| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of
*
Cell Biology and
Pediatrics, Duke University Medical Center, Durham, NC 27710
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although SP-A and SP-D are present predominantly in the lung, another collectin named mannose-binding lectin (MBL) is found in serum and is capable of activating complement (3, 4, 5, 6). Like the collectins, complement also aids in the recognition and clearance of microorganisms from the body, although the mechanisms are much more complicated and highly regulated. The complement system is a group of >30 proteins present in serum that participates in innate immunity through a variety of mechanisms, such as activating immune cells, promoting inflammation at the site of infection, opsonizing pathogens, and directly lysing susceptible microorganisms. Within the complement system, there are three highly ordered activation pathways through which pathogens can be recognized and targeted for clearance: the classical pathway, the lectin pathway, and the alternative pathway (reviewed in Ref. 7). The classical pathway of complement is the most extensively studied pathway and is initiated by the binding of a subcomponent of the first complement protein, C1q, to the target microorganism. The lectin pathway is activated when microorganisms are recognized by MBL via its calcium-dependent lectin domains. Finally, activation can occur through the alternative pathway, which is viewed as somewhat less specific than the other two pathways, since there is no pathogen-specific recognition molecule in this pathway.
SP-A, MBL, and C1q all have a similar overall structure that resembles a bouquet of flowers. SP-A and MBL have globular C-terminal lectin domains that bind to sugars on microorganisms in a calcium-dependent manner. These globular lectin domains are contiguous with long triple-helical collagen-like domains that appear to be the stalks of the flowers. C1q also has collagen stalks, and the three proteins have sequence homology throughout this region where they share a common Gly-X-Y repeating sequence that forms the helical structure of their collagen-like domains (8). C1q, however, does not have the lectin domains that SP-A and MBL have and is not classified as a collectin. Instead, C1q has globular Ab binding domains that recognize Ag-bound IgG and IgM.
As predicted by similar overall protein structures, SP-A, MBL, and C1q have some functional similarities as well. For example, all three molecules have been shown to function in innate immunity by enhancing the phagocytosis of microorganisms (9, 10, 11, 12). All three proteins interact with the same C1q receptors (13, 14, 15, 16), further demonstrating their functional similarities. Recently, SP-A was shown to enhance the clearance of apoptotic cells (17), a function previously described for C1q (18, 19). Because of the significant structural and functional similarities between the lung collectins and the known complement activators, MBL and C1q, we investigated the hypothesis that SP-A may interact with and regulate proteins of the complement system.
Complement proteins are present in the lung, where they may play a role in host defense. Lung-specific cells, including alveolar macrophages and type II epithelial cells, synthesize and secrete in culture specific complement proteins of the classical and alternative pathways (20, 21, 22). In addition, functional classical pathway activity was demonstrated in human (23) and rabbit (24) bronchoalveolar lavage fluid, although the level of functional complement activity is less than that predicted by the complement protein levels in the lavage fluid. The reduced activity is at least partly explained by a complement inhibitor in rabbit lavage fluid as proposed by Giclas et al. (24). We investigated whether SP-A, the most abundant of the surfactant proteins, may be able to regulate complement activity in such a way in the lung.
Previous studies have confirmed that SP-A does not directly substitute for C1q to activate complement, since SP-A cannot restore classical pathway activity to C1q-depleted serum (23, 25). However, an interaction between SP-A and C1q has been reported, although the significance of this interaction has not yet been elucidated (26). We hypothesized that SP-A can bind to C1q and prevent it from activating complement, thereby preventing inflammation and damage to the delicate pulmonary epithelium.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Purified complement proteins, including C1q, activated C1s, and
proenzyme C1r were purchased from Advanced Research Technologies (San
Diego, CA). Human serum albumin (HSA) and BSA were purchased from Sigma
(St. Louis, MO). SP-A was purified from the alveolar lavage fluid of
pulmonary alveolar proteinosis patients (27). SP-A was
treated with polymyxin agarose to reduce endotoxin contamination, and
SP-A preparations used in these studies contained 
0.2 pg
endotoxin/µg protein. Recombinant rat SP-D (RrSP-D) was purified as
previously described (28). The state of multimerization of
RrSP-D has been characterized previously and consists primarily of
dodecamers and a small subpopulation of multimers (29).
Rat MBL was purified from rat serum (30).
Buffers
Isotonic veronal buffers were prepared as previously described (31). Normal ionic strength veronal-buffered saline contained 1 mM MgCl2, 0.15 mM CaCl2, and gelatin (GVBS++), and low ionic strength dextrose-GVBS++ contained either 1 mM MgCl2 and 0.15 mM CaCl2 (DGVBS++) or 10 mM EDTA (EDTA-DGVBS).
Iodination of proteins
Proteins were labeled with 125I using Iodobeads (Pierce, Rockford, IL). Two beads were washed with phosphate-buffered water (pH 7.0) and placed in the reaction test tube along with 50 µl of phosphate-buffered water (pH 7.0). Two microliters of 125I (0.2 mCi) was added to the reaction tube and incubated for 5 min at room temperature, after which 50 µl of the protein to be labeled was also added. The reaction proceeded for 30 min at room temperature. To remove the free unconjugated iodine, the entire volume was applied to a prewashed Micro Bio-Spin 6 chromatography column (Bio-Rad, Hercules, CA) and centrifuged over a collection tube for 4 min at 1500 x g. Three microliters of the collected sample was counted to determine the specific activity of the labeled protein. In addition, the concentration of the labeled proteins was determined by the bicinchoninic acid protein assay (Pierce) to calculate the specific activity of the labeled protein.
SP-A binding studies
A microtiter plate binding assay was used to compare SP-A binding to C1q vs intact C1. All incubations in microtiter wells were performed with a volume of 200 µl. Intact C1 (Advanced Research Technologies) was bound to immune complexes coated onto microtiter plate wells. Immune complexes were formed near equivalence on the wells of a 96-well MaxiSorp BreakApart microtiter plate (Nunc, Rochester, NY) as described below. BSA was dissolved in Dulbecco’s PBS (DPBS; Life Technologies, Gaithersburg, MD) at a concentration of 10 mg/ml and heat-aggregated at 56°C for 30 min. Heat-aggregated BSA was diluted 200-fold in 0.1 M sodium bicarbonate buffer (pH 9.75) to a final concentration of 50 µg/ml. Two hundred microliters (10 µg) of heat-aggregated BSA was added to each well and incubated for 1 h at 37°C. Following the incubation, wells were washed three times in DPBS containing 0.05% Tween 20 (DPBS-T). Immune complexes were formed on the plate by adding anti-BSA Ab at 18 µg/ml in DPBS to the wells and incubating for 1 h at 37°C. Again, wells were washed three times with DPBS-T.
Purified C1 was diluted 100-fold to 2 µg/ml in
DGVBS++ and incubated with the preformed immune
complexes, which resulted in C1 bound to the microtiter wells. Some
wells were then incubated for 30 min at 37°C in
DGVBS++ buffer to retain intact C1 on the wells.
Other wells were incubated in EDTA-DGVBS buffer to release the
associated C1r and C1s molecules that require calcium for binding to
C1q, resulting in wells coated with C1q only. All wells were washed
twice in the same buffer and then once in
DGVBS++. Iodinated SP-A was diluted 1/500 (
1
µg/ml final concentration) in DGVBS++ and
incubated with the wells for 30 min at room temperature as described
above. Alternatively, iodinated anti-human C1q Ab (The Binding
Site, Birmingham, U.K.) was diluted 1000-fold in
DGVBS++ and incubated with the wells. All wells
were washed twice in DGVBS++ and counted in the
gamma counter to determine how much SP-A or C1q was present.
To determine the molar ratio of SP-A binding to C1q, we used a microtiter plate binding assay. Iodinated C1q was diluted into unlabeled C1q as a tracer and incubated with microtiter plate wells at a concentration of 10 µg/ml (2 µg/well) in 0.1 M sodium bicarbonate buffer (pH 9.5). Wells were washed three times in DPBS-T. Using the specific activity of the iodinated protein, the amount of C1q bound was calculated to be 0.39 µg/well. To determine the maximal SP-A binding to wells coated with 0.39 µg unlabeled C1q under the same conditions as those described above, radiolabeled SP-A was diluted into unlabeled SP-A and added to C1q-coated wells at increasing concentrations in DGVBS++ buffer. Wells were incubated with radiolabeled SP-A for 30 min at room temperature. Wells were washed three times in the same buffer. Finally, wells were broken apart and counted in a gamma counter to determine the amount of SP-A associated with the wells. The maximal amount of SP-A bound to wells was calculated based on the specific activity of the protein. SP-A binding to MBL in DGVBS++ was also determined using the same microtiter plate binding assay in which wells were coated with 2 µg of rat MBL.
C1r2C1s2 binding studies
To determine whether
C1r2C1s2 binds to SP-A as
well as to C1q, microtiter plate wells were coated with 2 µg of
purified C1q or SP-A in 0.1 M sodium bicarbonate buffer (pH 9.75) for
1 h at 37°C. Wells were washed three times in DPBS-T and then
incubated with a mixture of excess unlabeled C1r (4 µg/ml; 1/250
dilution of stock) and iodinated C1s (2 µg/ml; 1/500 dilution of
stock) in DGVBS++ buffer for 30 min at room
temperature. Wells were washed three times with
DGVBS++ buffer and counted in the gamma counter
to determine the amount of bound
C1r2C1s2.
The ability of SP-A to inhibit C1 formation was tested using a modified C1r2C1s2 binding assay. Microtiter plate wells were coated with 2 µg of purified C1q in 0.1 M sodium bicarbonate buffer (pH 9.75) for 1 h at 37°C. Wells were then washed three times in DPBS-T and incubated with increasing concentrations of either SP-A or HSA in DGVBS++ buffer for 30 min at 37°C. The wells were washed once with DGVBS++, and a mixture of excess unlabeled C1r (1/250 dilution) and iodinated C1s (1/500) was added to all wells for 30 min at room temperature. Alternatively, SP-A or HSA was added to the wells at the same time as the C1r2[125I]C1s2 in some experiments. Wells were washed three times and counted in the gamma counter as described above.
C1 dissociation assay
Radiolabeled C1s was preincubated with a 2-fold molar excess of unlabeled C1q and C1r for 90 min in DGVBS++ buffer at room temperature to form intact C1. Two hundred microliters of the mixture was incubated with microtiter plate wells coated with immune complexes for 1 h at room temperature. Wells were washed three times in DGVBS++ buffer, and 200 µl of buffer containing no protein or containing 100 µg/ml HSA, C1 inhibitor, or SP-A was added to wells with C1. Following a 2-h incubation at room temperature, the supernatants were removed. The wells were washed twice with DGVBS++, and the washes were combined with the supernatants and counted in the gamma counter to determine how much 125I-labeled C1s dissociated from the wells. The wells were also broken apart and counted to determine how much C1r2125I-labeled C1s2 remained bound to C1q. The data were used to calculate the percentage of C1 dissociation.
C1q/C1 binding to immune complexes
We tested whether the binding of SP-A to C1q interfered with C1q binding to immune complexes. Purified C1q at 10 µg/ml was preincubated with SP-A in DGVBS++ buffer or with buffer alone for 90 min at room temperature before being added to microtiter plate wells coated with immune complexes. The C1q mixtures were incubated with immune complexes for 1 h at room temperature and subsequently washed three times with DGVBS++ buffer. The wells were incubated with radiolabeled anti-C1q Ab (1/500) at room temperature for 30 min to determine the amount of C1q bound to the immune complexes.
Similar studies were performed using purified C1 to determine whether SP-A also inhibited C1 binding to immune complexes. C1 was diluted 200-fold to 1 µg/ml in DGVBS++ and preincubated with increasing concentrations of either SP-A or HSA for 30 min at 37°C. Because of the high background binding of purified C1 to microtiter plate wells, wells were blocked for 2 h at 37°C with 1% gelatin in DPBS before addition of C1 to the wells in the presence or the absence of SP-A. The C1 mixtures were then added to microtiter plate wells coated with immune complexes and incubated for 30 min. C1 binding to immune complexes was detected with iodinated anti-C1q Ab (1/1000). Wells were washed three times with DGVBS++ buffer and counted in the gamma counter.
Complement activation assay
The ability of C1q to activate the classical pathway of complement in the presence and the absence of SP-A was tested in a microtiter plate assay using immune complexes as the complement target. This assay was developed and validated by Miletic et al. (32). Immune complexes were formed on microtiter plate wells as described above. Purified C1q (1 mg/ml stock) was diluted 1/8000 in DGVBS++ and coated onto the immune complexes by incubating 200 µl in each well for 30 min at 37°C. Control wells were incubated with DGVBS++ alone. Following the incubation, wells were washed three times with DGVBS++ and incubated with SP-A, HSA, or no protein in DGVBS++ for 30 min at 37°C. After the incubation, the buffer was removed, and the wells were washed three times in DGVBS++. C1q-depleted serum (Advanced Research Technologies) was diluted 1/100 in DGVBS++, and 200 µl was added to all wells. Complement activation by the C1q-depleted serum was allowed to proceed for 30 min at 37°C, after which time the wells were washed five times with DPBS-T.
The magnitude of complement activity in the wells was determined by measuring the amount of C4 bound to each well by ELISA using a biotinylated anti-C4 Ab (32). Biotinylated anti-C4 Ab was diluted 1/1000 in DPBS and added to all wells for 1 h at 37°C. Wells were washed three times in DPBS-T. Streptavidin peroxidase (1 mg/ml) was diluted 1/50,000 in DPBS and added to all wells for 1 h at 37°C. Wells were washed three times in DPBS-T, and the plate was developed for 1 h using Sigma FAST tablets according to the manufacturer’s instructions. The OD of all wells was read at 405 nm. Negative control wells coated with BSA, but no anti-BSA Ab, failed to activate complement. Also, wells coated with BSA/anti-BSA immune complexes did not activate complement when incubated with C1q-depleted serum.
Statistical analysis
All data are expressed as the mean ± SEM of three or more
separate experiments unless otherwise stated in the figure legend.
Statistical analysis was performed using two-tailed Student’s
t test for unpaired samples with unequal variances or using
a Tukey test. Differences were considered statistically significant at
p 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
SP-A has previously been reported to bind directly to C1q, but the
functional significance of this interaction has not been investigated.
Experiments were conducted to determine whether SP-A also binds to C1q
that is complexed with two molecules each of C1r and C1s as part of the
intact C1 molecule. Two molecules each of C1r and C1s interact to
produce a tetramer that binds to the collagenous region of C1q, forming
the intact C1 molecule (33). Microtiter plate wells were
coated with intact C1 and incubated either in
DGVBS++ buffer to retain intact C1 or in
EDTA-DGVBS buffer to dissociate C1r and C1s from C1q. This treatment
resulted in some wells coated with intact C1 and some wells coated with
C1q only. Microtiter plate binding assays using radiolabeled SP-A
showed that SP-A bound to C1q to a much greater extent (>5-fold) than
to intact C1, indicating that SP-A binding to C1q is greatly reduced by
the presence of C1r and C1s (Fig. 1
A). Duplicate wells incubated
with radiolabeled anti-C1q Ab showed that the same amount of C1q
was present on all wells (Fig. 1B) and that the decreased
binding of SP-A to C1 is due solely to the presence of C1r and C1s.
SP-A did not bind to wells coated with C1r and C1s alone (data not
shown). To characterize the specificity of the interaction of SP-A with
C1q, SP-A binding to the C1q homologue, MBL, was also examined. SP-A
bound to MBL, although the level of binding was only 15% that of
binding of SP-A to C1q (data not shown).
|
C). SP-A
binding to C1q-coated microtiter plate wells saturated at an SP-A
concentration of 32 µg/ml. The maximal amount of SP-A bound to wells
was 0.32 µg, based on the specific activity of the protein. This
corresponds to a molar ratio of 1 SP-A molecule bound to 1.7 C1q
molecules. Because the coating of C1q to the plate may obscure some
SP-A binding sites, this ratio is approximate. The actual ratio may be
closer to 1:1. Surfactant protein A blocks C1 complex formation
Because SP-A binding to C1q was greatly reduced in the presence of
the C1r2C1s2 tetramer, SP-A
may bind to C1q in the same region that C1r and C1s bind to C1q.
Therefore, we performed experiments to determine whether C1q that is
bound to SP-A could also bind to C1r and C1s. C1q-coated microtiter
wells were incubated with either SP-A or HSA and then tested for the
ability to bind radiolabeled
C1r2C1s2 tetramer
(C1r2125I-labeled
C1s2). Although preincubation of C1q-coated wells
with HSA did not affect the ability of C1r2125I-labeled C1s2 to
bind to C1q, SP-A significantly inhibited the association of
C1r2 125I-labeled
C1s2 with C1q in a dose-dependent manner,
indicating a likely competition between SP-A and
C1r2C1s2 for the same
region of C1q (Fig. 2A). SP-A
even inhibited C1 formation when it was added to C1q-coated wells at
the same time (without preincubation) that C1r2125I-labeled C1s2
was added, although slightly higher SP-A concentrations were needed to
achieve the same level of inhibition (Fig. 2B). RrSP-D,
another pulmonary collectin, did not inhibit C1 formation in the same
assay used in Fig. 2A at a concentration of 15 µg/ml (data
not shown).
|
).
|
Because SP-A does not interact with
C1r2C1s2 to activate
complement, and SP-A can block C1 formation from free C1q, C1r, and
C1s, SP-A was tested for its ability to inhibit classical pathway
complement activation. A microtiter plate assay for complement
activation was used instead of a standard hemolytic assay because
rather low concentrations of SP-A agglutinated target SRBC (data not
shown). Immune complexes were coated onto microtiter plate wells and
incubated with C1q. Wells were then incubated with SP-A or HSA, after
which a dilution of C1q-depleted human serum was added. After 30 min
the amount of C4 deposited onto the wells (a result of complement
activation) was quantified by ELISA using a biotinylated anti-C4
Ab. Wells incubated with 200 µg/ml SP-A showed a 50% decrease in the
amount of bound C4, demonstrating the ability of SP-A to inhibit
complement activation by C1q (Fig. 4).
|
To determine whether SP-A can dissociate preformed C1,
radiolabeled C1 was prepared by incubating radiolabeled C1s with excess
C1q and C1r. Radiolabeled C1 was bound to immune complexes on
microtiter plate wells. Various proteins were incubated with the
C1-coated wells and tested for the ability to dissociate C1, releasing
radiolabeled C1s into the supernatant. C1 inhibitor was used as a
positive control because of its ability to dissociate C1. C1-coated
wells were incubated with 100 µg/ml HSA, SP-A, or C1 inhibitor in
DGVBS++ buffer. Following the incubation,
radioactivity associated with the supernatants and wells was determined
and used to calculate the percent dissociation of C1. With the addition
of no protein, 7% of the complexes were dissociated after 2 h.
Although C1 inhibitor dissociated 73% of the C1, SP-A did not
dissociate preformed C1 (Fig. 5).
|
The effect of SP-A on additional C1q interactions was
investigated. To investigate the ability of SP-A to affect the
interaction of C1q with immune complexes, purified C1q was preincubated
with SP-A and subsequently tested for the ability to bind to immune
complexes. SP-A significantly inhibited the binding of C1q to immune
complexes in a dose-dependent manner (Fig. 6). Effective inhibition occurred at a
molar ratio of SP-A:C1q of 2:1. Because of the reported interaction of
SP-A with IgG (34), we excluded the possibility that the
inhibition was due to SP-A binding directly to the immune complexes by
showing that preincubation of immune complexes with SP-A before the
addition of C1q had no effect on C1q binding. SP-A similarly inhibited
the recognition of immune complexes by intact C1 in a dose-dependent
manner (Fig. 7). Significant inhibition
could be seen at 3.2 µg/ml SP-A, which corresponds to approximately a
4-fold molar excess of SP-A compared with C1q present in intact C1. The
inhibition was specific to SP-A, since excess HSA did not affect C1
binding.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The complement system and the pulmonary collectins (SP-A and SP-D) both contribute to lung host defense. The role of complement in lung host defense is evident in complement-deficient animals that show a defect in the clearance of bacteria from their lungs (35, 36, 37). Similarly, SP-A- and SP-D-deficient mice are also more susceptible to both bacterial (38, 39, 40, 41) and viral (42) infections in the lung. SP-A and SP-D bind to numerous microorganisms and enhance their phagocytosis by alveolar macrophages (reviewed in Refs. 1 and 2) and neutrophils (43) and may minimize the need for complement activation in the lung under normal conditions.
Interaction between the surfactant proteins and the complement system
has been hypothesized based on structural and functional similarities
between SP-A and known complement activators, C1q and MBL. Despite
these similarities, SP-A cannot substitute for C1q in activating the
classical pathway of complement in C1q-depleted serum (23, 25), probably because SP-A lacks a binding site for
C1r2C1s2 (Fig. 3). However,
a direct interaction between SP-A and C1q has been described
(26), although the functional significance of this
interaction has not been investigated. We hypothesized that SP-A may
regulate complement activity in the lung through its association
with C1q.
In the present study, we present evidence that the interaction between
SP-A and C1q regulates C1q-mediated complement activation via two
different mechanisms. First, SP-A prevents C1q from associating with
C1r and C1s to form the intact C1 complex that is required to activate
complement (Fig. 2). This finding is supported by studies reported by
Tenner and colleagues, who showed that SP-A inhibited the efficient
assembly of hemolytic C1 by 60% when SP-A was in 20-fold excess of
C1q (25), which is easily within the physiological ratio
for these two proteins within the lung (23). It was
hypothesized that this inhibition was due to competition between SP-A
and C1q for binding to
C1r2C1s2, but our data
indicate that SP-A binds to C1q and not to
C1r2C1s2 (Fig. 3). Second,
SP-A interferes with immune complex recognition by both C1q and C1
(Figs. 6 and 7). This represents a novel, potentially important
mechanism for regulation of complement activation. We showed that SP-A
directly inhibited classical pathway complement activation by
preventing C1q from restoring activity to C1q-depleted serum. The
ability to prevent complement activation was specific to SP-A, because
neither HSA nor the other pulmonary collectin, SP-D, was able to
duplicate this effect (Fig. 4).
Prevention of C1 assembly appears to be a common mechanism for inhibiting complement activation at the earliest stage. For example, both soluble C1q receptor (44) and neutrophil defensin (45) have been shown to bind to the collagen-like region of C1q and inhibit C1q-mediated complement activation. Like SP-A, soluble C1q receptor binding to C1q inhibited the association of C1r and C1s with C1q and inhibited C1q-mediated complement activation.
Many regulatory mechanisms exist to minimize complement-induced damage to host tissues. Complement activation produces a number of inflammatory molecules, including the anaphylatoxins C3a, C4a, and C5a. These complement products promote inflammation and recruit phagocytic cells such as macrophages and neutrophils to the site of infection, both of which aid in the clearance of microorganisms but could have detrimental effects in the lung. Phagocytic infiltrates produce anti-microbial molecules that can also damage host tissues as well as target microorganisms. Both fluid phase molecules (C1 esterase inhibitor, factor H, and factor I) and membrane-bound molecules (complement receptor 1, decay-accelerating factor, and membrane cofactor protein) have been shown to regulate different stages of complement activation (reviewed in Ref. 46). The critical need for proper complement regulation can be seen in patients with inherited C1 inhibitor deficiency who develop the complement disorder hereditary angioedema (47).
SP-A may serve a protective role in the lung by preventing C1q-mediated complement activation and inflammation along the delicate alveolar epithelium. An anti-inflammatory role for SP-A has been reported previously. SP-A has been shown to serve a protective role in the lung by inhibiting the production of proinflammatory cytokines that damage lung tissue in response to LPS in vitro (48) and in vivo (49) or bacterial pathogens in vivo (40, 41). Similarly, SP-A may prevent inflammation in the lung by inhibiting complement activation.
The mechanism by which SP-A may regulate complement activity in vivo
depends on whether C1q is free or bound to C1r and C1s in the lung. For
SP-A to inhibit complement activation by preventing C1 formation, C1q
in the lung must be free from C1r and C1s, since our data indicate that
SP-A is unable to dissociate preformed C1 (Fig. 5). Very low levels of
C1q have been detected in bronchoalveolar lavage fluids from healthy
human volunteers (23), but there are no data concerning
the amount of C1r and C1s in the lung. The presence of at least some
intact C1 in the lung is suggested by the fact that there is classical
pathway activity in lung lavage fluid (23, 24). C1q, C1r,
and C1s are normally present in serum as a complex (50, 51), and it is estimated that only about 10% of C1 in serum is
in the form of free C1q and free
C1r2C1s2. However, intact
C1 complex is very dilution sensitive to dissociation
(52). In the lung the C1 concentration may be low enough
that dissociation of the complex is favored, leaving most of the C1q
available for regulation by SP-A. Our data also suggest another
mechanism for complement regulation by SP-A in vivo; SP-A may prevent
the recognition of immune complexes by C1. By this mechanism, SP-A may
inhibit complement activation in the lung even if all the C1q in the
lung is associated with C1r and C1s.
The specificity of the binding of SP-A to C1q was investigated by characterizing the binding of SP-A to the C1q homologue, MBL. SP-A bound to MBL, although the level of binding was only 15% that of SP-A binding to C1q. Under normal conditions in rats, MBL is absent from the alveolar airspaces where SP-A is predominantly found (23). SP-A binding to MBL may not be physiologically relevant, since the two proteins are not normally localized to the same compartment. However, it is possible that MBL levels in the alveolar compartment increase during inflammation as a consequence of enhanced alveolar epithelial or endothelial permeability, in which case the interaction may be physiologically significant.
The mechanism for the interaction between SP-A and C1q has not been determined. The collagen-like region of C1q contains the binding sites for C1r and C1s (53). The fact that SP-A readily binds to free C1q but only poorly to intact C1, containing C1r and C1s, is consistent with the hypothesis that SP-A binds to the collagen-like region of C1q. It is also possible that SP-A binds to the globular domain of C1q. Because SP-A binding to C1q and, to a lesser extent, C1 inhibits the binding of those molecules to immune complexes, it is likely that SP-A binds to the C1q globular domains or close enough to the globular domains to sterically hinder their interactions with Igs. The binding of SP-A to C1q could also induce a conformational change in the C1q molecule that no longer favors binding to Igs. Further studies that investigate the binding of SP-A to purified globular domains (45) or collagen-like regions of C1q (54) obtained by digestion of the molecule are necessary to define the SP-A binding site on C1q.
Like the binding of other collectins or C1q to ligands, SP-A binding to C1q is greater in low ionic strength DGVBS++ buffer that favors low-affinity interactions. Although weak, these interactions are physiologically important, since a defense collagen can bind to multiple motifs through its multiple binding domains to strengthen the overall affinity (55).
This study provides evidence that SP-A blocks the formation of the intact C1 complex required for classical pathway activation. In addition, SP-A inhibits immune complex recognition by C1. By these mechanisms, SP-A can inhibit detrimental complement activation along the delicate alveolar epithelium. These findings may explain why complement activity in bronchoalveolar lavage fluid is lower than predicted by complement protein levels in lung lavage fluid (23, 24); SP-A may be a lavage-specific complement inhibitor. Future studies will investigate whether SP-A regulates complement activity in the lung in vivo.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Jo Rae Wright, Box 3709, Department of Cell Biology, Duke University Medical Center, Durham, NC 27710. E-mail address: j.wright{at}cellbio.duke.edu
3 Abbreviations used in this paper: SP-A, surfactant protein A; SP-D, surfactant protein D; C1qR, C1q receptor; CRD, carbohydrate recognition domain; DGVBS++, low ionic strength GVBS++; DPBS, Dulbecco’s PBS; E IgG, Ab-sensitized sheep erythrocytes; GVBS++, normal ionic strength veronal-buffered saline with gelatin; HSA, human serum albumin; MBL, mannose-binding lectin; RrSP-D, recombinant rat SP-D.
Received for publication June 7, 2001. Accepted for publication September 19, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
activity from LPS-stimulated macrophages. Am. J. Physiol. 271:L310. activity from LPS-stimulated macrophages. Am. J. Physiol. 15:L310.
This article has been cited by other articles:
|
|
 |
|
  D. Rittirsch, M. A. Flierl, D. E. Day, B. A. Nadeau, S. R. McGuire, L. M. Hoesel, K. Ipaktchi, F. S. Zetoune, J. V. Sarma, L. Leng, et al. Acute Lung Injury Induced by Lipopolysaccharide Is Independent of Complement Activation J. Immunol., June 1, 2008; 180(11): 7664 - 7672. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. S. Bolger, D. S. Ross, H. Jiang, M. M. Frank, A. J. Ghio, D. A. Schwartz, and J. R. Wright Complement levels and activity in the normal and LPS-injured lung Am J Physiol Lung Cell Mol Physiol, March 1, 2007; 292(3): L748 - L759. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. M. Lin and J. R. Wright Surfactant protein A binds to IgG and enhances phagocytosis of IgG-opsonized erythrocytes. Am J Physiol Lung Cell Mol Physiol, December 1, 2006; 291(6): L1199 - L1206. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Liu, Y. Endo, D. Iwaki, M. Nakata, M. Matsushita, I. Wada, K. Inoue, M. Munakata, and T. Fujita Human M-Ficolin Is a Secretory Protein That Activates the Lectin Complement Pathway J. Immunol., September 1, 2005; 175(5): 3150 - 3156. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. S. Ferguson, J. J. Weis, J. L. Martin, and L. S. Schlesinger Complement Protein C3 Binding to Mycobacterium tuberculosis Is Initiated by the Classical Pathway in Human Bronchoalveolar Lavage Fluid Infect. Immun., May 1, 2004; 72(5): 2564 - 2573. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. G. Brinker, H. Garner, and J. R. Wright Surfactant protein A modulates the differentiation of murine bone marrow-derived dendritic cells Am J Physiol Lung Cell Mol Physiol, January 1, 2003; 284(1): L232 - L241. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Yang, C. Milla, A. Panoskaltsis-Mortari, S. Hawgood, B. R. Blazar, and I. Y. Haddad Surfactant Protein A Decreases Lung Injury and Mortality after Murine Marrow Transplantation Am. J. Respir. Cell Mol. Biol., September 1, 2002; 27(3): 297 - 305. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. S. Butler, D. Sim, E. Tam, D. Devine, and C. M. Overall Mannose-binding Lectin (MBL) Mutants Are Susceptible to Matrix Metalloproteinase Proteolysis. POTENTIAL ROLE IN HUMAN MBL DEFICIENCY J. Biol. Chem., May 10, 2002; 277(20): 17511 - 17519. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||
, where
n = 1.