| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Dermatology, Boston University School of Medicine, Boston, MA; and
Department of Dermatology, Columbia University, New York, NY, USA
1Correspondence: Department of Dermatology, Boston University School of Medicine, 609 Albany St., Boston, MA 02118, USA. E-mail: bgilchre{at}bu.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: melanocyte • stem cell • hair follicle • hair cycle
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2
3)
. HF regeneration is characterized by dramatic
changes in its microanatomy and cellular activity. The pigmented hair
shaft is produced only during anagen. Thereafter, the HF shortens in
length up to 70% as a result of apoptosis of cells constituting the
proximal HF epithelium (4
, 5)
.
Within the HF, neural crest-derived melanocytes produce and transport
melanin to the keratinocytes of the precortical zone that
differentiates to form the pigmented hair shaft. During postnatal life,
the HF pigmentary unit cyclically regenerates synchronously with the HF
transition through distinct hair cycle stages (6)
. The
melanogenic activity of the follicular melanocytes is strictly coupled
to the anagen stage, decreases during late anagen and early catagen,
and ceases during late catagen and telogen (7
, 8)
.
Melanogenically active melanocytes are located in the hair bulb,
adjacent to the upper part of the HF dermal papilla and contact
neighboring keratinocytes and dermal papilla fibroblasts with their
dendrites (6)
. Amelanotic melanocytes are located along
the HF outer root sheath and in the HF bulge (9
, 10)
, the
region known to contain HF stem cells (11
, 12)
.
Tyrosinase, an enzyme uniquely expressed in melanocytes, catalyzes the
rate-limiting initial events of melanogenesis (13
14
15
16
17)
.
The tyrosinase gene maps to the albino locus in mice, and
tyrosinase mutations lead to loss of pigment (18)
.
Tyrosinase-related proteins 1 and 2 (TRP1 and TRP2) share 40–45%
amino acid identity with tyrosinase and are also critically important
for melanogenesis, functioning as downstream enzymes in the melanin
biosynthetic pathway (13
, 19)
. TRP1 and TRP2 map to the
brown locus and the slaty locus in mice,
respectively, and mutations lead to a brown or gray rather than normal
black coat color (20
, 21)
.
The expression of tyrosinase, TRP1, TRP2, and tyrosinase activity are
hair cycle dependent (7
, 8
, 22)
. By western and northern
blotting, tyrosinase and TRP1 are undetectable in telogen HF
(7)
. Their expression increases rapidly during early
anagen and peaks in late anagen (7)
. During the
anagen-catagen transition, tyrosinase, TRP1, and TRP2 expression
rapidly decreases in skin homogenates (8)
, but the
expression of these melanogenic proteins in the distinct populations of
HF melanocytes during the hair cycle remains to be elucidated.
Hair pigmentation is tightly regulated by several factors (23
, 24)
. Stem cell factor (SCF) and its tyrosine kinase receptor
c-kit are critically important for the migration, proliferation, and
differentiation of melanoblasts during embryogenesis (25
, 26)
. In mice, melanocytes progressively disappear from the
epidermis shortly after birth and are present only in the HF
(27)
. SCF and c-kit map to the steel (Sl) and
the dominant white spotting (W) loci, respectively, and
mutations in these genes result in unpigmented hairs (25
, 28
, 29)
, suggesting that SCF/c-kit signaling is required for
maintenance of HF melanocytes. Administration of anti-c-kit antibody
(ACK2) during embryonic development leads to coat depigmentation
(27
, 30)
. In contrast, mice overexpressing SCF under the
control of the K14 promoter display hyperpigmentation and melanocyte
localization in atypical areas, including nose, mouth, ears, and
footpads (31
, 32)
.
Postnatally, SCF is expressed in dermal papilla fibroblasts of human HF
and in the hair matrix of murine HF (33
, 34)
, and c-kit is
expressed in murine hair matrix keratinocytes (34)
.
However, whether HF melanocytes express c-kit during the hair cycle and
whether expression correlates with melanocyte proliferation,
differentiation, and apoptosis are unknown. Although a possible
requirement of SCF for the HF melanocytes was demonstrated by using
ACK2 antibody during postnatal development (30)
, the
precise role for SCF/c-kit signaling in the control of cyclic
regeneration of the hair pigmentation unit remains to be elucidated.
