|
|
||||||||
Eijkman-Winkler Institute for Medical Microbiology, Utrecht University, Utrecht, The Netherlands
| Abstract |
|---|
|
|
|---|
| Introduction |
|---|
|
|
|---|
CD14, a phosphatidylinositol-anchored molecule found on monocytes, macrophages, and neutrophils, has been identified as a receptor that interacts with LPS. The interaction between LPS and phagocytes via CD14 is strongly enhanced by LPS-binding protein (LBP)2 (6).
Complement (C) receptor 3 (CR3) (CD11b/CD18) is a member of the
leukocyte integrin family of heterodimeric adhesion molecules,
consisting of a common ß-subunit and a unique
-subunit. (7). In
1986, a role for integrins in LPS-interaction was found by
Wright (8). Each of the three members of the integrin family (CR3,
LFA-1, and p150-95) were shown to mediate not only LPS
interaction but the binding of Escherichia coli to
macrophages as well (9).
CR3 is up-regulated and the adhesive capacity of leukocytes is
increased after the interaction of CD14 with LPS and LBP (10).
Monocytes from CD18-deficient patients or monocytes on which all three
members of the ß-integrin family were down-modulated were unable to
bind E. coli or E coated with LPS (ELPS) (8, 11, 12). From the CD18 receptor family, the binding specificity of CR3 is
the most well characterized. Two distinct binding sites on CR3
have been described: 1) a protein-binding-site that binds C3bi,
fibrinogen, and Leishmania glycoprotein 63, and 2) a
lipid-binding-site involved in the binding of LPS, lipid A, and lipid
IVA (13, 14, 15). Experiments with monocytes from CD18-deficient
patients suggest no role for CD18 in secretory functions, because these
cells were able to respond to LPS with the production of TNF-
and
IL-1ß; in addition, neutrophils from these patients could be primed
by LPS for an enhanced respiratory burst (12).
In an earlier report, we showed that LPS binding to monocytes is
completely dependent upon CD14 at concentrations of
100 ng LPS/ml
(16). With higher concentrations, other interactions, which are
LBP-independent, take over. This finding is in accordance with
similar studies of cell activation by us and others, for which the same
LPS concentration-dependent characteristics were found (17, 18).
To study the role of the surface properties of LPS, we constructed E coated with known amounts of LPS and compared the mechanism of neutrophil binding to free LPS and whole bacteria.
| Materials and Methods |
|---|
|
|
|---|
ReLPS from Salmonella minnesota strain Re595 and TNF were purchased from Sigma (St. Louis, MO). Highly fluorescent rLPS was prepared as described previously (16). BSA was obtained from Organon Technika (Turnhout, Belgium). Recombinant human LBP was a generous gift of Dr. H. Lichenstein (Amgen, Thousand Oaks, CA). The orange fluorescent membrane probe PKH-26 was purchased from Sigma. BODIPY-fMLP (green) and LDS-751 (red) were purchased from Molecular Probes (Eugene, OR).
Quantification of bacteria
S. minnesota strain Re595 was grown for 16 h in Mueller-Hinton broth (Oxoid, Hampshire, U.K.). The concentration of bacteria was determined by two completely independent methods. First, a suspension of bacteria was counted microscopically. Second, FITC-labeled standard beads of known concentration (Flow Cytometry Standards, San Juan, NC) were mixed with bacteria and measured in a FACScan (Becton Dickinson, Mountain View, CA). Beads and bacteria were separately gated on the basis of their forward and side scatter properties and fluorescence. As 10,000 events passed through the count-gate (selecting only beads), the number of bacteria was measured in the same time window. Thereafter, the exact number of bacteria in the starting suspension could be calculated.
Monoclonal Abs
Hybridoma cells secreting the murine mAbs 60bca (IgG1, directed against CD14) and 44a (IgG1, directed against CD11b) were obtained from the American Type Culture Collection (Manassas, VA). Culture supernatants were collected, and mAbs were purified over a protein G column (Pharmacia, Uppsala, Sweden). The production and characterization of mAb 8-2-c1, an IgM anti-LPS mAb, has been described previously (19). Goat anti-mouse FITC was obtained from Dako (Glostrup, Denmark).
