ARTICLE
Auteur(s) : Andrea Math*1,2, Jorge Frank4,5, Alessandra Handisurya1,
Pamela Poblete-Gutiérrez4,5, Katharina
Slupetzky1, Dagmar Födinger3, Dorian
Winter1, Georg Stingl1, Reinhard
Kirnbauer1,*
1Division of Immunology, Allergy and Infectious
Diseases (DIAID), Medical University of Vienna, Waehringer Guertel
18-20, A-1090 Vienna, Austria
2Division of Special and Environmental Dermatology
Medical University of Vienna, Vienna, Austria
3Division of General Dermatology, Department of
Dermatology, Medical University of Vienna, Vienna, Austria
4Department of Dermatology, University Hospital
Maastricht, Maastricht, The Netherlands
5Maasticht University Center for Molecular Dermatology
(MUCMD)
accepté le 24 Mai 2006
Epidermolytic hyperkeratosis (EHK) (OMIM 113800) is a generalized,
hyperkeratotic, occasionally blistering skin disease variably
involving the palms and soles [1, 2]. At birth, patients present
with erythroderma and blistering. Therefore this disease is also
referred to as “bullous congenital ichthyosiform erythroderma
(BCIE)”. Over time, generalized hyperkeratoses develop, which are
predominantly localized over the large joints, and blistering may
occur in mechanically stressed regions. The clinical manifestations
vary with regard to the extent of hyperkeratosis, the tendency to
form blisters, the presence of erythroderma, and the involvement of
palms and soles [1].EHK is inherited in a mostly autosomal dominant
fashion with apparently full penetrance, although about 50% of the
cases occur sporadically due to spontaneous mutation [3]. This
disease is caused by mutations in the keratin 1 (K1) or keratin 10
(K10) genes, which are expressed in the suprabasal layers of the
epidermis [2-6]. K1 and K10 gene mutations disrupt intermediate
filament assembly, which causes mechanical instability and, thus,
cytolysis of suprabasal keratinocytes with epidermal blistering.
Most of the missense mutations in keratin genes identified in EHK
patients reside within the highly conserved helix initiation and
termination motifs, the H1 domain and occasionally in the L12
linker region [4, 6-8].Here we report on the genetic basis of EHK
in a 24-year-old man from Kazakhstan suffering from a generalized
variant of the disease with involvement of palms and soles.
Although his parents and other family members were unavailable for
clinical examination in Austria, they were reported to be
unaffected, including the absence of epidermal nevi, and blood
samples from the parents, for isolation of lymphocytes and genetic
analysis, was obtained by mail.Histology, immunohistochemistry, and
electron microscopy revealed the typical features of EHK. Despite
the negative family history, the characteristic clinical and
ultrastructural findings led to the presumed diagnosis of EHK that
arose due to a spontaneous mutation in either the K1 or K10 gene.
Since EHK with involvement of palms and soles has been linked to
mutations in K1 rather than K10, we first examined the former gene
for the putative underlying molecular defect [1].
Materials and methods
Patient and Samples: Ethylenediamine tetraacetic acid (EDTA)
anticoagulated blood samples were drawn from the index patient and
his non-consanguinous parents, after informed consent for inclusion
in the study, in accordance with guidelines set forth by the local
institutional review board. For control purposes, blood samples
from 75 healthy, unrelated Caucasian control individuals, were
collected. Skin biopsies were taken from the patient and a healthy
control individual.
DNA extraction, sequencing, mutation detection and confirmation:
Genomic DNA was extracted from whole blood of the patient, his
non-consanguinous parents, and 75 controls. The coding regions and
adjacent splice sites of the K1 gene were further amplified by
polymerase chain reaction (PCR), and resulting amplimers were
automatically sequenced using dye-labeled primers on a 4200L
dual-dye DNA Analyser (LICOR Biosciences).
