ARTICLE
Auteur(s) : A Susanne Aalfs, DA Mira Oktarina,
Gilles FH Diercks, Marcel F Jonkman, Hendri H Pas
Centre for Blistering Diseases, Department
of Dermatology, University Medical Centre Groningen,
University of Groningen, P.O. Box 30.001, 9700 RB Groningen,
the Netherlands
accepté le 14 F�vrier 2010
Staphylococcal scalded skin syndrome is a skin disease that is
characterized by generalized superficial flaccid blisters and
denudation or desquamation [1]. The underlying cause of SSSS is a
Staphylococcus aureus infection which produces exfoliative toxins
(ETs) [1-4]. At first, patients present with fever, erythema
and tenderness of the skin, later followed by widespread
formation of fluid filled blisters that are thin walled and easily
rupture, especially in regions of friction. Gentle friction on
healed or uninvolved skin produces separation of keratinocytes of
the superficial epidermis, a blister will form within minutes; this
is called a positive Nikolsky sign, a characteristic phenomenon in
SSSS [5]. Within hours to days, large sheets of epidermis peel off,
due to the loss of the roof of the blisters, which resemble scalded
skin [1]. Denuded, sensitive and painful skin is left. Mucous
membranes are never affected [1, 5].
SSSS is a relatively uncommon disease that is mostly seen in
newborns and young children, but may also occur in immune
compromised patients or adults with renal failure [1-3]. The
mortality rate in children is low, with early diagnosis and
appropriate therapy, but is still 3%. Potentially fatal
complications include dehydration, hypothermia and the risk of
secondary infections [5, 6]. Long-term complications, such as
scarring, are hardly ever seen because of the superficial level of
the blister and the rapid healing after treatment [5]. In adults,
the mortality rate can reach over 50% despite aggressive antibiotic
management, because the majority of this group suffers from an
underlying disease or an immune compromised state [6, 7].
In the 18th and 19th centuries, SSSS in
infants was called ‘pemphigus neonatorum’ because of the clinical
similarities between SSSS and pemphigus; an autoimmune disease that
also causes blistering of the skin [4]. There are two different
subtypes of pemphigus, pemphigus foliaceus (PF) and pemphigus
vulgaris (PV). In the case of PF blisters develop in the
superficial layers of the epidermis, below the stratum corneum,
with an identical histopathology to that seen in SSSS [4, 8]. The
loss of keratinocyte adhesion in pemphigus is caused by
immunoglobulin G (IgG) autoantibodies. The target molecules
for these IgG autoantibodies in PF are recognized to be desmoglein
(Dsg) 1 [1, 4, 9].
Desmogleins are proteins found in desmosomes; intercellular
adhesion structures that interconnect the epidermal keratinocytes
by anchoring the intermediate filaments of one cell to another.
Dsg1 and -3 mediate this intercellular adhesion. Desmogleins are
cadherins, a family of calcium dependent adhesion molecules, of
which the members Dsg1 and Dsg3 and the desmocollins (Dsc) 1,
2 and 3 are expressed in skin. The cadherins demonstrate
a specific distribution within the epidermal layers; Dsg1 and Dsc1
are more intensely present in the superficial layers, whereas Dsg3
and Dsc3 are more restricted to the lower layers [10].
In 1970, exfoliative toxins (ETs) were for the first time
suggested to be involved in the pathogenesis of SSSS. Melish and
Glasgow isolated S. aureus from SSSS patients and injected it into
newborn mice. This resulted in intra-epidermal cleavage and
exfoliation, resembling the manifestation in human disease [11].
