al,1991)
at the basal layer of the epidermis (Denda et al, 1998). Glutamate
might play a role of signaling in the epidermal proliferative re-
sponse induced by barrier disruption. However, the mechanism
underlying these phenomena has not been clari¢ed. Like the sy-
naptic system, a transporter of glutamate might play an important
role of the localization and its release from epidermis by barrier
disruption. The secretion of endogenous glutamate might in-
crease the intercellular calcium concentration through an NMDA
receptor-like protein in the epidermal keratinocytes. An applica-
tion of MK 801 might reduce the autocrine e¡ect of glutamate
and consequently block the delay of barrier recovery. In neurons,
the glutamate-gated cation channel plays an important role of
depolarization of the cell membrane induced by a cation £ux.
The NMDA receptor passes calcium ions through its channel,
but the AMPA receptor does not (Hollemann et al,1991). In this
study, the antagonist and agonist of NMDA a¡ected the skin bar-
rier repair process, but AMPA did not a¡ect the recovery rate.
Not only an NMDA receptor, but also an AMPA receptor was
expressed in epidermal keratinocytes in a previous study (Gen-
ever et al, 1999). Lee et al (1992) demonstrated that topical applica-
tion of calcium chloride delayed barrier repair, but application of
sodium chloride did not a¡ect barrier recovery. These results sug-
gest that the calcium £ux through keratinocyte cell membrane is
important for the skin barrier homeostasis, as we described above.
The NMDA receptor is modulated by a voltage-dependent
magnesium ion block (Paoletti et al, 19 95). We previously demon-
strated that topical application of magnesium salts and a mixture
of magnesium and calcium salts accelerated the skin barrier re-
covery (Denda et al, 1999). The relative ratio of magnesium and
calcium ions in the epidermis might be a crucial factor in the
electric potential through the epidermis (Denda et al,2001a).An
external electric potential altered the magnesium and calcium lo-
calization in the epidermis and also a¡ected skin barrier recovery
(Denda et al, 2002a). The NMDA receptor-like protein we de-
monstrated in this study might be associated with these phenom-
ena we demonstrated previously.
In neurons, the polarity of cell membrane is regulated by a
variety of ligand-gated ion channels such as NMDA, acetylcho-
line or GABA receptor (Shepherd, 1994). Several receptors origi n-
ally found in nerve cells have been reported on keratinocytes.
Ndoye et al (1998) demonstrated that each muscrinic acetylcholine
receptor subtype was localized di¡erently in the human epider-
mis. Another report suggested that keratinocyte nicotinic choli-
nergic receptors related to calcium in£ux of the cell (Grando
et al, 1996). Stoebner et al (1999) reported the expression of benzo-
diazepine receptors. Recently, we demonstrated the expression
and function of the vanilloid receptor subtype 1 (VR1) in human
keratinocytes (Denda et al, 2001b; Inoue et al, 2002). Ectoderm de-
rived keratinocytes and neurons show similar expression of those
receptors.
The NMDA receptor in the mouse central nervous system is
basically constructed of a z1 subunit and esubunit (Moriyoshi
et al, 1991). Our ¢ndings suggested the expression of z1 in the epi-
dermal keratinocytes. The z1 subunit alone has a function of glu-
tamate-gated cation channel and a di¡erent type of esubunit add
variation to the channel function (Ikeda et al, 1992; Kutsuwada
et al, 1992). Previous studies suggested that autocrine acetylcholine
and paracrine acetylcholine both play an important role in epi-
dermal proliferation and di¡erentiation (Grando, 1997). They also
suggested that a di¡erence in the distribution of a variety of cho-
linergic receptors might be crucial for epidermal homeostasis. On
the other hand, Nguyen et al (2001) demonstrated that nicotinic
acetylcholine receptor, a calcium channel, are associated with ker-
atinocyte di ¡erentiation. Their report is not i nconsistent with this
study because previous reports suggested that di¡erentiation of
epidermal keratinocytes is not always dependent on barrier
homeostasis (Ekanayake-Mudiyanselage et al, 1998). Di¡erent
types of glutamate receptors might regulate skin barrier homeos-
tasis and other metabolic systems in the epidermis.
Recent ¢ndings suggest that an abnormality in the calcium
dynamics in the epidermal keratinocytes is a crucial cause of cu-
taneous diseases. For example, mutation of the calcium pump of
the keratinocyte is a cause of Darier’s diseases and Hailey-Hailey
disease (Sakuntabhai et al, 1999; Hu et al, 2000). Karvonen et al de-
monstrated that psoriatic keratinocytes have an inborn error in
calcium signaling (Karvonen et al, 2000). Abnormal calcium loca-
lization in the epidermis has been reported on aged skin, atopic
dermatitis and psoriasis patients (Forslind et al, 1999). Abnormal-
ities of the barrier function in psoriasis, atopic dermatitis (Grice,
1980) and aging (Ghadially et al, 1995) were reported previously.
