Experimental Dermatology 2005: 14: 498–508 Copyright # Blackwell Munksgaard 2005

Blackwell Munksgaard . Printed in Singapore

EXPERIMENTAL DERMATOLOGY
ISSN 0906-6705

Effective inhibition of melanosome transfer

to keratinocytes by lectins and niacinamide
is reversible
Greatens A, Hakozaki T, Koshoffer A, Epstein H, Schwemberger S, Amanda Greatens1,2, Tomohiro
Babcock G, Bissett D, Takiwaki H, Arase S, Wickett RR, Boissy RE. Hakozaki3,4, Amy Koshoffer1,
Effective inhibition of melanosome transfer to keratinocytes by lectins and Howard Epstein1, Sandy
niacinamide is reversible. Schwemberger5,6, George
Exp Dermatol 2005: 14: 498–508. # Blackwell Munksgaard, 2005 Babcock5,6, Donald Bissett7,
Hirotsugu Takiwaki4, Seiji
Abstract: Skin pigmentation results in part from the transfer of melanized
melanosomes synthesized by melanocytes to neighboring keratinocytes. Arase4, R. Randall Wickett2 and
Plasma membrane lectins and their glycoconjugates expressed by these Raymond E. Boissy1
epidermal cells are critical molecules involved in this transfer process. In 1
Department of Dermatology, University of
addition, the derivative of vitamin B3, niacinamide, can inhibit Cincinnati, Cincinnati, OH, USA;
2
melanosome transfer and induce skin lightening. We investigated the Department of Pharmaceutical Sciences,
effects of these molecules on the viability of melanocytes and University of Cincinnati, Cincinnati, OH, USA;
3
keratinocytes and on the reversibility of melanosome-transfer inhibition Research and Development Department, Procter
induced by these agents using an in vitro melanocyte–keratinocyte & Gamble Far East Inc., Kobe, Japan;
4
Department of Dermatology, School of Medicine,
coculture model system. While lectins and neoglycoproteins could induce University of Tokushima, Tokushima, Japan;
apoptosis in a dose-dependent manner to melanocytes or keratinocytes in 5
Department of Surgery, University of Cincinnati,
monoculture, similar dosages of the lectins, as opposed to Cincinnati, OH, USA;
neoglycoproteins, did not induce apoptosis to either cell type when treated 6
Shriners Hospitals for Children, Cincinnati, OH, USA;
7
in coculture. The dosages of lectins and niacinamide not affecting cell Skin Technology Division, Procter and Gamble,
viability produced an inhibitory effect on melanosome transfer, when used Cincinnati, OH, USA
either alone or together in cocultures of melanocytes–keratinocytes.
Cocultures treated with lectins or niacinamide resumed normal
melanosome transfer in 3 days after removal of the inhibitor, while Key words: Melanocyte, Pigmentation, Skin,
cocultures treated with a combination of lectins and niacinamide Complexion
demonstrated a lag in this recovery. Subsequently, we assessed the effect Dr Raymond Boissy
of niacinamide on facial hyperpigmented spots using a vehicle-controlled, Department of Dermatology
split-faced design human clinical trial. Topical application of niacinamide University of Cincinnati
resulted in a dose-dependent and reversible reduction in hyperpigmented PO Box 670592
lesions. These results suggest that lectins and niacinamide at Cincinnati, OH 45267-0592, USA
concentrations that do not affect cell viability are reversible inhibitors of e-mail: boissyre@ucmail.uc.edu
melanosome transfer. Accepted for publication 8 December 2004

Introduction
Skin coloration is a result of many complex pro-
cesses. Epidermal melanocytes synthesize pigmented
melanosomes that are subsequently transferred to
and retained by the surrounding keratinocytes.
Normal skin coloration is a result of both efficient
melanization of the melanosome by the melano-
cyte and proper transfer to and receipt of the
melanosome in the keratinocyte (1,2). The process
of melanin production has been well studied, and
the major enzymes involved in the melanization
pathway have been identified (3,4). However, little
is known about the process of intercellular mela-
nosome transfer.
Several theories have been proposed to explain
this transfer process. These theories consist of: (a)
release of melanosomes into the intercellular space,
with subsequent endocytosis by the keratino-
cytes; (b) cytophagocytosis, i.e. the phagocytosis
of dendritic tips of the melanocytes by keratino-
cytes; (c) direct inoculation of the melanosomes
into the keratinocytes; or (d) transfer of melano-
somes via a communication conduit between the
melanocytes and keratinocytes (5–8). These pro-
cesses may not be mutually exclusive; transfer of
melanosomes may occur through a combination
of these mechanisms (6).

