| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Laboratory of Cell Biology, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892, USA;
* Department of Biochemistry, Sapporo Medical University, Sapporo, Japan; and
Pathology Section, National Heart, Lung and Blood Institute, National Institutes of Health, Bethesda, Maryland 20892, USA
1Correspondence: Laboratory of Cell Biology, Building 37, Room 1B25, National Institutes of Health, Bethesda, MD 20892, USA. E-mail: hearingv{at}nih.gov
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: albinism • chaperones • pigmentation • melanogenesis
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
. OCA has
several distinctive phenotypes in humans, three of which have been
characterized and mapped to distinct pigment genes, as noted below.
Analogous phenotypes of OCA were recognized early in mice, and the
availability of inbred lines carrying characterized mutations in the
responsible genes has made them invaluable for studying genetic
disorders of pigmentation.
Mammalian melanogenesis is regulated directly or indirectly by
more than 95 distinct loci (3
4
5)
. Tyrosinase,
tyrosinase-related protein 1 (Tyrp1; also known as TRP1 and gp75), and
tyrosinase-related protein 2 (Dct; also known as TRP2 and DOPAchrome
tautomerase) are expressed specifically in melanocytes and function in
melanin synthesis within melanosomes. All three enzymes are type 1
membrane-bound melanosomal glycoproteins with similar structural
features and all three are believed to interact with one another in
melanosomes (6
7
8)
. These proteins are synthesized on
ribosomes and transported through the rough endoplasmic reticulum (ER)
and Golgi apparatus, where their post-translational processing and
glycosylation take place. They are then sorted to early melanosomes;
after delivery there, the synthesis of melanin usually ensues.
Tyrosinase catalyzes the critical, rate-limiting step of tyrosine
hydroxylation to DOPAquinone and the oxidation of 5,6-dihydroxyindole
(DHI) to indole-5,6-quinone (9
10
11)
. Since expression of
tyrosinase is necessary for pigment production, its malfunction results
in the reduction of melanin pigment by melanocytes in the skin, hair
follicles, and eyes, causing OCA type 1 (tyrosinase-negative)
(12)
. OCA1 in humans is comparable to the albino mouse,
which has a mutation (C103S) in Tyr that results in
amelanotic melanocytes that produce no melanin in the hair, eyes, or
skin (3
, 13
, 14)
. Melan-c melanocytes were established
from albino mice with the point mutation in Tyr, which
inactivates tyrosinase; these cells are suitable for studying OCA1.
Tyrp1, an abundant glycoprotein expressed specifically by
melanocytes, functions in mice as an enzyme that oxidizes
5,6-dihydroxyindole-2-carboxylic acid (DHICA) monomers into melanin
(15
, 16)
. Dct functions specifically to tautomerize
DOPAchrome to DHICA. Melanins generated from DHICA are brown in color
and poorly soluble, whereas these generated by DHI derivatives are
black and insoluble. Tyrp1 has an additional and important function to
stabilize tyrosinase in melanosomes, and thus is a multifunctional
melanogenic protein (17
18
19)
. We previously showed that
mutations in the gene encoding Tyrp1 not only accelerated the
degradation of mutant Tyrp1, but also the degradation of wild-type
tyrosinase coexpressed in these melanocytes, though it did not affect
other melanogenic proteins (8)
. Malfunction of Tyrp1
causes OCA type 3, which is associated with moderate hypopigmentation
of the skin, hair, and eyes (20
, 21)
. Melanocytes isolated
from an OCA3 patient produce reduced levels of brown melanin and had
relatively unstable tyrosinase (22)
. However, since OCA3
is a rare disease, its etiology has not been further clarified. OCA3 is
the human homologue of the brown mouse (which has a C110Y mutation in
its Tyrp1 gene) that produces brown rather than black
melanin (5
, 14)
. Thus, melan-b melanocytes that carry a
spontaneous mutation of the Tyrp1 gene are suitable for
investigating the etiology of OCA3 and to characterize how Tyrp1
stabilizes tyrosinase.
Mutations in general affect the biogenesis and function of proteins
(23)
. Correct folding is necessary for proteins to be
properly transported to their final destinations. For this purpose,
cells have a rigorous quality control system via molecular chaperones
such as Bip, calnexin, or calreticulin whereby newly synthesized
proteins are folded quickly and misfolded proteins are diverted to be
degraded or rescued (24)
. In such an environment, calnexin
functions as a molecular chaperone for the folding of glycoproteins
(25)
and retains proteins misfolded due to mutation(s) in
the ER. This abnormal ER retention of mutant proteins is responsible
for some inherited diseases such as cystic fibrosis, Wilson’s disease,
and von Willebrand disease (26
27
28)
. Calnexin has been
shown to play an important role in the folding of tyrosinase (29
, 30)
. Recent studies have suggested that mutant tyrosinase
synthesized in OCA1 melanocytes and mutant OA1 (ocular albinism type 1)
protein synthesized in OA1 melanocytes are not transported to
melanosomes, but are retained in the ER bound to calnexin (31
, 32)
; other factors, such as a lack of glycosylation, can also
affect the maturation and transport of tyrosinase (30
, 33
, 34)
. Although the etiology of OCA1 is becoming clearer, that of
OCA3 remains unknown. Another molecular chaperone, termed Bip, binds
preferentially to hydrophobic arrays of amino acids, recognizing
hydrophobic patches of unfolded proteins (35
, 36)
.
