Báo cáo khoa học: Effect of 5-lipoxygenase inhibitor MK591 on early molecular and signaling events induced by staphylococcal enterotoxin B in human peripheral blood mononuclear cells doc

Effect of 5-lipoxygenase inhibitor MK591 on early
molecular and signaling events induced by staphylococcal
enterotoxin B in human peripheral blood mononuclear
cells
Chanaka Mendis
1
, Katherine Campbell
1
, Rina Das
2
, David Yang
3
and Marti Jett
2
1 Department of Chemistry and Engineering Physics, University of Wisconsin-Platteville, WI, USA
2 Department of Molecular Pathology, Walter Reed Army Institute of Research, Silver Spring, MD, USA
3 Department of Chemistry, Georgetown University, Washington, DC, USA
Staphylococcal enterotoxin B (SEB) is one of the many
exotoxins produce by Staphylococcus aureus and is
implicated in inducing diarrhea, vomiting, muscle
numbness, possible involvement in autoimmune disor-
ders and lethal shock [1]. The massive impact of T cell
activation, proliferation, and cytokine production by
CD4
+
T cells via specific V
b
elements of T cell antigen
receptor [2] has prompted a number of investigations
to focus on the intricate signaling activities of SEB.
Even though the molecular events of SEB-induced

lethal shock in human peripheral blood mononuclear
cell (PBMCs) are not very apparent, the actual
response of lethal shock is expected to herald by
changes in signaling pathways [3,4]. In mammalian
cells, a variety of stimuli generate intracellular
responses that converge on a limited number of compo-
nents of multiple pathways [5]. Mitogen-activated pro-
tein kinase (MAPK) cascades together with arachidonic
Keywords
cross-talk; MK591; pathway inter-
connectors; signal transduction;
staphylococcal enterotoxin B
Correspondence
C. Mendis, Department of Chemistry and
Engineering Physics, University of
Wisconsin-Platteville, 308, Ottensman Hall,
1 University Plaza, Platteville, WI, USA
Fax: +1 608 342 1559
Tel: +1 608 342 1692
E-mail:
(Received 5 December 2007, revised 9 April
2008, accepted 11 April 2008)
doi:10.1111/j.1742-4658.2008.06462.x
Staphylococcal enterotoxin B (SEB) has been the focus of a number of stu-
dies due to its ability to promote septic shock and a massive impact on the
human immune system. Even though symptoms and pathology associated
with SEB is well known, early molecular events that lead to lethality are
still poorly understood. Our approach was to utilize SEB induced human
peripheral blood mononuclear cells (PBMCs) as a prototype module to
further investigate the complexity of signaling cascades that may ultimately

lead to lethal shock. Our study revealed the activation of multiple divergent
intracellular pathways within minutes of SEB induction including compo-
nents that interconnect investigated pathways. A series of performed inhibi-
tor studies identified a specific inhibitor of 5-LO (MK591), which has the
ability to block JNK, MAPK, p38kinase and 5-LO signaling-cascades and
drastically reducing the activity of pro-inflammatory cytokine TNF-a.
Further evaluation of MK591 utilizing cell proliferation assays in PBMCs,
human proximal tubule cells and in vivo studies (monkey) showed a
decrease in cell proliferation. The inhibitory effect of MK591 was recon-
firmed at a genetic level through the utilization of a set of SEB specific
genes. Signaling activities, inhibitor studies, cellular analysis and gene
expression analysis in unison illustrated the significance of pathway inter-
connectors such as 5-LO as well as inhibiting such inter-connectors (using
MK591) in SEB induced human PBMCs.
Abbreviations
5-LO, 5-lipoxygenase; HIF, hypoxia-inducible factor; IL, interleukin; JNK, c-Jun N-terminal kinase; MAPK, mitogen-activated protein kinase;
p38kinase, p38-mitogen activated protein kinase; PBMC, peripheral blood mononuclear cell; REPTC, renal epithelial proximal tubular cell;
SEB, staphylococcal enterotoxin B; TNF, tumor necrosis factor.
3088 FEBS Journal 275 (2008) 3088–3098 ª 2008 The Authors Journal compilation ª 2008 FEBS
acid metabolic cascades comprise one of the major sig-
naling systems efficiently utilized by cells to transmit
and integrate a plethora of intracellular activities [6].
Although the well regulated signal transduction is cru-
cial for typical cell behavior, aberrant signaling often
leads to diverse pathological consequences. Even
though cross-talk is not a novel concept, the induction
of multiple signal cascades and the inter-connectivity of
SEB-induced cascades in human PBMCs have not been
examined in detail. To gather evidence vital in identify-
ing early signaling events triggered by super-antigen

SEB in lethal shock, we examined the SEB-induced
signaling and cross-talk activities of 5-lipoxygenase
(5-LO), MAPK, c-Jun N-terminal kinase (JNK) and
p38-mitogen activated protein kinase (p38kinase)
pathways.
A number of signaling cascades have been docu-
mented in which a series of proteins activate and regu-
late one another in a sequential and cooperative
fashion [7,8]. One such protein kinase cascade known
as MAPK is activated in cells responsive to various
stimuli, predominantly growth factors [9]. MAPKs
have a molecular mass of 40–44 kDa and are activated
by the phosphorylation of both threonine and tyrosine
residues conserved in the threonine-glutamate-tyrosine
motif [10]. P38kinases are induced by a plethora of acti-
vators including and not limited to UV light, heat,
osmotic shock, inflammatory cytokines [tumor necrosis
factor (TNF)-a and interleukin (IL)-1] and growth fac-
tors (colony-stimulating factor-1) [11–13]. Similar to
p38kinase and MAPK, JNK contains a dual phosphor-
ylation motif and is induced by stress-inducing agents
or pro-inflammatory cytokines [14]. The arachidonic
acid signaling pathway, which has been studied in great
detail, is known to up-regulate inflammatory responses
and 5- and 15-hydroxyeicosatetraenoic acid metabolites
in human T cells and human lymphocytes, respectively
[6]. Products of the arachidonic acid metabolizing
enzyme 5-LO have been shown to stimulate the growth
of several types of cancers, whereas 5-LO activating
protein inhibitor MK-886 has shown to inhibit cell