Three isoforms of ET, exfoliative toxins A (ETA), B (ETB) and
D (ETD) have been identified as being capable of inducing human
SSSS [12-14]. In Europe and the USA the majority of cases involve
ETA producing strains, while in Japan the ETB strains dominate
[1, 5, 7, 15]. These ETs are glutamic acid-specific serine
proteases and, in 2000, Amagai et al. demonstrated that Dsg1
is the substrate, the same desmosomal protein to which the IgG in
PF is directed, which is specifically cleaved by ETA [8]. Two years
later, it was found that ETB had the same specificity [7]. Hanakawa
et al. showed that human and mouse Dsg1 are cleaved by ETA and
ETB at the same site, at glutamic acid 381, between extracellular
domain (EC) 3 and EC 4 [2, 16, 17]. ETs also cause bullous
impetigo, the localized form of SSSS. Here S. aureus gets through
the skin barrier and releases the toxin locally, causing blisters
at the site of infection [6, 8] ET producing S. aureus can be
isolated from the lesions, in contrast to the situation in SSSS. In
SSSS, the toxins are produced at a distant focus and get into the
circulation, causing blisters at remote sites. Therefore S. aureus
cannot be obtained from intact blisters. The diagnosis of SSSS is
therefore verified by isolation of ETA and/or ETB producing S.
aureus from other sites [17]. Commonly, these include the
conjunctivae, umbilicus, nasopharynx or blood [1, 2, 4, 7]. Proof
that ETs cause the lesions in SSSS and bullous impetigo through
cleavage of Dsg1 is overwhelming but, nonetheless, only obtained
through experimental studies that used S. aureus extracts or
purified ETA and ETB toxins in mouse model systems or tissue
sections of normal human skin. No study has addressed the fate of
Dsg1 in patients in order to confirm the current hypothesis of
pathogenesis. The aim of the present study therefore was to verify
that Dsg1 is indeed cleaved by ETs in vivo in patient skin.
Materials and methods
Patients
For this study we selected eight biopsies of eight different
patients with staphylococcal scalded skin syndrome. All the
patients, except for one, were children at the time of diagnosis,
ranging from an eight-day old neonate to the age of 7 years.
The only exception was an 82 year old woman. The diagnosis of
SSSS was based on clinical findings and histological aspects in the
haematoxylin-eosin (HE) staining. Histological aspects included
subcorneal blister formation, the degree of acantholysis and the
presence of a minor infiltrate. Biopsies of healthy skin obtained
from breast reduction were used as controls.
Biopsies
Punch biopsies were taken from lesional skin of patients with SSSS.
Three biopsies had been frozen immediately after collection in
liquid nitrogen and then been stored at – 80 °C. The other
five biopsies were formalin-fixed, paraffin-embedded and stored at
room temperature.
Immunofluorescence microscopy
From the frozen biopsies, cryostat sections with a thickness of
4 μm were cut and then mounted on Polysine™ glass slides. The
slides were cold air-dried before a fan for 15 minutes before
the staining procedure. 4 μm sections from the formalin-fixed,
paraffin-embedded tissue blocks were also cut and mounted on
Starfrost® glass slides. Sections were deparaffinised by
Xylol followed by microwave treatment for antibody retrieval.
Deparaffinised slides were placed in Tris HCl buffer (0.1 mol/L, pH
9.5) and placed in a 98 °C microwave for 10 minutes. This
procedure was followed three times, after which sections were ready
for staining.
The slides were washed with phosphate-buffered saline (PBS pH
7.2) followed by incubation with the primary antibody (mouse
monoclonals), diluted in PBS containing 1% (w/v) ovalbumin
(PBS/OVA). Incubation time was 30 minutes in a moist chamber
at room temperature. The sections were then washed with PBS for
15 minutes and, as a secondary step, incubated with
Alexa488-conjugated goat anti-mouse IgG (Molecular Probes Eugene,
OR, U.S.A) diluted in PBS/OVA 1% (dilution 1:600) for
30 minutes. For double staining with two different mouse
monoclonals, we used Zenon® Mouse IgG Labeling Kits
Alexa Fluor®488 and Alexa Fluor®568
(Molecular Probes, Invitrogen, USA), following technical protocols
from the company. Nuclear counterstaining was performed with
bisbenzimide diluted in PBS (8 μg/mL) for five minutes. After
a final five minute PBS wash, the sections were coverslipped under
SlowFade® antifade reagent (Molecular Probes,
Invitrogen, USA). Slides were examined with a Leica DMRA
fluorescence microscope at 40 times magnification. The
staining patterns were photographed using a Leica DFC 350FX digital
camera (Leica Microsystems AG, Wetzlar, Germany).