Although Bip binds stably to unfolded Tyrp1 when its disulfide bridges
have been chemically disrupted by dithiothreitol (DTT)
(37)
, it is not known whether Bip is involved in the
folding or retention of mutant melanogenic proteins in OCA.
In this study, we show the retention and premature degradation of mutant Tyrp1 or tyrosinase in the ER after binding to calnexin. We also show that tyrosinase and Tyrp1, and to some extent Dct, have an intermolecular association in that a mutation of one influences the maturation and degradation of the others. We suggest that abnormal retention and degradation of mutant tyrosinase are important etiologics factor of OCA1 and that similar effects on wild-type tyrosinase elicited by mutant Tyrp1 are important etiologic factors of OCA3.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, melan-b
(14)
, and melan-c (14)
cells were gifts from
Dr. Dorothy Bennett (St. George’s Hospital, London, UK) and were
cultured in RPMI 1640 Medium (Gibco BRL, Rockville, MD) containing 5%
fetal calf serum, 2 mM glutamine, 100 µM 2-mercaptoethanol, 200 nM
phorbol 12-mystrate 13-acetate (Sigma, St Louis, MO), 100 U/ml
penicillin, and 0.1 mg/ml streptomycin (Gibco BRL). NIH3T3 mouse
fibroblasts were obtained from the ATCC (Rockville, MD) and cultured in
Dulbecco’s modified Eagle medium (Gibco BRL) containing 10% fetal
calf serum, 2 mM glutamine, 100 U/ml penicillin, and 0.1 mg/ml
streptomycin (Gibco BRL). Calnexin (SPA-860) and KDEL (SPA-827)
polyclonal antibodies were purchased from StressGen (Victoria, Canada);
PEP1 against Tyrp1, PEP7 against tyrosinase, PEP8h against
Dct, and PEP13 against gp100 are rabbit antisera raised in our
laboratory against the carboxyl-terminal peptides of these proteins, as
described (39
40
41)
. Anti-BiP antibody was generously
provided by Drs. Taira and Yamashita (Iwate University, Iwate, Japan)
(42)
. Alexa 595-labeled goat anti-rabbit IgG
[F(ab)2] and Alexa 488-labeled goat anti-mouse
IgG [F(ab)2] were from Molecular Probes
(Eugene, OR). Anti-rabbit Ig horseradish peroxidase-linked whole
antibody was from Amersham Pharmacia Biotech (Piscataway, NJ).
Melanogenic assay
Tyrosinase activity was measured as described previously
(43)
. Cells were harvested with trypsin/EDTA, washed twice
with phosphate buffer saline without CaCl2 and
MgCl2 (PBS-), then
solubilized in lysis buffer A (1% Nonidet P40, 0.01% SDS, 100 mM Tris
HCl, pH 7.2, 1 µg/ml aprotinin, and 100 µM
phenylmethylsulfonyl-fluoride) overnight at 4°C. The samples were
centrifuged at 14,000 g for 10 min at 4°C and supernatants
were recovered for assay. Thirty microliters of each sample were taken
and added to 10 µl phosphate buffer containing L-DOPA cofactor and 10
µl 0.25 mCi/mmol [U-14C] tyrosine (NEN Life
Science Products, Boston, MA). The reactions were mixed and incubated
at 37°C overnight, then 40 µl of each sample was transferred to 3MM
filter paper discs (Whatman, Maidstone, UK) and air dried. Each filter
was washed three times in 0.1 N HCl, twice in 95% ethanol and once in
acetone, then air-dried. Radioactive melanin retained by the filters
was determined in a Beckman scintillation counter. Tyrosinase activity
is defined as pmol 14C-tyrosine incorporated into
melanin/µg protein/hr at 37°C.
Melanin content in cell extracts
Melanin content was measured using a modification of a
previously reported method (44)
. Melanocytes were cultured
until they became 100% confluent in 10 cm dishes and the cells were
solubilized in lysis buffer B (1% Nonidet P-40, 0.1% SDS, 0.5%
deoxycholic acid, 150 mM NaCl, and 50 mM Tris-HCl, pH 7.5) containing a
protease inhibitor mixture (Boehringer Mannheim, Mannheim, Germany).
Aliquots of each cell extract were transferred to 96-well plates, mixed
well, and immediately quantitated by absorbance at a wavelength of 405
nm using an automatic microplate reader (Molecular Devices, Sunnyvale,
CA) calibrated against synthetic melanin (Sigma, St. Louis, MO).
Immunohistochemical staining
Dual labeling, using immunofluorescence methods and laser
scanning confocal fluorescence microscopy, was used to evaluate the
localization of melanogenic proteins in melanocytes. The following
antibody dilutions were used: PEP1, 1:500; PEP7, 1:500; KDEL,
1:500; and HMB-45, 1:10. Melanocytes were plated in 4-well Lab-Tek
chamber slides (Nalge Nunc Intl., Naperville, IL) 24 h before each
experiment and fixed in 4% paraformaldehyde for 20 min at 23°C.