growth in a dose- and time-dependent manner in a gas-
tric cancer cell line [15]. Recent evidence of arachidonic
acid signaling activities has indicated possible cross-talk
of 5-LO and cyclooxygenase-2 through the cysteinyl
leukotriene receptor 2 in endothelial cells [16].
Superantigens are known to activate large families of
T cells based on expression of the Vb chain of the
T cell receptor, which in turn increases cell prolifera-
tion and pro-inflammatory cytokine secretion [1]. The
present study aimed to investige the early signaling
events induced by SEB in human PBMCs to determine
the pathway inter-connectors. We hypothesized that
such signaling inter-connectors can be effectively
targeted to terminate ⁄ reduce aberrant signaling and cel-
lular activities that may ultimately lead to SEB-induced
lethal shock. Our goal was to evaluate multiple signal-
ing cascades induced by SEB and target pathway inter-
connector 5-LO to inhibit unwanted cellular activities.
As multiple investigations have effectively utilized
systematic examination of gene expression profiles by
stimulants to reveal qualitative and quantitative differ-
ences that may ultimately lead to possible mechanisms
of action [17,18], we further evaluated the inhibitory
effect of 5-LO specific inhibitor (MK591) at a genetic
level by analyzing the gene expression pattern of a set
of SEB specific genes that explain some of the
SEB-induced symptoms via gene functions.
Results
Effect of SEB on multiple signal transduction
pathways in human PBMCs

Even though the ability of SEB to induce lethal shock
is known, the mechanism of its action remains unclear.
In the present study, we analyzed the activation of
various proteins kinases in response to SEB aiming to
better understand the intricacy of signaling activities.
Effect on MAPK
Human PBMCs at a cell density of 2.5 millionÆmL
)1
were treated with 100 ngÆmL
)1
of SEB for different
time periods and MAPK phosphorylation was quanti-
tated as described in the Experimental procedures.
When stimulated with SEB, MAPK phosphorylation
(activation) was visible by as early as 1 min, showed
maximum activity after 5 min (2.8-fold) and finally
returned to control levels by 60 min (Figs 1 and 2).
Effect on p38kinase
Human PBMCs were treated with SEB and the activa-
tion of p38kinase was measured as described in the
Experimental procedures. An initial burst of activation
of this stress-induced kinase was seen by as early as
1 min after exposure (two-fold), followed by a slight
increase up to 5 min (2.25-fold) and reached control
levels by 60 min (Figs 1 and 2).
Effect on JNK
As SEB is known to induce a stress response in human
PBMCs, we examined the effect of SEB on the activity
of stress-induced kinase (JNK), also known as stress
C. Mendis et al. Effects of SEB on human PBMCs

FEBS Journal 275 (2008) 3088–3098 ª 2008 The Authors Journal compilation ª 2008 FEBS 3089
activated protein kinase, using two different techniques
(immunoblots and kinase assays). Our immunoassay
results indicated a rapid five-fold activation of JNK
within 5 min of SEB exposure (Figs 1 and 2). A two-
fold JNK activity was observed even at 60 min. Even
though SEB-induced samples analyzed for JNK activa-
tion using kinase assays did show higher activation
than control levels, high background levels interfered
with any proper quantification of the activity (Fig. 1).
Effect on 5-LO
Our experiments confirmed a time-dependent activa-
tion of 5-LO, showing a maximum activity of two-fold
within 5 min and sustaining activity levels slightly
above control levels at 60 min (Figs 1 and 2).
Comparison of SEB triggered signaling cascades
Comparison of multiple signal transduction pathways
revealed that all investigated pathways showed a rela-
tively similar time-dependent protein activation pat-
tern, with a few differences unique to each key
pathway (Fig. 2). All four analyzed components
(p38kinase, 5-LO, JNK and MAPK) showed maxi-
mum activity by 5 min. At 5 min, JNK activity was at
least two-fold higher than the activity of the rest of the
components and continued to show activity over con-
trol levels (compared to other components) throughout
the study period (0–60 min). JNK and 5-LO showed
2.25-fold and 1.5-fold activity, respectively, at 60 min.
Effect of various inhibitors on SEB triggered
signaling cascades

Previous experiments carried out in our laboratory
have shown the involvement of eicosanoids in SEB-
induced shock in human lymphoid cells. In the present
study, we expanded the investigation to include inhibi-
tors of various signal transduction pathways such as
lipoxygenase inhibitors (MK886, MK591 and curcu-
min), p38kinase inhibitor (SB203580), JNK inhibitor
(SP600125) and MAPK inhibitor (PD98059) to observe
the effect in human PBMCs. As all investigated protein
phosphorylation data indicated a maximum expression
at 5 min, we chose 5 min as the time point to investi-
gate further the effect of all inhibitors. All inhibitors of
the lipoxygenase pathway were able to block SEB-
induced MAPK and p38kinase activation but only had
a partial inhibitory effect on JNK activation (Table 1).
MAPK inhibitor PD98059 had a negligible effect
on SEB-induced p38kinase, JNK and 5-LO activation,
whereas p38kinase inhibitor SB203580 clearly blocked
the induced MAPK and JNK activity, but did not
effect the 5-LO activity. Interestingly, JNK inhibitor
SP600125 (1,9-pyrazoloanthrone) demonstrated attenu-
ation of SEB-induced MAPK expression yet had no
effect on p38kinase or 5-LO activation.
Fig. 2. Comparison of signaling pathway activation profiles. A com-
parison of the time-dependent activation of 5-LO, MAPK, JNK and
p38kinase is shown. Activation of samples at multiple time points
(1, 5, 30 and 60 min) was quantitated using
NIH IMAGE software and
data are shown as triplicates with the respective SD values.
Fig. 1. SEB-induced activation of signaling cascades in human

PBMCs. Human PBMCs at 2.5 · 10
6
cellsÆmL
)1
were treated with
SEB and the activation of each of the key elements was quantitated
by immunoassays as described in the Experimental procedures.
Phosphorylation of all samples except JNK were analyzed using im-
munoassays whereas JNK was quantitated using both kinase and
immune-assays; 0, 1, 5, 30 and 60 refer to the time (min). PBMCs
were exposed to SEB. 0, PBMCs not exposed to SEB but subjected
all other experiment parameters (control sample).
Effects of SEB on human PBMCs C. Mendis et al.
3090 FEBS Journal 275 (2008) 3088–3098 ª 2008 The Authors Journal compilation ª 2008 FEBS
Effect of inhibitors on SEB-induced TNF-a
induction
The investigation then focused on analyzing the
effect of the same set of inhibitors on TNF-a induc-
tion (Table 1). The observed induction in SEB trea-
ted cells (22-fold) was drastically reduced in cells
treated with various inhibitors. Of the different
inhibitors, SB203580 showed the highest inhibitory
effect (1.5-fold) whereas MK886 and MK591 (5-LO
inhibitors) showed a slightly lower inhibition than
SB203580. Even though all the inhibitors (Table 1)
used in this study had a somewhat inhibitory effect
on TNF-a expression, the effect of both PD98059
and SP600125 was minimal compared to the other
inhibitors.
Effect of inhibitors on SEB-induced PBMC

proliferation
Figure 3 illustrates the concentration effect of SEB on
human PBMC proliferation. We observed that prolifera-
tion was directly proportional to SEB concentration in
the range 10–110 ngÆmL
)1
. For SEB concentration in
the range 0–10 ngÆmL
)1
, a slight drop in cell prolifera-
tion was observed whereas SEB concentrations higher
than 110 ngÆmL
)1
showed an increase in cell prolifera-
tion at a much lower rate. Both MK591 and SB203580
were able to block SEB-induced cell proliferation
(Table 2).
Effect of SEB and 5-LO inhibitor MK591 on
human renal epithelial proximal tubular cells
(REPTC) proliferation
SEB-treated REPTC proliferation was partially inhib-
ited by MK591 (33%) whereas SB203580 had no effect
on REPTC proliferation (Table 2).
Table 1. Effect of various inhibitors on SEB-induced signaling pathways. Human PBMCs (2.5 · 10
6
cellsÆmL
)1
) were treated with the
respective inhibitor for 30 min at 37 °C prior to a 5 min stimulation with SEB at 37 °C. All inhibitors were used at 20 l
M except SB203580

and SP600125, which were used at 10 l
M. All values are shown as a percent of control samples. TNF-a induction is referred to as the acti-
vation of TNF-a when induced by100 ngÆmL
)1
SEB for 5 min. ND, not determined.
JNK p38kinase MAPK 5-LO TNF-a induction
SEB 500 ± 29.5 250 ± 15 279 ± 28.4 200 ± 35 2200 ± 22.1
SB-203580 178.4 ± 18.6 34 ± 2.2 38.5 ± 1.5 175 ± 7.6 150 ± 3.3
MK591 212.5 ± 78.5 76 ± 5.3 71.5 ± 4.5 65 ± 25 190 ± 5.0
MK886 173.5 ± 19.5 55 ± 3.5 52.5 ± 2.5 ND 220 ± 1.1
Curcumin 168 ± 22 43 ± 6.2 39.8 ± 4 ND ND
SP600125 74.2 ± 2.5 337 ± 4.3 199 ± 10 158 ± 4.5 1550 ± 40
PD98059 337 ± 25 225 ± 23 31.0 ± 13 320 ± 13 1500 ± 30
SEB (ng·ml
–1
)
0 50 100 150 200 250
3
H labeled thymidine
incorporation (in cpm)
0
2000
4000
6000
8000
10
000
12
000
14

000
16
000
Fig. 3. Concentration dependence of SEB on PBMC proliferation. A
typical SEB-induced time-dependent cell proliferation pattern is
shown. Human PBMCs at 2.5 · 10
6
cellsÆmL
)1
density were incu-
bated in 96-well plates in the presence of SEB at various concen-
trations (range: 1–200 ng) and incorporated radioactivity, which is
directly proportional to cell proliferation, was quantitated as
described in the Experimental procedures.
Table 2. Effectiveness of MK591 on SEB stimulated cellular activi-
ties. Each cell type was treated with MK591 for 30 min at 37 °C prior
to the 5 min stimulation with SEB at 37 °C. All in vitro experiments
were performed at concentrations of 10 l
M SB203580 and 20 lM
MK591. All in vivo experiments were performed at 10 mMÆkg
)1
(SB203580) and at 20 mMÆkg
)1
(MK591) and the values are shown
as a percent of control samples. ND, not determined. Monkeys were
exposed to saline or SEB (15 lgÆkg
)1
) by aerosol, with or without the
treatment of the respective inhibitor and whole blood was collected
after 30 min of exposure. All cell proliferation experiments were

carried out as described previously [31] and TNF-a experiments were
carried out as described in the Experimental procedures.
Stimulant
T cell
proliferation
in PBMC (%)
Proliferation
of REPTC
(%)
T cell
proliferation in
monkeys (%)
TNF-a in
monkeys
(%)
SEB 100 100 100 100
SB203580 0.59 ± 0.04 100 ± 3.6 ND ND
MK591 14.67 ± 1.22 67.5 ± 2.5 45.2 ± 2.8 38.5 ± 2.5
C. Mendis et al. Effects of SEB on human PBMCs
FEBS Journal 275 (2008) 3088–3098 ª 2008 The Authors Journal compilation ª 2008 FEBS 3091
Effect of 5-LO inhibitor MK591 on monkeys
challenged with SEB
Findings of in vitro studies of human PBMCs were veri-
fied in PBMCs isolated from aerosol SEB challenged
monkeys. The animals were treated with a sublethal
dose of the toxin, which caused incapacitation. Each
monkey was used as its own control in a saline sham
experiment. TNF-a and T cell proliferation was assayed
from blood samples as early as 30 min post exposure.
MK591 was able to inhibit the expression of TNF-a

and T cell proliferation in vivo samples (Table 2).
Effect of MK591 on the expression of a set of
SEB specific genes
To examine our hypothesis of effectively targeting a
pathway inter-connector to block the SEB-induced sig-
naling cascades, we evaluated MK591 at a genetic level
by analyzing a set of SEB specific genes. Genes that
were chosen based on functional significance were ana-
lyzed by performing RT-PCR at 2 h and 16 h
(Table 3). Three genes [for cathepsin L, IL-17 and gua-
nylate binding protein (GBP)-2] that are up regulated
by SEB were all down regulated by MK591 at 16 h.
All three genes showed a reduction in the activation as
early as 2 h whereas the gene for IL-17 showed a four-
fold down regulation at 2 h. CTAP-III, which is down
regulated by SEB both at 2 h and 16 h, showed an up
regulation at 16 h with MK591. The effect of MK591
on proteoglycan V
0
was minimal at both 2 h and 16 h.
Discussion
Despite many decades of extensive investigation, nei-
ther the exact pathomechanism, nor the intricate
nature of SEB-induced signaling activities are well
understood. The mode of signal transduction is vital in
properly understanding the multifunctional role of
staphylococcal enterotoxin as a super-antigen [1]. A
recent study has revealed phosphatase-mediated
crosstalk between MAPK signaling pathways in the
regulation of cell survival [19]. Although crucial to

