|
|
|
|||||||
|
|||||||||
(Received for publication, May 31, 1994; and in revised form, November 2,
1994) From the
Ichthyosis vulgaris is an autosomal dominant disorder of
keratinization characterized by mild hyperkeratosis and reduced or
absent keratohyalin granules in the epidermis. Profilaggrin, a major
component of keratohyalin granules, is reduced or absent from the skin
of individuals with ichthyosis vulgaris. In this report, we have
further characterized the molecular basis of low profilaggrin
expression, which occurs in this disease. In situ hybridization revealed little profilaggrin mRNA in ichthyosis
vulgaris-affected epidermis. In keratinocytes cultured from the
epidermis of affected individuals, the abundance of profilaggrin was
reduced to less than 10% of normal controls, while the mRNA level was
decreased to 30-60% of controls. Expression of K1 and loricrin,
other markers of epidermal differentiation, were not affected. Nuclear
run-on assays indicated that the decrease in mRNA levels was not caused
by aberrant transcription. Nucleotide sequencing of 5`-upstream,
3`-noncoding, and flanking regions of the profilaggrin gene from
ichthyosis vulgaris-affected individuals revealed only minor changes,
probably due to genetic polymorphisms. Our results indicate that
defective profilaggrin expression in ichthyosis vulgaris is a result of
selectively impaired posttranscriptional control. Ichthyosis vulgaris (IV) ( Filaggrin is a cationic protein that aggregates
keratin intermediate filaments in the stratum corneum of the epidermis
(for review, see (4) ). Profilaggrin, the precursor of
filaggrin, is first expressed in the granular layer and marks the
terminal stages of epidermal differentiation. The phosphorylated
profilaggrin accumulates in keratohyalin granules and later undergoes
dephosphorylation and proteolysis to filaggrin. Profilaggrin and
filaggrin are noticeably decreased or absent from the epidermis of
patients with IV(5) . We have previously shown that
keratinocytes cultured from affected individuals maintain structural
and biochemical phenotypic characteristics of the disorder(6) .
For example, very little profilaggrin is detectable by
immunohistochemical staining on Western blots of extracts obtained from
IV keratinocytes compared with controls. These data are consistent with
the absence of keratohyalin in this disorder. This suggests a
pronounced decrease in profilaggrin synthesis and/or accumulation as is
seen in the skin biopsies from affected individuals. Recently, the
structure of the human profilaggrin gene has been reported by two
different laboratories(7, 8) . The gene (see Fig. 3A) contains three exons interrupted by two
introns of 9,713 and 570 bp, respectively. The 5`-noncoding region (75
bp) is divided into two exons separated by the large intron. The coding
region begins in the second exon and continues in the third exon, where
10-12 highly repetitive filaggrin sequences of exactly 972 bp
reside. The number of filaggrin repeats varies between individuals and
is inherited in a Mendelian fashion(9) . The amino terminus of
profilaggrin contains a calcium binding domain consisting of two
EF-hands resembling those present in the S-100 family of
proteins(7) . It has recently been shown that the S-100 domain
of profilaggrin binds calcium(8, 10) . Hence,
profilaggrin may not only function as a keratin aggregating protein,
but it may also play a critical role in the regulation of
calcium-dependent events during epidermal differentiation(7) .
Figure 3:
Expression of profilaggrin mRNA in
IV-affected individuals and controls. A, diagram demonstrating
the regions covered by the probes (2, exon 2; 3, exon
3) used in hybridization are indicated. B, Northern analysis
of total RNA from the three IV-affected individuals and unrelated
controls studied in Fig. 2. The amount of RNA in each lane was
normalized based on the glyceraldehyde-3-phosphate dehydrogenase level.
RNA was fractionated on a 1% glyoxal gel, blotted, and hybridized with
a repeat of the third exon of the profilaggrin gene (proFG)
and glyceraldehyde-3-phosphate dehydrogenase. The same blot was also
hybridized to keratin 1 (K1) (3`-end), human loricrin
(3`-end), and HS26 probes. Much lower hybridization signal
from the loricrin probe in the middle pair is unexplained; all RNAs
were on the same blot. It should also be noted that the intensity of
the hybridization signal generated from each probe does not reflect the
relative abundance of the RNAs as the specific activities of the probes
were not identical.