Cells were subsequently permeabilized with methanol for 20 min at
4°C; all subsequent processing was done at 23°C. After washing in
PBS, cells were incubated in PBS containing 0.5% bovine serum albumin
and 50 mM Tris-HCl for 1 h, then washed three times in the same
buffer to wash out the antibody after each step. The cells were then
incubated with monoclonal antibodies KDEL or HMB-45 overnight at 4°C
and/or with polyclonal antibodies PEP1 or PEP7 for 1 h at
23°C. This was followed with the appropriate secondary antibodies:
goat anti-rabbit IgG labeled with Texas red (dilution, 1:500) or Alexa
flour 488 goat anti-mouse IgG [F(ab)2] labeled
with fluorescein (dilution, 1:500). Nuclei were counterstained with
DAPI (blue fluorescence). Immunoreactive cells were classified into
three categories according to whether they showed green, red, or yellow
fluorescence (the latter color indicating colocalization of the red and
green signals). All preparations were examined with a Model
TCS4D.DMIRBE confocal microscope (Leica, Heidelberg,
Germany) equipped with argon and argon-krypton laser sources.
Western blotting and endoH digestion
For Western blotting, cells were solubilized for 1 h at
4°C in lysis buffer A. After protein concentrations were measured and
equalized, 0.5 µg of each lysate was resuspended in 2.5 µl 50 mM
sodium acetate, 0.5% SDS, 1% 2-mercaptoethanol, pH 5.5, and heated at
95°C for 10 min. Five microliters of 50 mM sodium acetate, pH 5.5,
with or without 0.25 U endoH (Boehringer) were added to each sample and
incubated for 3 h at 37°C. Two or 5 µg of each lysate was used
to examine the expression of Bip and calnexin, respectively. endoH
digestions were terminated by adding 15 µl 0.9 M Tris-HCl, pH 8.45,
24% glycerol, 8% SDS, 0.015% Coomassie blue G, 0.005% phenol red
(NOVEX, San Diego CA) and heating at 95°C. Samples were separated in
8% SDS-polyacrylamide gel electrophoresis (PAGE) gels (NOVEX) and
electrophoretically transferred to Immobilon-P membranes (Millipore,
Bedford, MA). Blots were blocked in 5% nonfat milk in 20 mM Tris, 137
mM NaCl, 3.8 mM HCl, 0.1% Tween 20, pH 7.6 (TBS-Tween) for 2 h
and then incubated with PEP1 (1:1,000), PEP7 (1:1,000), Bip
(1:1,000), or calnexin (1:1,000) in TBS-Tween. After four additional
washes in TBS-Tween, blots were incubated in anti-rabbit or mouse Ig
horseradish peroxidase-linked whole antibody (1:1,000) in TBS-Tween,
and immunoreactivity was detected by ECL Western blotting detection
(Amersham Pharmacia Biotech).
Metabolic labeling and immunoprecipitation
Metabolic labeling and immunoprecipitation experiments were
performed as described previously, with some modifications
(29)
. Melanocytes were cultured in methionine-depleted
RPMI 1640 (Gibco BRL), then labeled at 37°C for the times noted with
50 µCi [35S]methionine (Amersham Pharmacia
Biotech). For pulse-chase experiments, melanocytes were labeled for 20
min, then chased for specific periods up to 20 h in RPMI 1640
medium supplemented with 1 mM unlabeled methionine. Cells were
harvested and washed twice with PBS at 4°C and solubilized for 1 h at 4°C in lysis buffer C (2% Triton-X 100, 400 mM KCl, 50 mM
triethanolamine acetic acid, pH 7.2, containing protease inhibitor
mixture); a concentration of 1% Triton-X 100 was used for protein
binding among melanogenic proteins. For immunoprecipitation, the
clarified supernatants were preincubated with 20 µl Pansorbin
(Calbiochem, San Diego, CA) for 1 h at 4°C with mixing. The
supernatants were collected by centrifugation at 4°C and incubated
with tyrosinase (1:200), Tyrp1 (1:200), Dct (1:200), gp100 (1:200), Bip
(1:400), or calnexin (1:400) antibodies at 4°C for 1 h with
mixing. Twenty microliters Pansorbin was then added to each sample,
which was then incubated at 4°C with mixing for 1 h. The immune
complexes were washed four times with 0.7 M KCl, 0.5 M NaCl, 0.05%
Triton X-100, 10 mM Tris-HCl, pH 7.5, by centrifugation. The final
pellets were directly separated by SDS-PAGE or kept for further
experiments, such as endoH digestion or sequential immunoprecipitation,
as detailed below. For SDS-PAGE, the pellets were resuspended in 2x
SDS sample buffer, heated at 95°C for 5 min, and centrifuged. The
samples were separated on 8% SDS-PAGE gels and the separated protein
bands were visualized by fluorography using EnLightning (NEN Life
Science Products) and Kodak X-OMAT AR X-ray film (Eastman Kodak,
Rochester, NY, USA).
For endoH digestion, immune complexes were resuspended in 3 µl of 50 mM sodium acetate, 0.5% SDS, 1% 2-mercaptoethanol, pH 5.5, and heated at 95°C for 10 min. Five microliters of 50 mM sodium acetate, pH 5.5, and 0.25 U endoH were added to each sample. Samples were incubated for 3 h at 37°C and digestions were terminated by adding 10 µl 2x SDS sample buffer and heating at 95°C for 5 min.