Antibodies
The ectodomain of Dsg1 was stained with Dsg1-P23 (dilution 1:20),
Dsg3 with Dsg3-G194 (dilution 1:40), Dsc1 with U100 (1:40) and Dsc3
with U114 (all from Progen Immuno-diagnostika, Heidelberg,
Germany). The endodomain of Dsg1 was stained with DG3.10 (1:10)
(Acris Antibodies, Herford, Germany). Clone DG3.10 also recognizes
Dsg2. Dsg2 mRNA is present in low amounts in epidermis and other
stratified epithelia, however protein expression in normal human
skin is so low that it is not detectable by immunofluorescence and
therefore DG3.10 staining represents only Dsg1 [18, 19].
Plakoglobin was stained with 5F11 (1:1000) (Sigma-Aldrich,
Missouri, USA). All antibodies, except Dsg3-G194, reacted on
paraffin embedded sections after microwave treatment.
Results
Immunofluorescence on normal human skin shows Dsg1 to be unevenly
expressed throughout the epidermis. Expression starts at low levels
in the basal cells and then gradually increases towards the upper
layers. When the uppermost layers are reached, expression fades
again. When staining the patient biopsies for the ectodomain of
Dsg1, we observed various staining patterns that we classified in
four groups (table 1). These were:
(1) staining as in normal healthy skin, (2) a staining pattern that
was unevenly distributed around the cells, (3) loss of staining
when nearing a blister and (4) complete loss of intraepidermal
staining throughout the biopsy although sometimes small
intracellular dots remained visible, which most likely represent
newly synthesized Dsg1 at the endoplasmatic reticulum (figure 1).
The disappearance of Dsg1 staining fits in with the current
hypothesis that Dsg1 is degraded by S.aureus ET. We used an
ectodomain binding monoclonal, and the Dsg1 ectodomain is attacked
and spliced off at position Glu381 by ET. To investigate if loss of
the ectodomain also results in loss of the complete protein, we
also stained our sections for the endodomain. The endomain remained
present in the sections, also at positions where the staining of
the ectodomain was lost (figure 2). Near the
blister, the staining could, however, appear distorted, becoming
more cytoplasmic, possibly as a secondary effect to blistering.
In addition, we also investigated our biopsies for aberrant
expression of other cadherins – Dsg3, Dsc1 and Dsc3 – and for the
desmosomal plaque protein plakoglobin. In all SSSS biopsies, Dsc3,
Dsg3, and plakoglobin kept their normal distribution, similar to
that in normal human skin (data not shown). Dsg3 could only be
investigated in frozen biopsies as this monoclonal was not suitable
for staining paraffin sections. Dsc1 also remained present in most
biopsies but in the greater part of biopsy #6, surprisingly,
the membrane staining of Dsc1 was lost and replaced by a diffuse
cytoplasmic staining. At the same time the Dsg1 molecule was
present with a normal distribution (figure 3).
Table 1 Staining patterns of the Dsg1
ectodomain in patient specimens
|
Biopsy #
|
Normal staining pattern
|
Uneven cellular distribution of Dgs1
|
Loss of Dsg1 towards the blister
|
Complete loss of Dsg1
|
|
1
|
|
|
|
+
|
|
2
|
|
+
|
|
|
|
3
|
|
+
|
+
|
|
|
4
|
|
+
|
+
|
|
|
5
|
|
+
|
|
|
|
6
|
+
|
|
|
|
|
7
|
|
|
+
|
|
|
8
|
|
|
|
+
|
Discussion
In this study we investigated whether the ectodomain of Dsg1 is
cleaved in the skin of patients with SSSS. In all biopsies but one,
the normal staining pattern of the ectodomain of Dsg1 was not
preserved. In contrast, the staining pattern of the Dsg1 endodomain
remained normal in the skin. Similar observations were made by
Amagai et al. when searching the target of ETs [7, 8]. They
incubated normal human skin sections with ETA and ETB, and
demonstrated a loss of cell surface Dsg1 but not of the cytoplasmic
domain of Dsg1 [7]. We only saw a total loss of the Dsg1 throughout
the whole biopsy, as in the in vitro experiments of Amagai, in two
of our biopsies. In three of our eight biopsies, Dsg1 was only lost
at sites near the blister. Four out of the eight biopsies showed a
pattern where Dsg1 partially disappeared, with an uneven tissue
distribution. The smooth pattern along the cell membrane no longer
exists, but instead smaller and bigger gaps without Dsg1 are
present in this pattern. In two biopsies, both uneven cellular
distribution as well as loss of Dsg1 near the blister was seen.
