INTERNATIONAL JOURNAL OF IMMUNOPATHOLOGY AND PHARMACOLOGY

Vol. 20, no. 3, 557-566 (2007)

IMMUNOHISTOCHEMICAL EXPRESSION OF APOPTOTIC MARKERS IN DRUGINDUCED ERYTHEMA MULTIFORME, STEVENS-JOHNSON SYNDROME

AND TOXIC EPIDERMAL NECROLYSIS
A.V. MARZANO,A. FREZZOLINP, M. CAPRONF,A. PARODP, D. FANONI, P. QUAGLIN04,
V. GIRGENTI, M. LA PLACA5, P. FABBRF, R. CAPUTO and E. BERTI 6

Institute ofDermatological Sciences, University ofMilan, IRCCS Ospedale Maggiore Polic/inico,

Mangiagalli e Regina Elena ofMilan; I Istituto Dermopatico dell'Immacolata, Istituto di Ricovero
e Cura a Carattere Scientifico (IDI-IRCCS), Rome; 'Depanment ofDermatological Sciences,
University ofFlorence, Florence; 'Section ofDermatology, University ofGenoa, Genoa;
"Department ofBiomedical Sciences and Human Oncology, Section ofClinics and Oncological
Dermatology, University ofTurin; 'Depanment of Clinical and Experimental Medicine, University
ofBologna, Bologna; "University ofMilan 'Bicocca', Milan, Italy
Received November 15,2006 -Accepted March 21,2007
Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN) are considered to be severity
variants of the same disease, which is almost always associated with drug intake. In contrast, erythema
multiforme (EM) is a disorder regarded as only rarely caused by drugs. Keratinocyte apoptosis has been
shown to play an important part in the pathogenesis of SJS and TEN, whilst its role in EM remains controversial. To determine the expression of apoptosis-associated molecules Fas, Fas ligand (FasL), BcI-2
and Bax in the above disorders, an immunohistochemical anatysis was performed. We studied both lesional skin from thirty patients having drug-induced EM and 5 cases classified within the SJS/TEN spectrum and normal skin samples. We found a keratinocyte overexpression of Fas antigen, an important
molecule mediating apoptosis, not only in SJS and TEN but also in EM. Another noteworthy finding was
the strong expression of BcI-2, a protein known as blocking apoptosis, along the basal layer and in the
dermal infiltrate both in SJS/TEN and in EM. Taken together, these findings suggest that Fas-dependent
keratinocyte apoptosis may playa part in the pathogenesis of both SJS/TEN and EM. Fas-mediated cell
death may be partially suppressed by the BcI-2 protein.
skin lesions, whereas SJS and TEN are only severity
variants of a single entity (1-4).
This last is almost always associated with drug
intake, whilst EM, both in its majus variant and in
the form so-called minor, is predominantly caused
by preceding infections, notably with herpes simplex
virus (HSV), and only rarely by drugs.

Until recently, most textbooks of medicine considered erythema multi forme (EM) to be a spectrum
of disorders that included EM majus, Stevens-Johnson syndrome (SJS), and toxic epidermal necrolysis
(TEN). A few years ago, some authors suggested that
EM majus is different from SJS and TEN, not only in
severity but also in the pattern and distribution of the

Key words: apoptosis, erythema multiforme, Stevens-Johnson syndrome. toxic epidermal necrolysis
Mailing address: Dr Angelo V. Marzano, MD
Institute of Dermatological Sciences,
University of Milan,
Via Pace 9,
20122 Milan, Italy
Tel: ++39-2-55035101 Fax: ++39-2-50320779
e-mail: daniele.fanoni@hotmail.it

0394-6320 (2007)

557

Copyright by BIOLlFE. 8.a.8.

This publication and/or article is for individual use only and may not be further
reproduced without written permission from the copyright holder.
Unauthorized reproduction may result in financial and other penalties

558

A.V. MARZANO ET AL.