Rabbit ELPS
Rabbit E were stained with PKH-26 as described previously (20). In short, rabbit E were washed three times with PBS; next, 5 x 108 E were suspended in 0.5 ml of PKH-26 diluent. Subsequently, 0.4 µl of PKH-26 was added to 0.5 ml of diluent, added to the E, and incubated for 20 min at room temperature. The reaction was stopped in HBSS containing 10% BSA for 5 min, and E were washed three times with HBSS/0.1% BSA. Subsequently, E were incubated with increasing concentrations of LPS (1270 µg/ml) for 60 min at 37°C while being gently shaken and were washed three times with HBSS/0.1% BSA. In parallel, E were incubated with FITC-LPS in equal concentrations and washed; next, the FITC-LPS load was evaluated by flow cytometry. A linear relationship with a correlation coefficient of 0.97 was found between the amount of FITC-LPS added and the observed amount of fluorescence that associates to the E.
Quantitation of LPS on ELPS and S. minnesota
As summarized in Table I
,
we evaluated the number of LPS molecules incorporated into the E
membrane by three principally different methods. In experiments
concerning the interaction of ELPS with neutrophils, the
LPS-coated surface of the E will be in contact with the neutrophil.
Therefore, the density of LPS molecules per surface is the most
relevant parameter in our study. We calculated the surface of a rabbit
E to be 39.3 µm2 (estimated radius of 2.5 µm) (21); the
surface of S. minnesota was determined to be 9.4
µm2 (estimated length of 2 µm, estimated radius of 0.5
µm) (22). The LPS density of a Gram-negative bacterium has
been the subject of research for several groups and is supposed to be 3
mg of LPS per 1012 CFU (23, 24, 25). The estimated molecular
mass of LPS is 3200 g/mol (26, 27); from this, we calculated
that
60,000 molecules of LPS are present per µm2 on
S. minnesota.
|
The LAL assay is a well-accepted method to liberate LPS from membranes or vesicles for LPS quantitation in heated samples (28). The assay was performed according to the manufacturers instructions (Chromogenics, Mölndal, Sweden). A standard curve was generated using known concentrations of rLPS. Test samples containing either ELPS coated with different amounts of LPS or S. minnesota were analyzed. A density of 97,000 LPS molecules per µm2 was found for S. minnesota, which is somewhat higher than the density of 60,000 LPS molecules per µm2 that has been described in other literature (23, 24, 25). We found a 2.6 times higher LPS density on S. minnesota than on ELPS coated with 270 µg/ml.
Flow cytometric analysis of ELPS or bacteria using an anti-LPS mAb
mAb 8-2-C1 directed against rLPS was incubated for 30 min at 37°C in a concentration of 10 µg/ml with either S. minnesota or ELPS. After washing with HBSS/0.1%BSA, samples were incubated for 15 min at 37°C with goat anti-mouse FITC in a saturating concentration. After washing, samples were analyzed in a FACScan. The data obtained with this method represent LPS molecules that were available for binding to the mAb and consequently are of interest with respect to the possibility of interacting with cellular receptors. With this method, we found that LPS density was 16.2 times higher for S. minnesota compared with ELPS270.
Flow cytometric analysis of ELPS using FITC beads
A series of beads coated with different numbers of FITC molecules (7,0001,800,000) were used as standards. The mean fluorescence of the FITC beads was measured by flow cytometry; a standard curve was obtained. E were coated with 130 µg FITC-LPS as described previously, and mean fluorescence was determined. Using the standard curve, the amount of FITC molecules per E could be calculated. As our FITC-LPS was labeled with a labeling efficiency of one FITC-molecule per LPS molecule (16), the amount of LPS per E was obtained. LPS load on S. minnesota cannot be obtained using this method. If our calculations use 3 mg LPS/1012 CFU, as found in other literature, an LPS density ratio of 2.3 times more LPS per surface on S. minnesota compared with ELPS270 was found.