Figure 2:
SDS-PAGE and immunoblotting analyses of
profilaggrin in IV-affected individuals and controls. Equal protein
loadings of extracts from IV-affected keratinocytes cultured from three
unrelated individuals (IV)and normal controls (C)
were analyzed by PAGE. Controls in the first two pairs were unaffected
offspring, while the control in the third pair was an unrelated
unaffected age- and sex-matched individual. Lane F contains
foreskin epidermal extract. The proteins were blotted onto
nitrocellulose paper and reacted with antibodies directed against human
profilaggrin (proFG) and filaggrin (FG) (a),
neutral-basic keratins (K1, K5, and K6) (b), and loricrin (LORI) (c). Identity of
the 30 kDa loricrin band was confirmed in parallel studies with
antiserum generously supplied by Dennis Roop (Houston, TX) (data not
shown).
To investigate further the association between decreased
profilaggrin expression and IV, we studied profilaggrin mRNA levels in vivo by in situ hybridization and utilized a human
epidermal keratinocyte culture system as an in vitro model. We
analyzed profilaggrin expression at the protein, steady-state mRNA, and
transcriptional levels in keratinocytes cultured from individuals with
IV as well as from appropriate, unaffected family members and age- and
sex-matched normal controls. Our data indicate that selectively
impaired posttranscriptional regulation results in reduced profilaggrin
mRNA and protein in IV.
2) The profilaggrin 5` upstream region (885 bp) was
generated from human genomic DNA using the following
oligonucleotides 3) The profilaggrin-3`-noncoding
region (517 bp) was generated from human genomic DNA using an upstream
oligonucleotide that included the stop codon (underlined) of human
profilaggrin gene(9) : 5`-GATACTATTACTATGAGTAAGA-3`
and 5`-GACATCTAATTCTGGCCATGG-3`. Amplification using PCR was performed
using Taq polymerase (Promega Corp.). Conditions for PCR were
94 °C for 1 min, 42 °C for 1 min, and 72 °C for 1 min, for
30 cycles. 4) The K1 cDNA clone (a gift of Dr. D. Roop, Baylor
College of Medicine, Houston, TX) is a subclone of pK456 (110 bp) and
contains the BamHI-PstI fragment of pK456, which
encodes the carboxyl-terminal end domain and 3`-noncoding region of the
human K1 gene(19) . 5) The loricrin-specific probe (120 bp)
was prepared based on the 3`-noncoding sequence of human loricrin (20) using the following oligonucleotides:
5`-GTACCACGGAGGCGAAGGAGT-3` and 5`-GGTTGGGAGGTAGTTGTACAG-3`. Conditions
used for PCR were 94 °C for 1 min, 59 °C for 1 min, and 72
°C for 1 min, for 30 cycles. 6) The glyceraldehyde-3-phosphate
dehydrogenase, pHcGAP plasmid was obtained from ATCC. 7) The cDNA
for human S26 ribosomal protein (HS26) was a gift of Dr. P. Fort,
Montpellier Cedex 2, France. 8) Genomic DNA was prepared from
cultured fibroblasts of normal and IV-affected individuals as described
previously(21) .
Figure 1:
In situ hybridization of skin
from IV-affected individuals and controls. Biopsies of unaffected (C) and affected (IV) individuals were hybridized
with
The basis of
comparison relied on the use of glyceraldehyde-3-phosphate
dehydrogenase to normalize RNA on the blots. However, recent evidence
suggests that glyceraldehyde-3-phosphate dehydrogenase expression may
vary as a function of proliferative state and physiological conditions
within the cell(25, 26, 27) . Keratinocytes
used in this study were at postconfluence, when proliferation is
reduced(28) . Our assumption was that the rate of proliferation
of normal and IV-affected keratinocytes is similar. However, it is
possible that development, differentiation, and/or metabolic activity
of keratinocytes from IV-affected individuals differ from normal.
Hence, regulation of glyceraldehyde-3-phosphate dehydrogenase may be
altered, and the RNA profile may be inaccurate. Therefore the
same blots were hybridized with a second probe made from a cDNA for
HS26 ribosomal protein RNA, which shows invariant expression among
several cell lines(29) . The phosphorimaging analysis indicated
that the levels of profilaggrin mRNA normalized to HS26 were 24, 67,
and 40% of the corresponding controls. The result is similar to that
obtained using glyceraldehyde-3-phosphate dehydrogenase as a standard
for normalization, indicating profilaggrin mRNA levels in IV
keratinocytes are about 45% of normal controls.