498

The amount and regulation of melanosome
transfer contributes to the ultimate pigmentation
of human skin. Alterations in this melanosome-
transfer process may be the basis for several pig-
mentary/complexion diseases and disorders. This
is supported by several pigmentary disorders
involving containment of the melanosome within
the melanocyte, such as hair-graying (9) and nevus
depigmentosus (10). Cutaneous hyperpigmenta-
tion occurs in numerous conditions such as aging
(11), pregnancy (12), postinflammation (13), and
scarring (14). Inhibition of melanosome transfer
by the protease-activated receptor 2 (PAR-2) (15)
or niacinamide (16) results in lightening of cuta-
neous pigmentation. In addition to skin health
issues related to altered pigmentation, a cosmetic
desire for light and even skin tone is prevalent in
many areas of the world, especially among Asian
populations. Approximately 60% of Japanese
women and 75% of Chinese women desire lighter
skin (16). Therefore, the process of melanosome
transfer presents unique opportunities to ther-
apeutically and cosmetically influence human
skin pigmentation.
Plasma membrane lectins and their correspond-
ing glycoconjugates, putatively expressed by mela-
nocytes and keratinocytes, may play a critical role
in melanosome transfer for a variety of reasons.
These molecules are responsible for cell–cell recog-
nition in general (17), and phagocytosis by macro-
phages in specific (18). Previous studies have
indicated increased production of these molecules
in melanocytes after UVA irradiation (19), sug-
gesting that they facilitate melanosome transfer.
In addition, symptoms of the rare metabolic dis-
ease, carbohydrate-deficient glycoprotein syn-
drome type III, which alters carbohydrate moiety
synthesis and processing, include pale and unpig-
mented skin (20). Most importantly, certain lectins
and neoglycoproteins have been shown to impede
the melanosome-transfer process when used in a
coculture model system that recapitulates melano-
some transfer (21). These various lectins and neo-
glycoproteins interact with galactosyl, fucosyl, and
mannose residues (21) and had been previously
implicated in studies of melanin transfer between
transformed cells (19). However, it was unknown
whether these lectins or neoglycoproteins would
have any negative effects on cell viability or
whether the transfer inhibition would be perman-
ent. In this study, we investigated the effect of
various concentrations of lectins, neoglycopro-
teins, and niacinamide, a known skin-lightening
agent (16), both alone and in combination, on
melanocyte and/or keratinocyte viability and on
melanosome transfer between these two epidermal
Melanosome transfer to keratinocytes
cell types. We also examined the reversibility of
melanosome transfer impeded by these com-
pounds both in a culture model and in a clinical
trial. Further insight into these compounds as well
as their underlying mechanism of action will indi-
cate potential inhibitors and eventual stimulators
to manipulate human skin pigmentation.
Materials and methods
Cell cultures
Melanocytes and keratinocytes were obtained from Cascade Biolo-
gicals (Portland, OR, USA). Melanocytes were maintained in M154
basal medium (Cascade Biologicals) and supplemented with 5%
fetal bovine serum, 1% antibiotic/antimycotic (Gibco, Grand
Island, NY, USA), 1 mg/ml of transferrin, 1 mg/ml of vitamin E,
5 mg/ml of insulin, 0.5 ng/ml of basic fibroblast growth factor, 10
8 M
melanocyte-stimulating hormone, and 10
9 M endothelin-1 (all
from Sigma Chemical Co., St Louis, MO, USA). Keratinocytes were
maintained in M154 basal medium supplemented with keratinocyte
growth supplements (Cascade Biologicals) and 1% antibiotic/
antimycotic (Gibco). Cocultures of keratinocytes/melanocytes were
established as previously described (22). In short, established cultures
of keratinocytes and melanocytes were trypsinized, washed, and
reseeded at a keratinocyte : melanocyte ratio of 10 : 1 in a media
containing equal parts of keratinocytes and melanocytes.
Test compounds
The lectins used in these experiments (and the corresponding
carbohydrate moiety they bind to) consist of Narcissus pseudo-
narcissus (NP) L5650 (a-D-mannose), Pisum sativum (PS) L5380
(a-mannose), Psophocarpus tetragonolobus (PT) L2138 (GalNAc,
gal), Lycopersicon esculentum (LE) L2886 (GalNAc), and Tetra-
gonolobus purpureas (TP) Lp254 (a-L-fucose). All lectins were
purchased from Sigma-Aldrich (St Louis, MO, USA).
The two neoglycoproteins used in these experiments (and the
conjugate lectin they interact with) consist of a-1, 3a-1,6-manno-
triose–bovine serum albumin (BSA) (N402) (Con A) and galac-
tose a-1,3-galactose–BSA (N403) (Euonymus europaeus). Both
neoglycoproteins were purchased from Calbiochem/Novabio-
chem (La Jolla, CA, USA).
Niacinamide (nicotinic acid amide) was purchased from Sigma
Cell Culture.
Apoptosis detection
Cultures of melanocytes or keratinocytes and melanocyte/kera-
tinocyte cocultures were treated with test compound or vehicle [i.e.
phosphate-buffered saline (PBS)] alone, once or twice daily for
6 days and assayed for apoptosis/necrosis as described below.
Experiments were repeated three times with triplicate samples
per experiment. In experiments involving melanocyte–keratino-
cyte cocultures, we developed a differential trypsinization proced-
ure whereby melanocytes and keratinocytes were separated at
the termination of the experiment and evaluated individually for
apoptosis (see Fig. 1a,b). In short, cocultures were treated briefly
with 0.25% trypsin (approximately 3 min) to selectively remove
the melanocytes. The remaining keratinocytes were obtained by
further trypsinization.
Isolated cells were then processed for detection of non-viable
cells (i.e. cells in early and late stages of apoptosis and cells
undergoing necrosis) by flow cytometry using the Annexin V
apoptosis-detection kit (Trevigen, Gathersburg, MD, USA).
Counterstaining with propidium iodine (PI) was used to identify
499
Greatens et al.
a b
Figure 1. Differential trypsinization of melanocyte–keratinocyte
cocultures. Melanocyte –keratinocyte cocultures (a)
demonstrating dendritic melanocytes (arrows) and keratinocytes
(arrowheads) were treated with 1% trypsin for approximately
3 min after which (b) melanocytes were selectively detached and
collected, while the remaining keratinocytes (arrowheads) were
subsequently re-trypsinized and collected. Bar ¼ 50 mm.
cells in late apoptosis and/or necrosis. Cells were analyzed for
Annexin V–fluorescein isothiocyanate (FITC) and propidium
iodide (PI) reactivity by flow cytometry using an EPICS XL
flow cytometer (Beckman Coulter, Hialeah, FL, USA). Data
were analyzed using COULTER XL software (Coulter).
Melanosome-transfer inhibition
Transfer between melanocytes and keratinocytes in cocultures
was evaluated as previously described (21). In short, melanocytes
were prelabeled with succinimidyl ester of carboxy fluorescein
diacetate (CFDA) (Molecular Probes, Eugene, OR, USA) and
used to establish cocultures with keratinocytes. Cocultures were
treated with fresh media containing test compound or vehicle, i.e.
PBS, every 12 or 24 h during the treatment period. Experiments
were repeated twice with triplicate samples per experiment. Fol-
lowing treatment, cells were prefixed in 4% paraformaldehyde,
permeabilized with fluorescence-activated cell-sorter-permeabiliz-
ing solution (FACS perm, Becton Dickinson, San Jose, CA,
USA), and washed. The cells were then treated with a monoclonal
mouse anticytokeratin primary antibody (1 : 300), washed, and
incubated with goat antimouse IgG secondary antibody conju-
gated to phycoerythrin (PE) (Caltag, Burlingame, CA, USA)
(1 : 10). Cells were postfixed in 1% paraformaldehyde and trans-
fer quantitated by evaluating the mean CFDA fluorescence
(Mean X) in PE-positive keratinocytes in the presence or absence
of inhibitor using an EPICS XL flow cytometer (Coulter Cyto-
metry, Coulter). Data were analyzed using COULTER XL software.
Description of statistics
To determine whether the test compound affected the treated cells
as compared with control populations of cells/cocultures in a
statistically significant manner, we tested the population mean
and variance. We constructed a 100 99.5% confidence interval
for the mean. The relevant test statistic was Z, and this parameter
was calculated for each test sample, compared to the control
sample. Accordingly, P-values for each Z-statistic value were
calculated. All P-values for the Z-statistic of the one-tailed test
from each sample population were assessed for whether they were
less than a ¼ 0.05. All calculations were performed using Micro-
soft Excel 2000 (Microsoft, Redmond, WA, USA).
Human clinical trial
A double-blinded, randomized, vehicle-controlled, split-face
design human clinical study was performed from October 2002
through August 2003 (Kobe, Japan). Seventy-nine Japanese
women aged 28–54 years with multiple types of brown hyperpig-
mentation on both sides of the face were assigned to one of two
500
groups of either 39 or 40 subjects. Most subjects presented with a
combination of slight-to-moderate solar lentigines, melasma, or
freckles on both sides of the face. All subjects were given basic
skin care products (facial cleanser, lotion, and sunscreen, none of
which contained skin-whitening ingredients) to use during the
first 16 weeks of the study. In a split-face design, Group 1
(n ¼ 39) utilized 5% niacinamide in a moisturizer to one side of
the face vs. the vehicle moisturizer (without niacinamide) to the
other side of the face and Group 2 (n ¼ 40) utilized 2% niacin-
amide moisturizer vs. the vehicle moisturizer. Subjects applied
their assigned test products to the assigned sides of their faces in
a dosage-controlled manner (0.3 ml/application/entire half face)
twice daily (morning and evening) for 8 weeks (treatment period).
The group that received 5% niacinamide (Group 1) was moni-
tored for an additional 34 weeks after the treatment period to
determine reversibility of the lightening effect (recovery period)
(n ¼ 37 completed). To confirm that no side effects occurred, a
dermatologist evaluated the subjects’ images captured at baseline
and at 8 and 42 weeks (Group 1). All subjects gave written
informed consent. The protocol had been approved by the reg-
ulatory and safety review committee for human testing of The
Procter & Gamble Company (Kobe, Japan). All subjects
were given basic skin care products [facial cleanser, lotion, and
sunscreen (SPF 15)], none of which contained skin-whitening
ingredient(s), to use during the first 16 weeks of the study, while
no instruction was provided during the remaining 26 weeks (for
Group 1), i.e. they followed their normal skin care habits and
practices.
Image analysis of facial hyperpigmented spots
Facial hyperpigmented spot area on each side of the face was
objectively measured using a customized image analysis technique
similar to the one previously described (16,23). Specifically,
images of the right and left sides of each subject’s face were
captured at baseline and at 4 and 8 weeks (also at 16, 26, and
42 weeks for Group 1) by a high-resolution digital camera (Fuji
SC430 CCD Digital Camera) equipped with a cross-polarized
filter (Tiffen glass polarizing filter and Kenko polarizing filter
for light source). Facial illumination was provided by National
fluorescence light Twin 1 (FPL27EX-N, 6200–6400 K) positioned
to the right and left sides of the camera to provide even lighting.
Before image capture, subjects were equilibrated in a temperature-
controlled room (24 + 2 C) for 30 min. At the 4-, 8-, 16-, 26-, and
42-week visits, accurate repositioning of the subjects was facili-
tated by comparing side by side the live image with the digitally
stored image obtained at baseline. The system was calibrated by
white balancing the camera each study day. Computer analysis of
the digital images allowed quantification of total area of hyper-
pigmented spots in the selected region of interest (around the
cheek to the temple) as a spot area fraction (percentage of total
spot area in the area of the selected region of interest).
Visual assessment of hyperpigmented spot
Subjective visual grading of the captured images was carried out
to compare pretreatment (baseline image) vs. post-treatment at
the 8-, 26-, and 42-week time points as previously described (16).
Briefly, paired blinded pre- and post-treatment images appeared
side by side randomly on the monitor. Eight treatment-blinded
judges independently viewed pairs of images for each subject. The
judges indicated which of the two images had fewer areas of
hyperpigmented spots around the eye and the cheek, then rated
the magnitude of the difference between images on a scale of 1–4:
1, I think there is a small difference; 2, I know there is a small
difference; 3, I know there is a moderate difference; and 4, I know
there is a big difference. There was no ‘no difference’ option to
increase the sensitivity of the judges’ evaluations. The magnitude
rating was assigned a positive (þ) value when the post-treatment