Higher expression of NMDA receptor was observed in glioneur-
onal tumours from intractable epilepsy patients (Aronica et al
2001). Although the relationship between the expression of the
NMDA receptors and abnormal metabolism of the cells has not
been clari¢ed, NMDA receptor might also be associated with
these skin diseases.
We previously demonstrated that a psychological stress delayed
the skin barrier recovery and that it was mediated by glucocorti-
coid (Denda et al, 2000b). No report has demonstrated the rela-
tionship between psychological stress and NMDA receptor
expression in the epidermal keratinocytes. However, in lung
epithelial cells, glucocorticoids increased mRNA of glutamine
synthetase (Abcouwer et al,1996). On the other hand, psychologi-
cal stress also induces glutamate release in the brain (Gilad et al,
1990). In this study, we observed the release of glutamate by tape
stripping. Although the mechanism of the glutamate secretion
has not been clari¢ed, glutamate might be another mediator be-
tween psychological factors and epidermal homeostasis.
Acceleration of barrier repair after its damage could prevent in-
£ammatory responses induced by barrier disruption (Denda et al,
1997; Ashida et al, 2001b). The skin barrier homeostatic system of
human skin is similar to that of hairless mouse (Elias et al,1991).
Accelerations of the barrier repair by topical application of same
agents were observed in both hairless mice and human skin
(Denda et al, 1997; Zettersten et al, 1997) Thus, a new strategy for
improving the barrier repair may be used in human skin. This
strategy could result in novel therapeutic approaches to treat the
cutaneous disorders caused by barrier damage or abnormal ion
dynamics in keratinocytes.
1028 FUZIWARA ET AL THE JOURNAL OF IN VESTIGATIVE DERMATOLOGY
REFERENCES
Abcouwer SF, Lukaszewics GC, Souba WW: Glucocorticoids regulate glutamine
synthetase expression in lung epithelial cells. Am J Physiol 270:L141^L151, 1996
Aronica E,Yankaya B, Jansen GH, Leenstra S, vanVeelen CW, Gorter JA, Troost D:
Neuropathol Appl Neurobiol. 27:223^237, 2001
Ashida Y, Ogo M, Denda M: Epidermal interleukin-1ageneration is ampli¢ed at
low humidity: Implications for the pathogenesis of in£ammatory dermatoses.
Br J Dermatol 144:238^243, 2001a
Ashida Y, Denda M, Hirao T: Histamine H1 and H2 receptor antagonists accelerate
skin barrier repair and prevent epidermal hyperplasia induced by barrier dis-
ruption in a dry environment. J Invest Dermatol 116:261^265, 2001b
Denda M, Ashida Y, Inoue K, Kumazawa N: Skin surface electric potential induced
by ion-£ux through epidermal cell layers. Biochem Biophys Res Commun
284:112^117, 2001a
Denda M, Fuziwara S, Inoue K, Denda S, Akamatsu H,Tomitaka A, Matsunaga K:
Immunoreactivity of VR1 on epidermal keratinocyte of human skin. Biochem
Biophys Res Commun 285:1250^1252, 2001b
Denda M, Hosoi J, AshidaY: Visual imaging of ion distribution in human epidermis.
Biochem Biophys Res Commun 272:134^137, 20 00a
Denda M, Inoue K, Fuziwara S, Denda S: P2X purinergic receptor antagonist accel-
erates skin barrier repair and prevents epidermal hyperplasia induced by skin
barrier disruption. J Invest Dermatol 119:1034^1040, 2002b
Denda M, Inoue K, Inomata S, Denda S: GABA(A) receptor agonists accelerate cu-
taneous barrier recovery and prevent epidermal hyperplasia induced by barrier
disruption. J Invest Dermatol 119:1041^1047, 2002c
Denda M, Katagiri C, Maruyama N: Some magnesium salts and a mixture of mag-
nesium and calcium salts accelerate skin barrier recovery. Arch Dermatol Res
291:560^563, 1999
Denda M, Kitamura K, El ias PM, Feingold KR: Trans- 4 -(aminomethyl) cyclohex-
ane carboxylic acid (T-AMCHA), an anti-¢brinolytic agent, accelerates barrier
recovery and preve nts the epidermal hyperplasia induced by epidermal injury
in hairless mice and humans. J Invest Dermatol 109:84^90, 1997
Denda M, Kumazawa N: Negative electric potential induces epidermal lamellar
body secretion and accelerates skin barrier recovery after barrier disruption.