Apparently the toxins do not diffuse evenly into the epidermis and
Dsg1 is lost at those positions where sufficient active ET is
present. In vitro studies have demonstrated that cleavage of Dsg1
by ETs is both a time- and dose-dependent process [7, 8]. Thus,
both toxin load and duration of infection will likely affect the
degree of Dsg1 loss, as observed in our patient biopsies.
The biopsy of one of our patients demonstrated affected Dsc1
instead of Dsg1. This particular patient was an 82 year old
woman and also the only adult patient included in this study. She
was hospitalized because of both heart failure and mild renal
failure. During her hospitalization she developed blisters and
epidermolysis that were diagnosed as SSSS. Because of her renal
impairment she was already at risk of developing SSSS. Skin
cultures were taken and subsequently treatment was started with
flucloxacillin. As a result the lesions healed, which clinically
confirmed the diagnosis. The cultures however did not demonstrate
S. aureus and no expression of ETA and/or ETB genes could be
demonstrated by PCR. Instead, the skin cultures and one of the
blood cultures turned out to contain a coagulase negative
staphylococcus (CNS). Coagulase negative staphylococci are ordinary
commensals of the skin [20]. They are non-virulent and are mainly
considered as contaminants when found in blood or other cultured
specimens. A few strains of CNS isolated from skin lesions
have been demonstrated to produce superantigenic exotoxins [21].
But, to our knowledge, it has never been reported that these toxins
can produce skin infections like SSSS or bullous impetigo.
Interestingly, very recently, a single case of a Staphylococcus
sciuri scalded skin syndrome was reported [22]. S. sciuri
constitutes 0.79 to 4.3% of CNS and although they are
associated principally with animals, they may also colonize humans.
In humans, infection with S. sciuri is most frequently associated
with wound infections [23]. In our case, it can not be ruled out
that CNS toxins were responsible for the disappearance of Dsc1. As
at that time no investigation was performed to identify toxins, we
cannot retrieve any information which might support possible toxin
activity. The fact that no S. aureus was isolated from the cultures
also does not by definition exclude its presence. If it was
present, it remains, however, unexplained how Dsc1 became affected
and not Dsg1. No studies of the possible involvement of Dsc1 in
SSSS exist, although Amagai et al. did investigate if
desmocollin was affected by S. aureus ETs, and found no evidence
for this [7]. We also demonstrated that Dsg3, Dsc3 and plakoglobin
were not affected by the loss of the Dsg1 ectodomain or the ETs
themselves. Therefore desmosomal adhesion seems preserved in the
lower layers, as predicted by the desmoglein compensation
hypothesis [24]. Blisters will thus appear when Dsg3 is no longer
expressed, i.e. in the subcorneal layers.
Conclusion
Where previous studies demonstrated that ETs could induce the
histology of SSSS in neonatal mouse skin or normal human skin, in
this study we confirmed that the immunofluorescence pattern of Dsg1
in vivo appeared gone or partially gone in SSSS patient skin. In
addition, we found the loss of Dsc1 instead of Dsg1 in an adult
patient.
Acknowledgements
Financial support: none. Conflict of interest: none.
References
1 Nishifuji K, Sugai M, Amagai M. Staphylococcal
exfoliative toxins: ‘Molecular scissors’ of bacteria that attack
the cutaneous defense barrier in mammals. J Dermatol Sci 2007; 49:
21-31.
2 Hanakawa Y, Stanley JR. Mechanisms of blister
formation by Staphylococcal toxins. J Biochem 2004; 136:
747-50.
3 Anzai H, Stanley JR, Amagai M. Production of
low titers of anti-desmoglein 1 IgG autoantibodies in some patients
with staphylococcal scalded skin syndrome. J Invest Dermatol 2006;
126: 2139-41.
4 Stanley JR, Amagai M. Pemphigus, bullous impetigo,
and the staphylococcal scalded skin syndrome. N Engl J Med 2006;
355: 1800-10.