To study the regeneration of the hair pigmentary unit and its dependence on SCF/c-kit during the hair cycle, we have analyzed the expression of TRP1, TRP2, tyrosinase, and c-kit in proliferating, differentiating, and melanogenically active melanocytes during depilation-induced HF telogen–anagen transition in normal C57BL/6 mice, in SCF overexpressing mice (promoter: K14), and after administration of anti-c-kit antibody to C57BL/6 mice. Our study provides evidence that SCF/c-kit signaling is indeed required for the generation and migration of functional melanocytes during each new hair cycle and that different populations of the HF melanocytes show differential dependence on SCF during cyclic regeneration of the hair pigmentary unit.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
were housed in the animal facility of Columbia University (New York,
N.Y.). All mice were fed water and murine chow ad libitum
and were kept under 12-h light/dark cycles. To induce hair cycle,
depilation of the back skin was performed, as described previously
(35)
.
In C57BL/6 mice, the next hair cycle stages were studied, using at
least five mice per time point: telogen (unmanipulated skin), anagen II
(3 days after depilation), anagen IV (5 days), and anagen VI (8–12
days). In SCF transgenic mice, only anagen IV stage was studied, and
three transgenic mice and three wild-type animals were used for
analyses 5 days after depilation. For the immunohistochemical analyses,
back skin was harvested parallel to the vertebral line, embedded
quickly, and frozen in liquid nitrogen, using a special technique for
obtaining longitudinal cryosections through the HF from one defined
site (36)
.
Immunohistochemistry, TUNEL, and multicolor confocal microscopy
For immunohistochemical analyses, acetone-fixed cryostat
sections (8 µm) of adolescent C57BL/6 mouse back skin were used. To
study the expression of tyrosinase, TRP1, or TRP2, we performed
immunofluorescence protocol using the corresponding primary rabbit
antisera (Table 1
) and secondary tetramethylrhodamine-isothiocyanate (TRITC)-conjugated
goat-anti-rabbit IgG (Jackson ImmunoResearch, West Grove, Pa.), as
described previously (37)
. Briefly, cryosections were
incubated with primary antisera overnight at room temperature, followed
by application of secondary antibody (diluted 1:200) for 45 min at
37°C. Incubation steps were interspersed by four washes with tris
buffer-saline (TBS, 5 min each).
|
For double immuno-visualization of melanogenic proteins and the
proliferative marker Ki-67, cryosections were first
incubated with rat monoclonal antibody against murine Ki-67
(Table 1)
overnight at room temperature, followed by incubation with
TRITC-labeled goat ant-rat IgG (Jackson ImmunoResearch; 45 min,
37°C). After subsequent washing in TBS, sections were incubated with
primary rabbit antiserum against one of the melanogenic proteins (TRP1,
TRP2, or tyrosinase) overnight at room temperature followed by
incubation with fluorescein-isothiocyanate (FITC)-conjugated secondary
goat-anti-rabbit rat IgG (Jackson ImmunoResearch; 45 min, 37°C).
Finally, sections were washed three times with Tris buffer and then
counterstained by TO-PRO3-iodide (Molecular Probes, Eugene, Oreg.) for
visualization of cell nuclei (38)
.
For the double immunodetection of c-kit, and one of the melanogenic
proteins (TRP1, TRP2, or tyrosinase), the tyramide-amplification method
was used, as described before (39)
. Briefly, after
blocking of endogenous peroxidase and nonspecific avidin/biotin
binding, sections were incubated in TNB buffer (DuPont/NEN, Boston,
Mass.) followed by application of a rat monoclonal antibody against
murine c-kit (1:1000) overnight. Then the corresponding biotinylated
goat anti-rat antiserum (1:200, 30 min), diluted in TNB blocking buffer
(DuPont/NEN), was applied. The reaction product was developed using
commercial tyramide-amplification kit (DuPont/NEN). Sections were
incubated in streptavidin-horseradish peroxidase (HRP; 1:100 in TNB, 30
min), washed with TNT buffer (DuPont/NEN), followed by a 10-min
application of TRITC-tyramide (1:50 in Amplification Diluent,
DuPont/NEN). After blocking nonspecific binding by 10% normal goat
serum, sections were incubated with rabbit antiserum against one of the
melanogenic proteins (TRP1, TRP2, or tyrosinase), washed in TBS (3x5
min), followed by incubation with FITC-conjugated secondary
goat-anti-rabbit IgG (45 min, 37°C). Finally, sections were washed
three times with Tris buffer and then counterstained by TO-PRO3-iodide
for visualization of cell nuclei.
For the double immunodetection of TRP1 and TRP2, a method of
simultaneous detection of tissue antigens using antibodies raised from
the same species was applied (40)
, as described before
(39)
. Rabbit anti-TRP1 antibody was used in the dilution
1:10000 and then was visualized by the biotinylated goat anti-rabbit
IgG/streptavidin-HRP, followed by the application of TRITC-tyramide, as
described above. Then rabbit anti-TRP2 antibody was applied at a
dilution of 1:1000, followed by incubation with FITC-conjugated
goat-anti-rabbit IgG. Therefore, anti-TRP1 antibody used in a very low
concentration (1:10000) was undetectable for the FITC-conjugated goat
anti-rabbit secondary antibody and could only be visualized by highly
sensitive tyramide amplification assay. FITC-labeled secondary antibody
was used consecutively only for the visualizing rabbit anti-TRP2
antibody (diluted 1:1000).