For sequential immunoprecipitation, immune complexes were resuspended in 20 µl 1% SDS, 5 mM EDTA, 10 mM Tris-HCl, pH 7.5, and heated at 95°C for 5 min. Supernatants were collected by centrifugation and 180 µl 2% Triton X-100, 0.15 M NaCl, 10 mM Tris-HCl, pH 7.5 were added to each. Samples were again preincubated with 20 µl Pansorbin for 1 h at 4°C with mixing. The collected supernatants were incubated with the second antibody (as described in the figure legends) for 1 h at 4°C with mixing. Samples were again incubated with 20 µl Pansorbin for 1 h at 4°C with mixing to collect immune complexes. After centrifugation, the final pellets were resuspended in 2x SDS sample buffer, heated at 95°C for 5 min, separated on 8% SDS-PAGE gels, and visualized by fluorography as described above.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Although the tyrosinase activity of melan-b melanocytes is actually
somewhat higher than that of melan-a melanocytes, the melanin content
of melan-b melanocytes was only 25% that of melan-a cells, which
provides an important clue to the etiology of OCA3. Tyrosinase activity
and melanin content of melan-c cells were at background level, as
expected.
|
Subcellular distribution of wild-type and mutant tyrosinase and
Tyrp1
To investigate the subcellular localization of wild-type and
mutant tyrosinase and Tyrp1, we used immunohistochemical staining to
compare the distributions of these proteins with HMB-45, a marker for
early melanosomes (45)
, and KDEL, a marker for the ER
(46)
. Tyrosinase and Tyrp1 were detected by red
fluorescence, KDEL and HMB-45 by green fluorescence; in the merged
images shown in Fig. 1
, yellow indicates colocalization of the two signals.
|
In melan-a melanocytes (Fig. 1
, left column), the majority of
wild-type tyrosinase was found in granules distributed throughout the
cytoplasm (mainly in dendrites) and in a reticular pattern around the
perinuclear rim. Although some tyrosinase colocalized with the ER
marker (KDEL), most of it was in the particulate structures. The
staining pattern of early melanosomes (HMB-45) was indistinguishable
from most of the granular staining of wild-type tyrosinase in melan-a
melanocytes, although some tyrosinase-positive melanosomes were found
in the periphery of the cells in mature melanosomes that no longer
stained with HMB-45 (cf discussion below). Staining for wild-type Tyrp1
gave similar results. Colocalization of some Tyrp1 with the ER marker
was observed in perinuclear areas of some melanocytes, and
colocalization of Tyrp1 and HMB-45 was more extensive than that
obtained for tyrosinase. These results con-firm that the wild-type
forms of tyrosinase and Tyrp1 distribute primarily in melanosomes in
melan-a melanocytes.
In melan-b melanocytes (Fig. 1
, middle column), mutant Tyrp1 was found
only in the ER reticular network. Localization of wild-type tyrosinase
was also affected in that the majority colocalized with the ER marker.
In contrast, HMB45 antibody stained granular structures similar to that
seen in wild-type melan-a cells, showing that gp100 was still sorted
correctly to early melanosomes. These results show clearly that a
mutation in Tyrp1 not only affects the sorting of mutant Tyrp1, but
also that of wild-type tyrosinase.
We then examined the localization of these proteins in melan-c
melanocytes (Fig. 1
, right column). In melan-c cells, mutant tyrosinase
was exclusively confined to the ER network and colocalized with the
KDEL marker; mutant tyrosinase colocalized with HMB-45 in the
perinuclear region (gp100 is also processed through the ER), but HMB-45
staining of early melanosomes in the periphery of the melan-c cells was
devoid of mutant tyrosinase. Although some of the wild-type Tyrp1 in
these melan-c melanocytes was sorted to granular structures, much was
retained in the ER. Again, gp100 detected by HMB-45 was sorted
correctly to early melanosomes.
Mutations render tyrosinase and Tyrp1 sensitive to endoH
To more closely examine the processing of mutant and
wild-type tyrosinase and Tyrp1 in these different lines of melanocytes,
we assessed their sensitivities to endoH, an enzyme that removes high
mannose-type carbohydrates from N-linked glycoproteins. This
conversion occurs at the medial Golgi region, and when proteins are
correctly processed through the ER and Golgi, they become resistant to
endoH. Melan-a, melan-b, and melan-c melanocytes were solubilized,
incubated with or without endoH, and analyzed by Western blotting
(Fig. 2
). In melan-a melanocytes, the majority of wild-type tyrosinase and
Tyrp1 can be seen to be resistant to endoH. Some digestion of normally
processed proteins by endoH occurs due to heterogeneous processing of
sugars, and even normally processed proteins suffer minor changes in
mobility, as seen for wild-type tyrosinase and Tyrp1 in melan-a
melanocytes. In contrast, in melan-b melanocytes, mutant Tyrp1 was
completely sensitive to endoH; even though tyrosinase in these cells is
of the wild-type, its sensitivity to endoH was markedly increased.
Conversely, in melan-c melanocytes, mutant tyrosinase was completely
sensitive to endoH whereas more of the wild-type Tyrp1 was sensitive to
endoH.
|
Note that although the processing of wild-type tyrosinase is
dramatically compromised in melan-b melano-cytes, these cells
had higher levels of tyrosinase activity than did melan-a melanocytes,
as noted above. Although this might seem contradictory, previous
studies (47)
have shown that the catalytic functions of
tyrosinase are not dependent on their sugar modification. These data
clearly show that a mutation in tyrosinase or Tyrp1 caused not only
their retention in the ER, confirming the immunohistochemical results
described above, but also altered steady-state localization of the
wild-type protein in these same cells.