understanding cell survival, the investigation does not
provide any information about the importance of tar-
geting pathway inter-connectors. The present study
aimed to investigate a group of multiple signal trans-
duction pathways in human PBMCs (i.e. the first line
of defense encountered by foreign substances) using a
single stimulant (SEB) to better comprehend and visu-
alize the complexity of signal transduction pathways.
Protein activation experiments indicated the ability of
SEB to induce multiple signaling pathways as well as
high activation levels of JNK. This result lead us to
hypothesize that SEB may utilize a single pathway to
transmit majority of the signal but can use multiple
cascades at varying strengths depending on the time,
stimulant and availability of pathways. The concept
has many interesting consequences; specifically,
whether an extracellular stimulant may lead to aber-
rant cellular behavior. In such a case, it is important
to look at all possible signal transduction pathways
when deducing the key element or elements that may
abolish such a signal.
To further investigate the activation of multiple sig-
nal transduction pathways and to explore the exis-
tence of pathway inter-connectors, we performed a
series of inhibitor studies targeting MAPK, p38kinase,
5-LO and JNK. Previous investigations carried out in
our laboratory have indicated the inhibition of SEB-
induced arachidonic acid and MAPK activation in
human lymphoid cells by 5-LO inhibitors such as
curcumin, NDGA and MK881 (C Mendis, R Das,

D Yang and M Jett, unpublished results) whereas
Table 3. Effect of MK591 on the expression of a set of SEB specific genes. A set of SEB specific genes previously identified by differential
display-PCR and RT-PCR (18) were further examined using 5-LO inhibitor MK591. After designing specific primers for each gene of interest,
RT-PCR reactions were performed on samples treated with SEB (100 ngÆmL
)1
) with or without the inhibitor (MK591) for 2 h and 24 h as
described in the Experimental procedures. Identical total RNA samples were used for all analysis, and the bands of PCR products were digi-
tized after normalizing with a house keeping gene (18S rRNA), and quantitated using
NIH IMAGE software. All reactions were repeated twice
and the results are reported as mean ± SD values relative to the control. CTAP-III, connective tissue activating protein III; CTSL, cathepsin L
transcript variant1 mRNA; Prot-V
0
, chondroitin sulfate proteoglycan versican V
0
splice-variant precursor peptide. ND, not determined.
Gene name Sequence ID SEB (2 h) SEB (16 h) MK591 (2 h) MK591 (16 h)
CTAP-III BC028217 )1.73 ± 0.01 )2.21 ± 0.03 1.18 ± 0.03 1.42 ± 0.10
CTSL NM-001912 6.13 ± 0.22 4.9 ± 0.15 )1.01 ± 0.10 )1.20 ± 0.01
HIF-1 AF050127 2.3 ± 0.05 2.79 ± 0.03 1.00 ± 0.02 )1.45 ± 0.25
GBP-2 M55543 7.57 ± 0.08 3.26 ± 0.07 )2.0 ± 0.005 )4.80 ± 0.10
IL-17 NM_002190 2.2 ± 0.03 4.97 ± 0.23 1.12 ± 0.02 )4.75 ± 0.03
IL-6 M29150 52.26 ± 1.0 31.6 ± 0.26 ND ND
Prot-Vo U16306.1 ND )2 ± 0.2 )2.45 ± 0.06 )1.65 ± 0.02
Effects of SEB on human PBMCs C. Mendis et al.
3092 FEBS Journal 275 (2008) 3088–3098 ª 2008 The Authors Journal compilation ª 2008 FEBS
cross-talk of p38kinase and MAPK has been
observed in d-glucose-induced cell death [20]. Our
results demonstrated the ineffectiveness of targeting
MAPK inhibitor PD98059 because the inhibitor had
no effect on the activities of JNK, 5-LO or

p38kinase, which is somewhat similar to the results
obtained for PD98059 in IgG-opsonized sheep eryth-
rocyte-stimulated polymorphonuclear leukocytes [21].
The ability of p38kinase inhibitor SB203580 to selec-
tively and markedly inhibit SEB stimulated MAPK
and JNK activation suggested interconnectivity of
p38kinase, MAPK and JNK pathways, but no indi-
cation of any effect on the 5-LO pathway. Even
though SEB-induced JNK phosphorylation far
exceeded the activation of other pathways, JNK spe-
cific inhibitor SP600125 was only able to inhibit
MAPK activation.
The above results led us to continue our inhibitor
study targeting the 5-LO pathway, specifically
MK591. Targeting the 5-LO pathway using MK591
attenuated all previously observed activation of the
MAPK, p38kinase and JNK pathways. The multiple
pathway inhibitory effects of MK591 confirmed the
importance of 5-LO as a pathway inter-connector in
SEB-induced human PBMCs. Similar results were
observed in vascular smooth muscle cells, in which
JNK-1 and MAPK were induced by arachidonic acid
in a time- and concentration-dependent manner [22].
Figure 4 illustrates the inter-connectivity of signaling
cascades that participat7e in SEB-induced human
PBMCs.
Both human and animal models of endotoxin-
induced shock are similar, and both show an elevation
of pro-inflammatory cytokines (e.g. TNF-a levels)
within a few hours of induction, followed by a decline

to undetectable levels [23]. Inflammatory cytokine
TNF-a is produced in SEB-induced human PBMCs to
a 50-fold greater extent than in untreated cells [24],
modulating a wide variety of cellular processes such as
organ dysfunction and systematic shock [25–28]. High
induction of TNF-a SEB in human PBMCs prompted
us to utilize TNF-a as a cellular marker to further
investigate the effect of MK591. TNF-a levels that
were drastically elevated by 100 ngÆmL
)1
SEB in
human PBMCs were reduced back to control levels by
5-LO activating protein specific inhibitor MK591
(Table 1). Although intriguing, these results were based
on in vitro experiments and did not reveal any infor-
mation under physiological conditions. Further investi-
gation of MK591 using an in vivo model (monkey) did
show a similar pattern to that observed in vitro experi-
ments, but not at similar levels of inhibition.
One of the major characteristic of SEB-induced
human PBMCs is the ability to show massive T cell
proliferation. Once exposed, PBMCs show a minimum
30% increase in T cell production. We took advantage
5-LO
X
X
X
MAPK
X
X