For triple immuno-visualization of the c-kit, one of the melanogenic
proteins (TRP1, TRP2, or tyrosinase), and Ki-67, a method of
simultaneous detection of tissue antigens using antibodies raised from
the same species was applied (40)
. Briefly, skin
cryosections were first immunostained with rat monoclonal antibody
against c-kit (diluted 1:1000; Table 1
) followed by the application of
biotinylated goat anti-rat IgG, streptavidin-HRP, and TRITC-tyramide,
as described above. Then sections were processed for visualization of
Ki-67 with the corresponding rat monoclonal antibody
(diluted 1:50; Table 1
) overnight at room temperature followed by the
application of Cy-5-conjugated secondary goat-anti-rat IgG (Jackson
ImmunoResearch; 1:50, 45 min at 37°C). According to the original
protocols (40)
, rat anti-c-kit antibody used in the
dilution 1:1000 could only be detected by the tyramide amplification
assay. Therefore, anti-c-kit antibody was undetectable for the
Cy5-conjugated goat anti-rat secondary antibody used consecutively for
the visualizing rat anti-Ki-67 antibody (diluted 1:50).
After visualization of Ki-67 antigen, rabbit antiserum
against one of the melanogenic proteins (TRP1, TRP2, or tyrosinase) was
applied followed by the incubation with FITC-labeled goat anti-rabbit
IgG, as described above. Thus, this method allowed us to visualize in
skin sections simultaneously c-kit receptor on cell membrane
(TRITC-coupled red fluorescence), Ki-67 in cell nucleus
(Cy5-coupled blue fluorescence), and one of the melanogenic proteins in
cytoplasm (FITC-coupled green fluorescence).
For the double immuno-visualization of one of the melanogenic proteins
and TUNEL, cryostat sections were processed for the detection of
TUNEL-positive (TUNEL+) apoptotic cells using commercially available
Apo-Tag kit (Intergen, Purchase, N.Y.). Reaction product was visualized
by FITC-labeled anti-digoxigenin antibody. Then, one of the melanogenic
proteins was visualized using the corresponding primary antisera and
TRITC-conjugated goat anti-rabbit IgG, as described above
(4)
, and nuclei were counterstained by TO-PRO3.
For the analysis of SCF expression, the rat monoclonal antibody against
murine SCF was used (Table 1)
, and the tyramide-amplification assay was
applied, as described above. Sections were counterstained by TO-PRO3
for visualization of cell nuclei. Negative controls of immunostaining
were performed by omitting primary antibodies. Skin cryosections of the
SCF-overexpressing transgenic mice, which showed ectopic expression of
SCF in the basal epidermal keratinocytes and abundantly contained
melanocytes in the epidermis and mast cells in the dermis
(32)
, were used as positive controls for antibodies
against SCF, melanogenic proteins, and c-kit, respectively. HF in late
anagen phase displaying a high number of proliferating cells in the
proximal hair matrix were used as an internal positive control for the
antibody against Ki-67. All slides were analyzed using a
Zeiss confocal microscope and were photo-documented using a digital
image analysis system (Pixera, San Diego, Calif.).
Western blot analysis
Total tissue proteins were obtained from extracts of full
thickness C57BL/6 back skin and collected in a buffer consisting of
0.25 M Tris HCl (pH-7.5), 0.375 M NaCl, 2.5% sodium deoxycholate, 1%
Triton X-100, 25 mM MgCl2, 1 mM phenylmethyl
sulfonyl fluoride, and 0.1 mg/ml aprotinin, as described previously
(41
, 42)
. Protein concentrations were determined by the
Bradford method, and 100 µg of proteins was processed for Western
blot analysis as described (41)
. Antibody reaction was
performed with rabbit polyclonal anti-tyrosinase antibody (1:200)
obtained from Dr. V. J. Hearing (43)
. Horseradish
peroxidase-tagged donkey anti-rabbit IgG was used as secondary antibody
(Santa Cruz Biotechnology, Santa Cruz, Calif.; 1:2000). The ECL
detection kit (Amersham) followed by autoradiography (Kodak X-Omatic
AR) detected antibody binding.