Mutations in tyrosinase or Tyrp1 induce degradation of mutant Tyrp1
or tyrosinase but do not affect processing of other melanogenic
proteins
The results of the immunostaining and Western blotting experiments
reported above thus demonstrated that mutant tyrosinase and mutant
Tyrp1 were retained in the ER. Mutations often affect the 3-dimensional
structure of proteins and eventually cause their misfolding
(23)
. These misfolded proteins are confined and eventually
eliminated by the quality control mechanism (24
, 25)
. To
examine the dynamics of mutant and wild-type melanogenic proteins in
these cells, melanocytes were radiolabeled with
[35S]methionine; extracts of the labeled
proteins were then immunoprecipitated with specific antibodies. The
sensitivities of the wild-type or mutant proteins to endoH were also
assessed.
The synthesis and processing of wild-type and mutant tyrosinase are
shown at the top of Fig. 3
. In melan-a melanocytes (left column), immature (I) and mature (M)
forms of wild-type tyrosinase are detected as 69 kDa and 75 kDa
proteins, respectively. Wild-type tyrosinase was rapidly converted from
the immature form to the mature form within 1.5 h of chase and
degraded slowly thereafter with a half-life greater than 6 h. Both
forms of tyrosinase became highly resistant to endoH at 1.5 h of
chase and thereafter. In contrast, in melan-c melanocytes (right
column), the mutant tyrosinase was detected only as the 69 kDa immature
protein and was completely sensitive to endoH throughout the chase
period. As expected, its rate of degradation was much quicker, being
almost completely degraded within 3 h of chase. In melan-b
melanocytes (middle column), even though tyrosinase is wild-type, its
rate of conversion to the mature form from the immature form was
significantly retarded and its degradation was markedly accelerated.
Note that although most of the wild-type tyrosinase expressed in the
melan-b melanocytes is not correctly processed, its catalytic function
is actually higher in these cells than in the melan-a melanocytes.
Although surprising, this occurs because the catalytic function of
tyrosinase is not affected by its glycosylation; previous studies have
shown that mature tyrosinase can be fully deglycosylated back to the
naive form and that catalytic function is not impaired
(47)
.
|
We next performed a similar analysis to determine how Tyrp1 expression is regulated in these cells. In contrast to tyrosinase, the turnover rate of Tyrp1 in melan-a cells was faster and its half-life was only 3 h. Nascent Tyrp1 was expressed as an immature 69 kDa form that was quickly processed to the mature 75 kDa form. As seen with tyrosinase, newly synthesized Tyrp1 was completely sensitive to endoH treatment, but became almost completely endoH resistant within 1.5 h of chase. In contrast, in melan-b melanocytes, mutant Tyrp1 was detected only as the immature 69 kDa form and had almost completely disappeared within 6 h of chase without becoming endoH resistant. In melan-c melanocytes, where the tyrosinase is mutant but the Tyrp1 is wild-type, the conversion of nascent Tyrp1 to the 75 kDa form could be seen, but at a greatly reduced level. After 3 h of chase, nearly half of Tyrp1 remained in the immature form that was sensitive to endoH. However, the degradation rate of Tyrp1 did not significantly differ in either type of cell. Thus, mutations in Tyr or Tyrp1 significantly affect the maturation and degradation of wild-type Tyrp1 or tyrosinase expressed in these cells.
We then examined whether these mutations affect the processing of
melanosomal proteins in general. Tyrosinase, Tyrp1, and Dct share less
than 40% identity with each other at the amino acid level
(48)
. Since tyrosinase and Tyrp1 interact, the question
arose as to whether either or both of them are involved in the
processing of Dct or gp100, two other melanosomal proteins. Thus,
similar immunoprecipitation and endoH sensitivity analyses were
performed for Dct and gp100 (Fig. 4
). In melan-a, melan-b, and melan-c melanocytes, immature wild-type Dct
was expressed as a 66 kDa protein and the mature form was a 75 kDa
protein. The immature form was completely sensitive to endoH and the
half-life of the protein was comparable to that of Tyrp1. Mutations of
tyrosinase (melan-c cells) or Tyrp1 (melan-b cells) had little effect
on the maturation kinetics of Dct. Similarly, the cellular fate of
gp100, which is not sensitive to endoH and has the most rapid turnover
rate, was indistinguishable in the wild-type and mutant melanocytes.
|
Mutations in tyrosinase or Tyrp1 prolong the association of
calnexin with mutant and with wild-type tyrosinase, Tyrp1, and Dct
The above results demonstrated that the maturation of wild-type
tyrosinase or Tyrp1 could be severely affected by a single point
mutation. Extensive studies have revealed that the association of
proteins with ER chaperones directly or indirectly determines their ER
exit and pre-Golgi degradation. Calnexin plays an important role in the
pathogenesis of genetically inherited diseases such as cystic fibrosis,
Wilson’s disease, and von Willebrand disease (26
27
28)
. In
a previous study, we transfected various types of mutant tyrosinase
found in OCA1 patients into COS7 cells in order to examine their
dynamics and interaction with calnexin (31)
. These results
showed that calnexin efficiently trapped all mutant tyrosinases
examined and was involved in their targeted early degradation by
proteasomes. Thus, it was of great interest to investigate whether
calnexin was similarly involved in the processing of wild-type and
mutant tyrosinase (and/or Tyrp1) in melanocytes, where they are
naturally expressed in the context of the melanogenic environment,
rather than in an artificial transfection model that lacks melanosomes.