JInvestDermatol118:65^72, 2002a
Denda M, Sato J, Tsuchiya T, Elias PM, Feingold KR: Low humidity stimulates epi-
dermal DNA synthesis and ampli¢es the hyperproliferative response to barrier
disruption: Implication for seasonal exacerbations of in£ammatory dermatoses.
JInvestDermatol111: 87 3 ^ 87 8, 19 98
Denda M, Tsuchiya T, Elias PM, Fei ngold KR: Stress alters cutaneous permeability
barrier homeostasis. Am J Physiol Integr Comp Physiol 278:R367^372, 2000b
Denda M,Wood LC, Emma S, Claxon C, Brown BE, Elias PM, Feingold KR: The
epidermal hyperplasia associated with barrier disruption by acetone treatment
or tape stripping cannot be attributed to increased water loss. Arch Dermatol Res
288:230^238, 1996
Ekanayake-Mudiyanselage S, Aschauer H, Schmook FP, Jensen JM, Meingassner JG,
Proksch E: Expression of epidermal keratins and the corni¢ed envelope pro-
tein involucrin is in£uenced by permeability barrier disruption. JInvestDerma-
tol 111:517^523, 1998
Elias PM, Menon GK: Structural and lipid biochemical correlates of the epidermal
permeability barrier. In Elias PM (ed). Advances in Lipid Research. San Diego:
Academic Press, 1991; p 1^26
Feingold KR, Elias PM: The environmental interface: Regulation of permeability
barrier homeostasis. In: Loden M, Maibach HI (eds). Dry skin and moisturizers.
Chemistry and function. Boca Raton: CRC Press, 2000; p 45^58
Forslind B,Werner-Linde Y, Lindberg M, Pallon J: Elemental analysis mirrors epider-
mal di¡erentiation. Acta Derm Venereol Stockh 79:12^17, 1999
Genever PG, Max¢eld SJ, Kennovin GD, Maltman J, Bowgen CJ, Raxworthy MJ,
Skerry TM: Evidence for a novel glutamate-mediated signaling pathway in
keratinocytes. JInvestDermatol112:337^342, 1999
Ghadially R, Brown BE, Sequeira-Martin SM, Feingold KR, Elias PM: The aged
epidermal permeability barrier: Structural, functional, and lipid biochemical
abnormalities in humans and asenescent murine model. JClinInvest 95:2281^
2290, 1995
Gilad GM, Gilad VH, Wyatt RJ, Tizabi Y: Region-selective stress-induced increase of
glutamate uptake and release in rat forebrain. Brain Res 525:335^338, 1990
Grando SA, Horton RM, Mauro TM, Kist DA, Lee TX, Dahl MV: Activation of
keratinocyte nicotinic cholinergic receptors stimulates calcium in£ux and en-
hances cell di¡erentiation. J Invest Dermatol 107:412^418, 1996
Grando SA: Biological functions of keratinocyte cholinergic receptors. JInvestDer-
matol Symp Proc 2:41^48,1997
Grice KA: Transepidermal waterloss i n pathologic skin. In: Jarrett A (ed).The Physiol-
ogy and Pathophysiology of the Skin. London: Academic Press, 1980; p 2147^2155
Grynkiewicz G, Poenie M, Tsien RY: A new generation of Ca2 þindicators with
greatly improved £uorescence properties. J Biol Chem 260:3440^3450, 1985
Hollemann M, Hartley M, Heinemann S: Ca2 þpermeability of KA-AMPA-gated
glutamate receptor channels depends on subunit composition. Science 252:851^
853, 1991
Hu Z, Bonifas JM, Beech J, et al: Mutation in ATO2CI, encoding a calcium pump,
cause Hailey-Hailey disease. Nat Genet 24:61^65, 2000
Ikeda K, Nagasawa M, Mori H, et al: Cloning and expression of the epsilon 4 sub-
unit of the NMDA receptor channel. FEBS Lett 313:34^38, 1992
Inoue K, Koizumi S, Fuziwara S, Denda S, Inoue S, Denda M: Functional vanilloid
receptors in cultured normal human keratinocytes. Biochem Biophys Res Com-
mun 291:124^129, 2002
Karvonen SL, Korkiamaki T, Yla-Outinen H, et al: Psoriasis and altered calcium
metabolism: downregulated capacitative calcium in£ux and defective
calcium-mediated cell signaling in cultured psoriatic keratinocytes. JInvest
Dermatol 114:693^700, 2000
Koizumi S, Inoue K: Inhibition byATP of calcium oscillations in rat cultured hippo-
campal neurons. BrJ Pharmacol 122:51^58, 1997
Kutsuwada T, Kashiwabuchi N, Mori H, et al: Molecular diversion of NMDA recep-
tor channel. Nature 358:36^41, 1992
Lee SH, Elias PM, Proksch E, Menon GK, Man MQ, Feingold KR: Calcium and
potassium are important regulators of barrier homeostasis in murine epider-
mis. JClinInvest89:530^538, 1992
Lee SH, Elias PM, Feingold KR, Mauro T: A role for ions in barrier recovery after
acute perturbation. JInvestDermatol102:976^979, 1994
MauroT, Bench G, Sidderas-Haddad E, Feingold KR, Elias PM, Cullander C: Acute
barrier perturbation abolishes the Ca2 þand K þgradients in murine epider-
mis: Quantitative measurement using PIXE. J Invest Dermatol 111:1198^1201,
19 9 8
Meguro H, Mori H, Araki K, et al: Functional characterization of a heteromeric
NMDA receptor channel expressed from cloned cDNAs. Nature 357:70^74,
1992
Menon K, Price LF, Bommannan B, Elias PM, Feingold KR: Selective obliteration
of the epidermal calcium gradient leads to e nhanced lamellar body secretion.