Keratinocytes
cultured from individuals with IV and unaffected controls were
harvested 3 days after reaching confluence. Harvesting was done 24 h
after feeding, which increases profilaggrin mRNA transcription. (
Figure 4:
Nuclear run-on transcription analysis of
the profilaggrin gene in confluent keratinocytes cultured from
IV-affected individuals and normal controls. Nuclei were isolated from
control (C) and IV-affected keratinocytes (IV), and
run-on transcription assays were performed as described under
``Materials and Methods.'' The pair on the farright is from two of the individuals studied in Fig. 2. The two pairs on the left are not, but
expression of profilaggrin mRNA and protein was identical to that shown
in Fig. 2and Fig. 3. Equal counts/min of
The 3`-noncoding region of the profilaggrin gene was amplified from
three individuals with IV, two of whom were members of the same family.
The amplified 564-bp genomic fragment contained a sequence from 10 bp
upstream of the stop codon to 47 bp downstream of the poly(A) addition
site. Sequence analysis showed a few variations from that of normal
unaffected individuals or the published sequence(9) . The stop
codon and poly(A) addition site were intact and located as previously
reported. No alterations that might increase the instability of the RNA
were identified. We believe that these variations are due to genetic
polymorphism and do not contribute to the marked decrease in
profilaggrin mRNA seen in IV. We have previously demonstrated that keratinocytes cultured
from patients with IV exhibited several characteristics of the disorder
including decreased expression of profilaggrin protein(6) . In
the present paper, we show that profilaggrin mRNA levels are reduced in
IV epidermis and in keratinocytes cultured from IV-affected individuals
and that the abnormal profilaggrin expression is a specific defect and
does not reflect a more general defect in expression of markers of
epidermal differentiation. Our results also suggest that the defect
occurs primarily at the posttranscriptional level. The supporting
evidence for this conclusion is that the transcription level of the
profilaggrin gene between keratinocytes cultured from IV-affected
individuals and controls (as determined by run-on assays) is similar,
irrespective of the levels of corresponding steady-state cellular RNA. Decreased profilaggrin in IV is a result of deficient profilaggrin
mRNA, which in turn indicates either aberrant gene transcription or
posttranscriptional regulation. Northern analysis using a single
filaggrin repeat as probe consistently showed that total RNA isolated
from confluent IV keratinocytes contained less profilaggrin mRNA than
their normal counterparts (Fig. 3B). Our in vitro study of profilaggrin and loricrin expression agrees with the in situ hybridization (Fig. 1) and antibody staining of
skin biopsies (5, 32) . Noticeably, the reduction of
profilaggrin is more prominent at the protein than the mRNA level, with
keratinocytes of IV-affected individuals showing a 30-60%
reduction in steady-state mRNA and more than 90% reduction in protein ( Fig. 2and Fig. 3). This may reflect multiple subtle
controls of profilaggrin gene expression during terminal
differentiation. Nuclear run-on assays have been used to demonstrate
elongation of preinitiated RNA in various systems. Profilaggrin gene
transcription as measured by run-on analysis was normal in
keratinocytes from IV-affected individuals, while accumulation of the
mRNA was reduced. There is also no evidence to suggest that
transcriptional pausing or premature termination is relevant in the
control of the profilaggrin gene in IV. The results suggest that the
steady-state level of profilaggrin mRNA is not solely regulated at the
level of mRNA synthesis, but additional posttranscription controls
exist as well. This regulation appears to be specific to profilaggrin
mRNA because the other mRNAs studied (K1, loricrin,
glyceraldehyde-3-phosphate dehydrogenase, and HS26 ribosomal protein)
were similar in keratinocytes from IV-affected individuals and normal
controls. Our study eliminates the possibility that the defective
profilaggrin expression observed in IV occurs at the level of
transcription. The results, however, do not preclude the possibility
that IV results from mutation(s) in the profilaggrin gene. Mutations at
the cap site, splice junction, initiation, or termination codon and
coding region that result in low mRNA level and phenotypic aberrations
have been documented in a number of human genetic
diseases(33) . Results from sequence analysis of the 5`-, and
3`-noncoding, and flanking sequences suggest that these regions of the
profilaggrin gene from IV-affected individuals are similar to normal.