To determine the physical interactions of calnexin with tyrosinase,
Tyrp1, or Dct in melanocytes, cells were radiolabeled with
[35S]methionine, chased for various periods of
time, and initially immunoprecipitated with a calnexin antibody,
followed by a secondary immunoprecipitation with antibodies to
tyrosinase, Tyrp1, or Dct (Fig. 5
). In melan-a wild-type melanocytes, only barely detectable bands of
tyrosinase and Tyrp1 could be seen to be associated with calnexin, and
only in the initial pulse period. Other minor bands seen bound to
calnexin (other than the cluster at 69 kDa marked with an asterisk) may
be unrelated to melanogenic proteins. Significantly greater amounts of
tyrosinase or Tyrp1 were bound to calnexin, as seen by sequential
immunoprecipitation of extracts of melan-b and melan-c cells, and these
complexes could easily be seen even after 3 h of chase. Although
the rates of conversion and degradation of Dct did not seem to be
affected by mutations in tyrosinase or Tyrp1 (as shown in Fig. 4
above), the association of Dct with calnexin was slightly increased in
melan-b and melan-c cells. No association of gp100 with calnexin could
be detected by this sequential immunoprecipitation approach (data not
shown).
|
Tyrosinase and Tyrp1 form a stable complex
The results shown above support the concept that the maturation of
tyrosinase and Tyrp1 (and to some extent Dct) are tightly linked.
Indeed, it was previously proposed that Tyr, Tyrp1, and Dct form a high
molecular weight complex (7
, 18
, 19)
, although that
conclusion was based on biochemical observations that these proteins
copurified. Thus, it is unclear whether those large molecular weight
complexes might be nonspecific aggregates. Solid evidence for such
physical interactions in vivo has never been reported. Thus, we tested
the possibility that these three proteins form a complex.
We used sequential immunoprecipitation with specific antibodies to
directly investigate the interactions of tyrosinase, Tyrp1, Dct, and
gp100. Melanocytes were radiolabeled with
[35S]methionine for 1 h and labeled
proteins were solubilized in lysis buffer containing 1% Triton X-100.
The samples were first immunoprecipitated with antibodies to tyrosinase
or Tyrp1, the immune complexes were dissociated and subjected to a
second immunoprecipitation with normal IgG or with antibodies to
tyrosinase, Tyrp1, Dct, or gp100 (Fig. 6
). In all three types of melanocytes, tyrosinase and Tyrp1 were
sequentially immunoprecipitated whereas Dct and gp100 were not. When
the immunoprecipitates were solubilized with lysis buffer containing
2% Triton X-100, no bands could be sequentially immunoprecipitated
(data not shown). Thus, mutations in tyrosinase or Tyrp1 protein did
not significantly affect their ability to form a hetero-oligomer.
|
Mutations in tyrosinase or Tyrp1 prolong the association of Bip to
mutant and to wild-type tyrosinase, Tyrp1, and Dct
Since the results described show that the binding of calnexin to
the mutant and/or wild-type protein complex was prolonged by a mutation
in tyrosinase or in Tyrp1, we assume that the delay in maturation of
the wild-type protein and trapping of the mutant protein in the ER can
be attributed to the molecular chaperone function of calnexin. However,
we wanted to examine the possibility that another molecular chaperone
was associated with this delay or trapping. Bip is an obvious candidate
(24
, 49)
, and a recent study has reported that Bip assists
in the folding of Tyrp1 when its S-S bonds are chemically disrupted in
cells cultured with -glucosidase inhibitors or DTT
(37)
. Since the mutant tyrosinase and Tyrp1 expressed in
melan-c and melan-b melanocytes, respectively, naturally lack a single
S-S bond due to the point mutation in each encoding gene, it seemed
feasible that Bip might be involved. We also examined whether the
expression of calnexin and/or Bip was induced by the mutations in the
melanocytes used in this study. As shown in Fig. 7
(top), the expression of Bip was significantly augmented in melanocytes
carrying a mutation in Tyr or Tyrp1 whereas the expression of calnexin
was not significantly different among the three types of melanocytes.
|
To determine whether Bip interacted with tyrosinase, Tyrp1, or Dct,
melanocytes were radiolabeled with
[35S]methionine, chased for specific periods,
and immunoprecipitated with a Bip antibody. The Bip immune complex was
then dissociated and subjected to a second immunoprecipitation with
antibodies (Fig. 7
, bottom). When antibodies to tyrosinase, Tyrp1, or
Dct were used in the second round of immunoprecipitation, no band could
be detected in the wild-type melan-a cells. In contrast, enhanced and
prolonged association of tyrosinase or Tyrp1 with Bip was observed in
melan-b and melan-c cells. Tyrosinase and Tyrp1 remained associated
with Bip for up to 1 h in melan-b cells or for up to 2 h in
melan-c cells. The association of Dct with Bip was barely detectable in
melan-b and melan-c cells, but was not seen in melan-a cells. No
association of gp100 with Bip could be detected (data not shown).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
demonstrates the misrouting
of tyrosinase in OCA2, which suggests that such mechanisms may play a
role in all three known types of OCA.