TNF-α
X
GBP-2
Vasodilation
X
Cell
proliferation
X
Vascular
p
ermabilit
y

X
Tissue
degredation
X
HIF-1
Respiratory
distress
X
IL-17
Inflammation
SEB complex
p38
JNK
CA-L
CTAP-III
Fig. 4. Schematic diagram of the inhibitory
effect of MK591 on SEB-induced human

PBMCs. The inter-connectivity of SEB-
induced signaling cascades as well as the
effectiveness of 5-LO inhibitor MK591 is
shown, in addition to the inhibitory effect of
MK591 on a set SEB specific proteins and
genes. All genes are indicated by a double
outline around the name together with the
corresponding functionally related symptom
of SEB. All proteins are indicated by a single
outline around the name. Symbol ‘X’ indi-
cates inhibition of a protein or gene activity.
CTAP-III, connective tissue activating protein
III; CA-L, cathepsin L transcript variant1
mRNA.
C. Mendis et al. Effects of SEB on human PBMCs
FEBS Journal 275 (2008) 3088–3098 ª 2008 The Authors Journal compilation ª 2008 FEBS 3093
of the high activation of T cells and used it as our sec-
ond parameter to evaluate the effectiveness of MK591.
Our results clearly indicated the ability of MK591 to
efficiently inhibit SEB-induced T cell proliferation
in vitro, and the 50% T cell inhibition observed in a
monkey model further solidified the high potency of
the inhibitor (Table 2 and Fig. 4). Evidence from
experiments performed using animal models implicates
the kidney in general, and the REPTCs in particular,
as the major target of SEB uptake. It is also clear that
70% of injected SEB is accumulated in proximal
tubule cells within 2 h [29–31] of stimulation. Based on
the evidence of kidney involvement in blood pressure
regulation, it is not unreasonable to hypothesize that

damage to the renal epithelium caused by a vascular
shock-inducing agent, such as SEB (which possesses
the ability to interact directly with the vascular tone-
regulating kidney cells), may contribute to the develop-
ment of systemic shock (unpublished results). The
functional significance of REPTCs prompted us to
further investigate whether MK591 had any inhibitory
effect on REPTCs; if this is the case, we believe that
targeting 5-LO with MK591 may even help reduce sys-
temic shock. The ability of MK591 to effectively inhi-
bit T cell as well as REPTC (by 33%) proliferation
indicates for the first time the effectiveness of targeting
a pathway inter-connector (5-LO). It is possible that
the ability of the target to inter-connect multiple signal
pathways allowed it to influence cellular activity in the
two crucial cell types that are most effected by SEB
stimulation. SB203580, an inhibitor of pathway inter-
connector p38kinase, had a similar inhibitory effect on
T cell proliferation but did not inhibit REPTCs, which
is known to function as a major target of SEB uptake.
The result prompted us to focus further on evaluating
MK591 as a possible inhibitor.
We then investigated the effect of MK591 on the
expression pattern of a set of SEB specific genes that
somewhat explained the symptoms induced by SEB via
gene functions. We believed that the gene analysis
would further complement and validate the inhibitory
effects observed for MK591 at a protein level [18].
Some of the investigated genes are involved in func-
tions such as inflammation (IL-6 and IL-17), tissue

damage ⁄ cardiac dysfunction (cathepsin L), hypoxia
inducible conditions [hypoxia-inducible factor (HIF)-
1], alterations to the physiology of blood vessels (pro-
teoglycan V
0
) and vasodilatation (GBP-2). Our target
MK591 was able to alter the expression of each of the
above SEB specific genes at two time points (2 h and
16 h) as shown in Table 3, except proteoglycan V
0
(which did not show alteration at both time points)
and IL-6 (uncompleted experiments). All of the above
genes have been thoroughly investigated and contrib-
ute to a profile that is specific for SEB [18]. Even
though RT-PCR is considered to be a semi-quantita-
tive gene quantification method, the results observed
are significant for two reasons. The purpose of the
analysis was, first, to verify whether MK591 was able
to alter a SEB specific gene profile and, second, to
evaluate the effectiveness of MK591 at both protein
and genetic levels. Our investigation showed the abil-
ity of SEB to utilize multiple signaling cascades to
induce cell proliferation and TNF-a induction, which
ultimately would result in symptoms such as inflamma-
tion, respiratory distress, tissue degradation, vascular
permeability and vasodilation. Our model (Fig. 4)
shows the effect of targeting multi-pathway inter-con-
nector 5-LO by a specific inhibitor (MK591) and sum-
marizes the alterations observed at the genetic level
and the inhibitory effects observed at the protein

level. The model signifies the importance of targeting
components such as 5-LO that have the ability to
inter-connect multiple signaling cascades such as
JNK, p38kinase and MAPK. We believe that targeting
5-LO may have a wider impact on SEB-induced cellu-
lar events than targeting a component such as
p38kinase, which influences only JNK and MAPK and
not 5-LO. The inability of p38kinase inhibitor to block
5-LO activity may have played a role in the inability
of SB203580 to effectively block proliferation of
REPTC.
Potential signal blockers have to be examined
thoroughly because time and again these targets have
proven to be unsuccessful in clinical trials. This can
be due to a number of reasons, including the inability
of the targets to completely block the signaling
cascades due to the leaking effect or the ability of the
signal to use alternative signaling cascades upon the
inhibition of the primary signaling pathway. We are
confident that the ability of our target to function as a
pathway inter-connector may have a positive influence
on blocking the SEB-induced signaling activities. Fur-
thermore, it is important to reconfirm the inhibition
of signaling activities observed for SEB to somewhat
similar toxins such as lipopolysaccharides to better
understand whether the inter-connector has a universal
function. We have previously compared the gene
expression pattern of SEB and lipopolysaccharides and
know that each profile is remarkably different than the
other, even though both tend to induce similar

symptoms in exposed patients [18]. We are currently
analyzing the protein activation pattern of the two
toxins to differentiate specific signaling activities as
well as identify similar signaling activities (data not
included).
Effects of SEB on human PBMCs C. Mendis et al.
3094 FEBS Journal 275 (2008) 3088–3098 ª 2008 The Authors Journal compilation ª 2008 FEBS
Knowledge on the early molecular events investi-
gated in the present study will have a tremendous
impact on determining the effects of SEB on human
PBMCs and may complicate the finding of a therapeu-
tic target even more. However, the emergence of effec-
tively targeting convergence points (5-LO) in this
intricate web of pathways has helped with respect to
finding targets that may have potential therapeutic sig-
nificance. The early identification of these key elements
will also help in determining the potential exposure to
SEB before a patient experiences significant damage
and, undoubtedly, will be key in designing strategies to
block or even limit aberrant signaling cascades by
inhibiting pathway inter-connectors such as 5-LO.
Even though, collectively, the above data comprise an
attempt to elucidate the complexity of signal trans-
duction pathways induced by SEB, the cross-talk of
pathways, and the effectiveness of inhibiting a pathway
inter-connector utilizing trademark cellular and genetic
activation patterns associated with SEB, further work
is essential, both in vivo and in vitro, to pinpoint the
efficiency and the effectiveness of MK591.
Experimental procedures

Cells and cell cultures
Exposure of human PBMCs to SEB in vitro
Human PBMCs were collected from leukopacks from normal
donors as described previously [31]. Human PBMCs with or
without SEB and ⁄ or inhibitors of interest were used at a final
density of 2.5 · 10
6
cellsÆmL
)1
in RPMI 1640 medium sup-
plemented with 10% human AB serum. Prior to treatment
with 100 ngÆmL
)1
of SEB for different time periods (1, 5, 30
and 60 min for protein extraction and 2 h and 16 h for RNA
extraction), PBMCs were incubated with an inhibitor of
interest at room temperature for 30 min at 20 lm, except
SP600125 and SB203580, which were used at 10 lm under
5% CO
2
at 37 °C. Total RNA were isolated using Trizol
reagent (Life Technologies, Grand Island, NY, USA) accord-
ing to manufacturer’s instructions. 0; PBMCs not exposed to
SEB but subjected to all other experiment parameters.
Exposure of REPTC to SEB in vitro
Primary cultures of normal REPTCs from a 34-year-old
African-American male donor were obtained from Clonetics
(Walkersville, MD, USA). The cultures were maintained in
75 cm
2

or 162 cm
2
tissue culture flasks (Corning Corp.,
Corning, NY, USA) at 37 °Cin5%CO
2
atmosphere in renal
epithelial growth medium (Clonetics), supplemented with
20 unitsÆmL
)1
penicillin G sodium, 20 lgÆmL
)1
streptomycin
sulfate, 20 lgÆmL
)1
kanamycin sulfate, and minimal essential
vitamin solution (Gibco BRL, Grand Island, NY, USA).
The cultures were maintained for up to seven passages and
used for experiments after the second passage. Cells were
harvested and total RNA were isolated using Trizol reagent
(Life Technologies) according to manufacturer’s instructions.
Exposure of monkeys to aerosol challenge with SEB
Administration of SEB ⁄ saline (control), collection of
blood samples and processing of blood samples were
performed as described previously [31]. Monkeys were
exposed to saline or SEB (15 lgÆkg
)1
) by aerosol, with or
without the treatment of inhibitor SB203580 (10 mmÆkg
)1
)

and MK591 (20 mmÆkg
)1
), and whole blood was collected
30 min after exposure. PBMCs fractionated through a
Ficoll gradient were then utilized in proliferation assays,
TNF-a assays and total RNA isolations.
Toxin
SEB from S. aureus strain 10-275 purified by the method of
[32] was provided in the lyophilized form by the US Army
Research Institute of Infectious Diseases (Fert Detrick,
Frederick, MD, USA). The stock solution was prepared in
sterile, pyrogen-free deionized water at a concentration of
5mgÆmL
)1
and stored at –80 °C. When used, the stock
solution was diluted with cell culture medium to the desired
concentration.
Immunoblotting and western blotting
The cells were lysed after a brief sonication and incu-
bated in the lysis buffer containing 20 mm Hepes, 10 mm
EGTA, 40 mm b-glycerophosphate, 2.5 mm MgCl
2
,1mm
dithiothreitol, 150 mm NaCl, 2 mm sodium orthovana-
date, 1 mm phenylmethanesulfonyl fluoride, 10 lgÆmL
)1
leupeptin, 10 lgÆmL
)1
aprotinin, 1% NP40 and 0.5%
deoxycholate. Equal amounts of protein were either

resolved on 10% SDS ⁄ PAGE gels (Novex, Cleveland,
OH, USA) or immunoprecipitated overnight with an anti-
body of interest at 4 °C, incubated with protein-G sepha-
rose at 4 °C for 2 h, washed twice in lysis buffer
and resolved on 10% SDS ⁄ PAGE gels (Novex). Analyzed
gels were transferred to poly(vinylidene difluoride) mem-
branes (Invitrogen, Grand Island NY, USA), probed with
a specific antibody of interest (Santa Cruz Biotechnology,
Santa Cruz, CA, USA), except JNK inhibitor which was
purchased from Calbiochem (La Jolla, CA, USA). All
immune assays were detected using an ECL detection kit
(Amersham, Piscataway, NJ, USA), and quantitated using
nih image software, version 1.3 (NIH Image, Bethesda
MD, USA). All immunoblots and western blots were
repeated at least three times to check for reproducibility
and data are reported the mean ± SE.
C. Mendis et al. Effects of SEB on human PBMCs
FEBS Journal 275 (2008) 3088–3098 ª 2008 The Authors Journal compilation ª 2008 FEBS 3095
Proliferation assays
Human PBMCs were isolated and purified as described pre-
viously [31], except that human PBMCs were used at
2.5 · 10
6
cellsÆmL
)1
in RPMI 1640 medium supplemented
with 10% human AB serum with or without the inhibitors
MK591 (20 lm) and SB203580 (10 lm) at SEB concentra-
tions in the range 1–200 ngÆmL
)1

. The cultures were incu-
bated for 15 h with 1 lCu of methyl-[
3
H]thymidine per
sample. Incorporated radioactivity was determined using a
microplate scintillation counter (Packard, Meriden, CT,
USA). All proliferation assays were repeated at least three
times for reproducibility. Human REPTCs were also used
at 2.5 · 10
6
cellsÆmL
)1
. All in vivo experiments subjected to
TNF-a assays were carried out at 10 mmÆkg
)1
(SB203580)
and at 20 mmÆkg
)1
(MK591).
Kinase assays
Cell extracts were immunoprecipitated with p-JNK anti-
body (Santa Cruz Biotechnology) and washed with a kinase
buffer containing 20 mm Hepes (pH 7.6), 20 mm MgCl
2
,
20 mm b-glycerophosphate, 1 mm sodium orthovanadate,
2mm dithiothreitol and 20 lm ATP. Kinase assays were
performed using 2 lg of GST-c-Jun as substrate and
1 lCi Y [
32

P]ATP and quantitated using nih image
software. All assays were performed at least three times to
verify reproducibility.
ELISA immunoassays
Specific kits to quantitate TNF-a was purchased from
R&D systems and performed using equal amounts of pro-
teins obtained from human PBMCs or REPTCs with or
without treatment of 100 ngÆmL
)1
SEB and an inhibitor
according to the manufacturer’s instructions (Quantikine
R&D systems, Minneapolis, MN, USA). All in vitro experi-
ments subjected to TNF-a assays were performed at con-
centrations of 10 lm SB203580 and 20 lm MK591. All
in vivo experiments subjected to TNF- a assays were carried
out at 10 mmÆkg
)1
(SB203580) and at 20 mmÆkg
)1
(MK591). Proteins were quantified using a Ceres UV
900-Hdi plate reader (Bio-Tek Instruments Inc., Winooski,
VT, USA). All samples other than the control samples were
incubated with SEB for 5 min.
RT-PCR
RT-PCR analyses were performed using a Superscript
amplification kit (Life Technologies, Gaithersburg, MD,
USA). Housekeeping gene primers (18S) were obtained from
Clontech Corp. (Palo Alto, CA, USA) and all primers were
designed using a primer-design software program named
primer3. cDNA were amplified using AmpliTaq polymerase

(Applied Biosystems, Branchburg, NJ, USA). Genes were
analyzed on 1% agarose gels, visualized through an inhouse
imager and quantified using imagej (NIH Image). All cus-
tom primers except 18S were purchased from Invitrogen
(Carlsbad, CA, USA). All primers were designed to have an
annealing temperature of 60 °C and were subjected to PCR
for 35 cycles in a thermocycler (PerkinElmer, Waltham,
MA, USA). All sequences are indicated in the 5¢ to 3¢ direc-
tion and are abbreviated after the gene name as ‘L’ or ‘R’
for reverse or forward, respectively. CTAP-III-L; CAGCAA
CTCACCCTCACTCA, CTAP-III-R; GTTTGTCCTTTGG
TGGAGGA, CTSL-L; CATTTGCAGGCTCCTTTAGC,
CTSL-R; GGGCAAAGGTTTCCTCTTTC, HIF-1-L;
GAAAGCGCAAGTCCTCAAAG, HIF-1-R; TGGGTAG
GAGATGGAGATGC, GBP-2-L; GGTCCAGTTGCTGA
AAGAGC, GBP-2-R; TGACAGGAAGGCTCTGGTCT,
IL-6-L; TACCCCCAGGAGAAGATTCC, IL-6-R; TTT
TCTGCCAGTGCCTCTTT, PROT-V
O
-L; CCTTTCTGG
GGAAGAACTCC, PROT-V
O
-R; GGTCACATAGGAAG
CGTGGT, IL-17-L; ATTTGCCGAAGAGCCCTCAG,
AND IL-17-R; TTCCAAAGATGTAGCCGCCC.
Acknowledgements
We like to extend our gratitude to Michael Hartl and
Carla Sanchez [Walter Reed Army Institute of
Research (WRAIR), Silver Spring, MD, USA] for
their technical expertise in carrying out protein expres-

sion analysis, Boris Ionin (WRAIR) for his assistance
in working with REPTC and a monkey model and
Mark Hiner (University of Wisconsin Platteville) for
his assistance in optimizing the gene quantitative anal-
ysis technique. We would also like to thank the
University of Wisconsin-Platteville for its internal
funding opportunities (SAIF and PURF grants) and
Professor James Hamilton of the UWP Chemistry and
Engineering Department for developing an imager to
quantitate gene expression. Research was conducted in
compliance with the Animal Welfare Act and other
federal statutes and regulations relating to animals and
experiments involving animals and adheres to princi-
ples stated in the Guide for the Care and Use of
Laboratory Animals, NRC Publication, 1996 edition.
The opinions, interpretations, conclusions and recom-
mendations are those of the author and are not neces-
sarily endorsed by the US Army.
References
1 Jett M, Ionin B, Das R & Neil R (2001) The staphylo-
coccal enterotoxins. The Encyclopedia of Molecular
Medicine. New York, NY: John Wiley & Sons, Inc., 5:
p. 1089–1106.
Effects of SEB on human PBMCs C. Mendis et al.
3096 FEBS Journal 275 (2008) 3088–3098 ª 2008 The Authors Journal compilation ª 2008 FEBS
2 Torres BA, Perrin GQ, Mujtaba MG, Subramaniam
PS, Anderson AK & Johnson HM (2002) Superantigen
enhancement of specific immunity: antibody
production and signaling pathways. J Immunol 169,
2907–2914.

3 Marrack P & Kappler J (1990) The staphylococcal
enterotoxins and their relatives. Science 248, 705–711.
4 Cole B & Wells J (1990) Immunosuppressive properties
of the mycoplasma arthritis T-cell mitogen in vivo: inhi-
bition of proliferative responses to T-cell mitogens.
Infect Immun 58, 228–236.
5 Saxena M, Williams S, Tasken K & Mustelin T (1999)
Crosstalk between cAMP-dependent kinase and MAP
kinase through a protein tyrosine phosphatase. Nat Cell
Biol 1, 305–311.
6 Avis MI, Jett M, Boyle T, Vos DM, Moody T, Treston
MA, Martinez A & Mulshine L (1996) Growth control
of lung cancer by interruption of 5-lipoxygenase-medi-
ated growth factor signaling. J Clin Invest 97, 806–813.
7 Cobb MH, Boulton TG & Robbins DJ (1991) Extracel-
lular signal-regulated kinases: ERKs in progress. Cell
Regul 2, 965–978.
8 Blenis J (1993) Signal transduction via the MAP kinas-
es: proceed at your own RSK. Proc Natl Acad Sci USA
90, 5889–5892.
9 Das R & Vonderhaar BK (1996) Activation of raf-1,
MEK, and MAP kinase in prolactin responsive mam-
mary cells. Breast Cancer Res Treat 40, 141–149.
10 Mertens S, Craxton M & Goedert M (1996) SAP
kinase-3, a new member of the family of mammalian
stress-activated protein kinases. FEBS Lett 383, 273–
276.
11 Rouse J, Cohen P, Trigon S, Morange M, Alonso-
Llamazares A, Zamanillo D, Hunt T & Nebreda AR
(1994) A novel kinase cascade triggered by stress and