However, we have not sequenced the intron/exon boundaries, the
extensive first intron, the filaggrin repeats or the sequence
representing the unique leader or tail peptides. Nonsense or frameshift
mutations in a variety of human genes are associated with significant
reduction in the steady-state level of mRNAs. In numerous cases, the
cytoplasmic rather than nuclear mRNAs appear to be
unstable(34) . The mutation could also be allele-specific. It
is possible that IV-affected individuals have one mutant profilaggrin
allele in which a deletion or point mutation resides. However since IV
is an autosomal dominant disease, if this were the case it would also
mean that in the heterozygous condition the mutant allele would somehow
inhibit the expression of its normal counterpart. This situation has
been suggested in human diabetes insipidus (35) and type I
angioneurotic edema(36) , which are also autosomal dominant
diseases. Posttranscriptional control regulates a number of
eukaryotic mRNAs and may take place at nuclear RNA processing/turnover,
nucleocytoplasmic transport, cytoplasmic mRNA turnover, or at
translational efficiency (for reviews, see (37) and (38) ). Analysis of available noncoding sequences of vertebrate
genes demonstrates the presence of many highly conserved sequences in
5` and 3` noncoding regions, suggesting a possible involvement of such
sequences in posttranscriptional control (39) . To date, most
studies suggest the 3`-noncoding region contains the primary
determinants of mRNA instability. Rapid turnover of histone RNA is
mediated by a sequence capable of forming a short stem-loop structure
at the extreme 3`-end of the nonpolyadenylated histone mRNAs (40) . In the cases of lymphokine (i.e. granulocyte
macrophage colony stimulating factor), cytokine (i.e. human
interferon- The 3`-noncoding region of the human profilaggrin gene
contains a single AUUUUUA and two AUUUUA sequences, similar to the
consensus destabilizing AU pentamer (AUUUA). These sequences are
present in the genes from both normal and IV-affected cells. Therefore,
while this sequence may regulate profilaggrin mRNA degradation, the
sequence alone is unlikely to be the cause of lower profilaggrin mRNA
levels in IV. Recent studies identified at least three distinct
cellular AU-binding factors that may mediate the pathway of RNA
stability in other systems (49, 50) . Perhaps, a
similar factor(s) functions as a trans-destabilizing element
that specifically targets profilaggrin mRNA for increased degradation
in keratinocytes from IV-affected individuals. Alternatively, a
cellular factor(s) that selectively recognizes and stabilizes the
profilaggrin mRNA in trans may be defective in keratinocytes
from IV-affected individuals. The reduced profilaggrin mRNA level in
keratinocytes from IV-affected individuals may also be a result of
defective regulation through protein kinase C. Keratinocytes express a
variety of PKC isoforms, and PKC induces profilaggrin and loricrin
mRNAs in mouse keratinocytes (51) . Additionally, changes in
the expression of different PKC isoforms are observed during
keratinocyte differentiation (52) and in psoriasis (52, 53) . nPKC- The fact that profilaggrin probably functions as a calcium binding
protein in vivo suggests a close relationship with the known
calcium dependence of profilaggrin expression as well as epidermal
differentiation in
vitro(4, 56, 57) . Perhaps the putative
calcium binding domain of profilaggrin is involved directly with the
regulation of steady-state profilaggrin mRNA or translation. The
calcium binding domain of profilaggrin may also mediate autoregulation
at the posttranscriptional level similar to The decrease or absence of profilaggrin in
keratinocytes from IV-affected individuals would not only result in the
absence of profilaggrin as a keratin aggregating protein but also as a
calcium binding domain. The abnormal keratohyalin granules in IV
suggests that intracellular calcium may not be sequestered in
keratohyalin granules as is likely in normal cells. This could provoke
defective calcium-dependent regulation of terminal differentiation
resulting in the clinical phenotype. Alternatively, the clinical
findings in IV may relate to the proposed water binding properties of
free amino acids catabolized from filaggrin in the upper stratum
corneum(60) . Decreased water binding capacity may result in
abnormal desquamation and clinical scaling. Our current results
provide new molecular evidence that selectively impaired
posttranscriptional control intrinsic to keratinocytes from IV-affected
individuals is largely responsible for the reduction of the
profilaggrin mRNA and protein in the disorder. The finding of such a
molecular defect contributes to our understanding of the disorder as
well as gene regulation during normal epidermal differentiation.