Although the pathophysiology of SJS and TEN

is not completely understood, it has been shown that
dysregulation of keratinocyte apoptosis may play an
important part in the development of these diseases
(5-7).
Apoptosis is a form ofcell death occurring in several physiological conditions, most notably normal
development and tissue homeostasis as well as preventing pathological processes such as autoimmunity and cancer. Furthermore, programmed cell death
plays an important role during many inflammatory
and malignant pathological processes, including cutaneous disorders (8-10). Apoptosis is a sequence
of events based on cellular metabolism that lead to
cell shrinkage, nuclear and cytoplasmic condensation, chromatin fragmentation and phagocytosis.
Cell death by apoptosis can be triggered by several
stimuli, including intracellular stress and receptormediated signaling. The intracellular transduction of
these signals is mainly mediated by the members of
so-called caspase family, which are critical executioners of the apoptotic pathway. Indeed, the caspase-activated Dnase enters the nucleus and cleaves
DNA to produce DNA laddering.
Three major pathways can initiate apoptosis in
mammalian cells: one involving engagement of
death receptors, one involving the release of cytochrome c from mitochondria, and one involving
endoplasmic-reticulum (8-10). The mitochondrial
pathway is triggered both by external and internal
cues, such as DNA damage. Pro- and anti-apoptotic
members of the Bcl-2 family gather at the surface
of mitochondria, where they compete to regulate
cytochrome c release. In the event of apoptosis,
cytochrome c is released from mitochondria, leading to caspase activation. Members of Bcl-2 family
have been classified into three functional groups.
The most important are members of group I, such
as Bcl-2, possessing anti-apoptotic activity, and
members of group II, such as Bax, which are proapoptotic.(8-10). Death receptors are a subfamily of
transmembrane proteins which belong to the TNFfamily of receptors, including, among others, Fas
and the receptor for TNF-a, TNFR 1 (11). The best
known death-receptor signaling pathway to date is
that triggered by the binding of Fas ligand (FasL) to
Fas. Apoptosis upon binding of a death ligand to its
specific receptor is also dependent on the activation

of caspases.
The role of apoptosis in EM remains more controversial compared with the SJS/TEN spectrum;
moreover, analogies and differences between these
two conditions, concerning the immunological characteristics of the cutaneous inflammatory infiltrate,
need to be further elucidated. To fill this gap in
literature, we examined, retrospectively, 30 patients
diagnosed as having drug-induced EM and 5 cases
classified within the SJS/TEN spectrum performing
an immunohistochemical analysis. The expression
of apoptosis-associated molecules Fas, FasL, Bcl-2
and Bax playing a key part in orchestrating the apoptotic machinery (12), in addition to the infiltrate,
was evaluated.
MATERIALS AND METHODS
Patients
Thirty patients with EM, three with SJS and two with
TEN were chosen for investigation. These patients were
seen at several Italian departments of Dermatology from
1995 to 2004. Stringent criteria were used to include the
above-mentioned patients, according to the classification
for EM, SJS and TEN proposed by Bastuji-Garin et al
(I). Within the EM subset, 20 patients were diagnosed as
having EM minor, whereas 10 cases were defined as EM
majus. In all patients, histopathological examination confirmed the diagnosis of EM, SJS or TEN, respectively.
In all cases, medication was suspected of causing the
eruption. The most probable culprit drug was identified
according to the imputability method used for toxic drug
reactions (13). All patients had negative serology for HIV
and mycoplasma infection; moreover, the clinical lesions
had not been preceded by clinically identifiable HSV
episodes in the 4 weeks prior to our observation. In 10
patients with EM minor and in 5 patients with EM majus
polymerase chain reaction analysis for detection of HSVI and HSV-2 DNA sequences from lesional skin had been
performed, as previously described (14), with negative results. However, we must mention the possibility that some
of the patients may have had a cervical outbreak of their
HSV which would not be found on cutaneous exam and
possibly not noticed by the patients. Skin biopsy specimens were obtained on admission. They were taken from
"typical targets" or from "raised atypical targets" in EM,
performing the biopsy from the center to the edge of the
lesion. In SJS, where widespread small blisters were seen
on purple macules or on "flat atypical targets", they were
taken from these lesions near the bullae. In TEN, they
were taken from erythematous skin near the epidermal

101.J. Immuoopalhol. Pharmacol.

detachment. Negative control specimens were taken from

the clinically normal skin of 5 healthy volunteers during
plastic surgical operations after informed consent.
Immunohistochemical studies
Sections from formalin-fixed, paraffin-embedded tissue and from acetone-fixed cryostat sections were used
according to the specificity of the different antibodies
(Table I). A standard alkaline phosphatase anti-alkaline
phosphatase technique (Dakopatts, Glostrup, Denmark)
was carried out according to a previously described
method (15).
For APO-I IFas antigen detection, a two-step immunoperoxidase method was carried out using a polymerperoxidase complex-binding (En Vision+ TM, Peroxidase,
DAKO Corp., Carpinteria, CA; USA).
Two independent "blind" observers evaluated serial
sections. Stainings were quantitated using the following
nomenclature to indicate the cells number per field: weak
between 0 and 9, moderate between 10 and 19 and strong
greater than 20.
For statistical analysis, the stained cells were counted
in three consecutive microscopic fields (250 X) and statistical significance (p< 0.05) was assessed by Student's
t test.

RESULTS
The immunohistochemical findings concerning infiltrate phenotyping and apoptosis-associated
molecules both in SJS/TEN and in EM as well as
in normal skin are summarized in Tables II and III,
and stated in detail below. All the statistical data
regarding apoptotic markers are indicated in Table
III, while only statistically significant results are
reported in the present section.
Healthy skin
In normal skin the Fas antigen, which is one of
the most important molecules mediating apoptosis,
was very weakly expressed on keratinocytes of basal to granular layers of the epidermis. Fas was also
weakly expressed on the cells showing mononuclear
morphology in the dermis. In contrast, neither the
keratinocytes of the epidermis nor the cells of mononuclear morphology in the dermis expressed FasL,
the ligand whose interaction with Fas is responsible
for Fas-dependent cell death. Within the epidermis,
the keratinocyte expression of Bcl-2 oncoprotein, a
molecule known as blocking apoptosis, was almost

559

completely absent (Fig. 1). In contrast, Bcl-2 was

expressed by melanocytes present in the epidermis.
Single lymphocytes scattered in the dermis were
weakly Bcl-2 positive. Finally, all layers of the
epidermis, as well as the dermal mononuclear cells
were negative for Bax protein, which is a molecule
enhancing apoptosis.
Inflammatory infiltrate phenotyping
In the 3 cases classified as SJS and in those defined as TEN (two cases), the immunohistochemical findings proved to be very similar. Namely, the
inflammatory infiltrate in the epidermis was predominantly composed by T lymphocytes showing
a CD3/LAT +, CD8+ cytotoxic phenotype, with a
moderate number of CD68+ macrophages; CD4+ T
lymphocytes were slightly smaller than CD8+ (CD4/
CD8 ratio= 0.62). A more pronounced infiltrate was
found in the dermis and the nature of the inflammatory cells appeared to be similar.
In EM majus, a dense dermal infiltrate obscuring
the dermoepidermal junction and showing discrete
epidermotropism was observed. This was mostly
composed of CD3/LAT+, CD8+ cytotoxic T lymphocytes with a moderate number of CD68+ macrophages. CD4+ T lymphocytes were present in smaller
numbers than CD8+ cells, both in the epidermis and
within the dermal infiltrate (CD4/CD8 ratio= 0.55).
Interestingly, 5 patients with EM minor displayed
the same pattern of distribution of cellular subsets
as found in EM majus (Fig. 2). In all the other EM
minor cases, the CD4+ lymphocytic subpopulation
was predominant within the dermal infiltrate as well
as in the epidermis and at the dermoepidermal junction. A moderate number of CD68+ macrophages
and CD8+ lymphocytes were also disclosed in all
the above compartments (CD4/CD8 ratio=3.25).
Apoptosis-associated molecules
Epidermal Fas antigen expression was significantly increased in comparison with normal skin
(p=0.0006 for SJS and p=0.0004 for TEN, respectively). Interestingly, in the 2 patients with TEN, F~s
expression was seen in the whole epidermis (Fig.
3A), whilst in the 3 SJS cases anti-Fas staining was
restricted to the cells of basal layers (Fig. 3B). In
all SJS/TEN cases, FasL+ lymphocytes infiltrating
the perivascular and junctional dermis as well as the
lower epidermis were found, FasL expression being,

A.V. MARZANO ET AL.

560

Table I. Antibody reagents usedfor immunohistochemical studies.

Antigen

Antibody (sourcejt

Antigen distribution

CD3/LAT

3.8 (Novocastra)

Pan T lymphocytes

CD4

lF6 (Novocastra)

Memory/helper T-lymphocytes

CD8

C8/144B (Dako)

Suppressor/cytotoxic T-lymphocytes

CD68

Kpl (Dako)

Myeloid and macrophage cells

APOI/Fas (CD95)

APO-l (Dako)

Pro-apoptotic marker

Fas-L*

Fas-L (Novocastra)

Pro-apoptotic marker

Bc1-2

124 (Dako)

Anti-apoptotic marker

Bax

N-20 (Santa Cruz)

Pro-apoptotic marker

MelanA

PNL2 (Dako)

Melanocyte differentiation marker

* Fas-L indicates Fas ligand.

t Novocastra Laboratories Ltd, Newcastle upon Tyne,

UK; Dako Cytomation, Glostrup,

Denmark; Santa Cruz Biotechnology, Santa Cruz, CA, USA.

Table II. Inflammatory infiltrate phenotyping.

SJS
CD 3/LAT
CD4
CD8
CD 68

D
E

D
E

D
E

4.03 3.11
21.10 5.68
1.54 1.19
8.07 2.17
2.49 1.92
13.03 3.51
1.150.90
8.39 3.55

TEN
6.05 3.22
22.0 6.51
2.32 1.23
8.42 + 1.66
3.73 1.99
13.58 4.02
1.34 1.22
7.19 3.73

EM minor
6.15 4.02
38.52 16.15
4.70 3.07
29.46 12.35
1.45 0.95
9.06 3.80
1.66 + 0.69
9.184.08

EM malus
7.334.15
49.33 16.78
2.60 1.47
17.50 5.95
4.73 2.68
31.83 10.83
1.33 0.76
10.04 5.73

SJS: Stevens-Johnson Syndrome

TEN: Toxic Epidermal Necrolysis
EM' Erythema Multiforme
E: number ofpositive epidermal cells (mean standard deviation from the mean)
D: number ofpositive dermal cells (mean standard deviation from the mean)

albeit not strong, significantly higher than in healthy

skin (p=O.0498 for SJS and p=O.0434 for TEN, respectively). Interestingly, in the SJS/TEN spectrum
the basal keratinocytes and keratinocytes of the
involved hair follicles stained strongly for Bcl-2 oncoprotein (Fig. 4), sometimes there being Bcl-2 expression also at higher levels of the epidermis. Thus,
Bcl-2 was overexpressed if compared to healthy skin
specimens in basal epidermis (p=O.0046 for SJS and
p=O.0069 for TEN, respectively). In counting the
epidermal stained cells, both melanocytes and infil-

trating T lymphocytes positive for Bcl-2 were ruled

out. To distinguish the melanocytes, these cells were
stained by using the specific melanocytic marker
Melan-A. T lymphocytes composing the dermal inflammatory infiltrate also resulted as being strongly
positive for Bcl-2. As in normal skin, all epidermal
layers as well as the cells of the dermal infiltrate
were uniformly negative for Bax protein.
It is noteworthy that epidermal Fas expression in
the epidermis was significantly increased in EM as
compared with normal skin (p<O.OOO I both for EM

561

Int. J. Immunopathol. Pharmacol.

Table III. Immunohistochemicalfindings and statistical data concerning apoptosis-associated molecules.

SJS
Apo/Fas

0
FasL

Bcl-2

0
E

0
Bax

TEN

EM minor

EM maius

01

02

03

D4

0.0006

0.0004

<0.0001

<0.0001

NS
NS

NS
NS

0.0498

0.0434

NS
NS
NS

NS
NS
NS

0.0046

0.0069

<0.0001

0.0082

0.0008

0.0046

0.0005

<0.0001

NS
NS

NS
NS

NS
NS

NS
NS

HS

20.33
27.25
8.06
9.89
20.01 8.59 25.89 9.14 0.43 1.01
1.24 1.32 1.66 1.03 1.14 1.12 1.54 1.02 0.83 1.04
2,66
0
0
0
0
2.66 3.0
0
1.61 2.51 2.33 2.93 1.03 1.52
11.25
20.0
0
7.06
14.38
13.73 6.65 19.0 15.15
48.73
23.20
19.82
35.86
13.14
7.89
9.59
18.89
3.09 2.55
0
0
0
0
0
0
0
0
0
0

pI: Steven-Johnson Syndrome (SJS) versus Healthy Skin (HS)