Pharmacological manipulations in vivo
The rat monoclonal antibody ACK45 against mouse c-kit
(Pharmingen, San Diego, Calif.) was administered intracutaneously to
back skin of the 8-wk-old C57BL/6 mice in concentration 6 mg/kg at the
various time points after the depilation. The specificity and activity
of this antibody is similar to that of the published antibody ACK2 and
was reported to block c-kit signaling in vitro and in
vivo (27
, 30)
. Intradermal administration of the same
concentrations of mouse normal serum was used as a vehicle control.
Three different experiments were done. In the first experiment, ACK45
treatment was performed on days 1, 3, and 5 after depilation. In the
second experiment, ACK45 was administrated on days 1, 3, 5, 7, 9, 11,
and 13 after depilation. In both experiments, skin samples from 12
control mice and 12 ACK45-treated mice were analyzed in mid- and late
anagen (days 5–14 after depilation). In the third experiment, ACK45
antibody or vehicle control was administered intracutaneously to six
8-wk-old C57BL/6 mice (n=3 for each group) on days 1, 3, 5,
7, 9, 11, and 13 after depilation. After the completion of HF
anagen-catagen-telogen transition (day 30 after depilation), hair cycle
in these mice was induced again in the selected areas of their back
skin, and skin was harvested on days 12–14 after the second depilation
(days 42–44 after the beginning of the experiment).
Morphometry and statistical analysis
In the back skin, >50 HF at the distinct stages of hair cycle,
derived from three to five different mice, were studied for each
antigen analyzed. Only every 10th cryosection was used for analysis to
exclude the repetitive evaluation of the same HF, and two to three
cryosections were assessed from each animal. For the analysis of
melanocyte proliferation, cryosections of the early and mid-anagen skin
harvested on days 3 and 5 after depilation were used. At least 60–70
HF at anagen stages II and IV were analyzed, and the percentage of
Ki-67—positive (Ki-67+) melanocytes per total
number of the HF melanocytes was assessed. All sections were analyzed
at x200–400 magnification, and means and SE
were calculated from pooled data. Differences were judged as
significant if the P value was <0.05 as determined by the
independent Student’s t test for unpaired samples.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. We performed
double immunostainings for these melanogenic proteins (TRP1, TRP2, or
tyrosinase) and the proliferative marker Ki-67 at distinct
hair cycle stages in C57BL/6 mice to determine the relation between
differentiation and proliferation in these cells. To visualize whether
different subpopulations of the HF melanocytes express the c-kit
receptor during defined hair cycle stages, triple immunofluorescence
for c-kit, one of the melanogenic proteins, and Ki-67 was
used.
In resting (telogen) HF, melanocytes were found in the secondary hair
germs adjacent to the dermal papilla. These cells did not express TRP1
and could only be visualized by TRP2, known to be expressed by non-
differentiated melanocytes (Fig. 1A
, B
). In telogen HF, melanocytes were also
tyrosinase negative (tyrosinase-) (not shown) and nonproliferating, as
assessed by the absence of Ki-67 immunostaining (Fig. 1A
, 1B
; Fig. 2A
). However, single proliferating keratinocytes were seen in
the epidermis and in the HF infundibulum (Fig. 1A
, 1B
), consistent with the accepted patterns of cell
proliferation in murine telogen skin (5
, 45)
. Double
immunostaining for c-kit and TRP2 revealed that most of the
TRP2-positive (TRP2+) follicular melanocytes also express c-kit,
whereas some of the TRP2 immunoreactive melanocytes were c-kit negative
(c-kit-) (Fig. 1C
).
|
|
During early anagen (anagen II), TRP2+ melanocytes located in the
proximal part of the secondary hair germ also expressed TRP1, whereas
single cells located in distal part of the secondary hair germ (bulge)
remained only TRP2+ (Fig. 1D
). TRP1-positive (TRP1+) and
TRP2+ cells located in the proximal part of the secondary hair germ of
anagen II HF were also c-kit–positive (c-kit+) (Fig. 1E
, 1F
). However, TRP2+ melanocytes located in the HF bulge were
c-kit- (Fig. 1F
). The latter also were Ki-67
negative (Ki-67-), whereas 56.7 ± 7.3% of TRP1/c-kit
double-positive cells and 40.7 ± 5.4% of TRP2/c-kit
double-positive cells located in the proximal part of the secondary
hair germ also showed Ki-67 immunostaining in their nuclei,
indicating that they were proliferative (Fig. 1E
, 1F
; Fig. 2B
). In addition, c-kit was also
expressed in single cells of the secondary hair germ that were TRP1
negative (TRP1-) and TRP2- (Fig. 1E
). Because c-kit
expression was found on HF keratinocytes previously (34
, 46)
, we interpret these cells to represent a subset of c-kit+
keratinocytes of the HF.
