| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Institute of Medical Microbiology and Hygiene and
IIIrd Department of Internal Medicine, Johannes Gutenberg University, Mainz, Germany; and
Medical Molecular Biology Unit, Office of Research and Development, Faculty of Medicine, Siriraj Hospital, Mahidol University, Bangkok, Thailand
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The human umbilical cord vein endothelial cell line (ECV304), obtained from American Type Culture Collection (ATCC CRL-1998), Manassas, VA, was grown in medium 199 (Life Technologies, Paisley, Scotland) supplemented with 10% FCS (Life Technologies), 2 mM L-glutamine (Life Technologies), 100 U/ml penicillin, and 100 µg/ml streptomycin (Life Technologies) at 37°C in humidified air containing 5% CO2. C6/36, a cell line from Aedes albopictus, and PscloneD, a swine fibroblast cell line, were cultured at 28°C and 37°C, respectively, in L-15 medium (Life Technologies) containing 10% tryptose phosphate broth (TPB) (Sigma, St. Louis, MO), 10% FCS, 100 U/ml penicillin, and 100 µg/ml streptomycin. Untransformed human umbilical vein endothelial cells and human dermal microvascular endothelial cells were purchased from Promocell (Heidelberg, Germany) and were cultured in endothelial cell culture medium with growth supplement and 2% FCS, provided by the supplier. Only cells from passages 2 to 5 were used for experiments.
Preparation of virus stock and virus titration
Dengue-2 virus strain 16681 was propagated in C6/36 cells. Monolayer of cells in 75 cm2-tissue culture flasks (Greiner, Frickenhausen, Germany) received L-15 medium containing 1% FCS and 10% TPB prior to DV infection. The virus culture medium was first harvested after 5 days of incubation and fresh medium was then added. Later, virus culture fluid was harvested every second day until infected cells fully expressed the cytopathic effect. After removal of cell debris by centrifugation at 900 x g, the virus supernatant was aliquoted and stored at -70°C until used. Virus was titrated in plaque formation assays on PscloneD cells. Monolayers of cells were trypsinized and resuspended in L-15 medium containing 3% FCS and 10% TPB and plated at 1 x 105 cells/well in a volume of 0.5 ml in 24-well plates (Nunc, Roskilde, Denmark). Subsequently, dilutions of virus supernatant were added and the mixtures were incubated at 37°C for about 2–3 h; then 0.5 ml of L-15 medium containing 3% FCS, 10% TPB, and 2% (w/v) carboxymethylcellulose (Sigma) was added to each well. After 5 days of incubation at 37°C, the plaques were visualized by staining with a dye solution composed of 0.1% (w/v) napthalene black 10B (Serva, Entwicklungslabor, Heidelberg, Germany), 1.36% (w/v) sodium acetate (Carl Roth, Karlsruhe, Germany), and 6% (v/v) glacial acetic acid (Roth). Virus concentrations are given as plaque-forming units/milliliter.
Infection of endothelial cells
Monolayers of EC were trypsinized and resuspended in growth medium. About 2–3 x 105 or 8 x 104 ECV304 cells were seeded into each well of 24-well tissue-culture plates or 8-well glass chamber slide (Nunc), respectively. Primary cells were seeded at a density of 1–2 x 104/well in 96-well tissue-culture plates. After overnight incubation, virus culture fluid or heat-inactivated virus suspension (80°C, 20 min) was added to confluent monolayers of cells at the multiplicity of infection (MOI) of 0.1 and incubated at 37°C for 2 h. The virus supernatant was then removed and fresh growth medium was added to each well. Culture media and infected cells were harvested at various times after infection for further experiments.
Flow cytometry analysis
At 24, 48, and 72 h after infection, DV-infected and control cells were harvested from 24-well plates. Flow-cytometric assessment of percentage of dead cells was done after pooling populations of cells in suspension and trypsinized adherent cells of a given sample. Propidium iodide (Sigma) was added to give a final concentration of 0.2 µg/ml and samples were analyzed in a FACScan (Becton Dickinson Immunocytometry Systems, San Jose, CA). For flow-cytometric determination of DV infection, the harvested cells were washed twice with medium 199 and then fixed with 2% formaldehyde (Roth) in PBS for 1 h at room temperature (RT). After two washing steps with PBS, the fixed cells were permeabilized with 0.1% Triton X-100 in PBS. Permeabilized cells were incubated with dengue hyperimmune mouse ascitic fluid, generously provided by Dr. A. Nisalak (Armed Forces Research Institute of Medical Sciences, Bangkok, Thailand), at a final dilution of 1:400 for 1 h at RT. The cells were then washed twice with permeabilization solution and incubated with dichlorotriazinyl amino fluorescein (DTAF)-conjugated F(ab')2 fragment of goat anti-mouse IgG (Dianova, Hamburg, Germany) at a final concentration of 3.75 µg/ml for 30 min at RT in the dark. The cells were washed once with an excess volume of permeabilization solution and analyzed by flow cytometry.