heat shock that stimulates MAPKAP kinase-2 and
phosphorylation of the small heat shock proteins. Cell
78, 1027–1037.
12 Han J, Lee D, Bibbs L & Ulevitch J (1994) A MAP
kinase targeted by endotoxin and hyperosmolarity in
mammalian cells. Science 265, 808–811.
13 Freshney NW, Rawlinson L, Guesdon F, Jones E,
Cowley S, Hsuan J & Saklatvala J (1994)
Interleukin-1 activates a novel protein kinase cascade
that results in the phosphorylation of Hsp27. Cell 78,
1039–1049.
14 Kyriakis JM & Avruck J (1996) Sounding the alarm:
protein kinase cascades activated by stress and inflam-
mation. J Biol Chem 271, 24313–24316.
15 Fan XM, Tu SP, Lam SK, Wang WP, Wu J, Wong
WM, Yuen MF, Lin MC, Kung HF & Wong BC
(2004) Five-lipoxygenase-activating protein inhibitor
MK-886 induces apoptosis in gastric cancer through
upregulation of p27kip1 and bax. J Gastroenterol Hepa-
tol 19, 31–37.
16 Lo
¨
tzer K, Jahn S, Kramer C, Hildner M, Nu
¨
sing R,
Funk CD & Habenicht AJ (2007) 5-Lipoxygenase ⁄
cyclooxygenase-2 cross-talk through cysteinyl lenko-
triene receptor 2 in endothelial cells. Prostaglandins and
other Lipid Mediators. 84, 108–115.
17 Boldrick JC, Alizadeh AA, Diehn M, Dudoit S, Liu

CL, Belcher CE, Botstein D, Staudt LM, Brown PO &
Relman DA (2002) Stereotyped and specific gene
expression programs in human innate immune responses
to bacteria. Proc Natl Acad Sci USA 99, 972–977.
18 Mendis C, Das R, Hammamieh R, Royaee A, Yang D,
Peel S & Jett M (2005) Transcriptional response signa-
ture of human lymphoid cells to staphylococcal entero-
toxin B. Genes Immun 6, 84–94.
19 Junttila MR, Li SP & Westermarck J (2008) Phospha-
tase-mediated crosstalk between MAPK signaling path-
ways in the regulation of cell survival. FASEB J 22,
954–965.
20 Nakagami H, Morishita R, Yamamoto K, Yoshimura
S, Taniyama Y, Aoki M, Matsubara H, Kim S, Kaneda
Y & Ogihara T (2001) Phosphorylation of p38 mitogen-
activated protein kinase downstream of bax-caspase-3
pathway leads to cell death induced by high D-glucose
in human endothelial cells. Diabetes 50, 1472–1481.
21 Lenczowski JM, Dominguez L, Elder AM, King LB,
Zacharchuk CM & Ashwell JD (1997) Lack of a role
for Jun kinase and AP-1 in Fas-induced apoptosis. Mol
Cell Biol 17, 170–181.
22 Hii CS, Ferrante A, Edwards YS, Huang ZH, Hartfield
PJ, Rathjen DA, Poulos A & Murray AW (1995) Acti-
vation of mitogen-activated protein kinase by arachi-
donic acid in rat liver epithelial WB cells by a protein
kinase C-dependent mechanism. J Biol Chem 270, 4201–
4204.
23 Martich GD, Danner RL, Ceska M & Suffredini AF
(1991) Detection of interleukin 8 and tumor necrosis

factor in normal humans after intravenous endotoxin:
the effect of antiinflammatory agents. J Exp Med 173,
1021–1024.
24 Yan Z, Yang DC, Neill R & Jett M (1999) Production
of tumor necrosis factor alpha in human T lymphocytes
by staphylococcal enterotoxin B correlates with toxin-
induced proliferation and is regulated through protein
kinase C. Infect Immun 67, 6611–6618.
25 Hale ML, Margolin SB, Krakauer T, Roy CJ & Stiles
BG (2002) Pirfenidone blocks the in vitro and in vivo
effects of staphylococcal enterotoxin B. Infect Immun
70, 2989–2994.
26 Fast DJ, Schlievert PM & Nelson RD (1989)
Toxic shock syndrome-associated staphylococcal and
streptococcal pyrogenic toxins are potent inducers of
tumor necrosis factor production. Infect Immun 57,
291–294.
27 Miethke T, Duschek K, Wahl C, Heeg K & Wagner H
(1993) Pathogenesis of the toxic shock syndrome: T cell
C. Mendis et al. Effects of SEB on human PBMCs
FEBS Journal 275 (2008) 3088–3098 ª 2008 The Authors Journal compilation ª 2008 FEBS 3097
mediated lethal shock caused by the superantigen
TSST-1. Eur J Immunol 23, 1494–1500.
28 Chan KF, Siegel MR & Lenardo JM (2000) Signaling
by the TNF receptor superfamily and T cell homeosta-
sis. Immunity 13, 419–422.
29 Rapoport MI, Hodoral LF & Beisel WR (1967) Influ-
ence of thorotrast blockade and acute renal artery liga-
tion on disappearance of staphylococcal enterotoxin B
from blood. J Bacteriol 93, 779–783.

30 Normann SJ, Jaeger RF & Johnsey RT (1969) Pathol-
ogy of experimental enterotoxemia. The in vivo localiza-
tion of staphylococcal enterotoxin B. Lab Invest 20,
17–25.
31 Jett M, Neil R, Welch C, Boyle T, Bernton E, Hoover
D, Lowell G, Hunt RE, Chatterjee S & Gemski P
(1994) Identification of staphylococcal enterotoxin B
sequences important for induction of lymphocyte prolif-
eration by using synthetic peptide fragments of the
toxin. Infect Immun 62, 3408–3415.
32 Schantz EJ, Roessler WG, Wagman J, Spero L, Dunnery
DA & Bergdoll MS (1965) Purification of staphylococcal
enterotoxin B. Biochemistry 4, 1011–1016.
Effects of SEB on human PBMCs C. Mendis et al.
3098 FEBS Journal 275 (2008) 3088–3098 ª 2008 The Authors Journal compilation ª 2008 FEBS