For mutation verification, restriction enzyme analysis of the
PCR product containing exon 1 of the K1 gene was performed with
Tsp45I (New England Biolabs; recognition site GTSAC).
mRNA analysis: mRNA was extracted from skin biopsies using an
mRNA isolation kit and cDNA reverse transcribed by a first strand
synthesis kit according to the manufacturer’s instructions (all
Roche). Semi-quantitative PCR was performed with sense primer 5′-
GGAGAATGTGCCCCG-3′ (HK3, exon 8) and anti-sense primer
5′-GCCATAGCTGCCACG-3′ (HK4, exon 9) [9]. To normalize for mRNA
levels, PCR with primers specific for beta-actin was used as a
control. Amplimers were analysed on a 1% agarose gel stained with
ethidium bromide.
Light and electron microscopy: Full-skin biopsies of the patient
were formalin-fixed and paraffin-embedded, and tissue sections were
stained with hematoxylin/eosin. For electron microscopy, tissue was
fixed overnight in 2% paraformaldehyde/2.5% glutaraldehyde, and
rinsed in 0.1 M cacodylate buffer. After treatment with 2% osmium
tetroxide and uranyl acetate, dehydrated samples were embedded in
EPON 812. Ultrathin sections were collected on copper grids and
stained with uranylacetate/lead citrate. Finally, sections were
examined and micrographed using a JEOL 1200 EX electron microscope
(JEOL).
Immunohistochemistry: Immunohistochemistry was performed on
paraffin-embedded tissue sections with mouse monoclonal antibodies
directed against K1 (AE1, PROGEN) or K10 (CK10, BioGenex) using an
immunoperoxidase system.
Results
A 24-year-old male patient, who originated from Kazakhstan,
presented with generalized hyperkeratosis with involvement of the
palms and soles. At birth, he had photographically documented
erythroderma, which progressed within days into generalized
superficial blisters and erosions, predominantly on mechanically
stressed regions. When growing up, the patient progressively
developed generalized hyperkeratosis, predominantly on the trunk,
the large flexures and the ankles (( figure 1 )). Recurrent
blisters always healed without scarring. The palms and soles became
covered with thick hyperkeratotic plaques that extended to the back
of the hands and feet. Mostly during summertime, mechanical stress
and hyperhidrosis led to localized blistering with concomitant
bacterial superinfection. Fissures on the palms and soles caused
painful manual work and walking.
Histological examination of a biopsy from the axilla showed a
massively enlarged stratum corneum and extensive cytolysis of the
epidermal spinous layers ( (figure 2A) ).
Immunohistochemistry with monoclonal antibodies directed against K1
and K10 demonstrated abnormal clumping in suprabasal keratinocytes
(( figure 2C );
only shown for K1), in contrast to the homogeneous staining pattern
in normal skin ( (figure
2D) ). Ultrastructurally, keratin filaments in suprabasal
keratinocytes were aggregated into large perinuclear clumps ( (figure 2B) ). The
clinical, histological, and ultrastructural findings were in
accordance with the diagnosis of EHK.
In the patient, direct sequencing of the PCR fragment containing
exon 1 of the K1 gene revealed a heterozygous C-to-T transition at
nucleotide position 559 of the K1 cDNA sequence, counting the first
nucleotide of the initiating methionine codon as number 1 ( (figure 3A) ). This
base substitution leads to a missense mutation, consisting of an
amino acid substitution from leucine to phenylalanine at position
187 in the deduced amino acid sequence, designated L187F.
To verify the mutation detected in exon 1, we further studied
the parents as well as 75 healthy, unrelated Caucasian individuals
for absence of the mutation. Restriction enzyme digestion was
performed since the mutation abolishes a recognition site (GTSAC)
for the endonuclease Tsp45I. Whereas the control individuals as
well as both parents only revealed two fragments of 87 and 169 base
pairs, respectively, after restriction digestion, the index patient
displayed an additional undigested fragment of 256 base pairs (bp)
( (figure 3B) ),
indicative of a mutation and highly excluding this sequence
deviation as a common polymorphism.