Chemokine and cytokine quantitation
Supernatants from DV-infected, mock-infected, and
heat-inactivated virus-treated EC were quantitated for cytokine and
chemokine production. RANTES and monocyte chemotactic protein-1
(MCP-1) were measured using ELISA kits from Biosource International,
Camarillo, CA. ELISA kits for IL-8 were obtained from Innogenetics,
Zwijndrecht, The Netherlands; for IL-1ß from Immunotech/Coulter,
Hamburg, Germany; for IL-1
from Endogen, Biozol Diagnostica
Vertrieb, Eching, Germany; for IL-7, and granulocyte-macrophage-CSF
(GM-CSF) from R&D Systems, Wiesbaden, Germany; for IL-15 from
Cytoscreen, Laboserv Diagnostica, Giessen, Germany; and for TNF-
from Medgenix Diagnostic SA, Fleurus, Belgium.
Plasma and pleural fluid samples from DSS patients
EDTA plasma and PF samples obtained from six children dying of DSS were also assayed in ELISA for the chemokines MCP-1, RANTES, and IL-8. Diagnosis of DSS was based on the clinical criteria established by the World Health Organization (13); all patients were suffering from grade IV DSS. Bacteria were not found in smears and routine cultures of these samples. Albumin ratios for all plasma/PF pairs were above 0.5, indicating that these PF represented exudates. IL-8, MCP-1, and RANTES were determined using the same ELISA kits as described above. Albumin was quantitated at our clinical chemistry department.
RNA extraction and RT-PCR
Total RNA preparations from untreated cells or cells treated with inactivated or active virus were obtained by the method described by Chomczynski and Sacchi with minor modifications (14). RNA was quantitated by spectrophotometry (Pharmacia, LKB Biochrome, Little Chalfont, U.K.) at 260 nm. Approximately 1 µg of total RNA was used for the first-strand synthesis with oligo(dT) primer, AMV-RT, and reaction buffers as described by the manufacturer (reverse transcription system, Promega, Madison, WI). Subsequently, the newly synthesized first-strand cDNA was subjected to 20 rounds of PCR amplification of 95°C for 40 s, 62°C for 1 min, and 72°C for 3 min. Reaction mixtures contained primers at 1.5 µM each, MgCl2 (1.5 mM), dNTP (0.2 mM each), and 1 U of Taq polymerase (Life Technologies) in a total reaction volume of 50 µl. The amplification primers were 5'-ATGACTTCCTTCTGGCCGTGGC-3' (forward) and 5'-TCTCAGCCCTCTTCAAAAACTTCTC-3' (reverse) for IL-8; 5'-TGCCTCCCCATATTCCTCGG-3' (forward) and 5'-TCATGTTTGCCAGTAAGC-3' (reverse) for RANTES; 5'-CAAACTGAAGCTCGCACTCTCGCC-3' (forward) and 5'-ATTCTTGGGTTGTGGAGTGAGTGTTCA-3' (reverse) for MCP-1; 5'-CGGAGTCAACGGATTTGGTCGTAT-3' (forward) and 5'-AGCCTTCTCCATGGTGGTGAAGAC-3' (reverse) for glyceraldehyde-3-phosphate dehydrogenase (GAPDH). GAPDH served as a control to exclude variations between samples. Under these conditions, linear amplification of a cloned MCP-1 template was achieved as described elsewhere (15). RT-PCR products were electrophoresed in 1.8% agarose (FMC BioProducts, Rockland, ME) and stained with ethidium bromide. Results were documented with a Biometra imaging system (BioDocII, Göttingen, Germany).
Transient transfection and luciferase assay
Reporter plasmids pIL8(-420/+102)LUC and pRANTES(-935/+73)LUC were constructed by standard procedures. Fragments of the human IL-8 promoter region comprising position -420 to +102 and of the human RANTES promoter region between -935 and +73 were generated from human genomic DNA by PCR amplification with primers 5'-GGATCCATTGGCTGGCTTATCTTCACC-3' (IL-8, forward) and 5'-GGATCCTTTACACACAGTGAGAATGGT-3' (IL-8, reverse) or 5'-TGAAGCTTTCATATTCTGTAA-3' (RANTES, forward) and 5'-CTAAGCTTGGTACCTGTGGGAGAGGCT-3' (RANTES, reverse). The PCR products were cloned into pCR3.1 (Invitrogen, De Schelp, The Netherlands) and the insert was subcloned via restriction sites incorporated into the primers. The IL-8 promoter fragment was cloned into the BamHI site of the multiple cloning site in the promoterless pGL2 basic plasmid, carrying the firefly luciferase gene (Promega Deutschland, Mannheim, Germany). The RANTES promoter fragment was excised from pCR3.1 with HindIII and subcloned into pGL2 basic via the unique HindIII site contained in this plasmid. Final constructs were verified by Taq dye terminator cycle sequencing using an Applied Biosystems 373A automated sequencer. Plasmids for transfections were prepared by two rounds of cesium chloride density centrifugation. Transient transfection of ECV304 cells was achieved with Lipofectin (Life Technologies) according to the manufacturer’s protocol. Approximately 3 x 105 cells were seeded per well of 6-well tissue-culture plates (Nunc) and transfected 16 h later with 2 µg of supercoiled reporter-plasmid. Cells were washed 16–24 h later and treated with virus at a MOI of 0.1 or with an equal amount of inactivated virus. Luciferase assays with cellular lysates were performed 48 h after treatment using assay reagents from Promega and a Biolumat LB9500 instrument from Berthold (Wildbad, Germany). Calculations of the degree of reporter gene induction were based on the relative numbers of viable cells in the samples, as assessed by ATP determinations.
Bioluminescence assay for determination of cellular ATP
DV and mock-infected cells cultured in 24-well plates for various times after infection were harvested for the determination of intracellular ATP by chemiluminescence measurements with luciferase (Boehringer Mannheim, Mannheim, Germany) as described previously (16). Intracellular ATP of virus-infected cells is given as percentage of luminescence relative to that of control cells.
Determination of DNA strand breaks by TdT-mediated dUTP nick-end labeling (TUNEL)
TUNEL assays were performed on DV-infected and control cultures
in 8-well glass chamber slides. At 24, 48, and 72 h after
infection, cells were fixed with 2% paraformaldehyde (Merck) in PBS
for 30 min at RT. The fixed cells were processed for the detection of
free 3'OH termini due to DNA strand breaks using the TUNEL kit
(Boehringer Mannheim) according to the manufacturer’s protocol. In
some experiments, mouse anti-human TNF- neutralizing Ab (R & D
Systems, Wiesbaden, Germany) at a concentration of 5 or 10 µg/ml was
added 2 h after virus infection and the cells were processed for
the determination of apoptosis as described above.
Detection of nuclear translocation of nuclear factor-
ß
(NF-B)
Confluent ECV304 cells in eight-well glass chamber slides were
fixed after dengue or mock infection with cold 70% ethanol for 20 min
on ice, washed twice with PBS, and double stained with undiluted
culture fluid of mAbs specific for dengue nonstructural protein-1 (NS1)
and rabbit Abs against NF-ß p65 (Santa Cruz Biotechnology) at the
final concentration of 5 µg/ml for 1 h at RT. Cells were washed
three times with PBS and then incubated with a mixture of
DTAF-conjugated F(ab')2 fragment of goat anti-mouse IgG
plus Cyt3-conjugated donkey anti-rabbit IgG (Dianova) at a
concentration of 3.75 µg/ml each for 30 min at RT. After three final
washes, the cells were covered with mounting fluid and visualized under
a fluorescence microscope (Axiophot, Zeiss, Oberkochen, Germany).
Electrophoretic mobility shift assays (EMSA)
Whole cell lysates of untreated or virus-infected ECV304 cells
were prepared by four cycles of freezing and thawing in a lysate buffer
composed of glycerol, 20% (v/v) KCl, 0.4 M Tris-HCl buffer, 20 mM DTT,
2 mM PMSF, 10µg/ml aprotinin. Complementary synthetic
oligonucleotides (5'-GATCCAGAGGGGACTTTCCGAGA-3',
5'-GATCTCTCGGAAAGTCCCCTCTG-3')
comprising the NF-B site (bold letters) of the murine Ig -chain
gene were annealed, and overhangs (italics) were radioactively labeled
with [32P]dATP using Klenow polymerase. Labeled,
double-stranded oligonucleotides were purified by nondenaturing PAGE.
Binding reactions containing whole cell lysate (10 µg of
protein/reaction) were done as described previously (17).