During mid-anagen (anagen IV), HF bulge melanocytes expressed only TRP2
(Fig. 1G
, Fig. 2C
), and did not expressed TRP1,
c-kit, or Ki-67. Melanocytes in the elongating outer root
sheath were TRP2+ and c-kit+, showed relatively low levels of TRP1, and
did not express tyrosinase (Fig. 1H
, I
, J
, K
, L
, Fig. 2C
).
Quantitative immunohistomorphometry showed that 35.6 ± 4.8% of
TRP2+ cells located in the outer root sheath of anagen IV HF were
proliferative. Melanocytes located in the forming hair bulb above the
dermal papilla expressed all three melanogenic proteins (TRP1, TRP2,
and tyrosinase) and were c-kit+, and 64.5 ± 7.1% of them were
proliferative, as indicated by Ki-67 expression (Fig. 1G
, H
, I
, J
, K
, L
, Fig. 2C
).
In late anagen HF (anagen VI), melanocytes located in the HF bulge,
outer root sheath, and proximal hair bulb again showed distinct
patterns in the expression of melanogenic proteins and c-kit. Single
melanocytes seen in the HF bulge showed only TRP2 expression and were
negative for TRP1, tyrosinase, and c-kit (not shown). Melanocytes
located in the outer root sheath expressed TRP2 and c-kit and were TRP1
and tyrosinase- (Fig. 1M
, Fig. 2D
). Melanocytes
in the hair matrix above dermal papilla that were actively producing
pigment showed prominent expression of all three melanogenic proteins
(TRP1, TRP2, tyrosinase) and expressed c-kit (Fig. 1N
, O
, P
,
Fig. 2D
). None of the melanocyte subpopulations of the
anagen VI HF showed Ki-67 immunoreactivity and, therefore,
were nonproliferative (not shown). However, c-kit expression was also
seen in hair matrix keratinocytes, which is consistent with the
expression patterns described before (34
, 46)
.
Thus, during cyclic regeneration of the hair pigmentary unit, follicular melanocytes can be divided into three distinct subpopulations characterized by expression patterns of distinct combinations of melanogenic proteins, c-kit receptor, and proliferative activity. First, HF bulge melanocytes show only TRP2 expression, are c-kit-, and do not proliferate. These most likely represent the stem cell melanocyte population of the HF. Second, HF outer root sheath melanocytes are both TRP2+ and c-kit+, weakly express TRP1, and proliferate during early/mid-anagen. We assume that this subpopulation is stem-cell derived and differentiates into the third subpopulation of melanogenically active melanocytes, located in the hair matrix, that express all three melanogenic proteins, c-kit, and proliferate only during mid-anagen. This population probably represents fully differentiated melanocytes.
SCF is prominently expressed in the HF connective tissue sheath and
dermal papilla during cyclic regeneration of the hair pigmentation unit
Our above data (Figs. 1
, 2)
suggest that during cyclic
regeneration of the hair pigmentation unit, melanocytes undergoing
differentiation and fully differentiated melanocytes are both dependent
on the SCF/c-kit signaling, whereas stem cell melanocytes are probably
SCF independent.
To define SCF expression during anagen-associated regeneration of
the hair pigmentation unit, the highly sensitive
tyramide-amplified immunofluorescence method was applied. We
found that in telogen HF, SCF was prominently expressed in the
follicular connective tissue sheath and in dermal fibroblasts
(Fig. 3A
) and that this expression pattern was maintained throughout
anagen (Fig. 3B
, 3C
). However, during mid- and
late anagen, SCF expression also appeared in the HF dermal papilla
(Fig. 3B
, 3C
). This correlates with maximal
melanin-producing activity of the c-kit+ melanocytes located in the HF
matrix above the dermal papilla. Therefore, our data suggest that
proliferation, differentiation, and melanogenic activity of the c-kit+
follicular melanocytes depends on the c-kit ligand SCF, locally
produced by mesenchymal cells of the HF connective tissue sheath and
dermal papilla during HF anagen development.
|
SCF overexpression increases HF melanocyte proliferation,
differentiation, and ectopic expression of TRP1 in follicular bulge
melanocytes
To determine the role of SCF in the control of the melanocyte
proliferation and differentiation during cyclic regeneration of the
hair pigmentation unit, we have compared the number, melanogenic
protein expression, and proliferative activity of the HF outer root
sheath melanocytes in SCF overexpressing transgenic versus wild-type
mice. This subpopulation of the HF melanocytes was selected as having
the maximal number of the c-kit+ proliferating and differentiating
cells during the anagen IV stage of the induced hair cycle (Figs. 1
, 2)
. SCF expression in the transgenic mice used for this study was also
targeted to the HF outer root sheath by the K14 promoter
(32)
, which would maximize effect of SCF on c-kit+ outer
root sheath melanocytes.