Nonetheless, whether the low level of profilaggrin expression is a
cause or a result of ichthyosis vulgaris remains in question and serves
as the basis for further investigation. Autoradiographic analysis in this study was carried out by the
Phosphorimager Facility of the Markey Molecular Medicine Center at the
University of Washington. The biopsies were conducted at the
Clinical Research Center facility of the University of Washington
supported by the National Institutes of Health National Center for
Research Resources Grant M01RR00037.
Volume 270,
Number 2,
Issue of January 13, 1995 pp. 871-876
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
)is an autosomal dominant
skin disorder reported to occur in as many as 1 in 250 of the normal
population(1) . Affected skin appears scaly and is
characterized histologically by hyperkeratosis and a decreased or
absent granular layer(2) . In addition, keratohyalin granules,
an ultrastructural landmark in the granular layer, are either absent or
reduced and structurally abnormal(3) . Other clinical symptoms,
including hyperlinear palms and soles, a personal or family history of
atopy, and keratosis pilaris are often associated with IV(1) .
Although the disease is well described clinically, the etiology is
poorly understood.
Diagnosis of Ichthyosis Vulgaris
Probands were
identified from patients seen in the University of Washington Medical
Center Dermatology Clinics and from individuals referred by clinical
dermatologists from the community. Subjects who met published clinical
criteria for IV (1, 11) were biopsied from the
extensor surface of the arm after obtaining informed consent. Biopsies
were fixed in methyl Carnoy's for light microscopy and
immunocytochemistry and fixed in half-strength Karnovsky's for
electron microscopy as described previously(12) . Individuals
with the clinical criteria for IV who had one or fewer layers of
granular cells in hematoxylin and eosin-stained sections(1) ,
who had absent or attenuated staining with the anti-profilaggrin
antibody AKH1 with normal staining with the anti-keratohyalin antibody
AKH2 (13) and who had no keratohyalin granules when examined by
electron microscopy (3, 5) were considered to be
affected. In this report, results from affected, unrelated individuals
from three different families are presented. Related, unaffected family
members served as controls for two cases, and an unrelated, age- and
sex-matched subjects for the other.In Situ Hybridization
Biopsies were obtained from
the extensor surface of the arm and snap frozen in Tissue-Tek OCT (Miles, Inc.) embedding medium. Ten-micron frozen sections
were fixed briefly in paraformaldehyde and processed by standard
techniques (14) with the following modifications; proteinase K
digestion was omitted, an initial 1-h wash at room temperature with 4
SSC (1 SSC: 0.15 M sodium chloride, 0.015 M sodium citrate, pH 7.0) was added after hybridization, and
10 mM dithiothreitol was added to all rinses. Hybridization
was with a S-labeled riboprobe generated from a human
filaggrin repeat cloned into pGEM-1 (Promega Corp.). The plasmid was
linearized with EcoRI, and antisense riboprobe was generated
with the use of T7 RNA polymerase in the presence of
[
-S]UTP (1,000 Ci/mmol) (DuPont NEN).
Cell Culture
Adult human keratinocytes were
obtained from blister biopsies as described previously(6) .
Keratinocytes were cultured on mitomycin C-treated 3T3 cells and
maintained in a humidified 5% CO
atmosphere at 36.9 °C
in Dulbecco's modified Eagle's medium (Life Technologies)
containing 20% fetal calf serum, hydrocortisone, cholera toxin, and
epidermal growth factor. Cells were fed 3 times a week and 24 h before
harvesting. To ensure profilaggrin expression, cells were harvested 3
days after cells reached confluence.Protein Extraction, SDS-PAGE, and Western
Blotting
Total urea-Tris-soluble protein extracts (15) were obtained from keratinocytes cultured in parallel with
those used for the isolation of RNA and nuclei. Equal protein loadings
were separated on discontinuous 7.5-15% SDS-PAGE gels, and
proteins were blotted to nitrocellulose (Schleicher and Schuell).
Profilaggrin was detected with polyclonal anti-human
profilaggrin/filaggrin antiserum(16) , K1 with monoclonal
antibody AE3 (a generous gift of T.-T. Sun, New York University Medical
Center, (17) ), and loricrin with antibody raised against a
synthetic peptide corresponding to the 14 most carboxyl-terminal
residues of mouse loricrin known to cross-react with human
loricrin(18) .Plasmids, Probes, PCR Primers, and Genomic DNA
1)
The profilaggrin coding probe was a 972-bp filaggrin repeat of exon
3(7) . ((8) , GenBank M96943):
5`-TGGTAGGAGGCACAATGT-3` and 5`-GAGCCTGCTGGGTACTGA-3`. Amplification
using PCR was performed using Taq polymerase (Promega Corp.).
