 |
INTRODUCTION |
Pulmonary surfactant is a mixture of lipids and proteins that
functions to keep alveoli from collapsing at expiration (1). Surfactant
protein A (SP-A)1 is the
major protein constituent of the surfactant (2). SP-A belongs to the
collectin subgroup of the C-type lectin superfamily along with
surfactant protein D (SP-D), mannose-binding protein (MBP), and
conglutinin (3). SP-A is now recognized as playing an important role in
regulating innate immunity within the lung. This protein enhances the
phagocytosis of Staphylococcus aureus (4), herpes simplex
virus type I (5), type A Hemophilus influenzae (6),
Mycobacterium tuberculosis (7), and Klebsiella (8) by alveolar macrophages. SP-A can bind with broad specificity to a
variety of microorganisms including herpes simplex virus type I (9),
Pneumocystis carinii (10), and Aspergillus
fumigatus (11). The characteristics of transgenic mice with null
alleles for SP-A (12-14) provide compelling in vivo
evidence that SP-A is an important component of the innate immune
system within the lung. Animals lacking SP-A exhibit reduced bacterial
clearance and elevated pulmonary inflammation in response to microbial challenge.
Gram-positive bacteria including S. aureus cause infections
that can be life-threatening. Peptidoglycan (PGN), a major cell wall
component of Gram-positive bacteria, is a polymer of alternating N-acetylglucosaminyl and N-acetylmuramyl glycan
whose residues are cross-linked by short peptides (15). PGN, like
lipopolysaccharides (LPS) from Gram-negative bacteria, can elicit the
excessive release of proinflammatory cytokines from immune cells, which
contribute to many of the adverse clinical manifestations of bacterial
infections (16-18). CD14 and Toll-like receptors (TLRs) function as
pattern-recognition receptors for these bacterial ligands (19, 20).
TLRs possess an intracellular domain homologous to that of
interleukin-1 receptor (21) and participate in NF-
B signaling
cascades elicited by LPS and PGN. In vivo studies with mice
harboring null alleles for TLR2 provide strong evidence that TLR2 is
responsible for PGN-induced signaling (22, 23). Recent in
vitro studies with overexpression experiments also demonstrate
that TLR2 confers cell responsiveness to PGN (24, 25). Although CD14
alone appears incapable of signaling because it lacks a transmembrane
domain, it is still capable of enhancing PGN-induced NF-B signaling
mediated by TLR2 (24).
SP-A binds rough serotypes but not smooth serotypes of LPS (26, 27).
The protein inhibits TNF-
secretion induced by smooth LPS (26, 28)
but modestly enhances TNF- release induced by rough LPS (26) in
alveolar macrophages and U937 cells. We have shown that the direct
interaction of SP-A with CD14 is the likely mechanism for modulating
LPS-elicited cellular responses (26). In addition to SP-A, the
collectins, SP-D, and MBP also bind CD14 (29, 30), suggesting this may
be an important property of this protein family. The interaction of
SP-D with CD14 may be also accompanied by modulation of the cellular
response to ligands such as LPS. SP-A-deficient mice exhibit
significant increases in the production of TNF- and nitric oxide
after intratracheal instillation of smooth LPS when compared with wild
type mice (31). Intratracheal administration of SP-A to SP-A-deficient
mice diminished the production of the proinflammatory cytokines. Taken
together with the in vitro observations of the inhibitory
function of SP-A on smooth LPS-elicited TNF- secretion (26, 28),
there is growing evidence that SP-A promotes an anti-inflammatory
response to some bacterial ligands.
The role of SP-A in modulating innate immunity may be a key element to
understanding the dual requirement of the lung to remain relatively
quiescent in its inflammatory response to routine daily burdens of
inspired LPS and easily dispatched microorganisms while remaining
competent to mount a potent and vigorous response to specialized
pulmonary pathogens. Many inhaled pathogens that reach the alveolus are
thought to interact immediately with lung collectins because these
proteins are highly enriched at this biological interface.
PGN elicits many of the clinical manifestations of Gram-positive
organisms, and we focused on the interactions of SP-A with PGN and the
consequences of the interaction upon TNF- secretion by alveolar
macrophages and U937 cells. The specific objectives of this study were
to determine 1) the interaction of SP-A with PGN, 2) the role of SP-A
in modulating leukocyte cytokine responses to PGN, 3) the interaction
between SP-A and TLR2, and 4) the role of SP-A in altering PGN
interaction with TLR2. Our findings demonstrate that PGN is not a
ligand for SP-A and that SP-A reduces the PGN-elicited proinflammatory
cytokine release by reducing TLR2-PGN interactions.
| |
EXPERIMENTAL PROCEDURES |
Cells and Reagents--
The macrophage-like cell line U937
(JCRB9021) was obtained from the Health Science Research Resources Bank
(Osaka, Japan). L929 murine fibroblast cells were kindly provided by
Dr. Kazuko Kajiyama (Chugai Pharmaceutical, Tokyo, Japan). The cells
were maintained in endotoxin-free RPMI 1640 medium from Sigma with 10%
heat-inactivated fetal calf serum (FCS; Invitrogen). Human embryonic kidney (HEK) 293 cells (CRL-1573) were obtained from American
Type Culture Collection and grown in Dulbecco's modified Eagle's
medium containing 10% FCS. PGN derived from S. aureus was
purchased from Fluka. The PGN contained 66.84 ± 2.48 pg
(mean ± S.E., n = 3) of endotoxin/mg when
measured by a Limulus amebocyte lysate assay system
(ENDOSPECY; Seikagaku Kogyo, Tokyo, Japan). Smooth LPS
(Escherichia coli O26:B6) and rough LPS (Salmonella minnesota Re595) were obtained from Sigma.
Isolation of Rat Alveolar Macrophages--
Alveolar macrophages
were isolated from bronchoalveolar lavage fluids of Sprague-Dawley
rats. The lungs were lavaged with pyrogen-free saline (Otsuka
Pharmaceutical Co., Tokyo Japan), and alveolar macrophages were
sedimented by centrifugation at 150 × g. Isolated
macrophages were plated at 5 × 105 cells/well in
24-well plates (Falcon) in RPMI 1640 medium containing 10% FCS. The
cells were allowed to adhere for 2 h and then used for the
experiments after washing with PBS to remove the unattached cells.
SP-A--
Human surfactant was isolated from the bronchoalveolar
lavage fluids of patients with alveolar proteinosis as described
previously (32). After delipidation of surfactant with 1-butanol (33), SP-A was purified by affinity chromatography on mannose-Sepharose 6B
followed by gel filtration as described previously (34). Rat SP-A was
also purified from the lung lavage fluids of Sprague-Dawley rats that
had been given intratracheal instillation of silica (35) by the method
described above. Recombinant rat SP-A was expressed in Chinese hamster
ovary K1 cells and purified as described previously (11).
Removal of Endotoxin in SP-A Preparations--
Endotoxin was
removed from SP-A preparations using polymyxin B-agarose (Sigma) in the
presence of octyl-
-D-glucoside as described by McIntosh
et al. (28). The endotoxin-depleted human SP-A contained 0.159 ± 0.023 pg (mean ± S.E., n = 5) of
endotoxin/µg of protein when measured by the Limulus
amebocyte lysate assay system.
Binding of SP-A to PGN--
Two methods were used to evaluate
SP-A-PGN interactions. One method using microtiter wells was adapted
from that described for LPS binding (26, 29). Briefly, 5 µg/well PGN,
Re595 LPS, or O26:B6 LPS in 20 µl of ethanol was added onto
microtiter wells (Immulon 1B; Dynex Laboratories, Chantilly, VA), and
the solvent was evaporated in ambient air. Nonspecific binding to the
wells was blocked with PBS (pH 7.4) containing 0.1% (v/v) Triton X-100 and 3% (w/v) skim milk (buffer A). The indicated concentration of rat
SP-A (50 µl/well) in 20 mM Tris (pH 7.4) containing 0.15 M NaCl, 5 mM CaCl2, and 1 mg/ml BSA
was then added and incubated for 3 h at 37 °C. After the
incubation, the wells were washed with buffer A and were then incubated
with 10 µg/ml anti-SP-A IgG (50 µl/well) in buffer A for 1 h
followed by the incubation with horseradish peroxidase-labeled
anti-rabbit IgG (1:1000) for 1 h. The peroxidase reaction was
finally performed using o-phenylenediamine as a substrate
after washing the wells with PBS containing 0.1% (v/v) Triton X-100.
The binding of SP-A to PGN or LPS was detected by measuring absorbance
at 492 nm.
Because the PGN obtained was insoluble, we also carried out the binding
study by sedimentation based on methods used for the binding of SP-A to
phospholipid liposomes (36). We first tested whether insoluble PGN was
localized in the pellet after sedimentation. Ten micrograms of
insoluble PGN was centrifuged at 10,000 × g at room
temperature for 10 min and separated into the supernatant and the
pellet. The ability of each fraction to induce TNF- secretion from
U937 cells was examined by measuring secreted TNF- using an L929
cell cytotoxicity assay. For the binding study, the rat SP-A
preparation (200 ng/tube) in 50 µl of 20 mM Tris buffer
(pH 7.4) containing 0.1 M NaCl and 2% (w/v) BSA (buffer B)
was centrifuged at 10,000 × g at room temperature for
10 min. The PGN preparation was also centrifuged at the same time. The
supernatant of the protein solution was added to the PGN pellet. The
mixture of the protein and the PGN was suspended and incubated in the
presence of 5 mM CaCl2 or 5 mM EDTA
for 1 h at 37 °C. The mixture was then centrifuged at
10,000 × g at room temperature for 10 min. The supernatant was stored, and the resultant pellet was resuspended in 50 µl of buffer B and centrifuged again. The supernatants were then
combined, and the pellet was suspended in 100 µl of the buffer B. The
amount of SP-A in each fraction was determined by sandwich enzyme-linked immunosorbent assay using anti-SP-A IgG.
Induction of TNF- Secretion--
U937 cells (5 × 105/well) were placed on 24-well plates (Falcon) and
induced to differentiate by incubation with 10 nM PMA for
24 h. The cells were further incubated in the absence of PMA for
24 h in RPMI 1640 medium containing 10% FCS. Alveolar macrophages (5 × 105/well) were incubated on 24-well plates for
2 h after isolation from the bronchoalveolar lavage fluids of
rats. The indicated concentration of SP-A was preincubated with the
cells 30 min before adding PGN. The indicated amount of PGN was then
added into the well and incubated for 5 h at 37 °C with 5%
CO2. The cultured medium was collected and assayed for
TNF- concentrations using the L929 cell cytotoxicity assay as
described below.
Measurment of TNF- Concentration--
TNF- secretion into
medium from U937 cells or rat alveolar macrophages was measured using
L929 cell cytotoxicity assay performed by a modified method (26) based
on that described by Flick and Gifford (37). Briefly, the L929 cells
were seeded into 96-well plates (6 × 104/well) in 100 µl/well RPMI 1640 containing 10% FCS and 2 µg/ml actinomycin D
(Sigma). Dilutions of standard recombinant TNF- (1-50 pg/ml)
(PeproTech, Rocky Hill, NJ) or samples (1:10 for U937 cells or 1:80 for
rat alveolar macrophages) in a volume of 100 µl/well were added, and
the cells were incubated at room temperature for 15 min followed by
incubation at 37 °C overnight with 5% CO2. On the next
day the medium was removed, and the cells were stained with 0.2% (w/v)
crystal violet for 10 min. The wells were then washed with water, and
100 µl/well 33% acetic acid was added to extract the retained
crystal violet. The absorbance at 570 nm was finally measured.
NF-B Reporter Gene Assay--
The 2.6-kilobase cDNA for
human TLR2 was obtained by reverse transcription-polymerase chain
reaction using RNA isolated from U937 cells. NF-B activation was
measured as previously described (24, 38). HEK293 cells were plated at
1 × 105/well on 24-well plates on the day before
transfection. The cells were transiently transfected by
FuGENETM 6 transfection reagent (Roche Molecular
Biochemicals) according to the manufacturer's instruction, with 0.02 µg of TLR2 cDNA in pcDNA3.1(+) plasmid vector (Invitrogen),
0.1 µg of an NF-B reporter construct (pNF-B-Luc, Stratagene),
and 0.01 µg of a construct directing expression of Renilla
luciferase (pRL-TK, Promega). Forty-eight hours after transfection, the
cells were stimulated with 5 µg/ml PGN for 6 h in the absence or
the presence of SP-A (50 µg/ml), which was preincubated with the
cells for 2 h before adding PGN to the wells. Luciferase activity
was measured using the dual-luciferase reporter assay system (Promega)
according to manufacturer's instruction.
Binding of SP-A to sTLR2--
Expression and purification of a
soluble form of recombinant extracellular domain of TLR2 (sTLR2) will
be described elsewhere.2
sTLR2 consists of the putative extracellular domain
(Met1-Arg587) of TLR2 and a 6-histidine tag at
its C-terminal end and was expressed in baculovirus-insect cell
expression system. The sTLR2 protein was isolated from the culture
medium by an affinity column of nickel-nitrilotriacetic acid beads, as
previously described for the isolation of sCD14 (29).
The ligand blot analysis was performed by the method based on that
described for the binding of SP-A to sCD14 (26). Two micrograms of
sTLR2 was electrophoresed on 13% polyacrylamide gel in the presence of
SDS under denaturing conditions and transferred onto polyvinylidene
difluoride membrane. The membrane was incubated with 5 mM
Tris buffer (pH 7.4) containing 0.15 M NaCl, 5 mM CaCl2, 5% (w/v) BSA (buffer C) to block
nonspecific binding. The membrane was then incubated with 5 µg/ml
human SP-A in buffer C at room temperature for 3 h. The membrane
was washed and incubated with anti-SP-A IgG (10 µg/ml) for 90 min
followed by the incubation with horseradish peroxidase-labeled
anti-rabbit IgG (1: 1500) for 60 min. The binding of human SP-A to
sTLR2 was finally visualized by chemiluminescence (Super Signal, Pierce).
The binding of SP-A to sTLR2 was also examined using microtiter wells.
sTLR2 or BSA (10 µg/ml, 50 µl/well) was coated onto microtiter
wells. The wells were incubated with buffer A to block nonspecific
binding. The indicated concentrations of human or rat SP-A in 10 mM HEPES (pH 7.4) containing 0.15 M NaCl, 5 mM CaCl2, and 1% (w/v) BSA was then incubated
at 37 °C for 6 h. Anti-SP-A polyclonal IgG (20 µg/ml) was
incubated at 37 °C for 60 min after washing the wells with the
buffer A followed by the incubation with horseradish peroxidase-labeled
anti-rabbit IgG (1:1000) at 37 °C for 60 min. The wells were washed
with PBS (pH 7.4) containing 0.1% (v/v) Triton X-100. The binding of
human or rat SP-A was finally detected by using
o-phenylenediamine as a substrate for the peroxidase
reaction, and the absorbance at 492 nm was measured.
125I-Labeled SP-A was used to analyze the binding
characteristics of SP-A to sTLR2. Human SP-A was iodinated by the
method of Bolton and Hunter (39) using the Bolton-Hunter reagent
(Amersham Biosciences, Inc.). The specific activities of the
125I-labeled protein used ranged from 77 to 97 cpm/ng of
protein. In all preparations, more than 95% of the radioactivity was
precipitable by 10% (w/v) trichloroacetic acid. The binding of
125I-SP-A to sTLR2 was examined using microtiter wells.
sTLR2 or BSA (10 µg/ml, 50 µl/well) was coated onto microtiter
wells. The wells were incubated with 10 mM HEPES buffer (pH
7.4) containing 0.15 M NaCl, 5 mM
CaCl2, and 5% (w/v) BSA (buffer D) to block nonspecific
binding. The indicated concentrations of 125I-SP-A in
buffer D were incubated with the protein-coated wells at 37 °C for
6 h. The wells were washed with PBS (pH 7.4) containing 0.1%
(v/v) Triton X-100, and then the bound 125I-labeled
proteins were solubilized in 200 µl of 0.1 M NaOH, and the radioactivity was quantified using a 
-radiation counter. In some
experiments, 125I-SP-A (5 µg/ml) was preincubated with 50 µg/ml anti-human SP-A monoclonal antibody PE10 or control monoclonal
antibody 3C9 or mouse control IgG or 0.2 M carbohydrate in
the buffer D at 37 °C for 1 h. The mixture was then incubated
with sTLR2 coated onto microtiter wells at 37 °C for 6 h. The
wells were then washed, and the amounts of the labeled protein binding
to the wells were determined as described above. A binding reaction
with heat-treated (100 °C, 5 min) 125I-SP-A was also
performed. Five mM EDTA was also included in the binding
buffer instead of 5 mM CaCl2 to examine the
effect of Ca2+ on the SP-A binding to sTLR2.
Effect of SP-A on the Binding of sTLR2 to PGN--
sTLR2 was
iodinated by the method of Bolton and Hunter (39) using the
Bolton-Hunter reagent (Amersham Biosciences, Inc.) as described above
for the iodination of SP-A. Ten µg of PGN in 20 µl of ethanol was
added to the microtiter wells, which were subsequently dried in ambient
air. The wells were incubated with buffer D to block nonspecific
binding. 125I-sTLR2 protein (0.5-5 µg/ml) was then
incubated with PGN coated onto the microtiter wells at 37 °C for
6 h. The wells were washed with PBS (pH 7.4) containing 0.1%
(v/v) Triton X-100, and then the proteins were solubilized in 200 µl
of 0.1 M NaOH, and the bound radioactivity was quantified
using a -radiation counter. To examine the effect of SP-A on the
binding reaction with PGN, human SP-A (5 or 100 µg/ml) was
preincubated with 125I-sTLR2 (1 or 2 µg/ml) at 37 °C
for 2 h before addition to the wells.
| |
RESULTS |
PGN Is Not a Ligand for SP-A--
To examine the interaction of
SP-A with PGN, we first performed binding experiments using a solid
phase assay. We have previously shown that SP-A bound to rough LPS (Re
595) with high affinity but not to smooth LPS (O26:B6) (26). In the
current study we used these two different strains of LPS as controls.
SP-A avidly bound to Re595 LPS coated onto microtiter wells, but
the protein exhibited almost no binding to O26:B6 LPS. When 0-10
µg/ml SP-A was incubated with the solid phase PGN, no significant
binding was observed at any concentrations of the protein.
We next carried out the binding assay in solution by a sedimentation
method. The localization of PGN was determined by examining the ability
of the centrifuged fractions to induce TNF- secretion from U937
cells. The pellet fraction contained more than 99% TNF--inducing activity. Admixture of SP-A and PGN followed by centrifugation revealed
that no significant levels of the protein were recoverable in the
pellet fraction. The recovery of SP-A in the supernatant was
independent of the presence of Ca2+, EDTA, or PGN in
various combinations. These findings demonstrate that PGN from S. aureus is not a ligand for SP-A.
SP-A Attenuates the Leukocyte Response to PGN--
Next we
investigated whether SP-A alters cellular responses induced by PGN in
the U937 human macrophage-like cell line. The cells (5 × 105 cells/well) were first differentiated by treatment with
PMA and subsequently incubated with PGN for 5 h in the presence or
the absence of human SP-A. TNF- secretion into medium was quantified using the L929 cell cytotoxicity assay. PGN stimulated TNF-
secretion in a concentration-dependent manner (Fig.
1). When 25 µg/ml polymyxin B sulfate
was included in the medium in the presence of 10 µg/ml PGN, the level
of TNF- secreted was almost equivalent to that observed without
polymyxin B (8300 versus 8800 pg/ml, mean of two
experiments), indicating that the observed response is not due to
endotoxin contamination. Human SP-A alone also did not induce TNF-
secretion in the absence of PGN. Coincubation of SP-A with the cells
and 1-10 µg/ml PGN inhibited PGN-elicited TNF- secretion from
U937 cells (Fig. 1). Ten µg/ml recombinant rat SP-A also decreased
the level of TNF- secretion to 35.3% (mean of two experiments) of
that stimulated by 1 µg/ml PGN, indicating that the general effect is
not species-specific, but the potency may vary with protein source.
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 1.
SP-A inhibits peptidoglycan-induced
TNF- secretion by U937 cells.
Differentiated U937 cells (5 × 105) were preincubated
with (closed squares) or without (open squares)
human SP-A (10 µg/ml) for 30 min at 37 °C with 5%
CO2. PGN (1-10 µg/ml) was then added, and the cells were
further incubated for 5 h. The culture medium was collected, and
TNF- secretion into the medium was determined by L929 bioassay as
described under "Experimental Procedures." The data shown are mean + S.E. of three experiments. *, p < 0.05, when
compared with incubation without SP-A.
|
|
Stimulation of U937 cells with 100 and 1000 nM PMA
increased TNF- secretion to 2807 ± 861 and 6067 ± 1146 pg/ml (n = 3, mean ± S.E.), respectively. When 20 µg/ml human SP-A was coincubated with the cells in the presence of
PMA, there was no attenuation of TNF- secretion. The amounts of
TNF- secreted in the presence of SP-A were 2907 ± 1033 and
7120 ± 1028 pg/ml (n = 3, mean ± S.E.) for
100 and 1000 nM PMA, respectively. These results clearly indicate that the inhibitory effect of SP-A is not simply nonspecific for the activation of the cells used but specific for the PGN-induced cell responses.
To further confirm the inhibitory effect of SP-A on TNF- production
induced by PGN, we performed the experiments using rat alveolar
macrophages. One to ten µg/ml PGN was incubated with alveolar
macrophages in the presence or the absence of 10 µg/ml human SP-A
(Fig. 2). PGN induced very high levels of
TNF- release in a concentration-dependent manner in
these macrophages. Up to 157.2 ± 56.0 ng/ml (mean ± S.E.,
n = 3) TNF- was secreted at 10 µg/ml PGN. SP-A
markedly reduced TNF- secretion at all concentrations of PGN tested.
Rat SP-A also decreased TNF- secretion to 52.6% (mean of three
experiments) of the level observed at 2 µg/ml PGN in the absence of
the protein. When various concentrations of human SP-A were incubated
with alveolar macrophages stimulated with 5 µg/ml PGN, the protein
significantly reduced TNF- secretion at all concentrations tested
(Fig. 3). Half-maximal inhibition was
observed at ~20 µg/ml of human SP-A.
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 2.
SP-A attenuates peptidoglycan-induced
TNF- secretion by rat alveolar
macrophages. Rat alveolar macrophages (5 × 105)
were preincubated with (closed squares) or without
(open squares) human SP-A (10 µg/ml) for 30 min at
37 °C. PGN(1-10 µg/ml) was then added, and the cells were further
incubated for 5 h. The culture medium was collected, and TNF-
secretion was determined by L929 bioassay as described under
"Experimental Procedures." The results are expressed as percent of
PGN (10 µg/ml)-stimulated TNF- secretion in the absence of SP-A.
The mean value of PGN (10 µg/ml)-induced TNF- secretion in the
absence of SP-A was 157.2 ng/ml (100%). The data shown are mean + S.E.
of three experiments. *, p < 0.05, when compared with
incubation without SP-A.
|
|
View larger version (18K):
[in this window]
[in a new window]
|
Fig. 3.
Concentration-dependent
inhibition of peptidoglycan-elicited TNF-
secretion by SP-A. Rat alveolar macrophages (5 × 105) were incubated with 5 µg/ml PGN in the presence of
0-50 µg/ml human SP-A for 5 h. The culture medium was
collected, and TNF- secretion into the medium was determined by L929
bioassay as described under "Experimental Procedures." The results
are expressed as percent of PGN (5 µg/ml)-stimulated TNF-
secretion in the absence of SP-A. The mean value of PGN (5 µg/ml)-induced TNF- secretion in the absence of SP-A was 134.6 ng/ml (100%). The basal TNF- secretion without PGN (broken
line) was 13.6 ng/ml (n = 4). The data shown are
mean ± S.E. of three experiments. *, p < 0.01 and **p < 0.05, when compared with values obtained
without SP-A.
|
|
SP-A Attenuates PGN-induced NF-B Activation in TLR2-transfected
HEK293 Cells--
Because recent in vivo and in
vitro studies (22-25) demonstrate that PGN-induced signaling is
mediated by TLR2, we examined the effect of SP-A on PGN-induced NF-B
activation in TLR2-transfected HEK293 cells. PGN (5 µg/ml) stimulated
NF-B reporter activity in TLR2-transfected cells. SP-A treatment (50 µg/ml) alone did not affect the basal NF-B activity (Fig.
4). Coincubation of SP-A and PGN with the
cells significantly attenuated the activity of measured NF-B. SP-A
reduced the NF-B activity by ~34%. These data clearly indicate
that SP-A can alter TLR2-mediated signaling.
View larger version (20K):
[in this window]
[in a new window]
|
Fig. 4.
SP-A attenuates peptidoglycan-induced
NF-B activation in TLR2-transfected HEK293
cells. HEK293 cells (1 × 105/well) were
transfected with 0.02 µg of TLR2 cDNA in pcDNA3.1(+) together
with 0.1 µg of an NF-B reporter construct (pNF-B-Luc) and 0.01 µg of Renilla luciferase control reporter plasmid
(pRL-TK). Forty-eight hours after the transfection, the cells were
stimulated with 5 µg/ml PGN in the presence or the absence of human
SP-A (50 µg/ml) for 5 h. Luciferase activities were determined
as described under "Experimental Procedures." The results are
expressed as percent of PGN-stimulated NF-B activity in the absence
of SP-A. The data shown are the mean + S.E. of three experiments. *,
p < 0.05, when compared with PGN treatment without
SP-A.
|
|
SP-A Binds the Extracellular Domain of TLR2--
We constructed a
soluble form (sTLR2) of the extracellular TLR2 domain consisting of
Met1-Arg587 and a 6-histidine tag at its
C-terminal end. sTLR2 was expressed using a baculovirus-insect cell
expression system, and the protein was isolated by affinity
chromatography. When sTLR2 was electrophoresed and transferred onto a
polyvinylidene difluoride membrane, it was visualized as a band of
~75 kDa by Coomassie Blue staining (Fig.
5A). For the ligand blot
analysis, the membrane was incubated with human SP-A or BSA and probed
with anti-SP-A IgG. SP-A that had bound to the membrane was detected as
a band corresponding to that of sTLR2 (Fig. 5A),
demonstrating that SP-A binds to TLR2. We further examined the binding
of SP-A to sTLR2 coated onto microtiter wells. SP-A bound to the solid
phase sTLR2 but not BSA in a concentration-dependent manner
(Fig. 5B). From these results, we conclude that SP-A binds to the extracellular domain of TLR2.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 5.
SP-A binds to the extracellular domain of
Toll-like receptor 2. Panel A, ligand blot analysis.
Two µg of the soluble form of the recombinant extracellular domain of
the Toll-like receptor 2 (sTLR2) was electrophoresed and transferred
onto polyvinylidene difluoride membrane. sTLR2 on the membrane was
visualized by Coomassie Brilliant Blue staining (Coomassie
stain). The polyvinylidene difluoride membrane was also incubated
with human SP-A (5 µg/ml) or BSA as a control, and the binding of
SP-A to the membrane was detected by anti-SP-A IgG (ligand binding), as
described under "Experimental Procedures." st,
standards; Ab, antibody. Panel B,
concentration-dependent binding of human SP-A to sTLR2.
sTLR2 (10 µg/ml, 50 µl) (closed circles) or BSA
(open circles) was coated onto microtiter wells and
incubated with the indicated concentrations of human SP-A at 37 °C
for 6 h. The binding of SP-A to sTLR2 was detected using anti-SP-A
IgG, as described under "Experimental Procedures." The data shown
are mean ± S.E. of three experiments.
|
|
To characterize the SP-A binding to sTLR2, 125I-SP-A was
also used for the microtiter well binding. Human SP-A exhibited
concentration-dependent and saturable binding to solid
phase sTLR2 (Fig. 6). An analysis of the
binding as described by Klotz (40) (Fig. 6, inset) reveals that half-maximal binding occurs at 2.26 ± 0.53 µg/ml
(n = 3, mean ± S.E.). When calculated as an
oligomeric molecular mass (41), the binding of human SP-A to sTLR2 has
a K1/2 of 1.413 ± 0.331 nM
(n = 3, mean ± S.E.). Heat treatment (100 °C,
5 min) of 125I-labeled human SP-A and the inclusion of EDTA
in the binding buffer almost completely abolished the binding of SP-A
to sTLR2 (Fig. 7A). However,
coincubation of 125I-labeled human SP-A with excess
carbohydrates failed to reduce the SP-A binding to sTLR2. The effect of
anti-SP-A monoclonal antibody was also examined (Fig. 7B).
The control antibodies (3C9 and mouse IgG) showed some nonspecific
interference with the binding. Anti-human SP-A monoclonal antibody,
PE10, almost completely blocked the SP-A binding to sTLR2. These
results indicate that the binding of SP-A to sTLR2 is
Ca2+-dependent. Because the epitope for PE10
localizes to a region contiguous with human SP-A
Thr184-Gly194 (42), the data also strongly
implicate the carbohydrate recognition domain of the protein in TLR2
recognition.
View larger version (18K):
[in this window]
[in a new window]
|
Fig. 6.
Binding of 125I-SP-A to sTLR2 is
high affinity and saturable. The indicated concentrations of
125I-labeled human SP-A were incubated at 37 °C for
6 h with sTLR2 coated onto microtiter wells. The radioactivity
bound to the wells was determined as described under "Experimental
Procedures." The results represent the specific binding calculated by
subtracting the amounts of the labeled protein binding to BSA-coated
wells from total binding. The data shown are the mean ± S.E. of
three experiments. The inset contains a Klotz plot of the
binding data.
|
|
View larger version (23K):
[in this window]
[in a new window]
|
Fig. 7.
EDTA and monoclonal antibody PE10 block SP-A
binding to sTLR2. Panel A, 125I-Human SP-A
(5 µg/ml) was incubated in the absence (none) or the
presence of 5 mM EDTA, 0.2 M mannose, glucose,
galactose, or N-acetylglucosamine with sTLR2 coated onto
microtiter wells. The amount of 125I-SP-A binding to sTLR2
was determined as described under "Experimental Procedures."
Heat-treated (100 °C, 5 min) 125I-SP-A was also tested
(boiled). Panel B, 125I-human SP-A (5 µg/ml) was incubated with sTLR2 in the absence (none) or
the presence of anti-human SP-A monoclonal antibody (PE10), control
monoclonal antibody (3C9), or control mouse IgG. The results are
expressed as relative SP-A binding (%) compared with that obtained for
the control binding occurring in the absence of inhibitors
(none, 100%). The data are the mean + S.E. of three
experiments. *, p < 0.002, compared with control
binding (none).
|
|
SP-A Decreases PGN-TLR2 Interactions--
We next examined the
interaction between sTLR2 and solid phase PGN. When 0.5-5 µg/ml
125I-sTLR2 was incubated with 10 µg/well PGN coated onto
microtiter wells, the protein exhibited
concentration-dependent binding (Fig. 8A). We next investigated the
effect of human SP-A on the binding of sTLR2 to PGN. Human SP-A (0, 5, or 100 µg/ml) was preincubated with 1 µg/ml 125I-sTLR2,
and the mixture of SP-A and sTLR2 was further incubated with solid
phase PGN. The amount of sTLR2 binding to PGN in the presence of SP-A
was compared with that in the absence of SP-A. SP-A attenuated the
binding of 125I-sTLR2 to PGN in a
concentration-dependent manner (Fig. 8B). When
100 µg/ml SP-A was incubated with 1 µg/ml 125I-sTLR2,
the binding of the labeled sTLR2 to PGN was significantly decreased to
the level of ~35% of that in the absence of SP-A. Increasing the
125I-sTLR2 concentration to 2 µg/ml reduced the SP-A
inhibitory effect (32% inhibition at 100 µg/ml). The experiments
with heat-treated (100 °C, 5 min) SP-A showed that the denatured
SP-A failed to inhibit 125I-sTLR2 binding to PGN. These
results indicate that SP-A decreases the binding of sTLR2 to PGN and
that the inhibitory effect of SP-A on the sTLR2 binding to PGN is
dependent upon the relative concentrations of sTLR2 and SP-A. These
results are consistent with those observed for the inhibitory effect of
SP-A on NF-B activation in TLR2-transfected cells. Taken together,
these results demonstrate that SP-A alters the interaction of TLR2 with
PGN and down-regulates TLR-mediated signaling.
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 8.