After preincubation of lysates with 2 µg of dIdC, 30,000 cpm of
labeled probe were added per reaction mix. After an additional
incubation for 15 min at RT, samples were loaded onto 4% nondenaturing
PAGE. Complexes were revealed by autoradiography of the vacuum-dried
gel for 6–16 h using a reflection screen. In competition experiments,
a 200-fold excess of unlabeled double-stranded oligonucleotide was
added to the binding reaction. Polyclonal Abs directed to either the
p50 or the p65 subunit of NF-B were obtained from Santa Cruz
Biotechnology. For supershift experiments lysates were preincubated
with 0.2 µg of Ab/sample at 4°C for 1 h.
Indirect immunofluorescence for cell-bound C3dg
Experiments were performed in 8-well glass chamber slides. Human
AB dengue nonimmune serum and heat-inactivated pooled dengue
hyperimmune sera (hemagglutinin inhibition titer 
1/25,600) were used
as sources for complement and anti-dengue Abs, respectively.
DV-infected cell monolayers at 24 or 48 h were treated with 0.1 ml
medium 199 containing complement (at a dilution of 1:5), or
anti-dengue Abs (at a dilution of 1:100), or both, for 1 h at
37°C. Heat-inactivated normal human serum (56°C, 30 min) was used
as control. The cells were washed three times and were incubated for
1 h on ice with a 1:200 dilution of mAb specific for C3dg (clone
9), provided by Dr. P. J. Lachmann (Centre of Veterenary Sciences,
University of Cambridge, Cambridge, U.K.). Then, they were washed three
times and incubated with DTAF-conjugated F(ab')2 fragment
of goat anti-mouse IgG at a final concentration of 3.75 µg/ml in
the dark for 30 min on ice. The cells were washed twice, covered with
mounting fluid, and analyzed by fluorescence microscopy (Axiophot,
Zeiss). Medium 199 containing 10 mM NaN3 (Merck) was used
as a diluent and washing solution during the experiment.
Detection of C5b-9 complex formation
Cells grown in 24-well plates were used for complement-activation experiments at 24 or 48 h postinfection. Infected cells were washed twice with medium 199, and 0.2 ml of medium 199 containing complement (1:5) or anti-dengue Ab (1:100) or both was added and incubated for 1 h at 37°C. In some experiments, 10 mM EGTA and 10 mM MgCl2 (Sigma) were used for inhibition of the classical pathway of complement activation. Heat-inactivated serum was used as control. Thereafter, cells were washed twice with PBS. In pilot experiments we used an indirect immunofluorescence assay to detect C5b-9 complexes on cell surfaces and propidium iodide exclusion to assess membrane integrity of the treated cells. Upon microscopic inspection, C5b-9, formed on intact membranes, could be detected by this technique but gave very low fluorescence signals. Therefore, we used a sensitive capture ELISA technique for determination of C5b-9 formation. In this assay, cells were again first treated with Ab and complement. After deposition of C5b-9 complexes on the cell surface, cells were lysed by applying 0.2 ml of lysing buffer (1% Triton X-100 (Roth) in PBS) for a few minutes. Lysates were clarified by centrifugation at 10,000 x g for 5 min to remove cell debris and frozen at -20°C until used. The actual assay for the detection of C5b-9 complexes was performed according to the protocol of Hugo et al. (18) except that swine anti-rabbit IgG conjugated with horseradish peroxidase (Dako, Glostrup, Denmark) at a final concentration of 0.24 µg/ml was used instead of the biotinylated anti-rabbit IgG in the last step and tetramethylbenzidine (Medgenix Diagnostics, Fleurus, Belgium) was used as a chromogen. The absorbency was read at 450 nm in an ELISA reader (EAR 400, SLT Labinstruments, Salzburg, Austria).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
DV-infected cells express viral Ags in their cytoplasm
(19, 20, 21). Hyperimmune mouse ascitic fluid was used to
detect dengue viral proteins, and the number of infected ECV304 cells
was quantitated using flow cytometry. The analysis of green
fluorescence-positive cells was performed with 5000 events from each
sample. About 63.5% of the cells at 24 h after infection at MOI
of 0.1 expressed DV Ags and almost all cells were infected after 48 and
72 h of incubation (94 and 98.5%, respectively) (Fig. 1
). Approximately 25% of untransformed
EC were infected at 72 h as assessed by FACS or immunofluorescence
microscopy (not shown). For DV-infected ECV304 cells, we also
determined the release of infectious virions into culture supernatants
by plaque-forming assays. The peak titer, obtained 2 days after
infection of 2–3 x 105 ECV cells at an MOI of 0.1
was 19 x 105 PFU/ml (SD ± 2 x
105, n = 4).
|
To analyze cytokine production in response to DV, we collected
supernatants of infected ECV304 cells after 24, 48, and 72 h from
parallel cultures, and screened them by ELISA. No virus-mediated
induction was found for MCP-1, IL-1ß, IL-1, IL-7, IL-15, and
GM-CSF, and only minimal induction of TNF- was noted (data not
shown). In contrast, a marked time-dependent increase of RANTES and
IL-8 became detectable in supernatants of DV-infected ECV304 cells
(Fig. 2). Likewise, untransformed EC from
umbilical vein or dermal microvascular EC, cultured in media with 2%
FCS, produced significant amounts of RANTES only upon DV infection
(Fig. 3). IL-8 levels were increased
twofold in supernatants of HUVEC 96 h after infection as compared
with controls (not shown). On a per infected cell basis, ECV304 cells
produced roughly five times more RANTES than did primary cells.
|
|
RT-PCR for IL-8 and RANTES with total RNA from uninfected and
control cells revealed increased steady state levels of specific mRNAs
in infected cells (Fig. 4). We therefore
investigated whether IL-8- and RANTES-mRNA accumulation could be
attributed to transcriptional up-regulation of their promoters.
Transient transfections with promoter/reporter hybrid constructs were
performed. A 4.9-fold induction of transfected pIL-8(-420/+102)LUC was
observed 2 days after infection with DV of ECV304 cells and a 10- to
20-fold activation with pRANTES (-935/+73) (Fig. 5).
|
|
The results of the in vitro assays prompted us to analyze PF and
plasma samples of DSS patients for the presence of IL-8, MCP-1, and
RANTES. ELISAs were performed twice on each sample with virtually
identical results. As shown in Table I,
IL-8 in plasma and PF of DSS patients were all markedly increased over
the healthy donor controls; the difference exceeded 2 logs in some
cases. Combined treatment of cell-rich plasma from healthy donors with
inulin, 50 mg/ml at 37°C for 1 h (to activate complement), and
0.1% Triton X-100 for an additional 30 min at RT (to liberate any
cell-bound IL-8) led to increased plasma IL-8 levels of up to 627
pg/ml. MCP-1 levels in the patients’ samples were also markedly
elevated. Presumably due to the activation of platelets during
sampling, RANTES was detectable in all plasma samples of DSS patients
and healthy donors, with concentrations ranging between 750 and 1500
pg/ml. Plasma RANTES determinations were considered uninformative. In
contrast, RANTES concentrations in PF of DSS patients 1 through 6 were:
34, 1018, 530, 13, 12, and 60 pg/ml, respectively. Thus, the two
samples with the highest amounts of IL-8 also displayed the highest
levels of RANTES.
|
On microscopic inspection of infected ECV304 or primary cells we
regularly noted rounding, detachment, and nuclear condensation of many
cells after 48–72 h. Measurements of intracellular ATP and propidium
iodide-uptake were performed. DV led to approximately 50% ATP
reduction after 48 h and to >73% reduction by 72 h
postinfection (Fig. 6). The number of
propidium iodide-positive cells increased from 7% at 24 h to
almost 40% at 72 h. The morphology of dying cells in DV-infected
cultures suggested that they succumbed to programmed cell death.
Nicking of the DNA by endogenous endonucleases is another hallmark of
apoptosis (22), and the TUNEL method was next used to
detect intrachromosomal DNA strand breaks. As a control, cells were
treated with TNF-, and anti-TNF Abs were employed to detect
cytokine-dependent apoptosis. In the experiment of Fig. 7, controls incubated without (Fig. 7A) or with anti-TNF Abs (Fig. 7B) contained
less than 1% apoptotic cells. Treatment with TNF- plus
cycloheximide (10 µg/ml) for 16 h caused apoptosis in about
5–10% of cells (Fig. 7C), and this effect was inhibited by
anti-TNF Abs (Fig. 7D). DV infection led to apoptosis of
15–25% of the cells after 24 h (not shown) and 50–60% of the
cells after 48 h (E), and anti-TNF Abs were unable
to suppress this effect (Fig. 7F). Thus, apoptosis of
DV-infected cells appeared not to be dependent on TNF- secreted by
ECV304. Apoptosis of hepatoma cells following DV infection has been
reported recently (23).
|
|
B activation following
DV infection of ECV304 cells
The crucial role of transcription factor NF-B for the control
of immune responses, including the production of chemokines, is well
documented and NF-B has recently also been recognized to be a
central player in the regulation of cell death (24). We
therefore investigated whether DV-infected ECV304 cells activate
NF-B. First, double-immunofluorescence staining for viral Ag and
NF-B p65 was performed (Fig. 8).
Nuclei of untreated cells are essentially negative for p65, some
cytoplasmic staining is seen (Fig. 8A). Positive staining
with Ab to the viral Ag NS1 proved that virtually all cells were
infected (Fig. 8B). Double staining for NS1 and for p65
revealed that p65 was translocated to the nuclei in some of the cells
and to different degrees. Most of the cells exhibiting intense nuclear
staining for NF-B and complete translocation were characterized by a
strong and focal perinuclear staining for NS1 and appeared condensed
(Fig. 8C). Orange to yellow staining resulted from double
exposure with the two fluorescent dyes. Thus, NF-B p65 activation
occurred in infected cells.
|
B activation was assessed by EMSA. (Fig. 9A); extracts from cultures
infected for 24 and for 48 h showed increased binding activity, at
least in part due to increased p65 binding. No induction was seen at
12 h or earlier. While competition with a 200-fold molar excess of
unlabeled oligonucleotides showed that complexes with either extracts
from uninfected or DV-infected cells were specific for the probe,
immunosupershift experiments clearly demonstrated that binding of both
NF-B p65 and of p50 was markedly activated in virus-treated cells
(Fig. 9B). C1 disappeared after preincubation with
anti-p65 and thus probably represents a p65 homodimer while C2 was
reduced after either preincubation with anti-p50 or anti-p65
and thus may be a p50/p65 heterodimer. Preincubation with both Abs
(lane 6) led to a decrease of C2, SC1, and SC2,
suggesting that these complexes comprise p50 and p65, the major
activating form of NF-B.
|
A marked reduction of complement proteins and a concomitant
increase in complement fragments is observed in DHF/DSS, and the degree
of complement activation correlates with the severity of the disease
(5, 25, 26). Therefore, we investigated whether
DV-infected EC activate complement in vitro. DV-infected ECV304 cells
were incubated in the presence and absence of anti-dengue Abs with
human complement for 1 h at 37°C. Ab-dependent deposition of
C3dg fragments on the surface of infected cells was observed using a
mAb specific for a neoantigen that is exposed in C3bi and C3dg
(27) (Fig. 10A). This neoepitope is
expressed only after cleavage of substrate-bound complement C3b by
factor I, in the presence of factor H, upon binding to a
complement-activating surface (28). Positive staining for
C3dg was time dependent: staining was not observed when cells that had
been infected for only 12 h were analyzed. The latter negative
finding at the same time provided an important control. Deposition of
C3dg fragments was also not seen when infected ECV304 cells were
incubated with anti-dengue Abs and heat-inactivated nonimmune serum
(not shown), or with nonimmune serum alone (Fig. 10B).
Together, these results indicated that activation of complement by
dengue-infected ECV304 cells is Ab dependent.
|
C). Activation occurred via both
classical and alternative pathways since adding EGTA-MgCl2
only partially abolished the formation of C5b-9 complexes (data not
shown). Since the formation of C5b-9 complexes on the surface of cells may lead to cell lysis, measurements of cellular ATP were undertaken. No changes were observed after incubation of infected cells with anti-dengue Abs and complement (data not shown). This result indicated that no direct cytotoxic effect occurred after complement activation of dengue-infected ECV304 cells.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Based on the collective data, several pathways are envisaged that possibly converge to cause the massive but transient leakage syndrome. First is a selective action of cationic chemokines on the vascular endothelium in serosal tissues, which may be mediated via their interaction with heparan sulfate expressed at these sites. A second mechanism is the activation of complement on the surface of infected cells, and the third is apoptotic cell death.
The cellular source of chemokine production remains an open question. In line with the classic Ab dependent enhancement concept, which envisages non-neutralizing Abs to augment infection of macrophages, these cells may provide a source of these chemokines. However, the present study raises another possibility, i.e., that EC may also be major producers of IL-8 and RANTES. This hypothesis derives from two findings: first, EC are effectively infected by DV, and second, DV-infected EC selectively up-regulate transcription and secretion of IL-8 and RANTES.
That EC can be infected by DV has been shown previously, and we confirm
that the virus replicates to high titers in cultured EC, independent of
the presence of enhancing Abs. The reason for differential
susceptibility of ECV304 and HUVEC remains to be elucidated. Possibly
ECV304 express a higher density of DV receptor. That DV-infected EC
liberate large amounts of IL-8 and RANTES is a novel finding.
Transcriptional up-regulation was demonstrated with the use of reporter
gene constructs, and secreted proteins were quantified by ELISA. Quite
remarkably, chemokine production occurred selectively, and there was no
virus-mediated induction of MCP-1, IL-1ß, IL-1, IL-15, or GM-CSF.
The selective induction of cytokine synthesis is now emerging as a
common theme that possibly directs pathology in many diseases. The
underlying mechanism in the case of DV will be the subject of future
studies, but the very fact that the virus selectively induces
endothelial production of IL-8 and RANTES harbors potentially
significant consequences. It is known that IL-8 binds to heparan
sulfate (30) and to EC of serosal tissues
(31) that express this proteoglycan (32). It
is also known that both IL-8 and RANTES increase vascular permeability
via transient recruitment and local activation of neutrophils
(33). That the chemokines are indeed produced in quantity
during viral infection became evident from IL-8 measurements in plasma
and pleural fluids from DSS patients. Very high levels of circulating
IL-8 were detected in all cases. These could not have derived from
circulating Granulocytes, because maximal IL-8 levels in control blood
samples were two orders of magnitude lower despite
Granulocyte-stimulation via complement activation with inulin combined
with cell lysis with Triton X-100 to liberate any nonsecreted IL-8. We
believe it is reasonable to assume that the high chemokine levels in
DSS patients reflected production in infected cells, with EC
representing attractive candidates. While MCP-1 levels in
supernatants of ECV304 cells were not induced by DV, all plasma and PF
samples of DSS patients displayed elevated levels of this chemokine.
Levels of MCP-1 in PF from various clinical conditions have previously
been shown to correlate with the numbers of monocytes
(34). Thus, monocytes might also be an important source of
the chemokines in PF of DSS patients. Cytologic studies with PF from
DSS patients are in progress now. Of note, high levels of IL-8 have
been detected in patients with pleural effusion of other etiology,
whereby the values measured in our study markedly exceed those
previously published (34). The presence of large amounts
of RANTES in two out of six PF from DSS patients complements the
findings on IL-8.
Once infection of EC has occurred, cross-reacting, non-neutralizing Abs to DV will activate complement on these cells. This provides a second possible mechanism leading to vascular leakage. Complement attack on virus-infected cells is, of course, not conceptually novel, but its possible involvement in DSS has not received much consideration hithertofore. In an early hypothesis, one of us pointed out that such auto-attack would result in local liberation of anaphylatoxins and the generation of C5b-9 complexes (10). Both processes could contribute to disease. Here, the Ab-dependent deposition of activated C3 and C5b-9 on DV-infected cells was demonstrated. Our results indicate that the terminal complexes were not directly cytocidal, and it is known that subcytocidal C5b-9 attack increases vascular permeability of endothelial cell monolayers (35). C5b-9 generation can only occur in conjunction with the liberation of anaphylatoxins. In this context, high levels of C3a and C5a have indeed been detected in PF of DSS patients (P. Malasit, unpublished observation). C3a induces histamine release from mast cells, and thus would indirectly enhance vascular permeability. It is of interest that high histamine concentrations have actually been measured in urine samples of DSS patients (36). In addition to activating phagocytes, C5a induces shedding of heparan sulfate from EC (37). In the context of DV pathogenesis, this may be important because, if chemokines are targeted to serosal EC via binding to heparan sulfate, C5a-induced shedding might represent a counteracting factor. Furthermore, if heparan sulfate represents a binding site for DV on EC (38), shedding might also serve to limit infection.
The third putative pathway to vascular leakage could derive from apoptosis of infected cells. Massive reduction in cellular ATP, accompanied by positive TUNEL stainings for nicked DNA occurred after 3–4 days of infection. Apoptosis of DV-infected EC would acutely enhance local vascular leakage. Furthermore, apoptosis would be followed by rapid removal of infected cells, which would explain the difficulties in detecting the infectious agent in tissue specimens. Apoptosis of infected EC and complement-induced shedding of the viral receptor heparan sulfate from bystander cells would also nicely explain the abrupt termination of infection.
In summary, there are three major findings in this study that can be accommodated in a coherent hypothesis of the DSS leakage syndrome. First, EC are highly permissive to DV infection and respond by the selective production of chemokines IL-8 and RANTES. These may accumulate at serosal sites, causing local vascular leakage. Second, cross-reacting but non-neutralizing Abs to DV will activate complement on the surface of infected EC, causing liberation of anaphylatoxins and deposition of C5b-9. Of these, C3a may mediate histamine release, and C5a will induce shedding of heparan sulfate. The latter might counteract infection and reduce chemokine binding. Third, infected EC eventually undergo apoptosis. This would augment vascular leakage, and also provide a mechanism for the sudden disappearance of the infectious agent, explaining the short-lived duration of disease and the eradication of clues to the nature of the infected cells in vivo.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Matthias Husmann, Institute of Medical Microbiology and Hygiene, University of Mainz, Augustusplatz, D-55101 Mainz, Germany. E-mail address:
3 Abbreviations used in this paper: DHF/DSS, dengue hemorrhagic fever and dengue shock syndrome; DTAF, dichlorotriazinyl amino fluorescein; DV, dengue virus; EC, endothelial cells; EMSA, electrophoretic mobility shift assay; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; MCP-1, monocyte chemotactic protein-1; MOI, multiplicity of infection; PF, pleural fluid; RT, room temperature; TPB, tryptose phosphate broth; TUNEL, TdT-mediated dUTP nick-end labeling.
Received for publication March 25, 1998. Accepted for publication August 11, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
b which in turn induces apoptotic cell death. J. Virol. 71:3244.[Abstract]
B or die?. Curr. Biol. 7:R9.
This article has been cited by other articles:
|
|
 |
|
  G. Weissmann Dengue Fever in Rio: Macumba versus Voltaire FASEB J, July 1, 2008; 22(7): 2109 - 2112. [Full Text] [PDF]
|
|
|
|
 |
|
 S. Sangiambut, P. Keelapang, J. Aaskov, C. Puttikhunt, W. Kasinrerk, P. Malasit, and N. Sittisombut Multiple regions in dengue virus capsid protein contribute to nuclear localization during virus infection J. Gen. Virol., May 1, 2008; 89(5): 1254 - 1264. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. E. Dewi, T. Takasaki, and I. Kurane Peripheral blood mononuclear cells increase the permeability of dengue virus-infected endothelial cells in association with downregulation of vascular endothelial cadherin J. Gen. Virol., March 1, 2008; 89(3): 642 - 652. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Blackley, Z. Kou, H. Chen, M. Quinn, R. C. Rose, J. J. Schlesinger, M. Coppage, and X. Jin Primary Human Splenic Macrophages, but Not T or B Cells, Are the Principal Target Cells for Dengue Virus Infection In Vitro J. Virol., December 15, 2007; 81(24): 13325 - 13334. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Wati, P. Li, C. J. Burrell, and J. M. Carr Dengue Virus (DV) Replication in Monocyte-Derived Macrophages Is Not Affected by Tumor Necrosis Factor Alpha (TNF-{alpha}), and DV Infection Induces Altered Responsiveness to TNF-{alpha} Stimulation J. Virol., September 15, 2007; 81(18): 10161 - 10171. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Srikiatkhachorn, C. Ajariyakhajorn, T. P. Endy, S. Kalayanarooj, D. H. Libraty, S. Green, F. A. Ennis, and A. L. Rothman Virus-Induced Decline in Soluble Vascular Endothelial Growth Receptor 2 Is Associated with Plasma Leakage in Dengue Hemorrhagic Fever J. Virol., February 15, 2007; 81(4): 1592 - 1600. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y.-R. Lee, M.-T. Liu, H.-Y. Lei, C.-C. Liu, J.-M. Wu, Y.-C. Tung, Y.-S. Lin, T.-M. Yeh, S.-H. Chen, and H.-S. Liu MCP-1, a highly expressed chemokine in dengue haemorrhagic fever/dengue shock syndrome patients, may cause permeability change, possibly through reduced tight junctions of vascular endothelium cells J. Gen. Virol., December 1, 2006; 87(12): 3623 - 3630. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-P. Chen, H.-L. Lu, S.-L. Lai, G. S. Campanella, J.-M. Sung, M.-Y. Lu, B. A. Wu-Hsieh, Y.-L. Lin, T. E. Lane, A. D. Luster, et al. Dengue Virus Induces Expression of CXC Chemokine Ligand 10/IFN-{gamma}-Inducible Protein 10, Which Competitively Inhibits Viral Binding to Cell Surface Heparan Sulfate. J. Immunol., September 1, 2006; 177(5): 3185 - 3192. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M.-F. Hsieh, S.-L. Lai, J.-P. Chen, J.-M. Sung, Y.-L. Lin, B. A. Wu-Hsieh, C. Gerard, A. Luster, and F. Liao Both CXCR3 and CXCL10/IFN-Inducible Protein 10 Are Required for Resistance to Primary Infection by Dengue Virus J. Immunol., August 1, 2006; 177(3): 1855 - 1863. [Abstract] [Full Text] [PDF]
|
|
|
|
|
), cells treated with
inactivated DV (i.DV), or DV (DV). The constitutively expressed GAPDH
gene served as a control. The same results were reproduced in two
further experiments.
) ranged at
background levels whereas positive signals were observed in lysates of
infected cells (DV). No C5b-9 complexes were detected when control or
DV-infected ECV304 cells were incubated with complement without
anti-dengue Abs, in the presence of EGTA/MgCl2, or with
heat-inactivated complement (not shown). The data represent the average
of four experiments ± SD.