Cells
Neutrophils were isolated from peripheral blood obtained from normal human volunteers after informed consent. Blood was collected into heparin tubes (Greiner, Frickenhausen, Germany), diluted 1/1 with PBS (pH 7.4), loaded on a Ficoll (Pharmacia) and Histopaque (density 1.119, Sigma) gradient. After centrifugation for 20 min at 250 x g at room temperature, the neutrophil interface was collected and washed with RPMI 1640 (Life Technologies, Gaithersburg, MD) (pH 7.4) and 0.05% human serum albumin (CLB, Amsterdam, the Netherlands). The remaining E were lysed for 45 s with distilled water, after which concentrated PBS was added to reestablish the isotonic condition.
Cells were washed and resuspended in RPMI 1640 at 1 x 107 cells/ml and tested for viability and purity. Viability was always >95% using either trypan blue or propidium iodide exclusion. Purity was checked by microscopy of cytospin slides and was always >98%.
Flow cytometric assay for FITC-LPS binding
In a total volume of 50 µl, 5 x 105 neutrophils were incubated with increasing concentrations (1125 ng/ml) of FITC-LPS in the presence of LBP (100 ng/ml) or HBSS containing 0.1% BSA. For blocking experiments, cells were first incubated for 15 min on ice with 20 µg/ml of the mAbs 60bca (anti-CD14) or 44a (anti-CR3). The mean fluorescence of neutrophils was analyzed by flow cytometry.
Flow cytometric assay for ELPS binding and activation
In a total volume of 60 µl, 2 x 105
neutrophils were mixed with 2 x 106
PKH-26-ELPS in the presence of 300 ng/ml LBP or HBSS/0.1%
BSA, gently spun down for 5 min at 400 rpm, and incubated at 37°C for
30 min. For blocking or activation experiments, cells were preincubated
for 15 min at 37°C with 20 µg/ml of mAb or 0.57 nM of TNF-
.
After the reaction was terminated, cells were put on ice and subsequently incubated with 10-7 M BODIPY-fMLP to determine the level of fMLP receptor expression. As activated neutrophils express more fMLP receptors, we used this as a marker for cell activation. Thereafter, samples were stained with LDS-751 according to the manufacturers instructions (Molecular Probes) in a final concentration of 0.05 µg/ml per sample. LDS-751 is a selective red nuclear stain, and was used to distinguish neutrophils from unbound ELPS.
Flow cytometric assay for binding of S. minnesota to neutrophils
S. minnesota Re595 was grown overnight at 37°C in Mueller-Hinton broth, harvested, washed in PBS, resuspended in 0.1 M sodium carbonate buffer at pH 9.6 containing 47 µg/ml FITC (Isomer I; Sigma), incubated for 1 h at 37°C, and washed twice in HBSS. S. minnesota-FITC was incubated in a bacterium to neutrophil ratio of 10:1 or 30:1, and binding to neutrophils was analyzed by flow cytometry.
| Results |
|---|
|
|
|---|
Histograms of a representative experiment of neutrophils incubated
with 42 ng/ml FITC-LPS are shown in Fig. 1
A. Only in the presence of
LBP (100 ng/ml) was FITC-LPS able to bind to neutrophils.
Preincubating the neutrophils with anti-CD14 mAb resulted
in a complete inhibition of binding (Fig. 1
B). Neither
preincubation with anti-CR3 (Fig. 1
C) nor preincubation
with anti-HLA (data not shown) had any effect on binding,
confirming that the binding of FITC-LPS to neutrophils in the presence
of LBP is completely CD14-dependent. These results are in complete
accordance with literature describing LPS binding to monocytes and
macrophages as dependent upon CD14 and LBP (29, 30).
|
Low binding to isolated neutrophils was observed for
ELPS in the absence of LBP and was dependent upon LPS
density. A significant enhancement of this binding was seen in the
presence of LBP and was dependent upon the LPS load on the E (Fig. 2
A).
|
After preincubation of the neutrophils with either anti-CD14 mAb
(Fig. 2
B) or anti-CR3 mAb (Fig. 2
C), the
binding of ELPS was completely blocked. The finding that
anti-CD14 inhibited LBP-dependent binding is in accordance with
earlier results. The fact that the LBP-independent part of the binding
could be inhibited by anti-CD14 as well is in agreement with our
previous finding that for LPS concentrations from 100 to 300-1000
ng/ml, binding is no longer LBP-dependent but is still partially
mediated by CD14 (16).