SP-A attenuates the binding of sTLR2 to
peptidoglycan. Panel A, binding of sTLR2 to
peptidoglycan. Ten µg of PGN was coated onto microtiter wells and
incubated with 0.5-5 µg/ml 125I-sTLR2 at 37 °C for
6 h. The amount of 125I-sTLR2 binding to PGN was
determined as described under "Experimental Procedures." The data
shown are the mean ± S.E. of three experiments. Panel
B, effect of SP-A on the binding of sTLR2 to PGN. Human SP-A (0, 5, or 100 µg/ml) was preincubated with 125I-sTLR2 (1 µg/ml) at 37 °C for 2 h, and this mixture was further
incubated at 37 °C for 6 h with 10 µg/well PGN coated onto
microtiter wells. The amount of 125I-sTLR2 binding to PGN
was determined as described under "Experimental Procedures." The
results are expressed as percent of sTLR2 binding to PGN in the absence
of SP-A. The data shown are means + S.E. of three experiments. *,
p < 0.05, when compared with control binding performed
without SP-A.
|
|
| |
DISCUSSION |
This study demonstrates that SP-A inhibits PGN-induced TNF-
secretion by U937 cells and alveolar macrophages. In addition, SP-A
attenuates PGN-elicited NF-B activation in TLR2-transfected HEK293
cells. Direct binding studies also provide clear evidence that SP-A and
PGN bind the extracellular TLR2 domain. Binding competition analysis
demonstrates that SP-A can alter the interaction of TLR2 with PGN.
Direct interaction of SP-A with TLR2 and down-regulation of TLR
signaling by SP-A constitute the likely mechanisms by which SP-A
inhibits PGN-induced cellular responses. This study supports an
antiinflammatory role for SP-A in controlling the host response to
PGN from S. aureus.
Initial experiments performed in this study focused on interactions
between SP-A and PGN. Solid phase binding did not reveal any
interaction between SP-A and PGN. However, we were not able to quantify
the amount of PGN bound to the wells. We therefore conducted binding
assays in solution with sedimentable PGN. Consistent with the solid
phase binding results, SP-A failed to bind to PGN in solution. Taken
together, we conclude that PGN derived from S. aureus is not
a ligand for SP-A.
Previous studies from this (26) and other (28) laboratories demonstrate
that SP-A inhibits smooth LPS-elicited TNF- secretion by alveolar
macrophages and U937 cells. SP-A has also been shown to down-regulate
proinflammatory cytokine production elicited by Candida
albicans (44). In contrast to these studies, another study (45)
reports that SP-A stimulates the production of cytokines including
TNF-, interleukin-1, and interleukin-6. However, in vivo
studies (31) using SP-A (
/) mice give results consistent with an
inhibitory role of SP-A on inflammatory cytokine production. SP-A
(/) mice produced significantly increased TNF- and nitric oxide
in the lung compared with SP-A (+/+) mice after intratracheal administration of smooth LPS. Instillation of SP-A to SP-A (/) mice
restored regulation of proinflammatory cytokine production. Another
study (43) with SP-A (/) mice has shown that infection with group B
streptococcus and H. influenzae increased the
proinflammatory cytokines in the lung. In this report we sought to
determine whether the antiinflammatory role for SP-A was also
applicable to a PGN stimulus. This study demonstrates in
vitro that SP-A attenuates TNF- release induced by PGN derived
from S. aureus. The SP-A effect on PGN responsiveness occurs
with the U937 cell line as well as primary alveolar macrophages. The
results obtained from this and the previous in vivo (31, 43)
and in vitro (26, 28, 44) studies are consistent with the
idea that SP-A plays a role in modulating cytokine production and
inflammatory responses during bacterial infection within the lung.
We also sought to elucidate the mechanism by which SP-A inhibits
PGN-elicited TNF- secretion. Because this study has shown that SP-A
does not bind to PGN, the mechanism of the inhibitory effect must be
different from that by which MBP inhibits cellular responses caused by
streptococcal cell wall components (46). In the latter case, the direct
interaction of MBP with streptococcal rhamnose glucose polymer (RGP)
inhibits RGP-induced TNF- secretion. A previous study (26) from this
laboratory suggests that one of the possible mechanisms by which SP-A
modulates LPS-induced cytokine expression is likely to be due to the
interaction of SP-A with the LPS receptor, CD14. CD14 and TLR2 function
as pattern recognition receptors for PGN (19, 20). Because
TLR2-deficient mice were hyporesponsive to PGN (22, 23) and
transfection with a TLR2 cDNA conferred cell responsiveness to PGN
on HEK293 cells (24), TLR2 is concluded to be responsible for
PGN-induced cellular responses. Although CD14 enhances PGN-induced cell
signaling mediated through TLR2, TLR2 alone can induce significant
NF-B activation in response to PGN. In addition, because PGN- or
LPS-elicited TNF- expression is coupled with TLR-mediated NF-B
signaling, we examined the effect of SP-A on PGN-induced NF-B
activation in TLR2-transfected HEK293 cells. SP-A significantly reduced
the measured NF-B activity, indicating that SP-A can alter
TLR2-mediated NF-B signaling.
Next, we constructed a soluble form of recombinant extracellular TLR2
domain (sTLR2) and isolated sTLR2 protein expressed by the
baculovirus-insect cell expression system. The direct binding of PGN to
the extracellular TLR2 domain has now been demonstrated. Additional
details about this binding reaction will be described elsewhere.2 The preincubation of SP-A with sTLR2
significantly reduced the binding of sTLR2 to PGN. These results
obtained from the cell-free system are essentially consistent with
those obtained using U937 cells and alveolar macrophages, although the
magnitudes of the SP-A inhibition are different. The expression of TLR2
has been demonstrated in U937 cells (38), leukocytes (47), and lung (48). Taken together, these data support the idea that the direct interaction of SP-A with the extracellular TLR2 domain interfere with
PGN binding to TLR2, resulting in decreased TLR2-mediated NF-B
signaling and reduced TNF- secretion from immune cells. From the
present and previous (26) studies we now propose that SP-A modulates
cellular responses induced by bacterial ligands through direct
interactions with pattern recognition receptors, CD14, and/or TLR.
SP-A almost completely abrogated PGN-induced TNF- secretion from
U937 cells. However, this protein did not completely inhibit PGN-induced cytokine release in alveolar macrophages. The difference between SP-A's inhibitory effects on U937 cells and alveolar
macrophages may be due to the different capacity of these cells to
secrete TNF-. Ten µg/ml PGN induced secretion of more than 150 ng/ml TNF- in alveolar macrophages, whereas the same concentration of PGN applied to U937 cells produced 7.5 ng/ml TNF-. SP-A exhibited a greater inhibitory effect at lower concentrations of PGN than at
higher concentrations in alveolar macrophages, with 80% inhibition at
1 µg/ml versus 33% inhibition at 10 µg/ml. SP-A did
significantly attenuate, but did not completely abrogate, PGN-induced
NF-B activation in TLR2-transfected cells or sTLR2 binding to solid phase PGN. The absence of a one to one correlation of the SP-A effect
between TNF- secretion and NF-B activity or sTLR2 binding to PGN
may be a consequence of TLR2 overexpression or altered affinity for
sTLR2. The naturally occurring soluble form of mouse TLR4, which is
expressed by alternatively spliced mouse TLR4 mRNA, has been shown
to only partially block LPS-elicited NF-B activation (49),
indicating that the extracellular domain alone may not explain all the
activity of the receptor-ligand interaction. Clearly, more detailed
biochemical data regarding SP-A-TLR2 and PGN-TLR2 interactions and the
mechanism of signal transduction and the terminal inhibitory event are required.
We have previously shown that SP-A and SP-D bind CD14 by different
mechanisms (29). The SP-A neck domain and SP-D lectin domain
participate in CD14 binding. SP-A and SP-D recognize a peptide portion
and a carbohydrate moiety, respectively, of CD14. MBP also binds CD14
in a manner similar to that of SP-A (30). CD14 and TLRs possess
homologous structures consisting of leucine-rich repeats characteristic
of a short -sheet and -helix (50). CD14 and TLR2 contain 10 and
19 leucine-rich repeats, respectively. Because SP-A binds to the CD14
region containing leucine-rich repeats and also binds to deglycosylated
sTLR2,3 we infer that SP-A
may interact with the leucine-rich repeat region of TLR2. The binding
of rat SP-A to sCD14 was blocked by a monoclonal antibody that binds to
the SP-A neck domain but was not attenuated in the presence of EDTA
(29), indicating that the SP-A neck domain participates in CD14
binding. In this study the binding of human SP-A to sTLR2 was abolished
by a monoclonal antibody whose epitope is located at a region
contiguous to the human SP-A region
Thr184-Gly194 (42). In addition, it was
blocked by the presence of EDTA, indicating that the SP-A binding to
sTLR2 is Ca2+-dependent and that the
carbohydrate recognition domain is involved in sTLR2 binding. These
studies reveal the different mechanisms of the SP-A binding to the
pattern recognition receptors. The molecular and mechanistic details by
which SP-A and other collectins interact with TLRs are now under investigation.
It is relatively difficult to determine the actual SP-A concentration
in vivo since the epithelial lining fluid of the alveolus (alveolar hypophase) cannot be directly measured. Nevertheless, the
SP-A concentrations can be estimated based on the recovery of the
protein in the bronchoalveolar lavage fluids and the extrapolated hypophase volume (100-1000 µl/lung) (51, 52). The calculated SP-A
concentrations in the alveolar hypophase range from 180 µg/ml to 1.8 mg/ml (53-55). The levels of SP-A appear to vary in diseased states,
indicating complex responses under conditions of physiological stress.
In the rat model, the levels of SP-A mRNA and protein are elevated
in response to intratracheal administration of LPS (56). SP-A recovered
in the lavage fluid also increases in AIDS-related pneumonia (57).
However, in some situations (58), the SP-A concentration in lavage
fluid decreases in patients with bacterial pneumonia. Although one
cannot yet determine the exact concentrations of SP-A in the hypophase
of healthy and diseased human lungs, the SP-A concentrations used in
these studies are within the best estimates of the physiological ranges.
The respiratory system continually faces exposure to airborne LPS, PGN,
and microbes. SP-A may play an important role in the elimination of
these microbes by enhancing phagocytosis by alveolar macrophages.
Although SP-A enhances microbial clearance, it also appears to be an
important element in dampening the inflammatory response to some
organisms and their derivative cell surface components. Alveolar
macrophages and neutrophils produce proinflammatory cytokines through
the CD14/TLR pathway in response to microbial components including PGN,
LPS, and lipoteichoic acid. TNF- is a pivotal mediator of the host
responses to infections and triggers inflammatory responses. Because
overproduction of TNF- can cause chronic pathological states
especially in the lung, the inhibitory function of SP-A on TNF-
release may be crucial regulatory component for controlling pulmonary inflammation.
In conclusion, this study demonstrates that SP-A inhibits TNF-
secretion induced by PGN. The results also reveal that SP-A directly
binds TLR2, alters the interaction of TLR2 with PGN, and attenuates
downstream signaling events. These findings provide one mechanistic
framework by which SP-A can regulate inflammatory responses in the
alveolar compartment.
Secretion in U937 Cells and Alveolar
Macrophages by Direct Interaction with Toll-like Receptor 2*
§,
TOP
ABSTRACT
INTRODUCTION
To whom correspondence should be addressed: Dept. of
Biochemistry, Sapporo Medical University School of Medicine, South-1 West-17, Chuo-ku, Sapporo 060-8556, Japan. Tel.: 81-11-611-2111; Fax:
81-11-611-2236; E-mail: kurokiy@sapmed.ac.jp.