5 Ladhani S. Recent developments in staphylococcal scalded
skin syndrome. Clin Microbiol Infect 2001; 7: 301-7.
6 Ladhani S. Understanding the mechanism of action of the
exfoliative toxins of Staphylococcus aureus. FEMS Immunol Med
Microbiol 2003; 39: 181-9.
7 Amagai M, Yamaguchi T, Hanakawa Y,
Nishifuji K, Sugai M, Stanley JR. Staphylococcal
exfoliative toxin B specifically cleaves desmoglein 1. J Invest
Dermatol 2002; 118: 845-50.
8 Amagai M, Matsuyoshi N, Wang ZH, Andl C,
Stanley JR. Toxin in bullous impetigo and staphylococcal
scalded-skin syndrome targets desmoglein 1. Nat Med 2000; 6:
1275-7.
9 Payne AS, Hanakawa Y, Amagai M,
Stanley JR. Desmosomes and disease: pemphigus and bullous
impetigo. Curr Opin Cell Biol 2004; 16: 536-43.
10 Green KJ, Simpson CL. Desmosomes: new perspectives
on a classic. J Invest Dermatol 2007; 127: 2499-515.
11 Melish ME, Glasgow LA. The staphylococcal scalded
skin syndrome: development of an experimental model. N Engl J Med
1970; 282: 1114-9.
12 Kondo I, Sakurai S, Sarai Y. New type of
exfoliatin obtained from staphylococcal strains belonging to phage
groups other than group II, isolated from patients with impetigo
and Ritter's disease. Infect Immun 1974; 10: 851-61.
13 Ladhani S, Joannou CL, Lochrie DP,
Evans RW, Poston SM. Clinical, microbial and biochemical
aspects of the exfoliative toxins causing staphylococcal scalded
skin syndrome. Clin Microbiol Rev 1999; 12: 224-42.
14 Yamaguchi T, Nishifuji K, Sasaki M,
et al. Identification of the Staphylococcus aureus etd
pathogenicity island which encodes a novel exfoliative toxin, ETD,
and EDIN-B. Infect Immun 2002; 70: 5835-45.
15 Patel GK, Finlay AY. Staphylococcal scalded skin
syndrome: diagnosis and management. Am J Clin Dermatol 2003; 4:
165-75.
16 Hanakawa Y, Schechter NM, Lin C, et al.
Molecular mechanisms of blister formation in bullous impetigo and
staphylococcal scalded skin syndrome. J Clin Invest 2002; 110:
53-60.
17 Whittock NV, Bower C. Targetting of desmoglein
1 in inherited and acquired skin diseases. Clin Exp Dermatol
2003; 28: 410-5.
18 Schäfer S, Koch PJ, Franke WW. Identification
of the ubiquitous human desmoglein, Dsg2, and the expression
catalogue of the desmoglein subfamily of desmosomal cadherins. Exp
Cell Res 1994; 211: 391-9.
19 Iwatsuki K, Han GW, Fukuti R, et al.
Internalization of constitutive desmogleins with the subsequent
induction of desmoglein 2 in pemphigus lesions. Br J Dermatol
1999; 140: 35-43.
20 Roth RR, James WD. Microbial ecology of the skin.
Annu Rev Microbiol 1988; 42: 441-64.
21 Akiyama H, Yamasaki O, Tada J, Arata J.
The production of superantigenic exotoxins by coagulase-negative
staphylococci isolated from human skin lesions. J Dermatol Sci
2000; 24: 142-5.
22 Tang Y. A case report of Staphylococcus sciuri scalded
skin syndrome. Chinese Journal of Contemporary Pediatrics 2008; 10:
264.
23 Stepanovi S, Daki I, Morrison D, et al.
Identification and characterization of clinical isolates of members
of the Staphylococcus sciuri group. J Clin Microbiol 2005; 43:
956-8.
24 Mahoney MG, Wang Z, Rothenberger K,
Koch PJ, Amagai M, Stanley JR. Explanation for the
clinical and microscopic localization of lesions in pemphigus
foliaceus and vulgaris. J Clin Invest 1999; 103: 461-8.
|
Version imprimable
Version PDF