As neither parent revealed the mutation, verification of
paternity was obtained by examination of haplotypes using six
randomly selected microsatellite markers on different chromosomes,
D2S436, D8S298, D8S1786, D13S141, D16S3140, and D18S36. Since the
results of haplotyping unequivocally confirmed paternity, we
concluded that mutation L187F most likely represents a de novo
missense mutation in the K1 gene in our index patient.
To compare the K1 mRNA level from the patient’s skin with that
from the skin of an unaffected control individual, biopsies were
obtained. Comparable amounts of amplimers corresponding to spliced
K1 or beta-actin mRNA were detected in samples extracted from the
skin of the patient and the control individual ( (figure 4) ), indicating
that the mutation did not grossly affect steady state K1 mRNA
levels.
Discussion
The different missense mutations reported in EHK patients to date
result in amino acid substitutions within the H1-segment or within
particular sequences at the ends of the rod domains of K1 or K10,
respectively. The first 15 residues of the 1A-segment are highly
conserved between all types of intermediate filaments and are
therefore thought to be critical for filament assembly and function
[4]. Mutations that destabilize the integrity of assembled
filaments may lead to disintegration of the cytokeratin network and
to clumping of tonofilaments around the keratinocytes’ nuclei,
resulting in mechanical instability of the cytoskeleton.
The mutation at amino acid position 187 (amino acid 7 of the 1A
domain) in K1 identified in our patient affects a highly conserved
hydrophobic residue within the presumed alpha-helical heptad
repeats forming a coiled-coil domain. This most likely explains the
disruption of suprabasal keratin filament assembly and the severe
clinical phenotype. In spite of the fact that the substitution of a
leucine residue by a phenylalanine residue is a conserved change,
it is still very likely that any mutation at this position will
disrupt keratin filament assembly and cause disease, taking into
consideration the high degree of conservation of Leu187 among
different keratins. By contrast, amino acid polymorphisms that have
no functional consequence have been described in the highly
variable head and tail domains of different keratins expressed in
the epidermis [10-12]. Remarkably, the homologous mutation L7F in
the K5 [13] or K14 gene [14] encoding keratins expressed in the
basal layers of the epidermis causes epidermolysis bullosa simplex
(EBS). These patients have been described with either the
generalized Köbner type of EBS (K14 mutation), or the most severe
Dowling-Meara phenotype of EBS including oral erosions and nail
deformities (K5 mutation), respectively. During preparation of this
manuscript, Uezato et al. reported a newborn Japanese girl with
generalized bullous erythroderma, that evolved into epidermolytic
hyperkeratosis with involvement of palms and soles over three years
[15]. Strikingly, the patient’s severe phenotype was attributed to
the very same spontaneous L187F K1 mutation as described herein. In
addition, a polymorphism was ruled out by analysis of 57 unaffected
individuals including her parents. These results obtained in a
different ethnic background strongly corroborate the conclusion of
our study.
The clinical variation encountered in EHK may be explained by
the broad spectrum of different mutations in the K1 and K10 genes,
respectively [1]. Whereas a severe phenotype as seen in our patient
correlates with extensive tonofilament clumping upon electron
microscopy, milder forms of EHK usually reveal more subtle
perturbations in keratinocyte ultrastructure [8].
Keratin mutations occurring in the coding regions of the
respective genes usually cause disease via direct effects on
protein structure and function. By contrast, many diseases result
from mutations that have an effect on various aspects of mRNA
metabolism, including processing, stability, and translational
control. In the case presented herein, the mutant K1 protein most
likely exerts a dominant negative effect on K10, its suprabasal
“partner”-keratin, as well as on the residual amount of normal K1
encoded from the wild-type allele in trans. This is consistent with
the finding that K1 mRNA ( (figure 4) ) and protein
levels (data not shown) measured in the patient were comparable to
those detected in the skin of a healthy control individual,
although our data cannot distinguish between both wild-type and
mutant variants.