The hair cycle was induced by depilation, and the back skin of both
mouse strains was analyzed at anagen IV stage (i.e., 5 days after
depilation). We performed immunostaining for TRP1, TRP2, and
tyrosinase. TRP1 was prominently expressed by the numerous outer root
sheath melanocytes in the SCF transgenic mice, whereas in the wild-type
mice, TRP1 immunoreactivity was seen predominantly in HF matrix
melanocytes and only rarely in the outer root sheath (Fig. 4A
, B
). The number of TRP1+ cells in the HF
outer root sheath of SCF transgenic mice was significantly higher
(P<0.001) than in wild-type mice (Fig. 4C
). In
addition, TRP1+ cells were ectopically located in the HF bulge of SCF
transgenic mice, whereas such cells were not seen in the bulge of
wild-type HF.
|
Similarly, the number of TRP2+ cells in the HF outer root sheath of SCF
overexpressors was also significantly higher (P<0.001),
compared with that in wild-type mice (Fig. 4D
, E
, F
). However, no increase in the number of
TRP2+ cells was seen in the HF bulge of SCF transgenic mice, compared
with the wild-type HF (not shown). Also, no tyrosinase-positive
(tyrosinase+) cells were seen in the HF outer root sheath of SCF
overexpressors (not shown). Similarly, as in wild-type HF, tyrosinase
was exclusively expressed in the HF matrix melanocytes of SCF
transgenic mice.
Proliferative activity in outer root sheath melanocytes, assessed by double immunovisualization of the TRP2 and Ki-67, revealed a significant increase (P<0.01) in the percentage of double-positive (proliferating) cells in the HF outer root sheath of the SCF transgenic mice, compared with that in wild-type mice (53.7±5.3 vs. 34.1±4.9%, respectively). However, no proliferative TRP2+ melanocytes were seen in the HF bulge of SCF transgenic mice (not shown). This significant increase in the number of melanocytes and in their proliferative activity in SCF-overexpressing mice suggests that SCF is required for melanocyte proliferation during early and mid-anagen. Furthermore, ectopic expression of TRP1 in the HF bulge of SCF transgenic mice suggests that SCF/c-kit signaling also promotes melanocyte differentiation.
Blockade of the SCF/c-kit signaling during early and mid-anagen
results in the formation of gray, partially depigmented hairs
To determine the functional significance of the SCF/c-kit
signaling in the control of cyclic regeneration of the hair pigmentary
unit, anti-c-kit monoclonal antibody (ACK45) was intradermally
administered into the C57BL/6 mouse skin at various times points of the
depilation-induced hair cycle. This commercially available analog of
the ACK2 antibody blocks c-kit signaling in vitro and
in vivo (27
, 30)
. In the first experiment, the
requirement of SCF/c-kit signaling for melanocyte proliferation and
differentiation during early and mid-anagen was assessed by
administration of ACK45 antibody on days 1, 3, and 5 after depilation
(Fig. 5A
). This resulted in the formation of gray, partially
depigmented hairs in mice treated with ACK45, whereas control mice had
black, fully pigmented hair shafts (Fig. 5B
). On day 5 after
depilation, the number of TRP1+ and TRP2+ melanocytes in the HF outer
root sheath and hair matrix was significantly reduced in the
ACK45-treated skin (P<0.05), compared with the vehicle
control skin (Fig. 5C
, D
, E
, F
, I
, J
).
Importantly, no TUNEL+ (apoptotic) cells were seen in the HF after
ACK45 treatment, whereas as a positive control TUNEL cells were seen in
the HF during its transformation from anagen to catagen (Fig. 5G
, 5H
). Instead, proliferation of HF melanocytes
was significantly decreased (P<0.005) in the ACK45-treated
HF versus controls (Fig. 5K
).
|
This suggests that SCF/c-kit signaling is required for melanocyte
proliferation and differentiation during early and mid-anagen. This is
also consistent with the patterns of c-kit expression in proliferating
and differentiating melanocytes during the various hair cycle stages
and with our data demonstrating increased numbers of proliferating HF
melanocytes in SCF-overexpressing mice, described above (Figs. 1
, 2
, 4)
.
Continuous blockade of the SCF/c-kit signaling throughout anagen
leads to the complete hair depigmentation
To determine the dependence on SCF of the melanogenically active
follicular melanocytes located in the HF matrix above dermal papilla,
ACK45 was administered intradermally on days 1, 3, 5, 7, 9, and 11
after depilation (Fig. 6A
). On days 12–14 after depilation, control mice had black,
fully pigmented hairs, whereas ACK45-treated mice had white, completely
depigmented hairs over the entire back (Fig. 6B
, C
, D
),
associated with absence of melanin microscopically in skin
cross-sections (Fig. 6D
).
|
Western blotting of proteins extracted from full-thickness skin treated
with ACK45 showed no tyrosinase protein (molecular weight 75 kDa) as
early as day 8 after depilation, compared with the control skin (Fig. 6E
). SCF/c-kit blockade also dramatically decreased TRP1 and
TRP2 expression in the hair matrix melanocytes, compared with the
control HF (Fig. 6F
, G
, H
, I
). Therefore, these data suggest that
SCF/c-kit signaling is also critical for pigment production by HF
melanocytes during HF anagen development.
SCF/c-kit blockade does not affect the capacity of the hair
pigmentation unit to regenerate during the subsequent hair cycle
To assess whether blockade of the SCF/c-kit signaling once affects
hair pigmentation during subsequent hair cycles, a new hair cycle was
induced by depilation in mice that previously received ACK45 treatment.
A second depilation was performed in these mice after completion of the
HF anagen-catagen-telogen transition of the previous hair cycle (i,e.,
on day 30 after the first depilation) (Fig. 7A
). To compare the color of the new and old hairs, depilation
was performed only in selected areas of the back skin, and after the
second depilation mice were not treated with ACK45 (Fig. 7A
).
|
After the second (local) depilation, HF produced black, fully pigmented
hairs, in striking contrast to the fully depigmented hairs of the
previous hair cycle (Fig. 7B
). After the second depilation,
HF had completely normal patterns of expression of melanogenic proteins
in follicular melanocytes, as well as normal localization of melanin
granules in the HF (Fig. 7C
, D
). This suggests that blocking
SCF signaling did not permanently affect the stem cell melanocytes,
which repopulate the HF and regenerate the hair pigmentation unit
during subsequent hair cycles.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
We demonstrate that during HF telogen–anagen transformation, resting
melanocytes proliferate, differentiate, and migrate within the HF
synchronously with regeneration and construction of HF bulb. Whereas
resting melanocytes in the telogen HF may only be detected by the
expression of TRP2, their transition to melanogenic competence is
accompanied by the appearance of TRP1 and tyrosinase proteins (Figs. 1
, 2)
. These data are in line with the dynamics of the expression patterns
for these markers in melanoblasts and mature melanocytes during
embryonic development (44)
. In addition, hair
cycle–dependent expression of melanogenic proteins demonstrated in our
study is consistent with molecular data published previously (7
, 8)
. It was shown that only dopachrome tautomerase (TRP2)
activity was detected in telogen, whereas tyrosinase, dopa oxidase
activities, and gp75 protein (TRP1) become detectable in early anagen,
reaching the highest activity during late anagen. (7
, 8)
.
Based on melanogenic protein expression patterns and proliferative
activity during cyclic regeneration of the hair pigmentation unit, HF
melanocytes may be divided into three distinct subpopulations. The
first is located in the HF bulge, expresses only TRP2, does not
proliferate, and presumptively represents HF melanocyte stem cells. The
second is located in the HF outer root sheath, expresses TRP2 and
relatively weak TRP1, displays proliferative activity during early and
mid-anagen, and represents differentiating melanocytes. The third is
located in the hair matrix above the dermal papilla; expresses TRP2,
TRP1, and tyrosinase; proliferates only during mid-anagen; actively
produces melanin during mid- to late anagen; and progressively
disappears during catagen, presumably via apoptosis and/or
dedifferentiation (6
, 47
, 48)
.
We show here that the two latter melanocyte subpopulations express
c-kit and show dependence on SCF/c-kit signaling during anagen. In the
anagen HF, SCF is expressed by mesenchymal cells of the HF connective
tissue sheath and the dermal papilla (Fig. 3)
(i.e., in close vicinity
to the proliferating, differentiating, and pigment-producing
melanocytes; Figs. 1
, 2
).
We show that the HF melanocytes are maximally proliferative during
early and mid-anagen, which is associated with c-kit expression on
their surface (Figs. 1
, 2)
. Overexpression of the locally produced SCF
in transgenic mice (promoter: K14) (32)
significantly
increases the number of HF melanocytes and their proliferative activity
(Fig. 4)
. In contrast, the short-term administration of the ACK45
antibody blocking c-kit signaling dramatically reduces melanocyte
proliferation and cell number in mid-anagen HF (Fig. 5)
, which leads to
the partial hair depigmentation. These data are consistent with
previous reports demonstrating increase of the melanocyte proliferation
in the early anagen HF (6
, 9
, 49)
. This is also in line
with the observation that SCF/c-kit signaling is critically important
for the stimulation of melanocyte proliferation during embryonic and
postnatal development (31
, 50
51
52)
.
Here, we assume that SCF also promotes HF melanocyte differentiation
directly or indirectly via regulation of the melanogenic protein
expression. We demonstrate that overexpression of the locally produced
SCF amplifies expression of TRP1, a marker of more differentiated
melanocytes, in the outer root sheath, and leads to the aberrant
appearance of TRP1 expression in HF bulge melanocytes (Fig. 4)
. This is
consistent with the earlier induction of melanocyte differentiation in
the HF of transgenic mice, compared to wild-type mice. Also, we show
that mature melanocytes expressing c-kit in the HF matrix are
strikingly dependent on SCF to maintain their pigment-producing
activity. Continuous ACK45 treatment throughout anagen dramatically
decreases TRP1, TRP2, and tyrosinase expression in this subpopulation
of HF melanocytes and leads to the formation of completely depigmented
hair (Fig. 6)
. These data are consistent with other reports linking
regulation of the TRP1 and tyrosinase genes with SCF/c-kit signaling
(53)
. This regulation may occur via recruitment of the
Microphthalmia transcription factor (54
, 55)
, which is
known to strongly stimulate the transcriptional activities of the
melanogenic protein genes (56
, 57)
.
One group has reported that administration of an antibody blocking
c-kit signaling during embryonic or neonatal development induces
apoptosis in murine melanoblasts and melanocytes (58
, 59)
.
However, as we show here, no apoptotic cell death occurs in the HF
melanocytes of young adult mice after ACK45 treatment (Fig. 5)
. This
indicates that during cyclic regeneration of hair pigmentation unit,
SCF/c-kit signaling is involved in the control of melanocyte
proliferation, differentiation, and melanin production, rather than in
the inhibiting apoptotic cell death, at least in the postnatal period.
In our study, we also demonstrate that TRP2+ melanocytes located in the
HF bulge do not express c-kit receptor (Fig. 1)
and, most likely, are
not dependent on SCF/c-kit signaling during the anagen-coupled
regeneration of the hair pigmentation unit. In telogen HF, these cells
are located in the immediate vicinity of those melanocytes that
co-express TRP2 and c-kit. During anagen, TRP2+ cells in the bulge do
not proliferate and remain c-kit- (Figs. 1
, 2)
. We assume that these
HF melanocytes represent a stem cell-like subpopulation, most likely
responsible for the restoration of hair pigmentation unit during
subsequent hair cycles. It was shown previously that rapid termination
of anagen-associated melanogenesis by cyclophosphamide does not prevent
a new pigmentary activity during the following hair cycle
(60)
. Recently it was shown that melanocyte stem cells in
the skin of SCF transgenic mice are SCF independent (31)
.
Our data, showing that continuous blockade of the SCF/c-kit signaling
does not prevent the hair pigmentation unit from generating
melanogenically active melanocytes during the next hair cycle (Fig. 7)
,
support this concept.
However, we cannot exclude the importance of other factors, which may
be required for cyclic regeneration of the hair pigmentation unit. It
was shown that changes in anagen-associated melanogenesis were
accompanied by changes in gene expression of MC1-R (melanocortin
receptor-1) activated by POMC-derived ACTH and MSH peptides
(61)
. Furthermore, it has been recently demonstrated that
HF-associated ACTH, 
-MSH, and MC1-R expressions are hair cycle
dependent (62
, 63)
. In addition, other proteins known to
be involved in melanocyte biology (namely, Agouti signal protein,
endothelin family, FGF2, hepatocyte growth factor), may be important
for modulating the activity of HF melanocytes during hair cycle
(16
, 24
, 52
, 64
, 65)
.
Nevertheless, pharmacological blockade of SCF/c-kit signaling by ACK45
treatment during depilation-induced hair cycles appears to represent a
simple and fully reversible model for programming hair pigmentation in
mice. This fully supports previous reports of depigmented hairs 8 wk
after administration of anti-c-kit antibody to postnatal mice
(30)
. However, our model allows for partial or complete
hair depigmentation and, thus, for programing either gray or white
hairs in mice, depending on the time course of ACK45 administration
during the depilation-induced hair cycle. Finally, this depigmentation
is fully reversible and allows exploration of the effects of the
experimental modulators of pigmentation during several hair cycles.
Therefore, this model appears suited to studies of hair
pigmentation.
|
ACKNOWLEDGMENTS |
|---|
.
Received for publication June 29, 2000.
Revision received September 7, 2000.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