The activation of neutrophils was inhibited by the preincubation of
neutrophils with anti-CD14 (Fig. 2
E), whereas
preincubation with anti-CR3 had no effect on cell activation by
ELPS (Fig. 2
F). Preincubation with anti-HLA,
as a control, did not affect binding or activation (data not shown).
From these results, we speculated that ELPS interact with
neutrophils via CD14, leading to the binding of ELPS to the
activated CR3.
To evaluate whether neutrophil activation could indeed result in
ELPS binding, we activated the neutrophils with a different
agonist, TNF-
. Activation was confirmed by enhanced BODIPY-fMLP
binding (Fig. 3
D). TNF
activation of neutrophils resulted in ELPS binding that was
independent of LBP (Fig. 3
A); however, binding was
still dependent upon LPS density. The down-modulated ELPS
binding to neutrophils after preincubation with anti-CD14 could
indeed be restored by TNF-
activation of the neutrophils (Fig. 3
B). After preincubation with anti-CR3, the binding of
ELPS to TNF-
-activated neutrophils did not occur (Fig. 3
C). From this observation, we conclude that
ELPS interacts with the neutrophil via CD14, giving rise to
activation (confirmed by enhanced fMLP receptor expression), and that
the actual binding is mediated by activated CR3.
|
Because it is a particle that is naturally coated with LPS,
S. minnesota was used to investigate the receptors involved
in association with neutrophils. For E with the highest LPS load,
ELPS270, we had already observed some
LBP-independent binding to neutrophils. This trend was shown to be
followed by a significant association of S. minnesota-FITC
to neutrophils in the absence of LBP (Fig. 4
A). For a bacteria to
neutrophil ratio of 10:1 or 30:1, additional LBP enhanced binding by
two- to threefold. In this case, microscopic evaluation of our samples
revealed that this binding correlates with an association of
12 bacteria per cell. Preincubating neutrophils with anti-CD14
mAb inhibited the LBP-dependent part of the binding, whereas
anti-CR3 had no effect. These data suggest a role for CD14 in the
activation of another receptor that is involved in the actual binding
of particles with very high LPS load, such as S. minnesota.
Significant LBP-independent binding was observed that was mediated by
structures other than CD14 or CR3, as preincubation with mAbs directed
at these receptors as well as control Abs (anti-HLA) had no effect
on binding.
|
| Discussion |
|---|
|
|
|---|
Fluorescein-labeled LPS has been used to study interactions with
monocytes, and these interactions were shown to be dependent upon LBP
or serum and CD14 (31). The same conditions, as shown in this
paper, determine the binding of FITC-LPS to neutrophils. We have
shown previously that binding to monocytes was enhanced by LBP or
serum for LPS concentrations of
100 ng/ml and was completely
CD14-dependent and saturable (16). In chemiluminescence experiments
with isolated neutrophils, the production of oxygen radicals after
priming with 1 ng/ml LPS was also shown to be completely dependent upon
LBP and CD14 (16). Consequently, for free LPS in low concentrations, as
can be found in the clinical condition of Gram-negative sepsis (32),
only CD14 was found to play a role.
For LPS concentrations of >100 ng/ml, binding to monocytes is still partially CD14-dependent but is not LBP-dependent (16). For an E coated with LPS or a Gram-negative bacterium, one might expect a local LPS concentration or density that is relatively high; therefore, other receptors than CD14 might be involved. We found that S. minnesota has an LPS density that is 2.6 times higher than our E with maximal LPS coat, ELPS270, according to the LAL assay and 2.3 times higher according to the FITC beads method. According to the method using anti-LPS Abs, ELPS270 had an LPS density that was 16.2 times higher than S. minnesota. However, we think that the latter method is too dependent upon the form in which the LPS is presented, which may explain the large difference with the other two methods used. The results for S. minnesota (obtained by LAL assay) were not significantly different from the literature. We found a density of 97,000 molecules per µm2, whereas a density of 60,000180,000 is described in the literature (23, 24, 25, 33).