To date, morphological analysis of amniotic cells alone is not
considered to be a sufficiently reliable method for the detection
of a fetus affected with EHK. Thus, only the identification of the
underlying pathogenic mutation will be useful for genetic
counseling, prenatal diagnosis [16], and for studying putative
genotype-phenotype-correlations in this autosomal dominantly
inherited disorder in the near future.
Acknowledgements
The authors thank Andreas Ebner for excellent photoreprographical
assistance.
References
1 DiGiovanna JJ, Bale SJ. Clinical heterogeneity in
epidermolytic hyperkeratosis. Arch Dermatol 1994; 130: 1026-35.
2 Lacz NL, Schwartz RA, Kihiczak G. Epidermolytic
hyperkeratosis: a keratin 1 or 10 mutational event. Int J Dermatol
2005; 44(1): 1-6.
3 Muller FB, Huber M, Kinaciyan T,
Hausser I, Schaffrath C, Krieg T, Hohl D,
Korge BP, Arin MJ. A human keratin 10 knockout causes
recessive epidermolytic hyperkeratosis. Hum Mol Genet 2006; 15(7):
1133-41.
4 Porter RM, Lane EB. Phenotypes, genotypes and their
contribution to understanding keratin function. Trends Genet 2003;
19(5): 278-85.
5 DiGiovanna JJ, Bale SJ. Epidermolytic
hyperkeratosis: applied molecular genetics. J Invest Dermatol 1994;
102(3): 390-4.
6 Ishida-Yamamoto A, McGrath JA, Judge MR,
Leigh IM, Lane EB, Eady RA. Selective involvement of
keratins K1 and K10 in the cytoskeletal abnormality of
epidermolytic hyperkeratosis (bullous congenital ichthyosiform
erythroderma). J Invest Dermatol 1992; 99(1): 19-26.
7 Lane EB, McLean WH. Keratins and skin disorders. J
Pathol 2004; 204(4): 355-66.
8 Syder AJ, Qc Y, Paler AS, et al. Genetic
mutations in the K1 and K10 genes of patients with epidermolytic
hyperkeratosis. J Clin Invest 1994; 93: 1533-42.
9 Whittock NV, Eady RAJ, McGrath JA. Genomic
organisation and amplification of the human epidermal type II
keratin genes K1 and K5. Biochem Biophys Res Commun 2000; 274:
149-52.
10 Chipev CC, Korge BP, Markova N, et al. A
leucine – proline mutation in the H1 subdomain of keratin 1 causes
epidermolytic hyperkeratosis. Cell 1992; 70: 821-8.
11 Mischke D, Wild G. Polymorphic keratins in human
epidermis. J Invest Dermatol 1987; 88: 191-7.
12 Korge BP, Gan SQ, McBride OW, et al.
Extensive size polymorphism of the human keratin 10 chain resides
in the C-terminal V2 subdomain due to variable numbers and sizes of
glycine loops. Proc Natl Acad Sci USA 1992; 89: 910-4.
13 Nomura K, Shimizu H, Meng X, et al. A
novel keratin K5 gene mutation in Dowling-Meara epidermolysis
bullosa simplex. J Invest Dermatol 1996; 107: 253-4.
14 Yamanishi K, Matsuki M, Konishi K, et al.
A novel mutation of Leu (122) to Phe at a highly conserved
hydrophobic residue in the helix initiation motif of keratin 14 in
epidermolysis bullosa simplex. Hum Mol Genet 1994; 3: 1171-2.
15 Uezato H, Yamamoto Y, Yuwae C, Nonaka K,
Oshiro M, Kariya K, Nonaka S. A case of bullous
congenital ichthyosiform erythroderma (BCIE) caused by a mutation
in the 1A helix initiation motif of keratin 1. J Dermatol 2005;
32(10): 801-12.
16 Rothnagel JA, Longley MA, Holder RA,
et al. Prenatal diagnosis of epidermolytic hyperkeratosis by
direct gene sequencing. J Invest Dermatol 1994; 102: 13-6.
|
Version imprimable
Version PDF