OCA is an ER retention disease
Tyrosinase is unique in that its folding is predominantly due to
the chaperone function of calnexin (29
, 51)
. Since
calnexin was initially reported to play a critical role in the
maturation of tyrosinase, many groups have shown that calnexin is also
involved in regulating the processing of mutant tyrosinase and that
this is at least partly responsible for depigmentation (30
31
32
, 34
, 52)
. In humans, at least two different types of OCA1, termed
OCA1A (tyrosinase-negative) and OCA1B (‘yellow OCA’), have been
distinguished clinically. In OCA1A, tyrosinase activity is completely
absent and melanin pigment cannot be detected in the skin, hair, or
eyes. Seventy-eight distinct mutations of the TYR gene,
including 48 missense substitution mutations, have been reported in
OCA1A patients (2
, 53)
. In OCA1B, tyrosinase activity and
melanin synthesis are greatly reduced; some melanization may occur
during childhood and adult life, although little or no pigmentation is
evident at birth (2
, 21
, 54
55
56)
. Ten distinct mutations
of the TYR gene, including 9 missense substitutions, have
been reported in OCA1B patients (2
, 53)
. Missense
mutations in OCA1A and OCA1B tend to be located in four clusters in the
TYR gene (1
, 1
, 57
58
59
60
61)
; computer modeling
has suggested that some may alter the folding of tyrosinase and disrupt
binding of the two copper ions necessary for enzyme activity, thereby
destroying its catalytic function (30
, 62
, 63)
.
In contrast to the relatively frequently studied mutations of OCA1,
OCA3 has not been often studied clinically or experimentally. Only two
TYRP1 gene mutations—a truncated mutation S166X and
1104delA—have been described (20
, 21)
. Patients with
OCA3, so far reported only in African-American blacks, are usually born
with minimal pigmentation in the skin, hair, and eyes and have mild
visual problems; they develop increased melanin with age and the
clinical manifestations are not as severe as OCA1 (20
, 21)
. One reason for the rarity of OCA3 cases is probably the
less severe phenotype and the relatively mild clinical manifestations,
which are alleviated to some extent over time. Thus, OCA3 might be more
obvious clinically in darkly pigmented skin, such as African-American
blacks, than in lightly pigmented skin. Considering that Tyrp1 and
tyrosinase have conserved glycosylation, metal binding sites, and
transmembrane domains, we assume there must be many unreported cases of
OCA3 and numerous mutations in the TYRP1 gene, as has been
found for OCA1. Since truncated tyrosinases have maximum deformations
in protein structure and are abnormally trapped in the ER
(31)
, we presume that the truncated TYRP1
mutants described for OCA3 must have a similar fate.
How do missense mutations in the TYRP1 gene affect the
expression and function of its encoded protein and how do they impair
the function of tyrosinase? Like mutations in OCA1, missense mutations
that disrupt copper (or other metal ligand) binding probably elicit
abnormal retention of Tyrp1 because metal binding is critical to
maintain the correct structure of tyrosinase family proteins (30
, 62
, 63)
. To obtain further information about mutations that
affect the pigmentation of OCA patients, mutations in melan-b and
melan-c melanocytes are opportune models not only because they have
known specific mutations in tyrosinase or Tyrp1, but also because they
possess the full complement of enzymes and structural proteins
necessary for the synthesis of melanin. In this study, we showed that a
mutation in one of these melanogenic genes can differentially affect
the maturation of tyrosinase and Tyrp1. In melan-a cells (wild-type
melanocytes), the binding of tyrosinase and Tyrp1 to calnexin was
transient and binding to Bip was not observed. Melanogenic proteins
expressed in these cells were quickly converted to the mature type. On
the other hand, in melan-b (Tyrp1 mutant) and melan-c (tyrosinase
mutant) melanocytes, binding of mutant proteins to calnexin was
prolonged and binding to Bip was detectable. Acquisition of endoH
resistance of melanogenic proteins in melan-b cells was never observed
and tyrosinase and Tyrp1 were degraded rapidly. Transport of Tyrp1 to
the medial Golgi in melan-c cells was also markedly slower. Degradation
of Tyrp1 was not significantly changed. Since the rate of disappearance
of mutant proteins that bound calnexin correlated with the proteins
precipitated by each specific antibody, mutant tyrosinase and Tyrp1
were degraded immediately after their release from calnexin. Although
the double immunoprecipitation results obtained with melan-a cells (as
shown in Fig. 6
) could derive from the fully glycosylated forms of
tyrosinase and Tyrp1 (since the 60 min labeling time would allow for a
major part of the newly synthesized proteins to be processed), the
results obtained with the melan-b and melan-c cells (where none of the
mutant protein was processed correctly with an even longer pulse; see
Fig. 3
) clearly show that these initial interactions occur in the ER.
From these results, we conclude that OCA1 and OCA3 are unique ER
retention diseases because a mutation of one protein affects the
maturation and/or degradation of other melanogenic proteins.
Tyrosinase and Tyrp1 form a melanogenic protein complex
We found that tyrosinase and Tyrp1 bind each other. In a previous
study (8)
, we proposed that Tyrp1 stabilizes tyrosinase in
melanosomes, but we did not anticipate that tyrosinase could also
influence the processing of Tyrp1. Tyrosinase, Tyrp1, and Dct, which
belong to the tyrosinase gene family, have less than 40%
homology in their amino acid sequences (48)
. When
melanocytes are lysed, tyrosinase, Tyrp1, and Dct elute near the void
fraction of gel filtration columns (7
, 19)
. It has not
been reported in what subcellular compartment the physical associations
are made; we demonstrate in this study that tyrosinase and Tyrp1 form
complexes very early and before their processing in the Golgi. Although
direct binding of Dct to tyrosinase or Tyrp1 was not observed by
sequential immunoprecipitation, the retarded association of Dct with
calnexin or Bip when tyrosinase or Tyrp1 is mutant suggests that Dct
does bind to tyrosinase and/or Tyrp1 to some extent, forming an
oligomeric protein complex. The oligomerization that occurs even when
one of the components is mutant suggests that the formation of the
complex occurs in the ER. Indeed, without expression of Tyrp1,
tyrosinase can be properly folded so as to express catalytic activity
(e.g., in transfection experiments). However, such folding is far less
efficient than in the presence of Tyrp1, and the majority of
tyrosinases remain unfolded and are degraded (31)
.
Concerning the role of complex formation, at least two possible
interpretations have been considered. The first is that binding to
other molecules may facilitate folding by stabilizing the folding
intermediates, similar to the role of molecular chaperones. The second
possibility is that formation of oligomeric protein complexes might
reduce the cytotoxicity of melanogenic intermediates against
melanocytes. The pathway of melanin synthesis begins with the
hydroxylation of tyrosine, a step catalyzed by tyrosinase as the
critical and rate-limiting reaction. The DOPAquinone generated is
eventually auto-oxidized and cyclized spontaneously to produce DHI via
DOPAchrome (9
10
11)
. In melanogenesis, DCT functions as
DOPAchrome tautomerase to convert DOPAchrome to the carboxylated
intermediate DHICA, which is less toxic to the melanocyte, and Tyrp1
functions as a DHICA oxidase to oxidize and polymerize the DHICA
monomers into an even less toxic melanin. DHI, which is produced in the
absence of Dct, is significantly more toxic than DHICA
(64)
. When Dct and Tyrp1 are present, the synthesis of DHI
is minimized, and melanin is synthesized primarily through the less
toxic intermediate DHICA. Therefore, Tyrp1 and Dct are thought to be
critical to protect melanocytes from the inherent toxicity of
melanogenic reactions (64)
. When tyrosinase alone is
expressed by transfection in nonmelanocytic cells, these intermediates
harm the transfected cells, resulting in early cell death or retarded
growth of cells expressing the tyrosinase. Some mutations at the
Tyrp1 locus (e.g., Tyrp1lt)
cause the premature death of follicular melanocytes (65)
,
and melan-b cells grow far more slowly than do melan-a and melan-c
cells. Thus, it would be reasonable for melanogenic proteins in
melanocytes to be folded simultaneously and then transported together
in a complex to minimize unfavorable reactions. In such a fashion,
melanocytes might be protecting themselves from the potential toxicity
of melanin intermediates that might be generated by tyrosinase in the
absence of Tyrp1 or Dct.
We noticed that chaperone interactions, ER exit, and degradation of
melanogenic proteins may be differentially regulated. Transport and
degradation of Dct were not influenced by mutation although the
association of Dct with calnexin or Bip was increased; we assume this
reflects weak interactions of Dct with the tyrosinase/Tyrp1 complex, so
that Dct is not found in tyrosinase or Tyrp1 complex isolated from 1%
Triton X-100 lysis buffer. We observed an inconsistency when
degradation of Tyrp1 was not significantly augmented by mutation of
tyrosinase or Tyrp1 per se. Although associations with calnexin and Bip
apparently are prolonged in the case of tyrosinase, degradation was
accelerated by mutations in either protein. We previously showed that
various types of tyrosinase mutants were degraded by proteasomes
(31)
. If Tyrp1 in the mutant cells are degraded similarly,
the resistance to degradation of Tyrp1 may be explained by a recent
finding that disposal of misfolded proteins is directly regulated by
mannose trimming because Tyrp1 may be less susceptible to mannosidase
I. We recently identified an -mannosidase-like molecule, EDEM, that
has no catalytic activity but accelerates degradation of antitrypsin
mutant; we propose that this may be a receptor for the misfolded
protein (66)
.
The observed discrepancy of degradation kinetics in melan-a and in melan-c cells raises an interesting possibility that a subunit in the misfolded complex is dissociated before degradation. The underlying molecular mechanism is currently unknown. In conclusion, the oligomeric melanogenic protein formation is unique in that one mutation influences the maturation (transportation) and/or degradation of a wild-type counterpart in the melanogenic pathway. A mutation in tyrosinase resulted in the abnormal retention and early degradation of mutant tyrosinase and delayed the exit of Tyrp1 from the ER, yet did not affect the degradation of mutant Tyrp1. On the other hand, a mutation in Tyrp1 resulted not only in the abnormal retention and early degradation of mutant Tyrp1, but also elicited the retardation of transport and early degradation of wild-type tyrosinase. In mutant melanocytes, mutations of melanogenic proteins cause loss of regulation in the maturation and transport of melanogenic proteins that consist of oligomeric proteins in various degrees. In addition to the etiology of OCA1 wherein mislocalization of mutant tyrosinase in the ER by molecular chaperones causes a complete loss of pigmentation, the etiology of OCA3 that reduces pigmentation seems to result from a delayed maturation and early degradation of wild-type tyrosinase and the mislocalization of mutant Tyrp1 in the ER.
Received for publication April 16, 2001.
Revision received June 26, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