p2: Toxic Epidermal Necrolysis (TEN) versus HS
p3: Erythema Multiforme minor (EM minor) versus HS
p4: Erythema Multiforme majus (EM majus) versus HS
p5: SJS versus all the others disease= Not Significant (NS) at all
p6: TEN versus all the others disease= NS at all
p7: EM minor versus all the others disease= NS at all
p8: EM majus versus all the others disease= NS at all
E: number ofpositive epidermal cells (mean standard deviation from the mean)
D: number ofpositive dermal cells (mean standard deviation from the mean)

minor and EM majus). Intriguingly, Fas antigen was

very strongly expressed in the whole epidermis in
50% of EM cases (Fig. 5A), whilst Fas expression,
although marked, appeared to be confined to the basal cell layer in the other 50% of EM cases (Fig. 5B).
We did not note any difference concerning Fas expression between EM majus and EM minor. In contrast, the epidermis did not stain positively for FasL;
a weak anti-FasL staining, marking lymphocytes of
the infiltrate distributed around upper dermal vessels, was found. Interestingly, within the epidermis a
strong Bcl-2 expression confined to the basal keratinocytes was seen in all EM cases, resulting significantly higher than in normal skin (p<O.OOOI for EM
minor and p=0.0082 for EM majus, respectively).
Moreover, the lymphocytic dermal infiltrate showed
a marked positivity for Bcl-2 oncoprotein (Fig. 6A
and B). Finally, neither epidermis nor the dermal infiltrate stained positively for Bax protein.
DISCUSSION
TEN is a rare, potentially fatal disease, almost
always related to an adverse drug reaction. Clinical
findings consist of painful bullous or erosive lesions

that rapidly spread over large areas of the skin; several mucous membranes may also be involved and
this contributes to the severity of the process. SJS is
now regarded as the same, although less severe, disease (16). Prior reports that used immunolabeling to
characterize the effector cells ofthe epidermal injury
in SJS/TEN obtained conflicting results (13, 17-18).
It is conceivable that CD8+ T lymphocytes are important in initiating the death ofkeratinocytes in SJSI
TEN (19), while monocyte-macrophage populations
playa pivotal role at the late stage of SJS/TEN. In
our study, we found that the epidermal infiltrate was
predominantly composed of CD8+ T lymphocytes,
this finding confirming the skin biopsy specimens
taken at early stage of disease. EM is a much less
rare and severe disease, clinically characterized by
typical target lesions or raised atypical target lesions
mainly located on the extremities; blisters, usually
present in EM majus but rarely in EM minor, involve much less than 10% of the body surface (3).
Mucous membrane involvement occurs rarely in EM
minor, while it is usual and sometimes severe in EM
majus. From a pathogenetic point of view, CD4+ T
lymphocytes co-operating with monocytes-macrophages are regarded as the effector cells responsible

562

A.V. MARZANO ET AL.

Fig. 1. Expression of Bcl-2 protein in normal skin.

Keratinocytes appear to be completely negative (APAAP
method, original magnification X 400).

Fig. 2. Erythema multiforme minor. Strong reactivity for

CD8 of the T lymphocytes composing the upper dermal
infiltrate obscuring the dermoepidermal junction with
discrete epidermotropism. The CD8+ lymphocytic
infiltrate mainly invades the epidermal basal layers, while
only scattered CD8+ T cells in the upper epidermis are
seen (APAAP method, original magnification X 200).

for the interface damage, C08+ T lymphocytes playing a secondary role (20). In this study, it is of interest that C08+ T lymphocytes represented the major
cell population within the inflammatory infiltrate,
not only in EM majus butalso in 5 EM minor cases,
suggesting that a cytotoxic reaction may contribute
to the pathogenesis of both EM variants.

Fig. 3. A. Toxicepidermal necrolysis. Strong expression of

Fas antigen in the whole epidermis. B. Stevens-Johnson
syndrome. Anti-Fas staining is restricted to the cells of
basal layers, the remaining epidermis being negative
(Immunoperoxidase method, original magnification X
200 and X JOO, respectively).

On the other hand, it has been widely shown that

apoptosis, which is a form of cell death occurring in
both physiological (21) and pathological conditions
(22), plays a pivotal role in the development of SJS
and TEN (5-7). Recent data suggest that the activation of Fas through its ligand, FasL, is an initial
important step leading to diffuse apoptotic cell death
of epidermal cells in TEN (6). Fas (C095, APO-l)
is a cell membrane protein belonging to a family of
structurally related tumour necrosis factor (TNF)
receptors expressed almost ubiquitously in a variety
of cells, including keratinocytes (23). FasL is also a

Int. J. ImmnnopathoI. Pharmacol.

563

Fig. 4. Toxic epidermal necrolysis. Basal keratinocytes

and keratinocytes of a hair follicle are strongly positive
for Bcl-Z, A less intense Bcl-2 expression at higher levels
ofthe epidermis is also evident (APAAP method, original
magnification X 200).

cell membrane protein belonging to the TNF family

mainly expressed in activated T cells and natural killer cells, but also by keratinocytes in some pathological conditions (24). Thus, keratinocyte apoptosis in
SJS and TEN is predominantly induced by a suicidal
interaction between Fas and FasL, which are both
overexpressed by keratinocytes (6). Soluble Fas L
(sFasL) has also been investigated concerning its potential to mediate apoptosis. Significantly increased
amounts of sFasL secreted by peripheral blood
mononuclear cells have been demonstrated in TEN
and SJS, but not in drug-induced EM, suggesting that
the serum sFasL level may be a good indicator for the
early diagnosis of these disorders (7). In our study,
we found an overexpression of Fas in the epidermis
as compared with normal skin: namely, in TEN Fas
expression was observed in whole epidermis, whilst
in SJS anti-Fas staining was confined to the basal

Fig. 5. A. Erythema multiforme (EM). Fas antigen is very

markedly expressed in the whole epidermis. In contrast,
the entire dermis appears to be negative. B. EM Fas
expression confined to the basal cell layer. The other
epidermal layers as well as the dermis are uniformly
negative for Fas antigen (Immunoperoxidase method,
original magnification X 200).

layer cells. Intriguingly, we demonstrated a marked

increase in epidermal expression of Fas antigen also
in EM, both majus and minor: in 50% of cases the
Fas overexpression was limited to the basal keratinocytes, whilst in the other 50% was ubiquitous. These
findings suggest that Fas-mediated apoptosis may
be involved in the pathogenesis of EM as shown for
the pathogenesis of SJS and TEN (tV7)-, Moreover,
the FasL expression profile in SJs!tEN ',in the current study confirms the role of the Fas/FasL system .
Nevertheless, the findings in EM, where some FasL+
T lymphocytes infiltrating the dermis were seen, do

564

A.V. MARZANO ET AL.

Fig. 6. A. Erythema multiforme (EM). Strong Bcl-2

expression confined to the basal cell layer is evident. On
the contrary, the cells of all other epidermal layers do
not stainfor Bcl-2 oncoprotein. B. EM The mononuclear
dermal infiltrate also shows an intense reactivity for Bel2 (APAAP method, original magnification X400 and X
200, respectively).

not exclude that this mechanism is involved as apoptotic pathway also in this disorder. Another notewor.thy finding of the present study was the strong Bcl-2
expression along the epidermal basal layer as well as
in the mononuclear cells of the dermal infiltrate both
in EM andin SJS/TEN. The product ofthe protooncogene BCL72;i.e. Bcl-2 protein, has been shown to
block apop~~~UJ1der a variety of conditions and is
supposedtq:jVhibit a central step in the apoptotic cell
death pathway (25).
Although the anti-apoptotic effects of molecules
like Bcl-2 have clearly been established in vitro, their