J Invest Dermatol 102:789^795, 1994
Morhenn VB,Waleh NS, Mansbridge JN, Unson D, Zolotorev A, Cline P, Lawrence
T: Evidence for an NMDA receptor subunit in human keratinocytes and rat
cardiocytes. Eur J Pharmacol 268:409^414, 1994
Moriyoshi K, Masu M, Ishii T, Shigemoto R, Mizuno N, Nakanishi S: Molecular
cloning and characterization of the rat NMDA receptor. Nature 354:31^37, 1991
National Research Council (NRC) Guide, National Research Council, 1996
NdoyeA,BuchliR,GreenbergB,et al: Identi¢cation and mapping of keratinocyte
muscarinic acetylcholine receptor subtypes in human epidermis. JInvestDerma-
tol 111:410 ^ 416 , 1998
Nguyen VT, Ndoye A, Hall LL, et al: Programmed cell death of keratinocytes culmi-
nates in apoptotic secretion of a humectant upon secretagogue action of acet-
ylcholine. JCellSci114:1189^1204, 2001
Paoletti P, Neyton J, Asher P: Glycine-independent and subunit-speci¢c potentiation
of NMDA responses by extracellular Mg2 þ.Neuron 15:1109^1120, 1995
Proksch E, Feingold KR, Man MQ, Elias PM: Barrier function regulates epidermal
DNA synthesis. JClinInvest87:1668^1673, 1991
Rawlings AV, Scott IR, Harding CR, Browser PA: Stratum corneum moisturization
at the molecular level. JInvestDermatol103:731^741, 1994
Sakuntabhai A, Ruiz-Perez V, Carter S, et al: Mutations in ATP2A2, encoding a
Ca2 þpump, cause Darier disease. Nature Genetics 21:271^277, 1999
Schaeren-Wiemers N, Ger¢n-Moser A: A single protocol to detect transcript of var-
ious types and expression levels in neural tissue and cultured cells: in situ hy-
bridization using digoxigenin-labelled cRNA probes. Histochemistry 100:341^
440,1993
Shepherd GM: The nature of central system. In: GM Shephard (Ed). Neurobiology.
NewYork: Oxford University Press, 1994; p 515^539
Stoebner PE, Carayon P, Penarier G, et al: The expression of peripheral benzodiapine
receptors in human skin: The relationship with epidermal cell di¡erentiation.
Br J Dermatol 140:1010^1016, 1999
Wood LC, Jackson SM, Elias PM, Feingold KR: Cutaneous barrier perturbation sti-
mulates cytokine production in the epidermis of mice. J Clin Invest 90:482^487,
1992
Wood LC, Elias PM, Calhoun C, Feingold KR: Barrier disruption stimulates inter-
leukin-1 alpha expression and release from a preformed pool in murine epider-
mis. J Invest Dermatol 106:397^403,1996
Yamazaki M, Mori H, Araki K, Mori KJ and Mishina M: Cloning, expression
and modulation of a mouse NMDA receptor subunit FEBS Lett. 300:39^45,
1992
Zhang PC, Keleshian AM, Sachs F: Voltage-induced membrane movement. Nature
413:428^431, 2001
Zettersten EM, Ghadially R, Feingold KR, Crumrine D, Elias PM: Optimal ratios
of topical stratum corneum lipids improve barrier recovery in chronologically
aged skin. J Am Acad Dermatol 37:403^408, 19 97
NMDAR RELATES BARRIER HOMEOSTASIS 1029VOL. 120, NO. 6 JUNE 20 03