was selected to have less hyperpigmented spots, while it was
assigned a negative (–) value when the pretreatment was selected
to have less hyperpigmented spots. Thus, positive values indi-
cated efficacy and negative values indicated no efficacy. The
mean ratings of the eight judges were used for the statistical
analysis.
Data analysis
Image analysis data (hyperpigmented spot area fraction) were
compared for the change from the baseline by a paired t-test at
each subsequent time point. Visual assessment data (the mean
rating of the eight judges) were also compared by a paired t-test.
Statistical significances were assessed for whether P-values were
less than a ¼ 0.05 or not. All calculations were performed using
SPSS for Windows, version 11.5.
Results
Effect of lectins and neoglycoproteins on cell
viability of cultured melanocytes and keratinocytes
The viability of pure cultures of melanocytes or
keratinocytes exposed to lectins and neoglycopro-
teins was assessed. Test concentrations included
Melanosome transfer to keratinocytes
those previously demonstrated to reduce melano-
some transfer in cocultures of melanocytes and
keratinocytes (21) as well as concentrations that
were one-half and twice these reported concentra-
tions. Cells were then evaluated for Annexin V
and/or PI reactivity by flow cytometry as an indi-
cator of non-viable cells. Cells negative for both
Annexin V–FITC and PI were considered healthy.
The data demonstrated that the viability in pure
cultures of both melanocytes and keratinocytes
was significantly compromised by certain lectins
and neoglycoproteins generally in a dose-dependent
manner (Fig. 2a,b). Of exception were the lectin
PT and the neoglycoprotein alpha-galactose,
which exhibited no deleterious effect on melano-
cytes at the dosages tested.
The negative effects of lectins and neoglycopro-
teins on the viability of melanocytes or keratino-
cytes in monocultures were not apparent in
previous experiments where melanocytes and
keratinocytes were grown in coculture (21). The
negative effects on cell viability by several lectins

Figure 2. Effect of lectins or neoglyco

proteins on viability of melanocyte or
keratinocyte monocultures or cocultures. a Melanocytes in monoculture b Keratinocytes in monoculture
120 120
Pure cultures of (a) melanocytes and (b)
keratinocytes were treated twice daily for 6 100 100
days with the various concentrations of
*
lectins and neoglycoproteins listed. Cells 80
* * 80
* * * *
*
were then assessed for viability as
* * **
described in Materials and methods. (a)
60 60
* *
All lectins and neoglycoproteins except PT ** * *
40 40
and alpha-galactose, respectively, exhibited ** **
a negative effect on the viability of isolated 20 20
* *
cultures of melanocytes in a dose- *
0 0
dependent manner. (b) All compounds
0
4.4
8.8
17.5
4.4
8.8
17.5
0.9
1.8
3.5
0.9
1.8
3.5
0.9
1.8
3.5
0.5
1.0
2.0
0.5
1.0
2.0

0
4.4
8.8
17.5

8.8
17.5
0.9
1.8
3.5
0.9
1.8
3.5
0.9
1.8
3.5
0.5
1.0
2.0
0.5
1.0
2.0
4.4

exhibited a negative effect on viability of Lectin (µg/ml) Lectin (µg/ml)

isolated cultures of keratinocytes in a dose- 120 c Melanocytes in coculture 140 d Keratinocytes in coculture
dependent manner. Cocultures were
treated twice a day for 6 days with the 100 * * 120

various concentrations of lectins and 100 *

*
* ** *
80
neoglycoproteins listed. Subsequently, (c) * *
melanocytes and (d) keratinocytes were * ** 80
60
isolated by differential trypsinization and *
60

assessed for viability as described in 40

40
Materials and methods. All lectins 20
20
exhibited no or minimal effect on cell
3.5 NA

3.5 NA

viability of either melanocytes or 0 0

0
4.4
8.8
17.5
4.4
8.8
17.5
0.9
1.8
3.5
0.9
1.8
3.5
0.9
1.8

0.5
1.0
2.0
0.5
1.0
2.0

0
4.4
8.8
17.5
4.4
8.8
17.5
0.9
1.8
3.5
0.9
1.8
3.5
0.9
1.8

0.5
1.0
2.0
0.5
1.0
2.0

keratinocytes, especially at the lowest

concentration tested, except for PT that Lectin (µg/ml) Lectin (µg/ml)

exhibited a significant negative effect on Control Psophocarpus tetragonolbus

keratinocyte viability at the lowest Lycopersicon esculentum
Narassus pseudonarcissus
concentration tested. In contrast, both
Tetragonolobus purpeas α-1,3 α-1,6-mannitrose–BSA
neoglycoproteins affected cell viability of
melanocytes and keratinocytes. Error bars Pisum sativum α-1,3-galactose–BSA

represent SD; asterisks represent P < 0.05. Highest relative concentration (mg/ml) of lectin allowing normal cell viability
Table presents the highest relative
concentration (mg/ml) of lectin allowing NP TP PS PT LE
M K M K M K M K M K
normal cell viability in comparison
Monoculture 4.4 <4.4 4.4 8.8 0.9 0.9 >3.5 >0.9 >3.5 >3.5
between monoculture and coculture
Coculture 4.4 >17.5 >17.5 >17.5 >3.5 >3.5 1.8 <0.9 1.8 1.8
conditions.

501
Greatens et al.
(TP, PS, and LE) on melanocytes and keratino-
cytes in pure cultures were greatly reduced, and in
some cases eliminated, for both cell types when the
cells existed in coculture (Fig. 2c,d). However, PT
was the one lectin that demonstrated a negative
effect on the viability of keratinocytes even at the
lowest concentration. Therefore, all lectins except
PT were combined at the highest concentration
(i.e. NP ¼ 4.4; TP ¼ 17.5; PS ¼ 0.5; LE ¼ 0.5 mg/
ml) that did not affect the viability of melanocytes
and keratinocytes and were used as a cocktail. The
cocktail of lectins was used to treat cocultures
twice daily for 6 days and the viability of melano-
cytes and keratinocytes subsequently assessed. The
percentage of viable melanocytes and keratino-
cytes in cocultures was 97.6 + 1.8 and 94.3 + 1.5,
respectively. Therefore, no additional effects on
viability of either cell type were caused by the
treatment cocktail.
In contrast to the lectins, the neoglycoproteins
continued to induce significant loss of cell viabil-
ity to both melanocytes and keratinocytes main-
tained in coculture (Fig. 2c,d). Interestingly, both
neoglycoproteins were less toxic to melanocytes in
pure culture than melanocytes cultured with kera-
tinocytes. Conversely, both neoglycoproteins were
more toxic to isolated keratinocytes than kerati-
nocytes cultured with melanocytes. It is surprising
that neoglycoproteins are more cytotoxic to mel-
anocytes and keratinocytes than are lectins. Pos-
sibly, the binding efficiency and subsequent
interference with the homeostasis/function of
cell-surface glycoprotein is more effective with
neoglycoprotein than lectins. Regardless, both
neoglycoproteins were excluded from subsequent
experiments.
Effect of niacinamide on cell viability of
melanocytes and keratinocytes when maintained in
coculture
The viability of pure cultures of melanocytes
exposed to niacinamide was assessed. Cells were
cultured in the absence (i.e. vehicle alone) or pres-
ence of 0.1, 1.0, 10.0, 100.0, and 1000.0 mM niacin-
amide twice daily for 6 days. The data
demonstrated that the health of both cultured mel-
anocytes and keratinocytes was not compromised
at niacinamide concentrations tested (Fig. 3). We
next investigated the viability of each cell type to
treatment with niacinamide when existing together
in coculture. All concentrations of niacinamide
tested did not compromise viability of either mela-
nocytes or keratinocytes when both cell types also
existed in coculture (Fig. 3).
502
a
160
b
160
140
Monoculture
140
Monoculture
Percent healthy melanocytes
Coculture Coculture
Percent healthy kerotinocytes
120 120
100 100
80 80
60 60
40 40
20 20
0 0
0.0 0.1 1.0 10.0 100
Concentration (µM)
1000 0.0 0.1 1.0 10.0 100
Concentration (µM)
1000
Figure 3. Effect of niacinamide concentrations on viability of
melanocyte or keratinocyte monocultures or cocultures. (a)
Melanocyte and (b) keratinocyte monocultures were treated
twice daily for 6 days with niacinamide and assessed for various
stages of apoptosis/necrosis using Annexin V binding and
propidium iodine staining as described in Materials and
methods. Concentrations up to 1.0 mM produced no effect on
the viability (i.e. apoptosis or necrosis) of melanocyte or
keratinocyte monocultures. Cocultures were similarly treated
and (a) melanocytes and (b) keratinocytes separated by
differential trypsinization were assessed for viability.
Concentrations up to 1.0 mM also produced no effect on the
viability (i.e. apoptosis or necrosis) of melanocyte or keratinocyte
monocultures. Error bars represent SD.
Melanosome-transfer inhibition
To demonstrate melanosome-transfer inhibition,
melanocytes were labeled with a succinimidyl
ester of CFDA and cocultured with keratinocytes.
These cocultures were assessed by confocal micro-
scopy for CFDA transfer in the presence or
absence of 10 mM niacinamide after 6 days of
treatment (Fig. 4). Control cultures demonstrated
brightly fluorescing keratinocytes (arrowheads)
within a colony of keratinocytes (star) that also
contain CFDA-positive melanocytes (arrows)
(Fig. 4a). The corresponding differential contrast
image (Fig. 4b) demonstrates that the melanocytes
were darkly pigmented (arrows). In contrast, treat-
ment with niacinamide resulted in only weakly
fluorescing keratinocytes within a colony (star)
that also contained CFDA-positive melanocytes
of equal brightness as observed in the control
cultures (arrows) (Fig. 4c). The corresponding dif-
ferential contrast image (Fig. 4d) demonstrated
that the melanocytes were darkly pigmented
(arrows). These data indicate that a considerable
amount of the dye was transferred from melano-
cytes to keratinocytes in the control cocultures and
minimal dye was transferred in the presence of
niacinamide.
To quantitate melanosome-transfer inhibition,
melanocytes were labeled with a succinimidyl
ester of CFDA, cocultured with keratinocytes,
and subsequently quantitated for dye transfer in
the presence of inhibitors and concurrently
assessed for recovery of transfer after removal of
 Melanosome transfer to keratinocytes

Fluorescence DI-Contrast of treatment compounds for 3 days and then after

removal of treatment compounds for an additional
3 days as described in the table in Fig. 4. Cultures
were assessed for transfer at selected time points.
Control

These studies (Fig. 4) demonstrate that niacin-

amide inhibited melanosome transfer from the
melanocytes to the keratinocytes by 14% after
3 days of treatment, while our lectin cocktail
resulted in a 25% inhibition. Cocultures treated
with both the lectin cocktail and the niacinamide
resulted in a 29% inhibition of transfer. The inhi-
bitory effect of niacinamide and the lectin cocktail
Niacinamide

was reversible after 3 days, although the reversibil-

ity occurred more rapidly in the lectin-treated
cocultures. In contrast, the effect of treatment
with both lectins and niacinamide was not com-
pletely reversible in 3 days. In addition, the inhibi-
tory effect induced by the combination treatment
continued even after treatment ceased and was
Melanosome transfer inhibition and recovery using 10 µM niacinamide,
only 21% reversed after 3 days.
lectin Cocktail, or combination of both As previous experiments (21) did not indicate
Treatment group MnX fluorescence Inhibition (%) Recovery (%)
whether a twice-daily-dosing regimen was neces-
Control 1.19 ± 0.1 – – sary to achieve maximum melanosome-transfer
Niacinamide inhibition, we compared the effect of a single treat-
3 days 1.02 ± 0.05 14 –
1 day post 1.01 ± 0.06 14 0
ment to that of a twice-daily treatment for 6 days
2 day post 1.07 ± 0.02 10 29 with various concentrations of the test com-
3 day post 1.17 ± 0.04 2 86 pounds. Table 1 indicates a dose-dependent inhibi-
Lectin
tion of melanosome transfer with the compounds,
3 days 1.893 ± 0.02 25 –
1 day post 0.961 ± 0.06 18 28 and a greater inhibition was observed with the
2 day post 0.991 ± 0.01 17 29 twice-daily-dosing regimen.
3 day post 1.17 ± 0.04 2 92
Both
3 days 0.961 ± 0.06 19 – Human clinical trial
1 day post 0.85 ± 0.02 29 –14
2 day post 0.952 ± 0.06 20 –1.5
3 day post 0.01 ± 0.02 15 21
We next assessed the dose-dependent and reversi-
bility effect of melanosome-transfer inhibition on
in vivo using hyperpigmented facial spots. We
Figure 4. Cocultures of carboxy fluorescein diacetate (CFDA)-
labeled melanocytes and unlabeled keratinocytes were treated twice chose to assess niacinamide, which has demon-
daily for 6 days with vehicle only (a, b) or 10.0 mM niacinamide (c, d). strated effective percutaneous absorption (16), as
After 6 days, cocultures were viewed with fluorescence (a, c) or opposed to the lectins, which because of their
differential interference (DI) contrast (b, d) confocal microscopy relatively high molecular weights (i.e. between 26
simultaneously. (a) Control cocultures containing brightly
fluorescing keratinocytes (arrowheads) within a colony (star) that and 71 kDa) putatively would have poor percuta-
also contained CFDA-positive melanocytes (arrows). (b) Darkly neous absorption.
pigmented melanocytes (arrows) were apparent in the corresponding Image analysis results of the percentage change
DI contrast image. (c) In contrast, treatment with niacinamide
results in only weakly fluorescing keratinocytes within a colony (star)
from baseline in hyperpigmented spot area
that also contains CFDA-positive melanocytes of equal brightness
as observed in the control cultures (arrows). (d) The corresponding Table 1. Assessment of once vs. twice-daily treatment of inhibitors on
DI contrast image demonstrates that the melanocytes remain darkly melanosome transfer
pigmented. Table presents melanosome-transfer inhibition and
recovery using niacinamide (10.0 mM) and/or the lectin cocktail. Sample Inhibition (%)
Cocultures were treated for 3 days with inhibitor and assessed for
transfer as described in Materials and methods at the end of the 3-day Lectin 1 day 10
treatment period and daily for an additional 3 days after termination Lectin 2 day 33
of treatment with inhibitors. 0.1 mM niacinamide 1 day 0
0.1 mM niacinamide 2 day 20
1.0 mM niacinamide 1 day 16
inhibitor by flow cytometry. Melanocytes labeled 1.0 mM niacinamide 2 day 28
10 mM niacinamide 1 day 21
with CFDA were cocultured with keratinocytes 10 mM niacinamide 2 day 32
and initially maintained in the presence or absence

503
Greatens et al.

fraction were assessed after 4 and 8 weeks of treat- ment was stopped at week 8. Specifically, image
ment with either 2 or 5% niacinamide (Fig. 5a,b). analysis results of the percentage change from
After both 4 and 8 weeks of treatment, the side of baseline in spot area fraction were assessed at
the face receiving niacinamide demonstrated a weeks 16, 26, and 42 (i.e. 8, 18, and 34 weeks
higher hyperpigmented spot-reduction efficacy vs. post-treatment, respectively) (Fig. 5e). After the
vehicle for both concentrations; however, only 5% 18-week regression period, the spot-reduction effi-
niacinamide exhibited significant differences at cacy of 5% niacinamide was reduced from D4.7%
both time points. Visual grading performed on (at 8 weeks, significant difference vs. vehicle) to
test subjects at the 8-week time point also demon- D2.8% (at 26 weeks, no significant difference vs.
strated that 5% niacinamide treatment had signif- vehicle), and had leveled off through 42 weeks (no
icantly higher hyperpigmented spot-reduction significant difference vs. vehicle). Visual grading
efficacy vs. vehicle control (Fig. 5c,d), while 2% performed on test subjects also demonstrated
niacinamide treatment did not (Fig. 5c). These that regression from the 5% niacinamide
data confirm a dose–response effect in hyperpig- treatment resulted in no significant difference vs.
mented spot reduction by niacinamide. In vehicle at both the 26- and 42-week time points
addition, visual assessment of captured images by (Fig. 5f). After 18 weeks of regression, the spot-
a dermatologist confirmed that there was no side reduction effect of 5% niacinamide had regressed
effect due to niacinamide such as hypopigmenta- from D0.33 (at 8 weeks, significant difference vs.
tion or erythema after 8 weeks of treatment for all vehicle) to D0.24 (at 26 weeks, no significant dif-
subjects. ference vs. vehicle), and continued to regress to
Image capture was continued in the group that D0.12 at the 42-week time point (no significant
received 5% niacinamide (Group 1) after treat- difference vs. vehicle). In addition, visual assess-
Figure 5. Dose response of niacinamide
spot-reduction efficacy as assessed by
image analysis and visual grading plus
reversibility of 5 % niacinamide
a b hyperpigmented spot-reduction efficacy.
Group 1: Effect of 5% niacinamide moisturizer on Group 2: Effect of 2% niacinamide moisturizer on
facial hyperpigmented spot area facial hyperpigmented spot Area
Percentage change of hyperpigmented
4% 4% spot area fractions from baseline for (a)
5% and (b) 2% niacinamide and vehicle-
Percent change of spot area fraction from

Percent change of spot area fraction from

2% 2%
Week of use Week of use treated sides of the face during treatment
0% 0%
0 2 4 6 8 0 2 4 6 8 period of 8 weeks as assessed by image
–2% –2% analysis. Asterisks indicate significant
difference vs. vehicle at each
baseline

baseline

–4% –4%

–6% –6%
subsequent time point (*P < 0.05,
**P < 0.01). 5% niacinamide reduced
–8% * –8%
spot area fraction vs. vehicle by 4.4%
–10% –10% (P < 0.01), while 2% niacinamide did by
–12% Vehicle –12% Vehicle 2.6% at 8-week time point, indicating a
**
–14%
5% Niacinamide 2% Niacinamide dose dependency of niacinamide effect.
–14%
(c) Average of visual grading for the area
c Visual assessment of niacinamide effect on of hyperpigmented spots from baseline
hyperpigmented spot at 8-week time point d
1
for 5 and 2% niacinamide and vehicle-
Average of grading (–4 to +4: the
higher the more spot reduced)

0.9 Vehicle
treated sides of the face at the end of the
0.8 Niacinamide 8-week treatment period. Asterisks
0.7
0.6
indicate significant difference vs. vehicle
0.5 (*P < 0.05). (d) Representative facial
0.4
images of 5% niacinamide treatment of
0.3
0.2 the face at week 0, 4, and 8. (e)
0.1 Percentage change of hyperpigmented
0
2% 5% Week 0 Week 4 Week 8 spot area fractions from baseline for 5%
Niacinamide
niacinamide and vehicle-treated sides of
e f the face during treatment period of 8
Regression of 5% niacinamide spot reduction effect Regression of 5% niacinamide spot reduction effect
weeks and the regression period of 34
Percent change of spot area fraction from

by image analysis by visual assessment

Average of grading (–4 to +4; the

1.2
higher the more spot reduced)

0%
0 4 8 12 16 20 24 28 32 36 40
1
weeks. Asterisks indicate significant
Week of use
–5%
* difference (*P < 0.05, **P < 0.01). (f)
0.8
5% Niacinamide
Vehicle
Average of visual grading for the area
* of hyperpigmented spots from baseline
baseline

–10% 0.6

** 0.4 for 5% niacinamide and vehicle-treated

–15% 5% Niacinamide
0.2 Vehicle sides of the face at the end of the 8-week
* Weeks of use treatment period plus the 34-week
–20%
0
0 4 8 12 16 20 24 28 32 36 40 regression period (i.e. 42 weeks).
–25%
Treatment period Recovery period
Treatment period Recovery period Asterisks indicate significant difference
vs. vehicle (*P < 0.05).

504

ment by a dermatologist confirmed that there was
no side effect due to niacinamide such as hypopig-
mentation of erythema after 34 weeks of the
recovery period for Group 1.
Discussion
The molecular mechanisms involved in the inter-
cellular transfer process of melanosomes from
melanocytes to neighboring keratinocytes may be
harnessed for the amelioration of skin pigmentary
disorders and to develop novel and effective means
to lighten skin coloration. Initial events in melano-
some transfer most certainly involve recognition
between melanocytes and keratinocytes. The exact
mechanisms facilitating this interaction are cur-
rently unknown. Recently, it has been well demon-
strated that the PAR-2 on keratinocytes regulates
skin pigmentation (24) by increasing the phagocy-
tic activity of the keratinocyte (15,25). In addition,
PAR-2 expression on keratinocytes can be regu-
lated by ultraviolet light to increase melanosome
transfer (26).
Less-defined endogenous lectin receptors and
their glycoconjugate ligands also appear to be
involved in melanosome transfer (27). Initial stud-
ies demonstrated that melanosome transfer
between human melanoma cells and squamous
cell carcinoma-derived keratinocytes could be
mediated by certain lectins and neoglycoproteins
(19). Recent studies have demonstrated that the
addition of these agents to cocultures of normal,
untransformed human melanocytes and keratino-
cytes could also downregulate melanosome trans-
fer (21). In addition, it has been demonstrated that
exposure of melanoma cells to UV irradiation
results in the upregulation of these membrane lec-
tins and neoglycoproteins (27). Therefore, lectins
and their glycoconjugate ligands are excellent can-
didate molecules involved in melanosome transfer
for a variety of reasons. These molecules have been
well studied for their role in numerous cellular
processes, including endocytosis, intracellular traf-
ficking, and cell-to-cell recognition (17).
Niacinamide (28), also called nicotinamide or
3-pyridine carboxamide, is the physiologically
active amide of niacin, vitamin B3. It has multiple
medicinal effects on the skin, including anti-
inflammation, antioxidation, prevention of photo-
immunosuppression, and increasing intercellular
lipid synthesis (25). The lightening of skin color
has also been attributed to niacinamide. Recently,
we have demonstrated that niacinamide effectively
inhibits melanosome transfer by up to 68% in an
in vitro coculture model system (16). Niacinamide
is a biologically active form of niacin (vitamin B3)
Melanosome transfer to keratinocytes
that functions as part of coenzyme I (nicotinamide–
adenine dinucleotide, NAD) and coenzyme II
(reduced form of NADP) (29,30) and is involved
in over 200 enzymatic reactions (31). In the skin,
niacin can act as an anti-inflammatory agent in acne
(32), as an antioxidant preventing photoimmuno-
suppression and photocarcinogenesis (33), and a
stimulator for intracellular lipid synthesis (34).
However, the molecular mechanism of the inhibi-
tory effect of niacinamide or melanosome transfer is
unknown and requires further exploration.
Indeed lectins, neoglycoproteins, and niacin-
amide present interesting molecules that may mod-
ulate skin pigmentation. However, it was
unconfirmed in earlier studies whether the skin-
lightening effect induced by these agents was a
result of apoptotic events occurring in the kerati-
nocytes and/or melanocyte, as opposed to actual
inhibition of melanosome transfer. Our initial
results demonstrated that concentrations pre-
viously demonstrated to inhibit melanosome
transfer could induce dose-dependent apoptosis
in pure cultures of melanocytes and keratinocytes
(Figs 1c,d and 2a,b). However, we had not
observed any microscopic evidence of these apop-
totic effects using these concentrations of inhibi-
tors in our coculture model system (personal
observations). Therefore, we tested the apoptotic
effect of these agents on both keratinocytes and
melanocytes when existing together in coculture
(Fig. 2c,d). The apoptotic effects observed in cul-
tures of isolated melanocytes or keratinocytes
were, in most cases, greatly reduced, or completely
eliminated, when the same concentrations of nia-
cinamide or lectins were tested on melanocytes or
keratinocytes existing in coculture (Fig. 2c,d). This
suggests that cell-survival factors, produced by one
cell type and utilized by the other, may make both
cell types more tolerant to stressful stimuli. Many
cytokines, produced by epidermal cells such as
endothelin-1 (35), insulin (36), transferrin (37),
and basic fibroblast growth factor (38) may help
give cells added resistance. These compounds are
present in conditioned media, often added to
slowly proliferating cell monocultures in our
laboratory. Alternately, cell–cell contact between
melanocytes and recipient keratinocytes might
simply occupy receptors or molecules recognized
by the test inhibitors and/or alter gene expression
rendering cells more resistant to the apoptotic
effects of the test compounds. These results also
suggest that the lectins and their glycoconjugates
responsible for cell-to-cell interactions and that
play a vital role in the melanosome-transfer pro-
cess may also aid in the cell tolerance to the
inhibitory compounds. In addition, there was no
505
Greatens et al.
negative effect on cell viability induced by com-
bining the lectins in a cocktail. In contrast, all
concentrations of neoglycoproteins resulted in
compromised viability to melanocytes and
keratinocytes whether cultured in isolation or in
cocultures (Fig. 2c,d).
After effective and safe concentrations of niacin-
amide and lectins were determined, we then
assessed the reversibility of these agents on the
inhibition of melanosome transfer. As previously
stated, we had shown that both niacinamide and
our lectin cocktail were effective in inhibiting the
transfer of melanosomes from melanocytes to ker-
atinocytes (16,21). In addition, dosing the cocul-
tures twice a day vs. once a day was more effective
in inhibiting the transfer of melanosomes from the
melanocytes to the keratinocytes in coculture.
It was unknown whether inhibition by niacin-
amide or lectins was permanent and possibly
resulting in a toxic effect or reversible and able to
resume melanosome transfer after removal of the
inhibitory agents. To address this, we investigated
the residual effect of lectins and niacinamide after
their removal from the melanocyte–keratinocyte
cocultures. Cocultures treated with niacinamide,
our lectin cocktail, or both were then assayed by
flow cytometry for inhibition of melanosome
transfer as compared with the untreated control.
Cocultures treated with each inhibitor or both
were assayed after 3 days of treatment and com-
pared to the untreated control. Subsequently, sam-
ples were treated with inhibitors for 3 days and
then maintained in the absence of the compound
for 1, 2, or 3 days to assess the reversibility of
inhibition. Results demonstrated the following:
(a) the effect of niacinamide and the lectin cocktail
were significant, and the inhibitory effects were
reversible after 3 days, although the reversibility
appears to happen more rapidly in the lectin-
treated cocultures; (b) the effects of treatment
with both lectins and niacinamide are not comple-
tely reversible in 3 days; and (c) the effects caused
by the combined treatment appeared to persist
even after removal of the test compounds. It is
interesting to note the increased inhibition in the
coculture treated with both the lectin cocktail and
the niacinamide for 3 days and assayed after 1 day
with no treatment. Additionally, the transfer
recovery after 3 days was only 21%.
The skin- lightening effect of niacinamide on
hyperpigmented facial spots was demonstrated
with 2% and 5% niacinamide in an abbreviated
dose–response manner, with 5% being statistically
significant from the vehicle control. In addition,
the skin-lightening effect of niacinamide was tem-
porary since halting application of niacinamide
506
resulted in a normalization of hyperpigmented
facial spots toward the vehicle control. These clin-
ical results are consistent with the coculture results
and suggest that the inhibition of melanosome
transfer by niacinamide is reversible.
In this clinical trial, reduction of hyperpigmented
spots on the vehicle-treated sides of the subjects’
faces showed a somewhat different pattern between
Groups 1 and 2. This fluctuation speaks to the
variation between groups of subjects when the
base size is relatively small. Actually, statistical
test (Student’s t-test) between the two vehicle
groups revealed no significant differences
(P > 0.1) at both 4- and 8-week time points, while
significant correlations were observed on the reduc-
tion efficacy between paired treatments (niacina-
mide vs. Vehicle) in both groups.
In the clinical study, we had anticipated that the
hyperpigmented spot reduction observed in the
8-week treatment phase would have regressed back
to the baseline level by the week 16 time point
(8 weeks of niacinamide treatment followed by
8 weeks without niacinamide). But at the 16-week
assessment, regression had not occurred. This was
likely due to the continued usage of sunscreen pro-
duct during the 8–16-week time period. To ensure
that the affected melanocytes were still responsive
(i.e. ensure that a persistent or permanent
hypopigmentary effect had not been induced), the
clinical study was extended (without provided sunsc-
reen) to follow these subjects for an extended time
period. Thus, after week 16, subjects were not
instructed to use sunscreen but rather were allowed
to follow their normal skin care habits and practices.
We observed that with their normal sun exposure
habits during the ensuing 6-month (26-week) spring-
summer season, the reduction in hyperpigmented
spots had regressed back to baseline. Therefore,
the reduction in spots was a reversible effect and
not due to the induction of a hypopigmentary effect.
Vehicle-treated sides of the faces also demon-
strated a reduction of hyperpigmented spots from
baseline (October) to week 26 (May). This spot
reduction is at least partially attributed to the
instruction to strictly use the provided sunscreen
during 0–16 weeks (fall-winter). As noted above,
no sunscreen product was provided later in the
study (16–42 weeks), and thus the spot reduction
associated with vehicle þ sunscreen use also
regressed to the baseline by week 42, likely as a
result of normal sun-exposure habits and practices
of the subjects during the spring-summer season.
In conclusion, lectins and plasma membrane
glycoproteins of melanocytes and keratinocytes
are required for communication between these
cells, and to facilitate melanosome transfer.

Lectins and niacinamide may prove useful in
manipulating skin pigmentation. Further under-
standing of the underlying mechanisms of action
of these compounds may have important
implications for understanding the complexities
of skin coloration.
Acknowledgements
We thank Mr Ashish Budev and Ms Yasuko Inoue for their
excellent technical assistance. We thank the Society of Cos-
metic Chemists and The Procter and Gamble Company for
their support of this research.
References
1. Seiberg M. Keratinocyte–melanocyte interactions dur-
ing melanosome transfer. Pigment Cell Res 2001: 14:
236–242.
2. Boissy R E. Melanosome transfer to and translocation
in the keratinocyte. Exp Dermatol 2003: 13: 1–8.
3. Hearing V J. Biochemical control of melanogenesis
and melanosomal organization. J Invest Dermatol
Symp Proc 1999: 4: 24–28.
4. Garcia-Borron J C, Solano F. Molecular anatomy of
tyrosinase and its related proteins: beyond the histi-
dine-bound metal catalytic center. Pigment Cell Res
2002: 15: 162–173.
5. Topol B M, Haimes H B, Dubetret L, Bell E. Transfer
of melanosomes in a skin equivalent model in vitro.
J Invest Dermatol 1986: 87: 642–647.
6. Yamamoto O, Bhawan J. Three modes of melanosome
transfers in Caucasian facial skin: hypothesis based on an
ultrastructural study. Pigment Cell Res 1994: 7: 158–169.
7. Jimbow K, Sugiyama S. Melanosomal translocation
and transfer. In: Norlund J J, Boissy, R E, Hearing, V J,
King, R A, Ortonne, J-P, eds. The pigmentary system
physiology and pathophysiology. New York: Oxford
University Press, 1998: 107–114.
8. Lei T C, Virador V M, Vieira W D, Hearing V J. A
melanocyte–keratinocyte coculture model to assess
regulators of pigmentation in vitro. Anal Biochem
2002: 305: 260–268.
9. Westerhof W, Nanninga B. Graying of hair in humans:
morphological and biological data. In: Burgdorff W H C,
Katz S I, eds. Dermatology: Progress and perspectives
proceedings of the 18th World Congress of Dermatol-
ogy. Pearl River, NY: Parthenon Publishing Group,
1992: 508–512.
10. Jimbow K, Fitzpatrick T B, Szabo G, Hori Y. Con-
genital circumscribed hypomelanosis. A characteriza-
tion based on electron microscopic study of tuberous
sclerosis, nevus depigmentosus and piebaldism.
J Invest Dermatol 1975: 64: 50–62.
11. Gilchrest B A, Blog F B, Szabo G. Effects of aging and
chronic sun exposure on melanocytes in human skin.
J Invest Dermatol 1979: 73: 141–143.
12. Mosher D B, Fitzpatrick T B, Ortonne J-P et al. Dis-
orders of pigmentation. In: Fitzpatrick T B, Eisen A Z,
Wolff K, Freedberg I M, Austen K F, eds. Dermatol-
ogy in general medicine, 3rd ed., Vol. 1. New York:
McGraw-Hill, Inc., 1987: 794–876.
Melanosome transfer to keratinocytes
13. Nordlund J J, Abdel-Malek Z A. Mechanisms for
post-inflammatory hyperpigmentation and hypopig-
mentation. In: Bagnara, J, ed. Advances in Pigment
Cell Research Proceedings of the XIII International
Pigment Cell Society. New York, NY: Alan R. Liss,
Inc., 1988: 219–236.
14. Tyack Z F, Pegg S, Ziviani J. Postburn dyspigmenta-
tion: its assessment, management, and relationship to
scarring – a review of the literature. J Burn Care
Rehabil 1997: 18: 435–440.
15. Seiberg M, Paine C, Sharlow E et al. Inhibition of
melanosome transfer results in skin lightening. J Invest
Dermatol 2000: 115: 162–167.
16. Hakozaki T, Minwalla L, Zhuang J et al. The effect of
niacinamide on reducing cutaneous pigmentation and
suppression of melanosome transfer. Br J Dermatol 2002.
17. Monsigny M, Roche A C, Kieda C, Midoux P,
Obrenovich A. Characterization and biological impli-
cations of membrane lectins in tumor, lymphoid and
myeloid cells. Biochimie 1988: 70: 1633–1649.
18. Auger M J, Ross J A. The biology of the macrophage.
In: Lewis C E, McGee J O D, eds. The natural immune
system of the macrophage. Oxford: Oxford University
Press, 1992: 16–33.
19. Cerdan D, Redziniak G, Bourgeois C A, Monsigny M,
Kieda C. C32 human melanoma endogenous lectins:
characterization and implication in vesicle-mediated
melanin transfer to keratinocytes. Exp Cell Res 1992:
203: 164–173.
20. Stibler H, Westerberg B, Hanefeld F, Hagberg B. Car-
bohydrate-deficient glycoprotein (CDG) syndrome – a
new variant, type III. Neuropediatrics 1993: 24: 51–52.
21. Minwalla L, Zhao Y, Cornelius J et al. Inhibition of
melanosome transfer from melanocytes to keratino-
cytes by lectins and neoglycoproteins in an in vitro
model system. Pigment Cell Res 2001: 14: 185–194.
22. Minwalla L, Zhao Y, Le Poole I C, Wickett R R,
Boissy R E. Keratinocytes play a role in regulating
distribution patterns of recipient melanosomes in
vitro. J Invest Dermatol 2001: 117: 341–347.
23. Miyamoto K, Takiwaki H, Hillebrand G G, Arase S.
Development of a digital imaging system for objective
measurement of hyperpigmented spots on the face.
Skin Res Technol 2002: 8: 227–235.
24. Seiberg M, Paine C, Sharlow E et al. The protease-
activated receptor-2 regulates pigmentation via
keratinocyte–melanocyte interactions. Exp Cell Res
2000: 254: 25–32.
25. Sharlow E, Paine C, Eisinger M, Shapiro S, Seiberg M.
The protease-activated receptor-2 upregulates kerati-
nocyte phagocytosis. J Cell Sci 2000: 113: 3093–3101.
26. Scott G, Deng A, Rodriguez-Burford C et al. Protease-
activated receptor 2, a receptor involved in melano-
some transfer, is upregulated in human skin by
ultraviolet irradiation. J Invest Dermatol 2001: 117:
1412–1420.
27. Condaminet B, Redzininiak G, Kieda C. Ultraviolet
rays induced expression of lectins on the surface of a
squamous carcinoma keratinocyte cell line. Exp Cell
Res 1997: 232: 216–224.
28. Draelos Z D. Vitamins and their cutaneous effects.
Cosmetic Dermatol 1999: 12: 17–20.
29. Darby W J, McNutt K W, Todhunter E N. Niacin.
Nutr Rev 1975: 33: 289–297.
30. Horwitt M K. Niacin. In: Goodhart R S, Shils M E,
eds. Modern nutrition in health and diseases. Philadel-
phia: Lea & Febiger, 1980: 204–208.
507
Greatens et al.
31. Sheretz E F, Goldsmith L A. Nutritional influences on
the skin. In: Goldsmith L A, ed. Physiology, biochem-
istry and molecular biology of the skin, 2nd ed., Vol. 2.
New York: Oxford University Press, 1991: 1315–1316.
32. Shalita A R, Smith J G, Parish L C, Sofman M S,
Chalker D K. Topical nicotinamide compared with
clindamycin gel in the treatment of inflammatory
acne vulgaris. Int J Dermatol 1995: 34: 434–437.
33. Gensler H L. Prevention of photoimmunosuppression
and photocarcinogenesis by topical nicotinamide.
Nutr Cancer 1997: 29: 157–162.
34. Tanno O, Ota Y, Kitamura N, Katsube T, Inoue S.
Nicotinamide increases biosynthesis of ceramides as
well as other stratum corneum lipids to improve the
epidermal permeability barrier. Br J Dermatol 2000:
143: 524–531.
508
35. Imokawa G, Miyagishi M, Yada Y. Endothelin-1 as a
new melanogen: coordinated expression of its gene and
the tyrosinase gene in UVB-exposed human epidermis.
J Invest Dermatol 1995: 105: 32–37.
36. Graeven U, Herlyn M. In vitro growth patterns of
normal human melanocytes and melanocytes from dif-
ferent stages of melanoma progression. J Immunother
1992: 12: 199–202.
37. van Muijen G N, Ruiter D J, Hoefakker S, Johnson J P.
Monoclonal antibody PAL-M1 recognizes the transferrin
receptor and is a progression marker in melanocytic
lesions. J Invest Dermatol 1990: 95: 65–69.
38. Halaban R, Langdon R, Birchall N et al. Basic fibro-
blast growth factor from human keratinocytes is a
natural mitogen for melanocytes. J Cell Biol 1988:
107: 1611–1619.
 本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP 图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具