| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Production, Imparting Pathogenicity on Blastomyces dermatitidis1
,
,§
Departments of
*
Pediatrics,
Internal Medicine, and
Medical Microbiology and Immunology, and
§
Comprehensive Cancer Center, University of Wisconsin Medical School, University of Wisconsin Hospital and Clinics, Madison, WI
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
in lung alveolar fluids of mice infected with
B. dermatitidis was severalfold higher for WI-1 knockout
yeast compared with wild-type yeast, and in vitro coculture of
unseparated lung cells with these isogenic yeast disclosed similar
differences. Upon coculture with purified macrophages and neutrophils,
wild-type yeast blocked TNF- production, yet WI-1 knockout yeast
stimulated production. Coating knockout yeast with purified WI-1
converted them from stimulating TNF- production to inhibiting
production. Addition of purified WI-1 into stimulated phagocyte
cultures led to concentration-dependent inhibition of TNF-
production. Neutralization of TNF- in vivo exacerbated experimental
pulmonary infection, particularly for the nonpathogenic WI-1 knockout
yeast. Inducing increased TNF- levels in the lung by
adenovirus-vectored gene therapy controlled infection with wild-type
yeast. Thus, the WI-1 adhesin on yeast modulates host immunity through
blocking TNF- production by phagocytes, which fosters progression of
pulmonary infection. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Though the factors that account for virulence of B. dermatitidis are incompletely understood, a bona fide virulence determinant of the fungus has recently been identified (3). WI-1, a 120-kDa protein, is a major Ag and adhesion-promoting protein on B. dermatitidis. Tandem repeats of WI-1 display immunodominant B cell epitopes (4, 5) and also mediate attachment to CD18 and CD14 receptors on human macrophages (6). WI-1 null strains of B. dermatitidis created by gene targeting and mutation of the WI-1 locus exhibit greatly reduced pathogenicity (3). In contrast to wild-type yeast, WI-1 knockout strains are nonpathogenic in a murine model, even at high innocula of 105 organisms. These observations underscore the prominent role of WI-1 in pathogenicity of B. dermatitidis.
Mechanisms that underlie the virulence-promoting effect of WI-1 have been partially elucidated. Adherence is one of them. Yeast that lack WI-1 bind poorly to the lung ex vivo and to macrophages in vitro (3). Such findings imply that knockout yeast are unable to establish infection in the lower respiratory tract because they bind poorly to structures in the airway or alveoli. Nonadherent yeast might be more easily dislodged from the lung, or when lacking the capacity to enter inactivated lung macrophages, more readily recognized and killed by effector cells.
Recent results point to additional defects in B. dermatitidis yeast mutated at the WI-1 locus. Following i.v. inoculation of B. dermatitidis yeast into mice, wild-type yeast multiplied rapidly in the lung, killing the mice, whereas WI-1 knockout yeast multiplied poorly (B. Klein, M. Wüthrich, T. Brandhorst, B. Finkel-Jimenez, and H. Filutowicz, manuscript in preparation). This finding implies that WI-1 knockout yeast are defective in pathogenicity even when the route of infection does not require airway or alveolar adherence. This residual defect does not impair cell-wall integrity of the yeast, nor does it render the yeast more susceptible to phagocyte killing. Instead, immune deviation may be an additional attribute of WI-1 and provide an explanation. Histologic staining of lung sections demonstrates mature granulomas with only small numbers of sequestered yeast in mice that received the knockout, as compared with disorganized granulomas overrun by yeast in mice that received the wild-type (3). Similarly, the number and distribution of hematopoietic cells in lung lesions show corresponding differences. Lungs of knockout-infected mice contain relatively more CD3+ T-cells and fewer neutrophils, whereas the lungs of wild-type infected mice demonstrate the opposite. Thus, the profile of inflammatory response is heavily influenced by the presence of WI-1.
Based on above findings, we hypothesize that WI-1 modulates host
immunity early in the course of infection and thereby facilitates
establishment of B. dermatitidis in the lung. Many and
diverse microbes escape host elimination by modulation of host immunity
(7). Because cytokines and chemokines orchestrate and
shape the immune response, we investigated the influence of WI-1 on
production of pro-inflammatory cytokines known to be pivotal in innate
and acquired immunity. We demonstrate in this study that WI-1
interferes with host immunity by blocking production of a key
pro-inflammatory cytokine, TNF-. These findings provide a mechanism
by which WI-1 enhances virulence and offer further insight into how
medically important fungi induce human disease.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Isogenic strains of B. dermatitidis that contain and lack WI-1 were used in this study. American Type Culture Collection (Manassas, VA) strain 26199 is a wild-type virulent strain that expresses WI-1 and was originally obtained from a human patient. The isogenic, WI-1 knockout, designated strain no. 55, was recently derived from strain 26199 and is nonpathogenic (3). Isolates of B. dermatitidis were maintained in the yeast form on Middlebrook 7H10 agar (Ditco, Detroit, MI) slants with oleic acid-albumin complex at 37°C. Liquid cultures of yeast were grown in Histoplasma macrophage medium (8). Saccharomyces cerevisiae strain ECY36-3D was generously provided by Enrico Cabib (National Institute of Diabetes and Digestive and Kidney Diseases, National Institute of Health, Bethesda, MD). S. cerevisiae was maintained on yeast peptone dextrose agar at 37°C.
Mice
Inbred C57BL/6 and BALB/c strains of mice were obtained from The Jackson Laboratory (Bar Harbor, ME). Male mice 6–7 wk of age at the time of purchase were housed and cared for throughout these experiments according to guidelines of the University of Wisconsin Animal Care Committee, who approved all aspects of this work.
Reagents
Complete tissue culture medium consisted of RPMI 1640
supplemented with 10% heat-inactivated FBS (HyClone, Logan, UT), 25 mM
HEPES buffer, 2 mM L-glutamine, 1 mM sodium pyruvate, 100
U/ml penicillin, and 100 µg/ml streptomycin (BioWhittaker,
Walkersville, MD). Casein, LPS from Escherichia coli
(9), actinomycin D, and Histopaque 1077 and 1119 were
purchased from Sigma (St. Louis, MO). Recombinant murine IFN-
was
purchased from R&D Systems (Minneapolis, MN). Anti-TNF- mAbs (mAb)
2E2 and XT22-11 were kindly provided by Dr. Craig Reynolds at the
National Cancer Institute (Frederick, MD), Dr. George Deepe at the
University of Cincinnati (Cincinnati, OH), and DNAX Research Institute
(Palo Alto, CA). Anti-WI-1 mAb was DD5-CB4 (IgG2a) as described
previously (4).
Functionally active TNF- peptide of 11 aa residues (10)
(H-Pro-Ser-Thr-His-Val-Leu-Ileu-Thr-His-Thr-Ileu-OH) and a control
peptide without TNF- function
(H-Gly-Gly-Asp-Pro-Gly-Ileu-Val-Thr-His-Ser-OH) were synthesized at the
University of Wisconsin (Madison, WI) Biotechnology Center. Peptides
were isolated by HPLC and shown to be over 92% pure. Activity of the
peptides was verified using a TNF- bioassay described below.
Recombinant, replication-deficient adenoviruses expressing either
murine TNF- or 
-galactosidase
(LacZ)3 as
a control have been described (11) and were generously
provided by Dr. Theodore Standiford (University of Michigan, Ann
Arbor, MI).
Antigens
Secreted WI-1 was purified as described (12). Briefly, yeasts were grown in liquid Histoplasma macrophage medium in a gyratory shaker at 37°C for 1 wk. WI-1 was purified from supernatants in two steps using anion-exchange chromatography followed by hydrophobic interaction chromatography. Homogeneity of purified WI-1 was analyzed by SDS-PAGE and silver stain. Purified WI-1 was used to coat the surfaces of WI-1 knockout strain no. 55 as described (13).
Measurement of TNF-
TNF- was quantified using a bioassay that measures killing of
WEHI 64 var. 13 cells (American Type Culture Collection) as described
previously (14). Briefly, 50 µl of supernatant
containing known concentrations of rTNF- were mixed with 2 x
105/ml actinomycin D-treated WEHI
cells and incubated for 20 h at 37°C. After incubation, 20 µl
MTT (5 mg/ml in PBS; Sigma) was added and the cells were further
incubated for 4 h at 37°C. After removing 70 µl of supernatant
from the wells, 100 µl of isopropanol with 0.04 N HCl was added. OD
was measured at 570 nm with an automatic ELISA plate reader (Spectra
Max 190; Molecular Devices, Sunnyvale, CA).
A commercial ELISA kit was used to measure TNF- in lung homogenates
of mice (R&D Systems). ELISAs were developed with streptavidin HRP and
substrate tetramethylbenzidine (Sigma). OD (450 nM) of wells was
measured with an automatic plate reader as above.
Neutralization of TNF--mediated cytolytic activity by WI-1 was
quantified using a modification of the TNF- bioassay
(15). Briefly, 50 µl containing different concentrations
of soluble WI-1 were mixed with 50 µl of rTNF- diluted to 100
pg/ml and incubated for 2 h at 37°C. Then 2 x
105/ml of actinomycin D-treated WEHI cells were
added and incubated for 20 h at 37°C. OD values were measured as
above.
Bronchoalveolar lavage
Alveolar fluid was obtained from mice by bronchoalveolar lavage.
Briefly, alveolar fluid was harvested through a 20-gauge catheter
placed intratracheally (i.t.). A volume of 1.0 ml PBS containing 0.5%
EDTA was instilled and re-aspirated once, yielding a total volume of
0.8 ml lavage fluid per mouse. Individual fluid samples were
centrifuged at 2500 rpm for 10 min. Supernatants were collected, frozen
at -20°C, and thawed once for testing. TNF- content in individual
samples was quantified by ELISA.
Isolation of lung cells, peritoneal macrophages, and polymorphonuclear leukocytes (PMNs)
Lung cells were isolated as described (16). The
pulmonary vasculature was perfused with 3 ml of PBS to eliminate
peripheral blood cells. The lungs were removed, minced, and incubated
for 90 min at 37°C in digestion buffer containing 0.7 µg/ml
collagenase (Sigma) and 30 µg/ml bovine pancreatic DNase I (Sigma).
Tissue fragments and dead cells were removed by filtration through
40-µm cell strainers (Falcon; Becton Dickinson, Rutherford, NJ).
Cells were resuspended in RPMI 1640 and adjusted to a concentration of
2 x 106/ml. Lung cells and yeast were
cocultured at an E:T ratio of 20:1 for 48–72 h at 37°C, 5%
CO2 in a 24-well plate (Costar, Corning, NY).
Supernatants were collected and analyzed for TNF- content.
Peritoneal exudate cells (PEC) were isolated as follows. At 16 and 3 h before PEC isolation, mice were injected i.p. with 3 ml of 10% casein in PBS. Cells were harvested in 3 ml of cold PBS supplemented with 0.05% EDTA. After two washes with HBSS (Life Technologies, Rockville, MD), PEC were enriched for PMNs and macrophages on a two-step Histopaque gradient (Sigma) according to the method of Hilger and Danly (17). To establish that PMNs and macrophages were enriched, each cell fraction was stained with DiffQuick (Dade Behring, Newark, DE) and cellular composition was analyzed by light microscope. Purity of each fraction was >95%.
PMNs were adjusted to 2 x 106/ml in RPMI
1640 and 0.5 ml of the cell suspension was added to each well of a
24-well tissue culture plate (Costar). Preliminary experiments
demonstrated that a ratio of PMN:yeast of 1:1 was optimal for
production of TNF-; this ratio was employed throughout the study
with coculture for varying periods. Macrophages were plated at a
concentration of 2 x 106/ml (0.5 ml/well),
incubated for 1 h at 37°C, and washed to remove nonadherent
cells. In preliminary studies, a ratio of macrophages:yeast of 4:1 was
found to be optimal for eliciting TNF- production; this ratio was
used throughout the study with coculture for varying periods. Coculture
of peritoneal macrophages and neutrophils with yeast included the
addition of 2 µg/ml LPS. These experiments always included control
wells containing the cells together with LPS alone, and supernatants
harvested from such wells yielded values of TNF- that were <25
pg/ml. All in vitro coculture experiments were performed with live
yeast cells unless otherwise stated.
Experimental infection
Mice were infected i.t. with B. dermatitidis. Before infection, mice were anesthetized by i.p. injection of 30 mg/kg etomidate (Bedford Laboratories, Bedford, OH). Skin over the trachea was incised and underlying tissue was separated. A 30-gauge needle (Becton Dickinson) was bent and attached to a tuberculin syringe (Becton Dickinson) containing B. dermatitidis yeast. The needle was inserted into the trachea and 30 µl of inoculum was dispensed using a stepper device (Tridak, Brookfield, CT). Incised skin was closed with cyanoacrylate adhesive (Nexaband; Veterinary Products Laboratories, Phoenix, AZ). Mice recovered under a heating lamp. Two outcomes were measured in infected mice. Burden of lung infection was measured by plating homogenized lung on brain heart infusion (Difco), agar and enumeration of yeast CFU. The detection limit was 10 organisms. Alternatively, duration of survival was monitored.
Statistical analysis
Differences between wild-type yeast (strain 26199) and WI-1
knockout yeast (strain no. 55) in stimulation of TNF- were analyzed
using methods for standard ANOVA (18). Differences in the
number of CFU in tissue between groups of infected mice were analyzed
statistically using the Wilcoxon rank test for nonparametric data
(18). Kaplan Meier (19) survival curves also
were generated for mice that received infection. Survival times of mice
that were alive by the end of the study were regarded as censored. Time
data were analyzed by the log rank statistic and exact p
values were computed using the statistical packaged Stat Xact-3 by
Cytel (Cambridge, MA). Differences between groups were considered
significant if the two-sided p value is <0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Gene targeting and disruption of WI-1 greatly reduces the
pathogenicity of B. dermatitidis in a murine model of
pulmonary infection (3). Knockout yeast replicate poorly
in the lung, and tissue histopathology shows circumscribed granulomas,
whereas wild-type yeast multiply rapidly in the lung, and tissues show
only poorly organized granulomas. To establish temporally when
differences in multiplication of the two strains occurred in vivo,
burden of lung infection was analyzed at early timepoints after i.t.
administration of yeast. Lung CFU for the two strains were similar
until 48 h, when they diverged and mice infected with the
wild-type showed nearly 5-fold more CFU (Fig. 1
A).
|
in control of pulmonary
histoplasmosis and cryptococcosis (20, 21), we
investigated production of this proinflammatory cytokine in the
alveolar fluid of infected mice early in the course of infection as a
possible explanation for the difference in progression of infection of
the isogenic strains. At 6 and 24 h after infection, TNF- in
alveolar lavage fluid was 2- to 3-fold greater in mice infected with
WI-1 knockout yeast as compared with wild-type yeast (Fig. 1
B). These differences were detected at early time points
when the numbers of yeast were comparable for the two strains. In
contrast, no difference in TNF- was observed at 48 h, when the
number of lung CFUs increased sharply for the wild type and
significantly exceeded that for the knockout. To control for the number
of yeast interacting with host lung, we isolated lung cells from mice
and cocultured them in vitro with each strain. Under these conditions,
levels of TNF- again were severalfold higher in response to the
knockout as compared with the wild type, both at 48 and 72 h of
coculture (Fig. 1C). Priming of the cells in vitro with
rIFN- (100 ng/ml for 18 h) before addition of yeast yielded
similar results (data not shown). Differential counts on the explanted
lungs’ cells showed that they were comprised of 65% macrophages
and neutrophils and 35% lymphocytes.
To investigate whether the production of TNF- is needed for control
of infection, we neutralized the cytokine in vivo and analyzed the
impact on lung CFU 7 days postinfection. Depletion of TNF- enhanced
progression of lung disease, especially in mice that were infected with
the WI-1 knockout strain (Fig. 2). Lung
CFU after infection with WI-1 knockout yeast increased by 10-fold in
mice treated with anti-TNF- compared with rat IgG.
Neutralization of TNF- did not reduce survival following infection
with the knockout yeast (data not shown). Depletion of TNF- also
impaired control of infection with wild-type yeast, but to a lesser
degree than was observed with the WI-1 knockout strain. These results
demonstrate that TNF- contributes significantly to control of
B. dermatitidis infection, particularly to accelerated
clearance of the WI-1 knockout strain. The data also imply that rapid
clearance of the knockout strain is due to a robust TNF- response,
and that impaired clearance of the wild-type may be due to
down-regulation of TNF- production.
|
release following in vitro coculture with
B. dermatitidis differs sharply between isogenic
wild-type and WI-1 knockout yeast
Phagocytes are an important source of TNF-; they are prominent
in the cellular response to B. dermatitidis, and they are
major constituents of cells isolated from the collagenase-treated lungs
described in Fig. 1C. We analyzed the interaction of
wild-type and WI-1 knockout strains with phagocytes to determine
whether differences in TNF- production by these cells in vitro could
be correlated with differences observed in alveolar lavage fluid and
isolated lung cells. The knockout strain stimulated release of
increasing amounts of TNF- from both peritoneal macrophages and
neutrophils over the entire course of incubation, whereas the wild-type
yeast stimulated production initially, which was then followed by a
marked down-regulation of the TNF- response (Fig. 3). No differences in phagocyte viability
were evident when comparing cells cultured with wild type vs knockout.
About 80% of macrophages were viable after 96 h of incubation for
both strains, and about 75% of neutrophils were viable at 48 h of
incubation. Thus, induction of apoptosis or loss of viability of these
cells by another mechanism do not appear to account for differences in
TNF- production in response to wild-type vs knockout yeast.
Nevertheless, these observations with PEC recapitulate the results
above with lung cells and lavage fluid, suggesting that peritoneal
cells accurately model events in the lung with respect to phagocyte
interactions with B. dermatitidis yeast.
|
by phagocytes
Two formal possibilities could explain in vitro observations with
lung and peritoneal cells. Knockout yeast could expose "neo"
ligands on the WI-1 null surface, which engage phagocytes and evoke
robust TNF- release (i.e., active stimulation hypothesis).
Alternatively, WI-1 on the wild-type yeast could down-regulate the
TNF- response of phagocytes to B. dermatitidis, acting as
a virulence factor that interferes with host immunity (i.e.,
down-regulation hypothesis). To distinguish between the two models, we
performed a mixing experiment. A constant number of knockout yeast was
cocultured with phagocytes, and a varied and increasing number of
wild-type yeast was added. If the active stimulation hypothesis were
correct, phagocytes would maintain a robust TNF- response. If the
down-regulation hypothesis was correct, then addition of wild-type
yeast should interfere with the robust TNF- response to the knockout
yeast. Addition of wild-type yeast down-regulated the response to
knockout yeast at all concentrations of wild-type yeast tested (Fig. 4). Even at a ratio of 1000:1 for
knockout yeast vs wild type, the latter cells down-regulated TNF-
responses to knockout yeast. Furthermore, at low numbers of wild-type
yeast, phagocytes in the wells far exceeded wild-type yeast, yielding a
ratio of phagocytes:yeast of 1000:1 both for macrophages and
neutrophils. Hence, even interaction of yeast with a small proportion
of phagocytes led to marked down-regulation of TNF-.
|
production by phagocytes
The isogenic strains studied here differ only in expression of
WI-1 as previously described (3), pointing to the likely
role of WI-1 in down-regulation of TNF- production. To establish
formally that WI-1 is responsible for down-regulation of TNF-, we
used three independent approaches in in vitro experiments. First, we
coated the surfaces of the knockout yeast with WI-1 by addition of the
purified protein as described (13). Attachment of WI-1 to
yeast cell surfaces was analyzed using anti-WI-1 mAb DD5-CB4 and
FACS analysis, which confirmed that the presence of the adhesin was
quantitatively similar to the wild-type yeast (data not shown).
WI-1-coated knockout yeast behaved just like the wild-type yeast in its
ability to down-regulate TNF- production of macrophages and
neutrophils either when added alone to wells containing
phagocytes, or when "doped in" to wells in which uncoated
knockout yeast were being used to trigger robust TNF- production
(Fig. 5A).
|
production by phagocytes. Purified WI-1 inhibited TNF- production by
both macrophages and neutrophils in response to knockout yeast; it
inhibited responses in a dose-dependent manner, over a range of
concentrations, and abolished TNF- production at the maximal
concentration of 40 µg/ml (Fig. 5B).
These experiments have so far established that WI-1 down-regulates
TNF- production by phagocytes. However, they have not resolved
whether WI-1 can act in a soluble form or whether, upon introduction to
the well, it must first adhere to the surface of the knockout yeast to
mediate an inhibitory effect. This latter situation could result from
the novel mechanism of cell-wall biogenesis recently reported for
B. dermatitidis, whereby released WI-1 recycles back to the
yeast surface by binding exposed chitin fibrils (13). By
this theory, soluble WI-1 could bind directly to the surface of
knockout yeast before mediating inhibition of TNF- production.
To resolve whether WI-1 in soluble form inhibits TNF- production, we
used a third approach involving S. cerevisiae. These yeast
cells evoke TNF- production by macrophages and neutrophils, but they
do not have the exposed chitin fibrils on their surfaces required to
bind exogenously added soluble WI-1 (13). Phagocyte
TNF- production in response to S. cerevisiae yeast was
inhibited by soluble WI-1, in a dose-dependent manner (Fig. 5C). Concentrations of as little as 0.4 µg WI-1 protein
per ml inhibited the TNF- response of both macrophages and
neutrophils.
Addition of soluble WI-1 to wells containing either WI-1 knockout yeast
or S. cerevisiae did not reduce the viability of either
macrophages or neutrophils. Viability of macrophages at 96 h of
incubation was 80–84% in wells with soluble WI-1 as compared with
80% in wells without it. Similarly, viability of neutrophils at
48 h of incubation was 83–85% in wells with WI-1 as compared
with 85% in wells without it. Thus, WI-1 is not impeding phagocyte
production of TNF- by inducing cell death.
WI-1 does not directly bind or inactivate TNF-, but acts at the
level of the phagocyte
WI-1 can serve as an adhesion-promoting protein (6, 13). Therefore, we wondered whether WI-1 could inhibit TNF-
by directly binding it after production and release by phagocytes. In
this scenario, WI-1 would neutralize released TNF- rather than
inhibiting TNF- production by cells. To test this, we investigated
whether purified WI-1 could neutralize rTNF- detected by WEHI cells
in the bioassay. Increasing concentrations of WI-1, used in amounts
that inhibited TNF- production in the preceding assays, had no
effect on the cytotoxic effects of rTNF- for WEHI cells in vitro
(Table I). Even at high concentrations of
40 µg/ml of exogenous WI-1, there was no interference with the
activity of TNF- upon WEHI cells. WI-1 added, by itself, to WEHI
cells also did not influence their viability. These results imply that
WI-1 interferes with triggering of phagocyte TNF- production, and
does not neutralize released TNF- or compete for its receptor.
|
modifies the course of experimental
blastomycosis
In the above findings, we observed reduced TNF- levels in
alveolar lavage fluids of mice infected with wild-type yeast, and also
observed in vitro that WI-1 on wild-type yeast down-regulates TNF-
production. These results imply that B. dermatitidis may
circumvent host defense by suppression of TNF- production. If so, we
hypothesized that exogenous TNF- administered to infected mice
should ameliorate experimental B. dermatitidis infection
with wild-type yeast. We tested this possibility initially by i.t.
administration of TNF- peptide to mice 72 h before infection.
TNF- peptide treatment decreased burden of infection 7 days after
infection, but the difference compared with control peptide treatment
did not achieve statistical significance even when a dose of 10 µg of
TNF- peptide was administered (data not shown).
We reasoned that the peptide approach may have been limited technically
due to differences in the local distribution of peptide and yeast, lack
of synchrony between TNF- activation of cells and infection with
yeast, or limited half-life of the peptide. To circumvent these
technical limitations, we used a gene therapy approach
(11) in which production of murine TNF- reportedly
peaks 48 h after recombinant adenovirus is administered to mice.
In these experiments, mice were given rTNF- adenovirus or
LacZ control adenovirus i.t. at the same time that
they were infected with wild-type yeast. The dose of recombinant
adenovirus was optimized in preliminary experiments in which mice
received virus (or PBS) alone and were then analyzed for TNF- in
lung tissue serially over 2 wk. TNF- was detected as early as 2 days
postinfection with virus, and peaked at day 4. A dose of 1.5 x
108 PFU of TNF- adenovirus prompted production
of peak levels of 874 ± 50 pg/ml of TNF- in lung tissues of
TNF- adenovirus-treated mice. In contrast, TNF- levels were
19 ± 3 pg/ml in mice that received control LacZ
adenovirus, and 21 ± 4 pg/ml in mice that received PBS alone.
Following infection with wild-type B. dermatitidis, the
burden of lung infection was reduced by 10-fold in mice that were
treated with rTNF- adenovirus as compared with control adenovirus or
PBS alone (Fig. 6A). Levels of
TNF- in the lung were correspondingly greater in rTNF-treated mice
as compared with control mice at the time they were analyzed for burden
of yeast infection (Fig. 6B). Treatment with a higher dose
of 3 x 108 PFU of adenovirus resulted in
higher levels of TNF- in lung tissue, but did not improve the
outcome of infection over the lower dose (data not
shown).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, a proinflammatory cytokine, is produced chiefly by
mononuclear phagocytes and plays a prominent role in host defense
against microbes. TNF- has pleotropic immunoregulatory effects
including synergizing with IL-12 to increase the production of IFN-
by NK cells, synergizing with IFN- to enhance microbicidal activity
of macrophages through NO-dependent mechanisms, and facilitating
maturation of the adaptive immune response (21, 22, 23, 24).
Production of TNF- has been shown to be essential in control of
various infectious diseases, including those due to medically important
fungi. For example, TNF- is required for both nascent and memory
immunity in defense against Histoplasma capsulatum
(20). Naive mice depleted of TNF- fail to control
infection because alveolar macrophages generate diminished amounts of
reactive nitrogen intermediates. TNF- depletion renders immune mice
more susceptible to secondary histoplasmosis, but the mechanism derives
from an elevation of type 2 cytokines, IL-4, and IL-10. TNF- also is
essential for the control of Cryptococcus neoformans
infection (21). Mice depleted of TNF- at or before
infection fail to develop acquired immunity, and demonstrate
uncontrolled proliferation of the fungus in the lungs and at sites of
dissemination, including the spleen and brain.
In this study, we show that TNF- similarly contributes to host
defense against B. dermatitidis infection. Neutralization of
TNF- led to a 10-fold elevation in the number of organisms in the
lung 1 wk after infection. The effect of neutralization was most
pronounced in mice infected with the WI-1 knockout strain, which is
typically cleared from lungs in an accelerated fashion during the first
week after challenge. In mice depleted of TNF-, the number of CFUs
for the knockout at 7 days postinfection instead increased to levels
normally seen only with the wild-type yeast. Neutralization did not
shorten survival of these infected mice, suggesting either that TNF
does not have the exclusive role in this infection that it does in
other fungal infections, or that neutralization may not have been as
complete in our study. Nevertheless, interference with early TNF-
production or function impairs control of pulmonary infection with
B. dermatitidis.
It is noteworthy, in view of an established role for TNF- in host
defense, that virulent B. dermatitidis yeast block the
production of TNF-. In comparing the behavior of virulent wild-type
yeast with that of WI-1 knockout yeast, we found marked differences
between these two strains in evoking TNF- production in different
experimental systems. Mice infected with the knockout had severalfold
more TNF- in alveolar lavage fluid than did mice infected with the
virulent strain. In vitro coculture of lung cells with each of the two
strains, respectively, confirmed sharp differences between them in
stimulating lung TNF- production. Furthermore, although the knockout
strain evoked robust production of TNF- in peritoneal macrophages
and neutrophils over the entire time-course of in vitro coculture, the
virulent wild-type strain initially stimulated a small amount TNF-
production, but then down-regulated production. Lastly, the stimulatory
phenotype of the knockout could be abolished by the addition of
wild-type yeast, demonstrating active down-regulation of TNF- by
virulent B. dermatitidis.
Multiple lines of evidence demonstrate that WI-1 on wild-type yeast is
responsible for blocking production of TNF-. First, and most obvious
is the fact that the strains used in this study are isogenic, differing
only in the expression of WI-1 (3). Second, the coating of
stimulatory knockout yeast with WI-1, or addition of purified WI-1 into
wells with knockout yeast, reversed the effect on immune cells from
stimulatory to inhibitory. Despite these observations, it remained
possible that WI-1 through its adhesive property could bind the yeast
and phagocyte together, yet on apposition, a second or additional
factor could deliver a down-regulatory signal to the cell. However,
soluble, purified WI-1 was shown to inhibit S.
cerevisiae-induced production of TNF- by phagocytes even though
WI-1 is unable to bind the surface of that yeast. Hence, it would
appear that WI-1 itself down-regulates TNF- production by
phagocytes. Yersiniae outer membrane proteins, virulence
factors of Yersiniae, also have been reported to interfere
with TNF- production by phagocytes in vitro (7, 25, 26, 27).
The biological significance of TNF- down-regulation by wild-type
yeast was investigated using a gene therapy approach. We reasoned from
in vitro data and alveolar lavage fluid analysis that mice infected
with wild-type yeast might have insufficient TNF- levels in vivo at
the site of infection, which could be restored via gene therapy. In
vivo lung TNF- levels were enhanced early in the course of infection
by delivery of a recombinant adenovirus vector at the time of infection
with B. dermatitidis. Gene therapy resulted in local
pulmonary production of rTNF-, a significant reduction in the
progression of lung infection with wild-type yeast, and correspondingly
elevated levels of TNF- in vivo compared with infected control mice
that received control adenovirus expressing LacZ. These
observations indicate that restoration of TNF- in infected mice
circumvented a pathogenic mechanism of B. dermatitidis
yeast. Presumably, some or most of the TNF- in adenoviral-treated
mice was being synthesized and released by epithelial cells, which were
not subject to the inhibitory effects of WI-1. However, TNF- levels
in the lungs of treated mice were lower than those observed in another
study involving treatment of Klebsiella pneumonia
(5), where similar doses of virus led to peak levels of
about 2 ng TNF- 2 days after viral infection. In contrast, we
detected up to about 800 pg of TNF- at the peak of production 4 days
after infection. Differences between the studies in lung TNF- levels
and kinetics raise the possibility that wild-type B.
dermatitidis and WI-1 may target epithelial cells in addition to
phagocytes. Nevertheless, local lung production of rTNF- offered a
beneficial effect upon the response to infection.
WI-1 appears to down-regulate TNF- production by acting directly on
the phagocyte. We investigated the possibility that WI-1 neutralizes
released TNF-, but found no evidence for this using the bioassay
with WEHI cells. The precise mechanism of action for WI-1 on phagocyte
TNF- production remains a matter of speculation. One possibility is
that WI-1 interferes with the signaling apparatus involved in TNF-
production in response to B. dermatidis. In
Yersiniae species, Yersiniae outer membrane
proteins have been shown to down-regulate phagocyte TNF- production
through interference with elements of the cell-signaling apparatus,
including the mitogen-activated protein kinases p38 and Janus kinase
(7, 25, 26, 27). Similar mechanisms remain to be investigated
in B. dermatitidis.
Another possibility is that WI-1 itself does not interfere with TNF-
but instead stimulates phagocyte release of soluble factors that behave
in an autocrine fashion to down-regulate TNF-. The finding that
small numbers of wild-type yeast in cocultures with large numbers of
knockout yeast-inhibited TNF- is consistent with such a mechanism.
Based on E:T ratios, small numbers of wild-type yeast could interact
with only a minor proportion of phagocytes in the wells. It is possible
that such triggered cells released a soluble factor, for example a
potent anti-inflammatory cytokine, which in turn inhibited the
other phagocytes in the well. Such a mechanism is not exclusive of
interference with signaling and both of them could be operative.
In summary, our study demonstrates that WI-1 down-regulates TNF- and
that this property enhances the virulence of wild-type B.
dermatitidis yeast. These findings reveal a new mechanism of
action of WI-1 and its pathogenic effects. However, this mechanism
cannot account entirely for the profound difference in virulence
between wild-type and WI-1 knockout yeast. This is because resupply of
TNF- by gene therapy does not convert the virulence phenotype of
wild-type yeast to that of knockout yeast. Therefore, other
WI-1-mediated effects are likely to synergize with TNF-
down-regulation to explain the phenotype.
We propose that WI-1 may perturb the cytokine network, disrupting a
proper balance of pro- and anti-inflammatory cytokines needed for
clearance of B. dermatitidis. TGF- is one example of an
anti-inflammatory cytokine that opposes the effects of TNF-
(28). TGF- can interfere with the development of
acquired immunity by reducing the number and function of T-lymphocytes
recruited to sites of infection (29). Leshmania
species stimulate TGF- as a virulence factor, which leads to
downstream immunoregulatory disturbances that promote infection
(30, 31). We postulate that an imbalance in the ratio of
opposing cytokines such as TNF- and TGF- may impede development
of cell-mediated immunity and clearance of fungal infection.
Elucidation of the cytokine network evoked by infection with the
virulent, wild-type yeast, as compared with the nonpathogenic WI-1
knockout yeast should address this model and provide a comprehensive
picture of the host: fungal pathogen interaction.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Bruce S. Klein, University of Wisconsin-Madison, 600 Highland Avenue, K4/434, Madison, WI 53792.
3 Abbreviations used in this paper: LacZ, -galactosidase; PMN, polymorphonuclear leukocyte; PEC, peritoneal exudate cells; i.t., intratracheally.
Received for publication September 20, 2000. Accepted for publication December 6, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
in pulmonary host defense in murine invasive aspergillosis. J. Immunol. 162:1633. is required for the development of protective T cell immunity to Cryptococcus neoformans. J. Immunol. 157:4529.[Abstract]
exacerbates primary and secondary pulmonary histoplasmosis by differential mechanisms. J. Immunol. 160:6072., and others. In Fundamental Immunology,
4th Ed. W. Paul, ed. Lipincott-Raven, Philadelphia, pp. 775–811.
production and downregulation of the MAP kinases p38 and JNK. Mol. Microbiol. 27:953.[Medline]
: the opposite sides of the avenue?. B. Beutler, ed. Tumor Necrosis Factors: The Molecules and Their Emerging Role in Medicine 131. Raven, New York.
. Annu. Rev. Immunol. 16:137.[Medline]
as a virulence mechanism for Leishmania braziliensis. Proc. Natl. Acad. Sci. USA 90:3442. in Leishmanial infection: a parasite escape mechanism. Science 257:545.This article has been cited by other articles:
|
|
 |
|
  A. Koneti, M. J. Linke, E. Brummer, and D. A. Stevens Evasion of Innate Immune Responses: Evidence for Mannose Binding Lectin Inhibition of Tumor Necrosis Factor Alpha Production by Macrophages in Response to Blastomyces dermatitidis Infect. Immun., March 1, 2008; 76(3): 994 - 1002. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. L. Bohse and J. P. Woods RNA Interference-Mediated Silencing of the YPS3 Gene of Histoplasma capsulatum Reveals Virulence Defects Infect. Immun., June 1, 2007; 75(6): 2811 - 2817. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. L. Bohse and J. P. Woods Expression and Interstrain Variability of the YPS3 Gene of Histoplasma capsulatum Eukaryot. Cell, April 1, 2007; 6(4): 609 - 615. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Lekkala, A. M. LeVine, M. J. Linke, E. C. Crouch, B. Linders, E. Brummer, and D. A. Stevens Effect of Lung Surfactant Collectins on Bronchoalveolar Macrophage Interaction with Blastomyces dermatitidis: Inhibition of Tumor Necrosis Factor Alpha Production by Surfactant Protein D. Infect. Immun., August 1, 2006; 74(8): 4549 - 4556. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. T. Brandhorst, G. M. Gauthier, R. A. Stein, and B. S. Klein Calcium Binding by the Essential Virulence Factor BAD-1 of Blastomyces dermatitidis J. Biol. Chem., December 23, 2005; 280(51): 42156 - 42163. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. L. Bohse and J. P. Woods Surface Localization of the Yps3p Protein of Histoplasma capsulatum Eukaryot. Cell, April 1, 2005; 4(4): 685 - 693. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. C. Herring, N. R. Falkowski, G.-H. Chen, R. A. McDonald, G. B. Toews, and G. B. Huffnagle Transient Neutralization of Tumor Necrosis Factor Alpha Can Produce a Chronic Fungal Infection in an Immunocompetent Host: Potential Role of Immature Dendritic Cells Infect. Immun., January 1, 2005; 73(1): 39 - 49. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. T. Brandhorst, M. Wuthrich, B. Finkel-Jimenez, T. Warner, and B. S. Klein Exploiting Type 3 Complement Receptor for TNF-{alpha} Suppression, Immune Evasion, and Progressive Pulmonary Fungal Infection J. Immunol., December 15, 2004; 173(12): 7444 - 7453. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. J. Rooney and B. S. Klein Sequence Elements Necessary for Transcriptional Activation of BAD1 in the Yeast Phase of Blastomyces dermatitidis Eukaryot. Cell, June 1, 2004; 3(3): 785 - 794. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. D. Sullivan, P. J. Rooney, and B. S. Klein Agrobacterium tumefaciens Integrates Transfer DNA into Single Chromosomal Sites of Dimorphic Fungi and Yields Homokaryotic Progeny from Multinucleate Yeast Eukaryot. Cell, December 1, 2002; 1(6): 895 - 905. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. C. Herring, J. Lee, R. A. McDonald, G. B. Toews, and G. B. Huffnagle Induction of Interleukin-12 and Gamma Interferon Requires Tumor Necrosis Factor Alpha for Protective T1-Cell-Mediated Immunity to Pulmonary Cryptococcus neoformans Infection Infect. Immun., June 1, 2002; 70(6): 2959 - 2964. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Finkel-Jimenez, M. Wuthrich, and B. S. Klein BAD1, an Essential Virulence Factor of Blastomyces dermatitidis, Suppresses Host TNF-{alpha} Production Through TGF-{beta}-Dependent and -Independent Mechanisms J. Immunol., June 1, 2002; 168(11): 5746 - 5755. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) and
neutrophils (PMNs) in response to B. dermatitidis
wild-type strain 26199 and WI-1 knockout strain 55. Cells were isolated
from mice and supernatants from coculture of cells with yeast in vitro
were analyzed for TNF-