role in protecting against apoptosis initiated through

death receptor-dependent pathways, including that
Fas-dependent, has only rarely been investigated in
immune-mediated disorders (26).
We postulate that Fas-dependent cell death may
be partially suppressed by the Bcl-2 protein, strongly
expressed both in EM and in the SJS/TEN spectrum.
Indeed, it is conceivable that in a potentially fatal
disorder such as TEN the inhibition of Fas-mediated
cell death by Bcl-2 overexpression is only minimal
and that apoptosis via other systems plays a part in
inducing the progression of the disease. On the other
hand, the pattern of expression of Bcl-2 protein in
normal skin remains controversial. In fact, we found
lack of keratinocyte Bcl-2 positivity as previously
reported by some authors (27-28), but our results are
not consistent with findings of other authors, who
demonstrated Bcl-2 immunoreactivity in normal
basal keratinocytes (12, 29). The expression of Bax
protein, which is the product of a protooncogene of
the Bcl-2 family acting as death agonist (30), has
also been investigated, giving negative findings.
These data seem to suggest that Bax protein does
not play any part in the apoptosis of keratinocytes
either in EM or in SJS/TEN. On the other hand,
other authors found Bax immunoreactivity in normal
epidermis and its appendages (31). These results are
discrepant with our finding of negative Bax and Bel2 stainings in normal epidermis. Indeed, the ratio of
Bcl-2/Bax is crucial in determining whether apoptosis or cell survival will be promoted. Thus, it is
conceivable that the Bcl-2/Bax apoptotic pathway is
not pivotal for keratinocytes in physiological conditions. In EM as well as in the SJS/TEN spectrum, the
role of Bcl-2/Bax ratio seems to be directed towards
suppressing death receptor-or mitochondria-dependent apoptotic pathways.
In summary, although a strong expression of the
anti-apoptotic protein Bcl-2 along the basal layer
and in the dermal infiltrate has been found both in
EM and in the SJS/TEN spectrum, the concurrent
epidermal overexpression of Fas antigen suggests
that keratinocyte apoptosis may play a part in the
pathophysiology of all these diseases. However,
with the aim of studying drug-induced disorders, we
examined only EM cases associated with drug intake, excluding HSV-associated EM. Thus, the latter
needs to be the subject of future investigations.

101.J. Immuoopalhol. Pharmacol.

REFERENCES
I.

2.

3.

4.

5.

Bastuji-Garin S., B. Rzany, R.S. Stern, N.H.

Shear, L. Naldi and J.e. Roujeau. 1993. Clinical
classification of cases of toxic epidermal necrolysis,
Stevens-Johnson
syndrome,
and
erythema
multifonne. Arch. Dermatol. 129:92.
Assier H., S. Bastuji-Garin, J. Revuz and J.C.
Roujeau. 1995. Erythema multifonne with mucous
membrane involvement and Stevens-Johnson
syndrome are clinically different disorders with
distinct causes. Arch. Dermatol. 131:539.
Roujeau J.e. 1997. Stevens-Johnson syndrome
and toxic epidermal necrolysis are severity variants
of the same disease which differs from erythema
multifonne. J. Dermatol. 24: 726.
Auquier-Dunant A., M. Mockenhaupt, L. Naldi,
O. Correia, W. Schroder and J.e. Roujeau. 2002.
Correlations between clinical patterns and causes
of erythema multifonne majus, Stevens-Johnson
syndrome, and toxic epidermal necrolysis. Arch.
Dermatol. 138:1019.
Paul e., P. Wolkenstein, H. Adle, J. Wechsler,
H.J. Garchon, J. Revuz and J.C. Roujeau. 1996.
Apoptosis as a mechanism of keratinocyte death in
toxic epidermal necrolysis. Br. J. Dermatol. 134:
710.

6.

7.

Viard I., P. Wehrli, R. Bullani, P. Schneider, N.

Holler, D. Salomon, T. Hunziker, J.H. Saurat, J.
Tschopp and L.E. French. 1998. Inhibition of toxic
epidermal necrolysis by blockade of CD95 with
human intravenous immunoglobulin. Science 282:
490.
Abe R., T. Shimizu, A. Shibaki, H. Nakamura, H.
Watanabe and H. Shimizu. 2003. Toxic epidermal
necrolysis and Stevens-Johnson syndrome are
induced by soluble Fas Ligand. Am. J. Pathol. 162:
1515.

8.

Baetu T.M. and J. Hiscott. 2002. On the TRAIL to

apoptosis. Cytokine Growth Factor Rev. 13: 199.
9. Orrenius S. 2004. Mithochondrial regulation of
apoptotic cell death. Toxicol. Lett. 149:19.
10. Nakagawa T., H. Zhu, N. Morishima, E. Li, J.
Xu, B.A. Yankner and J. Yuan. 2000. Caspase-12
mediates endoplasmic-reticulum-specific apoptosis
and cytotoxicity by amyloid-B. Nature 403:98.

565

II. Wehrli P., I. Viard, R. Bullani, J. Tschopp and

L.E. French. 2000. Death receptors in cutaneous
biology and disease. J. Invest. Dermatol. 115: 141.
12. Baima B. and M. Sticherling. 2001. Apoptosis
in different cutaneous manifestations of lupus
erythematosus. Br. J. Dermatol. 144:958.
13. Moore N., M. Biour, G. Paux, E. Loupi, B. Beqaud,
F. Boismare and R.J. Royer. 1985. Adverse drug
reaction monitoring: doing it in the French way.
Lancet 2: 1056.
14. Lucotte G., e. Bathelier, V. Lespiaux, e. Bali and
T. Champenois. 1995. Detection and genotyping of
herpes simplex virus types I and 2 by polymerase
chain reaction. Mol. Cell Probes 9:287.
15. Cordell J.L., B. Falini, W.N. Erber, A.K. Ghosh,
Z. Abdulaziz, S. MacDonald, K.A. Pulford, H.
Stein and D.Y. Mason. 1984. Immunoenzymatic
labeling of monoclonal antibodies using immune
complexes of alkaline phosphatase and monoclonal
anti-alkaline phosphatase (APAAP complexes). J.
Histochem. Cytochem. 32:219.
16. Roujeau J.e. 2000. Necrolyse epidermique: les
avancees physiopathogeniques, Ann. Dermatol.
Venereol. 127:546.
17. Villada G., J.e. Roujeau, T. Clerici, I. Bourgault
and J. Revuz. 1992. Immunopathology of toxic
epidermal necrolysis. Arch. Dermatol. 128:50.
18. Miyauchi H., H. Hosokawa, T. Akaeda, H. Iba
and Y. Asada. 1999. T-cell subsets in drug-induced
toxic epidermal necrolysis. Arch. Dermatol. 1:851.
19. Le Cleach L., S. Delaire, L. Boumsell, M. Bagot,
I. Bourgault-Villada, A. Bensussan and J.C.
Roujeau. 2000. Blister fluid T lymphocytes during
toxic epidermal necrolysis are functional cytotoxic
cells which express human natural killer (NK)
inhibitory receptors. Clin. Exp. Immunol. 119:225.
20. Margolis R.J., M.G. Tonnesen, T.J. Harrist, A.K.
Bhan, B.I.J. Wintroub, M.C. Mihm Jr and N.A.
Soter. 1983. Lymphocyte subsets and Langerhans
cell/indeterminate cells in erythema multiforme. J.
Invest. Dermatol. 81(5):403.
21. Cohen J.J. and R.S. Duke. 1992. Apoptosis and
programmed cell death in immunity. Annu. Rev.
Immunol. 10:267.
22. Cohen J.J. 1993. Apoptosis. Immunol. Today 14:
126.

566

A.V. MARZANO ET AL.

23. Sayama K., S. Yonehara, Y.Watanabe and Y. Mild.

1994. Expression of Fas antigen on keratinocytes in
vivo and induction of apoptosis in cultured keratinocytes. J. Invest. Dermatol. 103:330.
24. Iwai K., T. Miyawald, T. Takizawa, A. Konno, K.
Ohta, A. Yachie, H. Seki and N. Taniguchi. 1994.
Differential expression of bcl-2 and susceptibility
to anti-Fas-mediated cell death in peripheral blood
lymphocytes, monocytes and neutrophils. Blood 84:
1201.
25. Reed J.e. 1996. Mechanism ofbcl-2 family protein
function and dysfunction in health and disease. Behring Inst. Mitt. 97: 72.
26. Liu H. and R.M. Pope. 2003. The role of apoptosis in
rheumatoid arthritis. Curr. Opin. Pharmacal. 3:317.
27. Itoh N., Y. Tsujimoto and S. Nagata. 1993. Effect
of bcl-2 on Fas antigen-mediated cell death. J. Im-

munol. 151:621.
28. Nakagawa K., K. Yamamura, S. Maeda and M.
Ichihashl, 1994. Bcl-2 expression in epidermal keratinocytes disease. Cancer 74:1720.
29. Hockenbery D.M., M. Zutter, W. Hickey, M.
Nahm and S.J. Korsmeyer. 1991. Bcl-2 protein is
topographically restricted in tissues characterized by
apoptotic cell death. Proc. Natl. Acad Sci. USA 88:
6961.
30. Oltvai Z.N., C.L. Miliman and S.J. Korsmeyer.
1993. Bcl-2 heterodimerizes in vivo with a conserved homolog Bax that accelerates programmed
cell death. Cell 74:609.
31. Tomknva H., W. Fujimoto and J. Arata. 1998.
Expression of the Bcl-2 homologue Bax in normal
human skin, psoriasis vulgaris and non-melanoma
skin cancers. Eur. J. Dermatol. 8:256.