Conditions for PCR were 94 °C for 1 min, 50 °C for 1 min, and
72 °C for 1 min, for 30 cycles.Northern Analysis
Total cellular RNA was prepared
from cultured keratinocytes using the guanidine thiocyanate-acid phenol
method(22) . Equal amounts of total RNA (10 µg) were
separated on 1% glyoxal gels, blotted onto GeneScreen Plus membranes
(DuPont NEN) and hybridized to a nick-translated or random primed probe
overnight at 60 °C in the buffer recommended by the manufacturer
except that 200 µg/ml sonicated denatured salmon sperm DNA was
included. Filters were washed 3 times in 2 SSC, 0.1% SDS,
followed by 3 times in 0.1% SDS/0.1 SSC at 65 °C. All
Northern blots were reprobed with glyceraldehyde-3-phosphate
dehydrogenase and HS26 cDNA to ascertain equal RNA loadings. The
autoradiographs were scanned using the PhosphorImager (Molecular
Dynamics, Sunnyvale, CA) and quantified using image analysis software
(ImageQuant Version 3.22, Molecular Dynamics). All comparison between
hybridization signals was based on equal glyceraldehyde-3-phosphate
dehydrogenase or HS26 signal.Isolation of Nuclei and Nuclear Run-on
Assays
First or second passage human keratinocytes were cultured
to 3 days post-confluence. Nuclei were isolated by a modification of
the method of Greenberg and Bender(23) . Keratinocytes were
harvested from five to six 100-mm plates, resuspended in 1% Nonidet
P-40 lysis buffer, vortexed briefly, and severed with sterile scissors.
Cells were broken using 40 strokes of a Dounce homogenizer, and nuclei
were isolated, snap frozen in liquid nitrogen, and stored at -70
°C. Nuclei were incubated with run-on reaction buffer in the
presence of 250 µCi of [-P]GTP (3,000
Ci/mmol) for 30 min. Labeled RNA was isolated and purified through a
SuperSelect column (5`
3`, Boulder, CO). Equivalent
P-labeled RNA (as determined by trichloroacetic acid
precipitation) was hybridized to plasmid DNA that had been linearized,
denatured, and immobilized onto nitrocellulose using a slot blotting
apparatus (Bio-Rad). Hybridization was carried out in 1 ml of TES
buffer with 100 mg/ml denatured sonicated salmon sperm DNA for
48-72 h at 65 °C, followed by two washes in 2
SSC,
0.1% SDS and two washes in 0.1 SSC, 0.1% SDS at 65 °C for
30 min. Exposure to x-ray film was for 3-7 days. Nuclear run-on
was also carried out in the presence of 0.5% sarkosyl (24) in
order to distinguish between polymerase initiation and elongation.
-Amanitin (2 µg/ml) was used to show that in vitro transcription was due to RNA polymerase II activity(24) .
In Situ Analysis of Profilaggrin mRNA in IV and Normal
Epidermis
It has been shown previously by immunohistochemical
techniques using an anti-human profilaggrin antibody that
immunoreactive profilaggrin and filaggrin are markedly decreased in the
epidermis of IV-affected individuals compared with normal
controls(5) . To determine if the decrease is due to a
reduction in profilaggrin mRNA, in situ hybridization using a S-labeled riboprobe transcribed from a human filaggrin
repeat was performed. A dramatic decrease in silver grains in the
granular layer of affected IV epidermis compared with normal skin was
seen (Fig. 1) showing reduced levels of profilaggrin mRNA in the
epidermis of IV-affected individuals and confirming studies at the
protein level (5) . No significant labeling of the epidermis
was seen with the sense strand riboprobe (data not shown).
S-labeled antisense filaggrin riboprobe. Reduced
silver grains were viewed by light microscopy. B, basal layer; S, spinous layer; G, granular layer; C,
cornified layer; Bar = 30
µm.