The binding of E coated with different amounts of LPS to isolated
neutrophils was dependent upon LBP and was prevented by either
anti-CD14 or anti-CR3 mAbs. Anti-CD14 preincubation prevented
activation by ELPS; preincubation with anti-CR3 did not
affect neutrophil activation. This observation led to our hypothesis
that ELPS activate neutrophils via CD14 and subsequently
bind to the activated CR3. This hypothesis was confirmed by the
observation that ELPS bound to neutrophils activated by
TNF-
(an LPS-independent activating agent) that were preincubated
with anti-CD14. ELPS did not bind to activated
neutrophils preincubated with anti-CR3. Therefore, an activating
role is proposed for CD14 in the presence of LBP, although the actual
binding of ELPS was mediated by the activated and
up-regulated CR3. These findings illustrate a difference in the
mechanism of interaction for free LPS and the LPS present on a
particle.
Our finding that ELPS binding was mediated by CR3 is in
accordance with the fact that monocytes or macrophages isolated from
CD18-deficient patients are defective in the binding of
ELPS or E. coli (12). LPS stimulation of
phagocytes from CD18-deficient patients still results in the production
of TNF-
and IL-1ß, and the neutrophils of these patients
can be primed for an oxidative burst. Furthermore, Zarewych et al.
recently showed that, indeed, LPS in the presence of serum or LBP
induced a physical association between CD14 and CR3, leading to a
CR3-mediated attachment of neutrophils to surfaces followed by a
dissociation of CD14-CR3 complexes (34). These investigators
suggested that CD14 serves as a scout for CR3, which, in turn,
is responsible for transmembrane signaling and attachment of the cell.
We demonstrated an association of S. minnesota with neutrophils in the presence of buffer only. This association could be enhanced twofold by additional LBP.
The LBP-dependent part of the binding was inhibited by preincubating
the cells with anti-CD14, indicating CD14-mediated binding.
LBP-independent binding was not mediated by CD14. This corresponds with
our previous finding that for high LPS concentrations, binding was only
partially dependent upon CD14 (16). However, in contrast to
ELPS, no reduction of LBP-dependent S. minnesota
interaction with neutrophils was seen after preincubation with
anti-CR3. LBP-independent binding was also not affected by
anti-CR3. This illustrates that most of the LBP-dependent binding
of S. minnesota is mediated by CD14. Recently, Grunwald et
al. also showed a CD14-dependent interaction of monocytes with E.
coli that was dependent upon serum or LBP that could be inhibited
by anti-CD14 mAb, excess soluble sCD14, or excess LPS. A
mAb directed at CR3 did not inhibit binding and ingestion. In the hands
of these investigators, a significant background binding of E.
coli was also seen (35). However, microscopic evaluation
of our samples revealed an association of
12 bacteria per cell. As
observed after classical opsonization with C and Abs, this association
is only marginal compared with the association of 78 bacteria per
cell (36).
Our data show that structures other than CR3 and CD14 account for the binding of S. minnesota on activated cells, and this binding is independent of LBP. This finding is in contrast to ELPS that bound to CR3 after LBP-dependent neutrophil activation via CD14 or via direct TNF stimulation of the neutrophil.
The results described in this paper suggest that the way in which LPS is presented determines the mechanism of binding to neutrophils. Interaction of free LPS with neutrophils is completely mediated via CD14 and LBP, whereas for whole bacteria, the interaction is mainly independent of LBP and CD14. For E coated with LPS, a two-step mechanism is proposed: neutrophils are activated via CD14 and subsequently binding is mediated by CR3.
| Footnotes |
|---|
2 Abbreviations used in this paper: LBP, LPS-binding protein; C receptor 3, CR3; ELPS, E coated with LPS; LAL, limulus amebocyte lysate. ![]()
Received for publication January 23, 1998. Accepted for publication December 22, 1998.
| References |
|---|
|
|
|---|
mß2, Mac-1): indirect effects of an Arg-Gly-Asp (RGD) sequence. J. Immunol. 151:3324.[Abstract]
This article has been cited by other articles:
![]() |
E. M. Behrens, U. Sriram, D. K. Shivers, M. Gallucci, Z. Ma, T. H. Finkel, and S. Gallucci Complement Receptor 3 Ligation of Dendritic Cells Suppresses Their Stimulatory Capacity J. Immunol., May 15, 2007; 178(10): 6268 - 6279. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. K. Behera, E. Hildebrand, S. Uematsu, S. Akira, J. Coburn, and L. T. Hu Identification of a TLR-Independent Pathway for Borrelia burgdorferi-Induced Expression of Matrix Metalloproteinases and Inflammatory Mediators through Binding to Integrin {alpha}3beta1 J. Immunol., July 1, 2006; 177(1): 657 - 664. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. L. Gioannini, D. Zhang, A. Teghanemt, and J. P. Weiss An Essential Role for Albumin in the Interaction of Endotoxin with Lipopolysaccharide-binding Protein and sCD14 and Resultant Cell Activation J. Biol. Chem., November 27, 2002; 277(49): 47818 - 47825. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. M. Ellerbroek, A. I. M. Hoepelman, F. Wolbers, J. J. Zwaginga, and F. E. J. Coenjaerts Cryptococcal Glucuronoxylomannan Inhibits Adhesion of Neutrophils to Stimulated Endothelium In Vitro by Affecting Both Neutrophils and Endothelial Cells Infect. Immun., September 1, 2002; 70(9): 4762 - 4771. [Abstract] [Full Text] [PDF] |
||||
![]() |
F. E. J. Coenjaerts, A. M. E. Walenkamp, P. N. Mwinzi, J. Scharringa, H. A. T. Dekker, J. A. G. van Strijp, R. Cherniak, and A. I. M. Hoepelman Potent Inhibition of Neutrophil Migration by Cryptococcal Mannoprotein-4-Induced Desensitization J. Immunol., October 1, 2001; 167(7): 3988 - 3995. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. H. Flo, L. Ryan, L. Kilaas, G. Skjak-Brak, R. R. Ingalls, A. Sundan, D. T. Golenbock, and T. Espevik Involvement of CD14 and beta 2-Integrins in Activating Cells with Soluble and Particulate Lipopolysaccharides and Mannuronic Acid Polymers Infect. Immun., December 1, 2000; 68(12): 6770 - 6776. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. Dellacasagrande, E. Ghigo, S. Machergui-El, Hammami, R. Toman, D. Raoult, C. Capo, and J.-L. Mege alpha vbeta 3 Integrin and Bacterial Lipopolysaccharide Are Involved in Coxiella burnetii-Stimulated Production of Tumor Necrosis Factor by Human Monocytes Infect. Immun., October 1, 2000; 68(10): 5673 - 5678. [Abstract] [Full Text] [PDF] |
||||
![]() |
Q. Jiang, S. Akashi, K. Miyake, and H. R. Petty Cutting Edge: Lipopolysaccharide Induces Physical Proximity Between CD14 and Toll-Like Receptor 4 (TLR4) Prior to Nuclear Translocation of NF-{kappa}B J. Immunol., October 1, 2000; 165(7): 3541 - 3544. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. van der Flier, F. Coenjaerts, J. L. L. Kimpen, A. M. Hoepelman, and S. P. M. Geelen Streptococcus pneumoniae Induces Secretion of Vascular Endothelial Growth Factor by Human Neutrophils Infect. Immun., August 1, 2000; 68(8): 4792 - 4794. [Abstract] [Full Text] [PDF] |
||||
![]() |
P. Hofman, M. Piche, D. F. Far, G. Le Negrate, E. Selva, L. Landraud, A. Alliana-Schmid, P. Boquet, and B. Rossi Increased Escherichia coli Phagocytosis in Neutrophils That Have Transmigrated across a Cultured Intestinal Epithelium Infect. Immun., February 1, 2000; 68(2): 449 - 455. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |