Download Complement Levels and Activity in the Normal and LPS - AJP-Lung

PageArticles
1 of 46 in PresS. Am J Physiol Lung Cell Mol Physiol (October 27, 2006). doi:10.1152/ajplung.00127.2006
Complement Levels and Activity in the Normal and LPS-Injured Lung
Molly S. Bolger1, DeAndre S. Ross1, Haixiang Jiang2, Michael M. Frank2, Andrew J.
Ghio3, David A. Schwartz4, Jo Rae Wright1
1
Department of Cell Biology, Duke University Medical Center, Durham, North Carolina
2
Department of Pediatrics, Duke University Medical Center, Durham, North Carolina
3
Human Studies Division, Environmental Protection Agency, Chapel Hill, North Carolina
4
NIEHS and NTP, Durham, North Carolina
Copyright © 2006 by the American Physiological Society.
Page 2 of 46
2
Abstract
Complement, a complex protein system, plays an essential role in host defense through
bacterial lysis, stimulation of phagocytosis, recruitment of immune cells to infected
tissue, and promotion of the inflammatory response. Although complement is most well
characterized in serum, complement activity is also present in the lung. Here we further
characterize the complement system in the normal and inflamed lung. By Western blot,
C5, C6 and Factor I were detected in bronchoalveolar lavage (BAL) at lower levels than
in serum, whereas C2 was detected at similar levels in BAL and serum. C4 binding
protein (C4BP) was not detectable in BAL. Exposure to Lipopolysaccharide (LPS)
elevated levels of C1q, Factor B, C2, C4, C5, C6 and C3 in human BAL, and C3, C5, and
Factor B in mouse and rat BAL. Message for C1q-B, C1r, C1s, C2, C4, C3, C5, C6,
Factor B, and Factor H, but not C9 or C4BP, was readily detectable by RT-PCR in
normal mouse lung. Exposure to LPS enhanced Factor B expression, decreased C5
expression, and did not affect C1q-B expression in mouse and rat lung. BAL from rats
exposed to LPS had a greater ability to deposit C3b onto bacteria through complement
activation than did BAL from control rats. In summary, these data demonstrate that
complement levels, expression and function are altered in acute lung injury and suggest
that complement within the lung is regulated to promote opsonization of pathogens and
limit potentially harmful inflammation.
Keywords
Inflammation, Bronchoalveolar Lavage, Lipopolysaccharide, C3b Deposition,
Complement expression
Page 3 of 46
3
Introduction
In spite of constant exposure to inhaled pathogens, allergens and noxious gases, the lung
is normally free of infection and inflammation primarily due to a variety of defense
systems that function locally to promote pathogen clearance. These defense systems
include mucus, the mucociliary system, phagocytes, and opsonic proteins including
surfactant proteins and complement (Wright 2005).
The complement system is a family of proteins, most well characterized in serum,
that contributes to host defense via a variety of mechanisms. For example, complement
proteins enhance bacterial clearance by opsonizing bacteria and enhancing ingestion by
phagocytes. Bacteria may also be directly lysed by complement proteins. In addition
complement cleavage products act as chemoattractants to recruit immune cells to the site
of infection where they can further contribute to pathogen clearance. It has also been
demonstrated that complement activation products can contribute directly to vascular
leakage, allowing serum proteins to enter the infected tissue from the bloodstream (Bossi,
et al. 2004).
The complement system is activated in response to invading pathogens through
three different pathways – classical, alternative and lectin. Complement protein C1q
binds to antibody on a pathogen surface to activate the classical pathway; cleaved C3
(C3b) binds directly to bacterial surfaces to activate the alternative pathway; and
mannose binding lectin and ficolin bind directly to pathogen surfaces to activate the
lectin pathway. The three pathways converge at C3, where the important opsonin, C3b,
is formed and the pathway continues to C5, where the important chemoattractant and
polymorphonuclear neutrophil (PMN) activator, C5a is formed. The terminal
Page 4 of 46
4
complement proteins C6, C7, C8, and C9 form the membrane attack complex that can
directly lyse bacteria.
The major site of complement synthesis is the liver, which secretes complement
proteins into the circulation (Alper, et al. 1980), although other tissues have been shown
to produce complement proteins (reviewed in (Colten and Garnier 1998)), and tissue
specific complement synthesis may be important for supporting an immediate, local
immune response. For example, human and rat lung epithelial cells and alveolar
macrophages have been shown to synthesize complement proteins in vitro (Ackerman, et
al. 1978;Cole, et al. 1983;Strunk, et al. 1988). In addition, complement proteins have
been detected in human lung washings (bronchoalveolar lavage, BAL), and the relative
BAL levels of the different proteins in the normal lung appear to differ from levels in
human serum (Reynolds and Newball 1974;Robertson, et al. 1976;Watford, et al. 2000).
Complement activity has also been observed in the lung. Classical pathway
activity was measured in human BAL using standard in vitro activation assays (Watford,
et al. 2000). The importance of complement in the lung in vivo is supported by the fact
that recurrent respiratory infections occur in many patients with deficiencies in
complement proteins or complement receptors (Figueroa and Densen 1991).
Additionally, mice infected intranasally with S. pneumonia are more likely to die if they
do not express C1q (Brown, et al. 2002), and mice treated with cobra venom factor to
deplete their complement system are less able to clear S. pneumonia or P. aeruginosa
from their lungs (Gross, et al. 1978). Furthermore, in adult respiratory distress syndrome
(ARDS), complement activation products measured in serum can be predictive of disease
outcome (Langlois and Gawryl 1988;Langlois and Gawryl 1988). Although complement
Page 5 of 46
5
has a role in lung host defense, its activity in the normal lung is reduced compared to
serum even at same total protein concentrations (17-40% of serum activity) (Watford, et
al. 2000), which may be due in part to complement inhibition by Surfactant Protein A
(SP-A), a lung specific immune molecule (Watford, et al. 2001).
Lung specific regulation of complement synthesis and activation may be
important in maximizing the host response to inhaled pathogens while minimizing
complement functions that could damage the delicate lung epithelium through promoting
an inflammatory response. In particular, C5a has been shown to promote inflammation
by causing a massive influx of neutrophils and protein into the alveolar space (reviewed
in (Guo and Ward 2005)). Complement mediated tissue damage is normally modulated
by a group of at least 11 regulatory proteins that control complement activation by
inhibiting formation of active complement complexes or by cleaving specific active
complement protein fragments (Liszewski and Atkinson 1998). C1inhibitor, which
inhibits classical pathway activation through interaction with the C1 complex, has been
detected in human BAL and is expressed in mouse lung (Vinci, et al. 2002;Watford, et al.
2000). However, little else is know about which complement regulatory proteins are
present in the lung and how they may function to protect this delicate tissue.
Complement is also regulated by changes in gene expression in response to infection or
injury. The expression of several complement components increases during the acute
phase, a period of systemic inflammation marked by increases in liver synthesis of certain
proteins (Schutte, et al. 1975). In vitro, cytokines, hormones, and lipopolysaccharide
(LPS) are able to modulate complement expression (Falus, et al. 1987;Falus, et al.
Page 6 of 46
6
1990;Lake, et al. 1994;Lappin and Whaley 1991;Rothman, et al. 1989;Strunk, et al.
1985;Zach, et al. 1992) (and reviewed in (Volanakis 1995)).
The goals of this study were to characterize further the complement proteins that
are present in the normal lung and in a LPS model of acute lung inflammation and to
determine if complement proteins are synthesized in the lung. Additionally, the
hypothesis was tested that complement is regulated in the lung to enhance phagocytosis
and to limit potentially harmful inflammation. Evidence is presented that levels of
complement proteins in the lung differ from levels in the serum and that the lung
responds to LPS-injury with changes in expression of specific complement components
and with an enhanced ability to opsonize bacteria.
Page 7 of 46
7
Materials and Methods
Buffers
Buffers were: isotontic VBS with metals and gelatin (GVBS++), prepared as previously
described (Wagner, et al. 2001;Watford, et al. 2002); EDTA-GVBS- -, made without
MgCl2 or CaCl2 and containing 0.01M EDTA; EGTA- GVBS++, containing 5 mM MgCl2
and 8 mM EGTA; Dulbecco’s Phosphate Buffered Saline (DPBS, Sigma- Aldrich, St.
Lois, MO); and Tris Buffered Saline (TBS; 20 mM Tris-base, 137 mM NaCl, 3.8 mM
HCl, pH 7.6).
Human LPS exposure and bronchoalveolar lavage (BAL)
Normal human bronchoalveolar lavage (BAL) was obtained from healthy human
volunteers as previously described (Watford, et al. 2000). BAL was obtained from
healthy human volunteers before and at various time points after exposure to aerosolized
corn dust extract (delivering a dose of 0.4 µg endotoxin/kg body mass) as previously
described (Deetz, et al. 1997). All BAL samples were aliquoted and stored at -80°C
before analysis by Western blots.
Serum collection
Blood was collected via cardiac puncture (mouse and rat) or phlebotomy (human). Blood
was allowed to sit at room temperature for 45 minutes and then on ice for 30 minutes.
Clotted blood and cells were removed from serum by centrifugation at 2095 g for 10
minutes at 4°C. Serum was stored at -20ºC.
Page 8 of 46
8
Animal models
Pathogen free, male, C57/BL6 mice (Jackson Labs, Bar Harbor, Maine) were
anesthetized by injection of ketamine and xylazine and a brief inhalation of isofluorane.
An incision was made to expose the trachea and 50 µl of lipopolysaccharide (LPS from
Escherichia coli 026:B6, Sigma-Aldrich) dissolved in sterile saline or only saline (as a
vehicle control) was injected into the trachea using a 1/2 cc insulin syringe (28G1/2,
Becton Dickinson, Franklin Lakes, NJ). A dose of 100 µg LPS/kg mouse body mass was
used. After injection the incision was sutured and the mice were allowed to recover.
Next, after various lengths of time post-LPS exposure, mice were euthanized by
intraperitoneal injection of pentobarbital and exsanguination.
Pathogen free, Sprague-Dawley, male rats (Taconic Farms, Hudson, NY) were
anesthetized by inhalation of halothane. The trachea was visualized using a
laryngoscope, a cannula was inserted, and 200 µl of sterile saline with or without LPS
was instilled through the trachea at a dose of 100 µg LPS/kg rat body mass. Rats were
allowed to recover, and then after various lengths of time, euthanized by intraperitoneal
injection of pentobarbital and transection of the renal artery.
Western Blots
Human, mouse or rat BAL, normal human, mouse or rat serum, purified human
complement proteins (C1q, C2, C4, C5, C6, Factor B, C4 Binding Protein (C4BP) and
Factor I from Advanced Research Technology (now Complement Technology, Inc.,
Tyler, TX)), and normal human, mouse or rat IgG (Sigma-Aldrich) were separated by
electrophoresis on a 15% SDS PAGE gel (7.5% for C4BP). For gels with normal human
Page 9 of 46
9
BAL, equal total protein was added in lanes containing normal human serum or BAL.
For human mouse or rat BAL from LPS-treated individuals, equal volumes of BAL fluid
were loaded onto the gel for each time point. Samples were transferred overnight from
the gel to nitrocellulose membrane (Protran®, Schleicher and Schuell BioScience, Keene,
NH). Blots were blocked with TBS-Tween (0.1% Tween -20 (Sigma-Aldrich a) in
TBS) containing 5% dry milk for 1 hour at room temperature. Blots were washed 3 times
with TBS-Tween and then incubated with primary antibody (at concentrations listed in
Table 1) in TBS-Tween containing 1% dry milk for either 2 hours at room temperature
(C4, C5 for human samples, Factor I and SP-A), 2 hours at 4°C (C1q, C6 and Factor B),
or 4°C overnight (C2, C3, C5 for mouse and rat samples, and C4BP). Anti- human C3,
C5, and Factor B antibodies were found to cross-react with mouse and rat samples;
treatment of blots was the same as for human samples unless otherwise stated. Blots
were washed 3 times with TBS-Tween or TBS-Tween containing 1% dry milk. Next,
blots were incubated with an HRP-conjugated secondary antibody diluted in TBS-Tween
(TBS-Tween containing 1% dry milk for C4BP) for 45 minutes at room temperature.
Blots were washed as above, plus an additional wash of 1-4 hours or overnight (C4BP)
and treated with ECL. Film was exposed to the blot for 10 seconds-18 hours and
developed. In the case of some proteins (C6, human Factor B, Factor I), samples were
denatured but not reduced for the SDS PAGE gel, because reduced IgG was similar in
size to the reduced complement protein.
Page 10 of 46
10
RT-PCR
Lung or liver tissue was removed from untreated mice or from mice or rats treated with
LPS or saline. Tissue was homogenized in lysis buffer (Biorad AurumTM Total RNA
Mini Kit) using the FastPrepTM FP120 with zirconium oxide grinding media (1.251.6mm, Glenmills, Clifton, NJ). RNA was collected using the manufacturer’s protocol
for the above Biorad kit, including DNase treatment. RNA was eluted from the Biorad
columns in 30 µl elution buffer. The concentration of the RNA was determined using an
OD 260/280 ratio and the quality of the RNA was evaluated by analysis on a 1% agaroseformaldehyde gel.
RNA was used in a reverse transcriptase reaction to synthesize cDNA. Total
RNA (7-8 µg) was used to synthesize 40 µl of cDNA with the Invitrogen (Carlsbad, CA)
SuperScript First-Strand Synthesis System for RT-PCR kit, per manufacturer’s
directions. Random hexamers were used at 30 ng primer/µg RNA. In order to control
for genomic DNA contamination, parallel samples were prepared without the addition of
reverse transcriptase. When cDNA was synthesized from RNA isolated from LPS and
saline treated mice, equal µg of RNA were added to each reaction. PCR for each
complement component was performed using either the BioRad iTaqTM DNA Polymerase
kit (used 10x RXN buffer with 1.6 mM MgCl2, 0.2 mM dNTPs, 0.34 µM of each primer,
and 0.08U iTaq/µl) or the BioRad iQTM SYBR Green Supermix (use 2x supermix with
0.15 µM of each primer). This basic thermocyle was followed using the annealing
temperature and primers listed in Table 2: 94°C-5 minutes and 30 cycles of 94°C-1
Page 11 of 46
11
minute, anneal-1.5 minutes, 72°C-1 minute. PCR products were then separated on a 1%
agarose, TAE gel containing ethidium bromide.
Quantitative real time RT-PCR was performed using the BioRad MyiQ Single
Color Real-Time PCR Detection System to detect amplification of the PCR products and
the accompanying Optical System Software (Version 1.0) was used to analyze data.
Samples were prepared with BioRad iQTM SYBR Green Supermix containing 0.2 µM of
each primer. In each case, reactions were prepared for both the complement gene of
interest and the negative control gene ( -actin for mice and -glucuronidase (GUSB) for
rats). Mock cDNA synthesis reactions (prepared with no reverse transcriptase, described
above) were also added to PCR reactions run in parallel, to control for genomic
contamination. Parallel samples were prepared containing serial dilutions of liver cDNA
to use as a standard curve. Cycle threshold (Ct) values for each sample were calculated
using the MyiQ program and melt curves were also examined to insure analysis of a
single PCR product. Standard curves were constructed using the Ct values and relative
starting quantities of message for the liver control samples. These curves were used to
calculate relative starting quantities of message for each gene of interest and for -actin
(mouse) or -glucuronidase (GUSB, rat) in each lung sample. Data are presented as the
fold change in RNA message in LPS-treated animals, compared to saline-treated animals.
Collection of mouse and rat BAL and concentration of rat BAL
After euthanasia, the chest cavity of each mouse was opened, the trachea was exposed
and a cannula was inserted. One mL of DPBS was injected through the cannula to wash
the lungs. Cells were removed from this BAL fluid by centrifugation at 233 g for 10
Page 12 of 46
12
minutes. For Western blots, aliquots of the BAL collected at various time points after
exposure to LPS were stored at -80°C.
After euthanasia, the rat lungs were removed and washed with 10-12 mL DPBS
(for Western blots) or 50 mL DPBS with 0.2 mM EGTA (for functional assays), by
passing the fluid through a cannula inserted into the trachea. Cells were removed by
centrifugation, and BAL was stored for Western blots, as above. For functional assays,
small debris was removed from the BAL by passing the supernatant over a membrane
(Pall Life Sciences (East Hills, NY)-GH Polypro, 90 mm, 0.45 µm). Because the process
of lavaging the lung unavoidably dilutes the proteins of the alveolar lining, it is necessary
to concentrate the lavage in order to measure complement function. A protocol
previously used for human BAL (Watford, et al. 2000) was modified. All steps were
performed at 4°C to prevent complement degradation. First BAL from 6 rats was pooled
and concentrated approximately 30-fold using a tangential filter pump system (Cole
Parmer Masterflex (Vernon Hills, IL) with Millipore (Billerica, MA) Pellicon XL filterPXC010C50). Concentrated BAL was then further concentrated using a centrifugal spin
concentrator (Apollo (San Diego, CA) – 3502.2) for a final concentration of
approximately 75-fold. Protein concentrations for each preparation were determined
using a BCATM Protein Assay Kit (Pierce). The initial and final volumes of saline and
LPS BAL were kept the same so that relative comparisons could be made for equal
volumes of BAL. The final protein concentrations of the saline and LPS BAL were
approximately 3 mg/mL and 5.4 mg/mL, respectively.
Page 13 of 46
13
C3b deposition assay
A modification of a previously described method for detecting deposition of C3b onto the
surface of bacteria was used (Watford, et al. 2000). Group B streptococcus (GBS) were
grown and prepared as previously described (Watford, et al. 2000). Bacteria (5x107
bacteria per condition) were incubated for 2 hours at 37°C in buffer (GVBS++, EDTAGVBS++, or EGTA- GVBS++) containing either normal rat serum or concentrated BAL
from LPS or saline-only treated rats. Prior to incubation with bacteria, CaCl2 and MgCl2
were added to concentrated BAL to inactivate the EGTA in the lavage buffer (final
concentration 0.550 mM CaCl2, 0.500 mM MgCl2). Bacteria were centrifuged and
resuspended in 100 µl DPBS and incubated for 20 minutes at 37°C. Bacteria were again
centrifuged, resuspended in DPBS containing either FITC-conjugated goat-anti-rat C3
(MP Biomedicals, Aurora, OH) or FITC-conjugated Chrom Pure goat IgG as a negative
control (Jackson Immuno Research, West Grove, PA), and incubated on ice for 45
minutes. Finally bacteria were washed twice in 1 mL DPBS and resuspended in 1%
formaldehyde in DPBS. Associated fluorescence was measured by Fluorescence
Associated Cell Sorting (FACS).
Statistics
For data shown in Figures 5 and 6, statistical significance was determined using paired
two-tailed Student’s t-tests. Error bars represent standard error of the means.
Page 14 of 46
14
Results
Complement proteins are present in the lung.
Previous studies from our laboratory reported that human lung washings
(bronchoalveolar lavage (BAL)) contain complement components and complement
activity (Watford, et al. 2000). While some complement proteins (C3 and C4) are present
at similar levels in BAL and serum samples, others (C1q and Factor B) are much less
abundant in BAL samples compared to serum. Classical pathway activity in BAL was
present but reduced compared to classical pathway in the serum. Alternative pathway
activity was not detected in the BAL. Based on these findings, we hypothesized that
regulation of complement activity in the lung may differ from that in the serum at least
partly as a consequence of the different relative amounts of complement proteins in the
lung and serum.
In order to characterize further lung content of complement components, Western
blots were performed on human BAL from three different individuals and normal human
serum (NHS). The process of lavaging the lung unavoidably dilutes the proteins of the
airway lining fluid. To make relative comparisons possible, NHS was diluted to the same
final protein concentration as the BAL (a 1:700 dilution of NHS). Western blots showed
detectable levels of C2, C5, C6 in the BAL (Figure 1). In order to determine the relative
amounts of each protein, densitometric analyses were performed on these Western blots
(Table 3). Although relative levels varied between individuals, on average C2 was
present at approximately the same level in BAL and serum. In contrast, C5 and C6 were
present in BAL at lower concentrations than in serum (41% and 32%, respectively).
Page 15 of 46
15
In the serum, complement activation is regulated by a group of proteins known as
complement regulatory proteins. These proteins could play an important role in the lung,
a delicate organ quite vulnerable to damage caused by excessive or uncontrolled
complement activation with extensive local inflammation. Little is know about levels of
complement regulatory proteins in the lung. Western blots were performed with human
BAL and serum in order to detect two complement regulatory proteins: 1) Factor I, which
inactivates C3b and C4b, and 2) C4-binding protein (C4BP), which regulates the classical
C3 convertase. Factor I was detected in BAL (Figure 1) at levels slightly less than in
serum (Table 3). C4BP was virtually undetectable in BAL (Figure 1). One sample
(BAL-2) had a very faint band for this protein. The presence of Factor I in BAL suggests
that complement activation in the lung is regulated, at least in part, by complement
regulatory proteins. However, virtually undetectable levels of C4BP in the BAL are
consistent with the possibility this regulatory system is different in lung than in the
serum.
Interestingly, C2 (Figure 1), C4 and C3 (previously reported (Watford, et al.
2000)), which are all downstream of C1 in the classical pathway, are present at similar
relative levels in the BAL and serum. However, C1q levels in BAL are lower than in
serum (Watford, et al. 2000). In fact, the relative levels of both C1q and Factor B, the
important initiating components in the classical and alternative pathways, are low in BAL
from the normal lungs compared to serum (Watford, et al. 2000). Therefore, we
hypothesized that levels of C1q and Factor B are regulated in the lung in order to control
complement activation and prevent lung damage.
Page 16 of 46
16
Levels of complement proteins change during lung injury.
In order to determine whether lung complement levels are altered during acute
inflammation, a well-characterized model of lipopolysaccharide (LPS) – induced lung
injury was used (Ulich, et al. 1991). LPS is a cell wall component of gram-negative
bacteria and is a potent mediator of inflammation (Brigham and Meyrick 1986).
Instillation of LPS into the lung leads to the production of inflammatory proteins, an
influx of immune cells (in particular neutrophils) into the alveolar space, and an increase
in the permeability of the alveolar epithelium leading to pulmonary edema.
BAL samples were obtained from human volunteers before and at various times
after exposure to aerosolized corn dust extract, containing a known quantity of LPS
(Deetz, et al. 1997). Western blots were performed on these samples to determine the
levels of complement proteins. Equal volumes of BAL fluid were loaded on the gel for
the different time points before and after exposure. Prior to LPS exposure (Pre expo),
C1q and Factor B were barely detectable in the BAL, but there was a dramatic increase in
the amounts of both proteins in the BAL as early as 4 hours after exposure (Figure 2).
The levels of both proteins began to decrease by 24 hours and were again barely
detectable by 168 hours (7 days). To test whether this increase in BAL protein was
specific to C1q and Factor B, Western blots were performed for C4, C2, C6 and C5 with
the same BAL samples. Levels of each of these complement components were also
increased after LPS exposure (Figure 2). A Western blot for C3 showed a less dramatic
increase after LPS exposure. Even though all of the complement components tested
increased in the BAL after LPS exposure, in some cases the time specific changes in the
Page 17 of 46
17
levels of the different complement components varied. For example, C1q is dramatically
increased at 4 hours after exposure and began to decline by 48 hours, whereas C4 appears
not to decline until 96 hours.
Similar experiments were performed using mouse and rat models of LPS-induced
lung inflammation. LPS suspended in sterile saline was directly instilled into the lungs of
these animals, and BAL was collected by lavaging the whole lung at various time points
after instillation. As above, equal volumes of BAL from treated or untreated animals
were separated on an SDS-Page gel and Western blots were performed. As with human
samples, levels of each complement protein tested (Factor B, C3 and C5) increased in the
mouse and rat BAL after exposure to LPS (Figure 3). However, the timing of the
response to LPS seemed to differ in the human and animal models. Whereas, the
response to LPS in the human BAL seemed to occur quite rapidly, peaking in most cases
just four hours after exposure, the rodent responses were more delayed. In the mouse
model, increased complement protein levels did not appear to change until about 18 hours
after exposure and did not peak until 24 hours after exposure (Figure 3A). In the rat
model, complement protein levels seemed to increase by 4 hours after exposure, but did
not peak until 12 hours after exposure (Figure 3B). The difference in the timing of the
LPS response in the human and animal models may be explained by the difference in the
way the LPS was administered. The human exposure was at a lower dose than the
animals, but the use of aerosolized LPS, rather than direct instillation of LPS in solution,
may have resulted in a higher effective dose reaching the alveolar spaces. Additional,
inherent differences between the species may have also contributed to the differences in
response time.
Page 18 of 46
18
Additionally, Western blots were performed on the human BAL samples for
Surfactant Protein A (SP-A) (Figure 2). SP-A has several important roles in the innate
immune system of the lung, and it has been shown to inhibit complement activation in
vitro (Watford, et al. 2001). After exposure to LPS, levels of SP-A in the BAL were also
increased. SP-A is an abundant component of the alveolar lining and is present in the
serum at very low levels. While the increases in complement proteins in the BAL could
be due to proteins entering the alveolar space from the blood stream or to a change in
local complement expression, the increase in SP-A levels after LPS exposure are unlikely
to be due to leakage from the serum. In rat lung, levels of RNA and protein for SP-A and
SP-D were previously shown to increase after exposure to LPS (McIntosh, et al. 1996).
Considered together, these data suggest that at least some LPS-induced changes in BAL
protein levels are due to changes in local synthesis.
Complement is synthesized in the lung and expression is regulated during lung
injury.
To determine which complement components were synthesized by the lung in vivo, RTPCR was used. RNA was collected from either mouse liver or mouse lung and cDNA
was synthesized. To demonstrate that the PCR products were from cDNA and not
genomic DNA contamination, control reactions with no reverse transcriptase (No RT)
were included. By RT-PCR, C1q-B, C1r, C1s, C2, C4, C3, C5, C6, Factor B, and Factor
H were detected in both lung and liver (Figure 4). Interestingly, little or no C9 or C4BP
message was detected in the lung, even though both were easily detected in the liver
(Figure 4).
Page 19 of 46
19
In order to determine if complement is locally regulated during acute lung injury,
quantitative real time RT-PCR was used to assess relative levels of mRNA for several
complement components. C1q, Factor B, C5, C3, and -actin or -glucuronidase
(GUSB) (as a negative controls for mouse and rat, respectively) genes were amplified
from lung tissue collected from mice or rats treated with either LPS dissolved in saline or
only saline (as vehicle control).
In mice, eighteen hours after LPS exposure, there was a 14-fold increase in the
expression of Factor B, as compared to the saline control (an 8-fold increase when
corrected for -actin) (Figure 5A). At eighteen hours after treatment, there was no
significant difference in C1qB expression between the saline and LPS-treated mouse
lungs as compared to -actin control. There was about a two-fold increase in the average
relative starting quantity of both -actin and C1q after LPS-treatment, probably as a
result of differences in the amount of starting total RNA (Figure 5A). There was also no
significant change in C1q expression 6 hours after LPS treatment (data not shown).
Additionally, there was no significant change in C3 expression after LPS treatment,
compared to -actin control (Figure 5A). Interestingly, there was a 3.4 fold decrease in
the amount of C5 expression 18 hours after LPS treatment, as compared to saline control
(a 6-fold decrease when corrected for -actin) (Figure 5B). In additional experiments,
increases in the cytokines TNF-alpha, MCP-1 and MIP-2 were detected after LPS
exposure (data not shown)– all of which are expected to rise during lung injury.
Significant changes were also observed in complement gene expression in rat
lung after exposure to LPS. Eighteen hours after exposure to LPS, there was a 3.2 fold
increase in Factor B expression, as compared to the saline control (a 2.3 fold increase
Page 20 of 46
20
when corrected for GUSB) (Figure 5C). At an earlier time point, 12 hours after LPS
exposure, there was a slightly greater increase in Factor B expression – 4.2 fold (2.6 fold
when corrected for GUSB) (data not shown). Eighteen hours after exposure to LPS, there
was a 2.9 fold decrease in C5 expression, as compared to the saline control (a 3.9 fold
decrease when corrected for GUSB) (Figure 5C). As in the mouse, expression of C1q did
not appear to be significantly altered after LPS exposure.
C3b opsonization of bacteria is greater in BAL from LPS-treated rats than BAL
from normal rats.
Our data demonstrate that the composition of the lung’s complement system changes
after LPS-induced inflammation. To determine how these changes might affect
complement function, in vitro activation assays were performed. We hypothesized that
the increase in complement proteins in the alveolar space would lead to an increase in the
ability of the lung to use complement to fight invading pathogens. One mechanism by
which complement may contribute to pathogen clearance is through opsonization of
bacteria with C3b. We hypothesized that the C3b opsonization in the lung would
increase after LPS treatment.
To test this hypothesis, a rat model of LPS lung injury was required in order to
obtain sufficient quantities BAL. As for the Western blot and RT-PCR analyses, rats
were treated using direct intratracheal instillation of LPS or saline. Eighteen hours later,
BAL was collected and pooled from 6 LPS-treated and 6 saline-treated rats. BAL was
concentrated to a volume closer to that of the alveolar lining, since the process of lavage
dilutes the alveolar protein. This process was previously shown to be necessary in order
Page 21 of 46
21
to detect complement activity in human BAL (Watford, et al. 2000). The concentrated
BAL samples (or normal rat serum (NRS) as a control) were incubated with Group B
streptococcus to allow C3b to be deposited on the surface of the bacteria. NRS and saline
BAL were added to the bacteria at the same total protein concentration; equal volumes of
saline and LPS BAL were added to the bacteria. Bacteria were washed and then stained
with fluorescently labeled antibody specific for rat C3 or labeled normal goat IgG at same
total protein concentration (as a negative control). Bacteria were analyzed using flow
cytometry. Six independent experiments were performed using 3 different sets of pooled,
concentrated BAL.
As expected, there was a significant increase in the amount of C3b deposited on
bacteria after incubation with NRS in GVBS++ buffer, which allows activation of
complement through the classical and alternative pathways (Figure 6). This increase was
also dose-dependant (data not shown). In all cases, control IgG binding was very low
and did not change upon addition of NRS or BAL (data not shown). C3b binding was
inhibited when bacteria were incubated with NRS in the presence of EDTA (which
inhibits both the classical and alternative pathways) and EGTA (which inhibits only the
classical pathway). Therefore, the C3b binding measured with NRS is complement
dependent and is probably acting through the classical pathway.
When bacteria were incubated with concentrated rat BAL in GVBS++,
significantly more C3b was detected in the LPS-BAL samples than in the saline controlBAL samples (Figure 6). Therefore, it can be concluded that LPS-treatment increases the
ability of BAL to deposit C3b onto bacteria, leading to the hypothesis that bacteria that
reach the lung after LPS injury may be more readily opsonized and cleared. C3b binding
Page 22 of 46
22
from rat BAL was inhibited by EDTA, which suggests that it is complement dependant
(Figure 6). Most of the binding was also inhibited by EGTA, which suggests that C3b
deposition is occurring primarily through the classical pathway in the BAL.
Page 23 of 46
23
Discussion
The purpose of this study was to compare the complement system in the lung and serum
and to determine if lung complement levels and function change in response to lung
injury. Data presented support an active role for complement in the lung and contribute
to the idea that the immune system in the lung is uniquely organized to both optimize
clearance of pathogens and modulate inflammation. Additionally, these data demonstrate
that the complement system in the lung changes during inflammation through changes in
local complement expression and presumably through movement of complement proteins
from the bloodstream into the alveolar space.
We first compared the levels of complement protein in human bronchioalveolar
lavage (BAL) to those in serum. In order to make these relative comparisons possible,
serum was diluted and serum and BAL samples were loaded onto SDS-Page gels at the
same total protein concentration before Western blot analysis. Western blots of normal
human BAL showed that complement proteins C2, C5 and C6 were present in the lung at
similar (C2) or lower (C5 and C6) relative levels compared to normal human serum
(Figure 1, Table 3). Additional Western blots were performed to detect complement
regulatory proteins, Factor I and C4BP. Levels of Factor I in the BAL were slightly less
than levels in the serum. However, C4BP could be detected in serum but was virtually
undetectable in BAL.
Analysis of relative levels of complement proteins in normal human BAL and
serum correlates with previous measurements of complement activity in the BAL.
Previous work showed that levels of C1q in normal human BAL were low compared to
serum but that levels of C4 and C3 were comparable in BAL and serum (Watford, et al.
Page 24 of 46
24
2000). We now show that C2 is also present at comparable levels in normal human BAL
and serum. Therefore all of the classical complement proteins necessary for C3b
deposition are present in the normal lung, but the first component is present only at low
levels. These data are consistent with previous experiments demonstrating that classical
pathway activity was present in concentrated normal human BAL, albeit at reduced levels
compared to serum activity. Previous work also showed that levels of Factor B in the
lung were low compared to serum, and correspondingly that alternative pathway activity
was undetectable in the BAL from normal lung (Watford, et al. 2000).
In order to determine how the lung complement system changes during lung
injury, samples of human, mouse and rat BAL were obtained from individuals before and
at various time points after exposure to LPS. At each time point, the lung was lavaged
with the same volume of fluid. Because an increase in protein within the alveolar space is
a physiological consequence of treatment with LPS, the BAL samples at different time
points contained different total protein concentrations. Therefore, in order to compare
levels of complement components present in the airspace at each time point, equal
volumes of BAL for each sample was loaded onto SDS-Page gels before Western blot
analysis. Western blots showed a marked, transient increase in several complement
proteins in human, mouse and rat BAL. Additionally, levels of Surfactant Protein-A (SPA), a lung specific immune molecule that has been shown to interact with complement,
increased in human BAL after LPS exposure.
In order to determine whether increases in BAL complement levels were due to
changes in local complement synthesis, experiments were first performed to determine
which complement components are expressed by lung tissue. RT-PCR was performed on
Page 25 of 46
25
mouse lung tissue for a panel of complement genes. C1q-B, C1r, C1s, C2, C4, C3, C5,
C6, Factor B, and Factor H were all readily detectable, but C9 and C4BP were not. The
absence of detectable C4BP message was particularly interesting, because this was the
only complement protein tested that was virtually undetectable in human BAL.
Quantitative real-time PCR analysis of mouse and rat lung tissue demonstrated a
significant increase in Factor B expression and a significant decrease in C5 expression
after animals were exposed to LPS. However, expression of C1q (mouse and rat) and C3
(mouse) were not significantly changed after LPS exposure.
The increases in lung complement protein levels after LPS exposure correlate to
changes in lung complement function. After exposure to LPS, concentrated rat BAL
contained significantly more complement activity, as measured by the ability to deposit
C3b onto bacteria. This increase could be due directly to an increase in C3 levels and/or
an increase in other complement components, which could increase complement activity.
C3 titer experiments demonstrated elevated levels of functional C3 protein in
concentrated rat BAL from LPS-treated animals in two out of three independent
experiments (unpublished observations) and levels of C3 protein increased in human,
mouse and rat BAL after exposure to LPS. Because C1q is at low initial levels in normal
BAL, C1 may be a limiting factor in lung complement activity in the unchallenged lung.
Therefore, the increase in C1q protein in human BAL after LPS exposure most likely
contributes to increased C3b opsonization of bacteria. However, levels of C2, C4
(human), and Factor B (human, mouse and rat) proteins are also increased after LPS
exposure and may contribute to the increased activity. The majority of the activity
measured in the concentrated BAL from LPS-treated rats is likely to be through the
Page 26 of 46
26
classical pathway, because most of the activity was inhibited by both EDTA and EGTA.
However, the inhibition by EGTA for these samples was somewhat variable and in some
cases there were two populations of bacteria within one sample – one where C3b
deposition was inhibited by EGTA and one where C3b deposition was mostly unchanged
by the addition of EGTA. These data show that the inhibition by EGTA is not complete
and suggest that some of the activity detected in the BAL from LPS-treated animals may
be through the alternative pathway. This finding is consistent with our observation that
there is an increase in both local Factor B expression and Factor B protein levels in the
BAL after LPS exposure.
Based upon our and other’s data, we speculate that the complement system is
regulated not only to maximize opsonization and clearance of pathogens as they enter the
lung, but also to limit harmful inflammation along the delicate lung epithelium. Previous
data have linked C5a and to a lesser extent the membrane attack complex (MAC) with
lung inflammation. In the lung, complement fragment C5a leads to an influx of
neutrophils and protein into the alveolar space, a hallmark of many types of lung
inflammation (Desai, et al. 1979;Henson, et al. 1979;Larsen, et al. 1980;Mulligan, et al.
1996). Additionally, fibrin formation and hemorrhage within the lung as well as damage
to the alveolar epithelium have been noted in animal lungs treated with C5 cleavage
products (Desai, et al. 1979;Larsen, et al. 1980). Either C5a or proteins of the
complement membrane attack complex (C5b-9, MAC) combined with IgG immune
complexes can lead to greatly enhanced production of CXC and CC chemokines by
alveolar macrophages, and MAC can also lead to enhanced production of MCP-1 by
cultured endothelial cells (Czermak, et al. 1998). Additionally, C5aR null mice and mice
Page 27 of 46
27
treated with an anti-C5 antibody have reduced lung damage in two different models of
lung injury (Bozic, et al. 1996;Buras, et al. 2004;Czermak, et al. 1999) . However, it is
also important to note that C5a may also have a protective role in the lung, since C5aR
null mice die from Pseudomonas aeruginosa lung infections that wild type mice can
easily overcome (Hopken, et al. 1996).
Restricting levels of C5 and MAC is a possible mechanism by which the lung
minimizes inflammation in response to normal levels of foreign particles. We have
demonstrated that the level of C5 protein in normal human BAL is approximately 41% of
the level in normal human serum (at the same total protein level). Additionally, C6 in
BAL is only about 32% of serum levels. These low levels may be important for
controlling complement activity in the normal lung. However, during an acute
inflammatory episode, which is modeled by the LPS exposure, levels of C5 and C6
proteins rise in the lung, which may contribute to lung damage that occurs with this type
of injury. Even though we measured an increase in C5 protein in the BAL of all three
species tested, in the mouse and rat models, we measured a significant decrease in C5
expression by lung tissue from LPS-injured animals. The increase in C5 protein within
the BAL is most likely due to an influx of serum proteins into the injured lung and not to
a change in local expression. Despite the apparent influx of serum C5, the decreased
local expression of C5 may serve a protective mechanism to help minimize C5 mediated
inflammation in the lung. We cannot rule out the possibility that decreased local
expression of C5 does not contribute to a physiologically significant effect in the face of
acute LPS-induced inflammation. However, the ability of the lung to decrease C5
expression suggests that there is a local mechanism to regulate complement activity that
Page 28 of 46
28
could be even more important in a different or less severe type of lung injury. It is also
interesting to note that C9 was one of only two complement genes for which there was no
detectable message in the lung tissue. Lack of C9 expression could be another way
complement mediated inflammation is controlled in the normal lung, by minimizing
formation of the membrane attack complex.
Although the majority of complement components are expressed by lung tissue,
the changes in expression of complement in response to LPS-induced lung injury seem to
be quite specific. We observed changes in expression of Factor B and C5, but not C1q,
C3, Factor I, Factor H, Mannose Binding Lectin A, C6, C1r or C1s (unpublished
observations). Changes in expression may be limited to specific complement
components, but these changes are not limited to only the LPS model of acute lung
injury. Changes in local complement expression have also been reported to be associated
with mouse models of human diseases including Listeria monocytogenes meningitis,
glomerulonephritis, and lupus (Passwell, et al. 1988;Stahel and Barnum 1997;Stahel, et
al. 1997;Welch, et al. 2000).
Because many changes take place in the lung during LPS-induced injury, it is not
clear if the changes in Factor B and C5 expression are a direct effect of LPS on lung
cells. These changes could be an indirect response to cytokines, to other inflammatory
mediators or possibly even to complement proteins leaking into the lung from the serum.
INF- , IL-6 type cytokines, IL-1 type cytokines, hormones, LPS, and C5a have all been
reported to effect complement expression both in vivo and by cultured cells (Falus, et al.
1987;Falus, et al. 1990;Lake, et al. 1994;Lappin and Whaley 1991;Rothman, et al.
1989;Schlaf, et al. 2004;Strunk, et al. 1985;Zach, et al. 1992). Detailed analyses of
Page 29 of 46
29
promoters and enhancers for Factor B and C3 have demonstrated the intricate control by
individual cytokines through specific regulatory sequences (Huang, et al. 2001;Huang, et
al. 2002;Kawamura, et al. 1992;Wilson, et al. 1990). Important future studies will be to
define the mechanism by which Factor B and C5 expression are modulated during LPSinduced lung injury.
Acute lung inflammation is a critical health issue in a variety of human diseases.
Because complement is intricately involved in the inflammatory response and serves as
an important first line of defense in removing infectious organisms, understanding how
this system is regulated in the lung has clear clinical implications. Here, by using both
human and rodent models of LPS-induced lung inflammation, we detail changes that are
occurring in the complement system of the lung both at the protein and mRNA level. We
also present evidence to support the idea that the complement system functions
differently in the lung than in the serum. The relative amounts of the complement
proteins and the presence of complement regulatory proteins are very important
determinants of complement function in the serum. The relative amounts of complement
proteins are different in the lung than in the serum and C4BP, a key complement
regulatory protein, is virtually undetectable in the normal lung. Maintaining specific
levels of each complement protein and complement regulatory protein, regulating
complement expression in response to inflammatory stimuli and using SP-A to regulate
complement activity all appear to be mechanisms used in the lung to help insure a
balance between the helpful and harmful activities of complement.
Page 30 of 46
30
Acknowledgements
We would like to thank Kathy Evans for performing Western blots for SP-A and for
general experimental support. We would also like to acknowledge Dr. John Whitesides,
Patrice McDermott, and Danielle King from the Human Vaccine Institute facility for the
FACS analysis.
Grants
This work was supported by a National Institute of Health-National Heart, Lung and
Blood Institute Grant RO1HL-51134 (J.R. Wright and M.M. Frank).
Contact Information
Address for reprint requests and other correspondence: J. R. Wright, Box 3709, Dept. of
Cell Biology, Duke Univ. Medical Center, Durham, NC 27710 (e-mail:
[email protected])
Page 31 of 46
31
Figure Legends
Figure1: Complement proteins are present in BAL at different levels compared to serum.
Normal human BAL from three different individuals, normal human serum diluted to the
same total protein concentration (NHS 1:700), purified complement proteins (C2, C5, C6,
Factor I, and C4BP), and human IgG and HSA (as controls) were analyzed by Western
blot using antibodies against complement proteins C2, C5, and C6, Factor I or C4BP.
Figure 2: Levels of complement proteins in human BAL change after exposure to LPS.
Human volunteers were exposed to aerosolized LPS at a dose of 0.4 µg/kg body mass.
Lungs were lavaged prior to exposure (Pre expo) and at various time points after
exposure (4, 24, 48, 96, and 168 hours). BAL samples, NHS, purified complement
proteins, IgG, and HSA were analyzed by Western blot using antibodies against C1q,
Factor B, C2, C4, C5, C6, C3 and Surfactant Protein A (SP-A). Equal volumes of BAL
were loaded for each time point. All Western blots were performed with samples from at
least two individuals. Blots shown are representative data. *individual number 1,
#individual number 2, **individual number 3.
Figure 3: Levels of complement proteins in mouse and rat BAL change after exposure to
LPS. Male mice or rats were treated with LPS via direct intratracheal instillation at a
dose of 100 µg LPS/kg mouse body mass. BAL was collected from untreated animals
and from animals at various time points after exposure. BAL samples, dilute mouse or
rat serum (NMS or NRS, respectively) and mouse or rat IgG (as a control) were analyzed
by Western blot using antibodies against C3, C5, and Factor B. Equal volumes of BAL
Page 32 of 46
32
were loaded for each time point. All Western blots were performed with samples from at
least two animals at each time point. Blots shown are representative data.
Figure 4: Complement protein mRNA is detectable in the mouse lung. RNA was
collected from whole male mouse lung or liver and cDNA was synthesized. RT-PCR
was performed with primers specific for a variety of complement genes. No reverse
transcriptase (No RT) controls serve as a control for genomic contamination. Water was
added to some reactions instead of cDNA as a no template control. Lane 1-liver cDNA,
lane 2-lung cDNA, lane 3-lung no RT control, lane 4-water only. All RT-PCR reactions
were performed with samples from at least two animals. Data shown are representative.
Figure 5: The lung regulates expression of specific complement components. Male mice
(A and B) or rats (C) were intratracheally instilled with LPS (at a dose of 100 µg/kg
body mass) or saline (as a control) and sacrificed 18 hours later. RNA was collected
from lung tissue and cDNA was synthesized. Real-time quantitative RT-PCR was
performed using the BioRad Sybr green system to detect amplification of each PCR
product. Factor B, C1q C3, and C5 were amplified and in each case -actin (for mice) or
-glucuronidase (GUSB, for rats) was amplified at the same time as a control for
differences in how much total cDNA was added. Also standard curves were constructed
to determine relative starting quantities of cDNA for each unknown sample.
Comparisons were then made between relative starting quantities in LPS and saline
control samples and are presented as a fold change for each gene. Fold changes in
expression for C5, Factor B, C1q or C3 can then be compared to fold changes in the
Page 33 of 46
33
control gene. Parts A and B: *significantly different from -actin control (student t-test,
p<0.05). N=4 for Factor B, C5, and -actin. N=3 for C1q and C3. Part C:
*significantly different from GUSB control (student t-test, p<0.05). N=4 for Factor B,
C5, and GUSB.
Figure 6: BAL from LPS treated rats supports greater C3b deposition onto bacteria than
does BAL from saline treated rats.
Freshly harvested Group B streptococcus were washed and then incubated with NRS,
BAL from male saline-treated rats (Sal BAL) or BAL from male LPS-treated rats at
37ºC, for 1 hour. Next bacteria were washed for 20 min at 37°C, stained with FITC
labeled goat-anti-rat C3 or FITC- labeled normal goat IgG as a control, washed again and
fixed. Incubations with NRS or BAL were performed in GVBS++ (which allows
complement activation though alternative and classical pathways), in EDTA buffer
(which inhibits all complement activation) or EGTA buffer (which inhibits only the
classical pathway). The same amount of total protein was added for NRS and Sal BAL
samples. The same volume was added for Sal BAL and LPS BAL. C3b deposition is
shown as the median fluorescence detected when FACS analysis was performed on
bacteria. *significantly different from bacteria only control, #significantly different from
saline BAL (student t-test, p<0.05), N=6 individual experiments, with three different sets
of pooled BAL samples.
Page 34 of 46
34
References
1. Ackerman SK, Friend PS, Hoidal JR, and Douglas SD. Production of C2 by human
alveolar macrophages. Immunology 35: 369-72, 1978.
2. Alper C, Raum D, Awdeh Z, Petersen B, Taylor P, and Starzl T. Studies of hepatic
synthesis in vivo of plasma proteins, including orosomucoid, transferrin, alpha 1antitrypsin, C8, and Factor B. Clin Immunol Immunopathol 16: 84-89, 1980.
3. Bossi F, Fischetti F, Pellis V, Bulla R, Ferrero E, Mollnes TE, Regoli D, and
Tedesco F. Platelet-activating factor and kinin-dependent vascular leakage as a novel
functional activity of the soluble terminal complement complex. J Immunol 173: 6921-7,
2004.
4. Bozic CR, Lu B, Hopken UE, Gerard C, and Gerard NP. Neurogenic amplification
of immune complex inflammation. Science 273: 1722-5, 1996.
5. Brigham KL, and Meyrick B. Endotoxin and lung injury. Am Rev Respir Dis 133:
913-27, 1986.
6. Brown JS, Hussell T, Gilliland SM, Holden DW, Paton JC, Ehrenstein MR,
Walport MJ, and Botto M. The classical pathway is the dominant complement pathway
required for innate immunity to Streptococcus pneumoniae infection in mice. Proc Natl
Acad Sci U S A 99: 16969-74, 2002.
7. Buras JA, Rice L, Orlow D, Pavlides S, Reenstra WR, Ceonzo K, and Stahl GL.
Inhibition of C5 or absence of C6 protects from sepsis mortality. Immunobiology 209:
629-35, 2004.
8. Cole FS, Matthews WJ, Jr., Rossing TH, Gash DJ, Lichtenberg NA, and
Pennington JE. Complement biosynthesis by human bronchoalveolar macrophages. Clin
Immunol Immunopathol 27: 153-9, 1983.
9. Colten H, and Garnier G. Regulation of Complement Protein Gene Expression. In:
The Human Complement System in Health and Disease, edited by Volanakis J and M
Frank. New York, NY: Marcel Dekker, Inc., 1998.
10. Czermak BJ, Lentsch AB, Bless NM, Schmal H, Friedl HP, and Ward PA. Role
of complement in in vitro and in vivo lung inflammatory reactions. J Leukoc Biol 64: 408, 1998.
11. Czermak BJ, Breckwoldt M, Ravage ZB, Huber-Lang M, Schmal H, Bless NM,
Friedl HP, and Ward PA. Mechanisms of enhanced lung injury during sepsis. Am J
Pathol 154: 1057-65, 1999.
12. Deetz DC, Jagielo PJ, Quinn TJ, Thorne PS, Bleuer SA, and Schwartz DA. The
kinetics of grain dust-induced inflammation of the lower respiratory tract. Am J Respir
Crit Care Med 155: 254-9, 1997.
13. Desai U, Kreutzer DL, Showell H, Arroyave CV, and Ward PA. Acute
inflammatory pulmonary reactions induced by chemotactic factors. Am J Pathol 96: 7183, 1979.
14. Falus A, Beuscher HU, Auerbach HS, and Colten HR. Constitutive and IL 1regulated murine complement gene expression is strain and tissue specific. J Immunol
138: 856-60, 1987.
15. Falus A, Rokita H, Walcz E, Brozik M, Hidvegi T, and Meretey K. Hormonal
regulation of complement biosynthesis in human cell lines-- II. Upregulation of the
Page 35 of 46
35
biosynthesis of complement components C3, factor B and C1 inhibitor by interleukin-6
and interleukin-1 in human hepatoma cell line. Mol Immunol 27: 197-201, 1990.
16. Figueroa JE, and Densen P. Infectious diseases associated with complement
deficiencies. Clin Microbiol Rev 4: 359-95, 1991.
17. Gross GN, Rehm SR, and Pierce AK. The effect of complement depletion on lung
clearance of bacteria. J. Clin. Invest. 373-378, 1978.
18. Guo RF, and Ward PA. Role of C5a in inflammatory responses. Annu Rev Immunol
23: 821-52, 2005.
19. Henson PM, McCarthy K, Larsen GL, Webster RO, Giclas PC, Dreisin RB,
King TE, and Shaw JO. Complement fragments, alveolar macrophages, and alveolitis.
Am J Pathol 97: 93-110, 1979.
20. Hopken UE, Lu B, Gerard NP, and Gerard C. The C5a chemoattractant receptor
mediates mucosal defence to infection. Nature 383: 86-9, 1996.
21. Huang Y, Krein PM, and Winston BW. Characterization of IFN-gamma regulation
of the complement factor B gene in macrophages. Eur J Immunol 31: 3676-86, 2001.
22. Huang Y, Krein PM, Muruve DA, and Winston BW. Complement factor B gene
regulation: synergistic effects of TNF-alpha and IFN-gamma in macrophages. J Immunol
169: 2627-35, 2002.
23. Kawamura N, Singer L, Wetsel RA, and Colten HR. Cis- and trans-acting
elements required for constitutive and cytokine-regulated expression of the mouse
complement C3 gene. Biochem J 283 ( Pt 3): 705-12, 1992.
24. Lake FR, Noble PW, Henson PM, and Riches DW. Functional switching of
macrophage responses to tumor necrosis factor-alpha (TNF alpha) by interferons.
Implications for the pleiotropic activities of TNF alpha. J Clin Invest 93: 1661-9, 1994.
25. Langlois P, and Gawryl M. Complement Activation Occurs through Both Classical
and Alternative Pathways Prior to Onset and Resolution of Adult Respiratory Distress
Syndrome. Clin Immunol Immunopathol 47: 152-163, 1988.
26. Langlois P, and Gawryl M. Accentuated Formation of the Terminal C5b-9
Complement Complex in Patient Plasma Precedes Development of the Adult Respiratory
Distress Syndrome. Am Rev Respir Dis 138: 368-375, 1988.
27. Lappin DF, and Whaley K. Modulation of complement gene expression by
glucocorticoids. Biochem J 280 ( Pt 1): 117-23, 1991.
28. Larsen GL, McCarthy K, Webster RO, Henson J, and Henson PM. A
differential effect of C5a and C5a des Arg in the induction of pulmonary inflammation.
Am J Pathol 100: 179-92, 1980.
29. Liszewski M, and Atkinson J. Regulatory Proteins of Complement. In: The Human
Complement System in Health and Disease, edited by Volanakis J and M Frank. New
York, NY: Marcel Dekker, Inc., 1998.
30. McIntosh JC, Swyers AH, Fisher JH, and Wright JR. Surfactant proteins A and
D increase in response to intratracheal lipopolysaccharide. Am J Respir Cell Mol Biol 15:
509-19, 1996.
31. Mulligan MS, Schmid E, Beck-Schimmer B, Till GO, Friedl HP, Brauer RB,
Hugli TE, Miyasaka M, Warner RL, Johnson KJ, and Ward PA. Requirement and
role of C5a in acute lung inflammatory injury in rats. J Clin Invest 98: 503-12, 1996.
Page 36 of 46
36
32. Passwell J, Schreiner GF, Nonaka M, Beuscher HU, and Colten HR. Local
extrahepatic expression of complement genes C3, factor B, C2, and C4 is increased in
murine lupus nephritis. J Clin Invest 82: 1676-84, 1988.
33. Reynolds HY, and Newball HH. Analysis of proteins and respiratory cells obtained
from human lungs by bronchial lavage. J. Lab. Clin. Med. 84: 559-573, 1974.
34. Robertson J, Caldwell JR, Castle JR, and Waldman RH. Evidence for the
presence of components of the alternative (properdin) pathway of complement activation
in respiratory secretions. J Immunol 117: 900-3, 1976.
35. Rothman BL, Merrow M, Despins A, Kennedy T, and Kreutzer DL. Effect of
lipopolysaccharide on C3 and C5 production by human lung cells. J Immunol 143: 196202, 1989.
36. Schlaf G, Nitzki F, Heine I, Hardeland R, Schieferdecker HL, and Gotze O. C5a
anaphylatoxin as a product of complement activation up-regulates the complement
inhibitory factor H in rat Kupffer cells. Eur J Immunol 34: 3257-66, 2004.
37. Schutte M, DiCamelli R, Murphy P, Sadove M, and Gewurz H. Effects of
anesthesia, surgery and inflammation upon host defense mechanisms. I. Effects upon the
complement system. Int Arch Allergy Appl Immunol 48: 706-20, 1975.
38. Stahel PF, and Barnum SR. Bacterial meningitis: complement gene expression in
the central nervous system. Immunopharmacology 38: 65-72, 1997.
39. Stahel PF, Frei K, Eugster HP, Fontana A, Hummel KM, Wetsel RA, Ames RS,
and Barnum SR. TNF-alpha-mediated expression of the receptor for anaphylatoxin C5a
on neurons in experimental Listeria meningoencephalitis. J Immunol 159: 861-9, 1997.
40. Strunk RC, Whitehead AS, and Cole FS. Pretranslational regulation of the
synthesis of the third component of complement in human mononuclear phagocytes by
the lipid A portion of lipopolysaccharide. J Clin Invest 76: 985-90, 1985.
41. Strunk RC, Eidlen DM, and Mason RJ. Pulmonary alveolar type II epithelial cells
synthesize and secrete proteins of the classical and alternative complement pathways. J
Clin Invest 81: 1419-26, 1988.
42. Ulich TR, Watson LR, Yin SM, Guo KZ, Wang P, Thang H, and del Castillo J.
The intratracheal administration of endotoxin and cytokines. I. Characterization of LPSinduced IL-1 and TNF mRNA expression and the LPS-, IL-1-, and TNF-induced
inflammatory infiltrate. Am J Pathol 138: 1485-96, 1991.
43. Vinci G, Lynch NJ, Duponchel C, Lebastard TM, Milon G, Stover C, Schwaeble
W, and Tosi M. In vivo biosynthesis of endogenous and of human C1 inhibitor in
transgenic mice: tissue distribution and colocalization of their expression. J Immunol 169:
5948-54, 2002.
44. Volanakis JE. Transcriptional regulation of complement genes. Annu Rev Immunol
13: 277-305, 1995.
45. Wagner E, Jiang H, and Frank M. Complement and kinins: mediators of
inflammation. In: Clinical Diagnosis and Management by Laboratory Methods, edited by
Henry J. Philadelphia, PA: W.B. Saunders Company, 2001.
46. Watford WT, Ghio AJ, and Wright JR. Complement-mediated host defense in the
lung. Am J Physiol Lung Cell Mol Physiol 279: L790-8, 2000.
47. Watford WT, Wright JR, Hester CG, Jiang H, and Frank MM. Surfactant
protein A regulates complement activation. J Immunol 167: 6593-600, 2001.
Page 37 of 46
37
48. Watford WT, Smithers MB, Frank MM, and Wright JR. Surfactant protein A
enhances the phagocytosis of C1q-coated particles by alveolar macrophages. Am J
Physiol Lung Cell Mol Physiol 283: L1011-22, 2002.
49. Welch TR, Frenzke M, and Witte D. Evidence of a role for local complement
expression in a murine model of progressive glomerulonephritis. Pediatr Res 48: 200-5,
2000.
50. Wilson DR, Juan TS, Wilde MD, Fey GH, and Darlington GJ. A 58-base-pair
region of the human C3 gene confers synergistic inducibility by interleukin-1 and
interleukin-6. Mol Cell Biol 10: 6181-91, 1990.
51. Wright JR. Immunoregulatory functions of surfactant proteins. Nat Rev Immunol 5:
58-68, 2005.
52. Zach TL, Hill LD, Herrman VA, Leuschen MP, and Hostetter MK. Effect of
glucocorticoids on C3 gene expression by the A549 human pulmonary epithelial cell line.
J Immunol 148: 3964-9, 1992.
Page 38 of 46
Page 39 of 46
Page 40 of 46
Page 41 of 46
Page 42 of 46
Page 43 of 46
Page 44 of 46
Page 45 of 46
Page 46 of 46