Decreased Profilaggrin Expression in IV
Keratinocytes
Western analysis of profilaggrin, K1, and loricrin
in protein extracts from keratinocytes obtained from IV-affected
individuals compared with normal controls is demonstrated in Fig. 2. It should be noted that cultured human keratinocytes
express profilaggrin at confluence but do not process it to filaggrin
as normally occurs during terminal differentiation in
vivo(6) . It typically appears as a broad smear on SDS
gels. The anti-human profilaggrin antibody detected very little
profilaggrin in extracts from affected individuals, in agreement with
previous findings(6) . We visually estimated the amount of
immunoreactive protein in keratinocyte extracts from affected
individuals to be less than 10% of normal controls. This result
paralleled the loss or reduction of keratohyalin granules observed in
the corresponding skin biopsies (data not shown). In contrast, the
expression of neither loricrin, another marker of the late stages of
epidermal differentiation also localized in keratohyalin granules, nor
K1, a marker of suprabasal differentiation, were reduced (Fig. 2). The results strongly suggest that the defect in
keratinocytes obtained from IV-affected individuals is specific to
profilaggrin.Steady State Level of the Profilaggrin mRNA in IV
Keratinocytes
To gain insight into the etiology of the disease,
profilaggrin mRNA levels in keratinocytes from affected individuals and
controls was determined. Northern analysis of total RNA harvested from
confluent keratinocytes of three unrelated IV subjects and controls was
conducted with a probe from a filaggrin repeat within exon 3 of the
profilaggrin gene (Fig. 3A). The level of profilaggrin
mRNA in affected cells was markedly reduced compared with the normal
counterparts (Fig. 3B). Similar results were obtained
when exon 2, which contains part of the EF-hand domain, was used as a
probe (data not shown). The intensity of each hybridization signal was
quantified using the PhosphorImager and the level of mRNA normalized to
the internal glyceraldehyde-3-phosphate dehydrogenase control. The
levels of profilaggrin mRNA were 30, 61, and 43% of their corresponding
controls (from left to rightpanel). When
loricrin- and K1-specific probes were used on the same blots, no
differences in RNA levels were seen between normal and affected (Fig. 3B). These results were consistent with those of
the Western analysis, although compared with the normal counterparts,
the levels of immunoreactive profilaggrin protein were much lower in IV
keratinocytes than the levels of profilaggrin mRNAs.Transcription of Profilaggrin mRNA in IV and Normal
Keratinocytes
Comparison of profilaggrin mRNA levels using
Northern blots determines the steady state level of RNA but does not
differentiate between decreased RNA levels resulting from defective
transcriptional or posttranscriptional events (see
``Discussion''). To delineate between the two, profilaggrin
transcription was examined by nuclear run-on assay.
)Nuclei and total RNA were then isolated. Hybridization of
the nascently transcribed labeled RNA to a human filaggrin repeat, K1,
and glyceraldehyde-3-phosphate dehydrogenase fragments revealed
virtually identical levels of transcription between keratinocytes from
normal and IV-affected individuals (Fig. 4). In contrast,
experiments using nuclei from subconfluent keratinocytes gave very low
signal for profilaggrin transcription (data not shown). Because the
assay utilized a probe of filaggrin repeat from exon 3, the result also
suggested that transcriptional pausing in either the first or second
exon was not a cause of reduced profilaggrin mRNA in IV-affected
individuals. In order to test more rigorously for possible
transcriptional pausing in profilaggrin expression, sarkosyl which
overcomes pausing (24) was added. It had no effect on signal
intensity in either IV-affected or normal individuals (data not shown),
suggesting that transcriptional pausing was not a factor.
P-labeled RNA were hybridized to filters loaded with
linearized plasmids (10 mg each) containing inserts of the following:
3` terminus of K1 cDNA (K1), cDNA of a filaggrin repeat (FG), and glyceraldehyde-3-phosphate dehydrogenase. Linearized
plasmid pGEM-3 was used for an estimation of the nonspecific
hybridization signal.
Analysis of 5`-Flanking Region from IV-affected
Individuals
The sequence of the 5`-upstream flanking DNA of
human profilaggrin from -1 to -869 bp obtained from three
IV-affected individuals was also determined to identify any promoter
mutations that might account for the lower profilaggrin mRNA
expression. Very few differences in nucleotide sequence were apparent
(data not shown); no changes were specific to IV. Such identified
changes could be due to allelic polymorphism or errors due to PCR. No
changes were seen in the immediate upstream region containing putative
regulatory motifs such as the TATA box, SP1, or AP1
sites(7, 8) .Analysis of 3`-Noncoding Region
A number of genes
whose expression is reduced under various conditions contain AU-rich
sequences that act as destabilizing signals in the 3`-noncoding region
of the mRNA(30, 31) . AU-rich sequences are found in
the 3`-noncoding region of the human profilaggrin gene(9) .
Therefore, the 3`-noncoding and flanking sequences of the profilaggrin
gene in IV-affected individuals were investigated for possible
alterations that could cause decreased profilaggrin mRNA stability.
), and proto-oncogene (i.e. c-myc), a
variable number of AU-rich domains of the 3`-noncoding region confer
mRNA instability (for review, see (37) and (41) ).
This domain may interact with the poly(A)
tail(31, 42) , a 32-kDa polypeptide(43) , or a
complex of proteins (37, 43) prior to undergoing rapid
mRNA degradation. For granulocyte macrophage colony stimulating factor
mRNA, the 3`-noncoding region contains two regions that are responsible
for mRNA stabilization mediated by calcium(44) . The main
region maps to the AU-rich sequences. A second region located upstream
of the AU-rich domain contributes to the overall calcium response.
Active ongoing translation also plays an important role in
posttranscriptional control (45, 46) . Regulation of
mRNA stability may also be mediated by specific sequences other than
the AU pentamer. The paradigm is the presence of specific recognition
sites that may play dual functions for endonucleolytic cleavage and
protection of mRNA in Xenopus and Drosophila(47) . In addition, a repeated
pyrimidine-rich motif present in the 3`-noncoding region of
15-lipoxygenase mRNA may play critical role in translational regulation
during red blood cell differentiation(48) . These studies
clearly indicate a complex array of interaction between cis- and trans-acting factors in posttranscriptional
regulation. is the predominant isoform
expressed in epidermis and has been localized to the granular layer of
human epidermis(54) . The isoform may be involved in a
regulatory pathway controlling profilaggrin mRNA stability. In a study
on tumor necrosis factor mRNA decay, inhibition of PKC affected an
early step in the process of mRNA degradation by increasing the rate of
poly(A) removal(55) . Similar events may occur in IV; failure
of an element in the PKC pathway may lead to increased RNA instability.
-tubulin (58) or at the transcriptional level as seen in the collagen
type I gene(59) .
)))
We are grateful to Paul V. Haydock for his early
contribution to the work and the riboprobes. We thank Barbara Hager for
her excellent technical assistance in tissue culture and PAGE, and Mary
Hoff for her expertise in in situ hybridization. T.-T. Sun and
Dennis Roop generously provided AE3 and loricrin antibodies,
respectively. We also thank Phillipe Fort for the cDNA for HS26.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  A. Sandilands, C. Sutherland, A. D. Irvine, and W. H. I. McLean Filaggrin in the frontline: role in skin barrier function and disease J. Cell Sci., May 1, 2009; 122(9): 1285 - 1294. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Takaishi, T. Makino, M. Morohashi, and N.-h. Huh Identification of Human Hornerin and Its Expression in Regenerating and Psoriatic Skin J. Biol. Chem., February 11, 2005; 280(6): 4696 - 4703. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J A Tlacuilo-Parra, E Guevara-Gutierrez, and M Salazar-Paramo Acquired ichthyosis associated with systemic lupus erythematosus Lupus, April 1, 2004; 13(4): 270 - 273. [Abstract] [PDF]
|
|
|
|
 |
|
 T. Makino, M. Takaishi, M. Toyoda, M. Morohashi, and N.-h. Huh Expression of Hornerin in Stratified Squamous Epithelium in the Mouse: A Comparative Analysis with Profilaggrin J. Histochem. Cytochem., April 1, 2003; 51(4): 485 - 492. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Makino, M. Takaishi, M. Morohashi, and N.-h. Huh Hornerin, a Novel Profilaggrin-like Protein and Differentiation-specific Marker Isolated from Mouse Skin J. Biol. Chem., December 7, 2001; 276(50): 47445 - 47452. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Marenholz, M. Zirra, D. F. Fischer, C. Backendorf, A. Ziegler, and D. Mischke Identification of Human Epidermal Differentiation Complex (EDC)-Encoded Genes by Subtractive Hybridization of Entire YACs to a Gridded Keratinocyte cDNA Library Genome Res., March 1, 2001; 11(3): 341 - 355. [Abstract] [Full Text]
|
|
|
|
 |
|
 R. B. Presland and B. A. Dale Epithelial Structural Proteins of the Skin and Oral Cavity: Function in Health and Disease Critical Reviews in Oral Biology & Medicine, January 1, 2000; 11(4): 383 - 408. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |