| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Centocor, Malvern, PA 19355;
Pulmonary and Critical Care Division, University of Pennsylvania Medical Center, Philadelphia, PA 19104;
SmithKline Beecham Labs, King of Prussia, PA 19406;
§
The Blood Research Institute, The Blood Center of Southeastern Wisconsin, Milwaukee, WI 53233; and
¶
Department of Pathology, Stanford University School of Medicine, Palo Alto, CA 94305
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-induced
intradermal accumulation of neutrophils in a transmigration model using
human skin transplanted onto SCID mice. Localization of the epitopes of
these Abs indicated that these function-blocking Abs mapped to specific
regions on either face of domain 1. This suggests that these regions of
the first Ig-like domain may contain or be close to binding sites
involved in PECAM-1-dependent homophilic adhesion, and thus may
represent potential targets for the development of antiinflammatory
reagents. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A number of leukocyte and endothelial cell adhesion molecules contribute to neutrophil emigration, including PECAM-13 (CD31) (reviewed in Refs. 6, 7). PECAM-1, a transmembrane glycoprotein of the Ig superfamily, is expressed on the surface of leukocytes and platelets and on the endothelium, where it concentrates at cell-cell borders. Its extracellular region is composed of six Ig-like domains, while its long cytoplasmic tail has several serine and tyrosine residues whose phosphorylation may be important in regulating the activity of the molecule. PECAM-1 functions as both an adhesion molecule and a signal transducer through binding interactions with itself (homophilic adhesion) or with other non-PECAM-1 molecules (heterophilic adhesion) (6, 7).
The involvement of PECAM-1 in in vivo inflammation (reviewed in Ref. 8) has been established in a number of studies that have demonstrated that soluble forms of PECAM-1 or anti-PECAM-1 Abs block neutrophil recruitment in several animal models of acute inflammation (9, 10, 11, 12, 13, 14, 15). The mechanism of the involvement of PECAM-1, however, is currently still being defined. In all of the studies noted above, neutrophils were observed to accumulate within the vessels and/or in subendothelial areas, but were strikingly absent from extravascular sites, suggesting that in vivo, PECAM-1 is involved in the migration of neutrophils across the endothelium and perivascular basement membrane. Leukocyte transendothelial migration involves two steps: initial passage through endothelial intercellular junctions (diapedsis), followed by movement across the basement membrane. PECAM-1 may be required for both steps of this process, as Liao et al. (16) have shown in in vitro transendothelial migration studies that mAbs that blocked PECAM-1 homophilic adhesion inhibited movement across the endothelium, while Abs that inhibited heparin-inhibitable PECAM-1 heterophilic adhesion blocked passage through extracellular matrix.
We and others have shown that the first Ig-like domain of
PECAM-1 is required for homophilic adhesion
(17, 18, 19). A number of residues in domain 1 have been
implicated in homophilic adhesion (19); however, their
importance in mediating PECAM-1 function during in vivo
inflammation is not known. Therefore, to define functionally important
regions in domain 1, a panel of murine mAbs against the first Ig-like
domain of huPECAM-1 was generated and characterized in terms
of their effects in in vitro adhesion studies as well as on in vivo
neutrophil accumulation in a model of human skin transplanted onto SCID
mice. It was noted that Abs that disrupted homophilic adhesion also
blocked TNF--induced intradermal accumulation of neutrophils. The
epitopes of these functionally active Abs mapped to specific regions on
either face of the first Ig-like domain. These regions may represent
potential targets that could be exploited for the development of
anti-inflammatory PECAM-1 antagonists.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A huPECAM-1 cDNA fragment encoding Ig-like domains 1 and 2 was ligated into cDNA encoding the human IgG4 C region using PCR amplification methodology. The resulting construct was expressed in SP2/0 murine hybridoma cells, and recombinant protein (designated 1,2PECAM-1-IgG) was purified to homogeneity from culture supernatants by protein A chromatography. Mice were immunized and boosted with the 1,2PECAM-1-IgG fusion protein and spleen cells were isolated and fused with murine hybridoma F/0 cells (American Type Culture Collection, Manassas, VA). Clones producing anti-PECAM-1 mAbs were selected by ELISA for their reactivity with immobilized recombinant 1–2PECAM-IgG4 and their lack of reactivity with control IgG4 fusion protein. These clones were subcloned, and single clones were isolated with subsequent purification of the secreted mAb.
Preparation of Fab fragments by papain digest
Whole Ig was digested in PBS with 10 mM cysteine and 5 mM EDTA at pH 8.8. Papain (Boehringer Mannheim, Indianapolis, IN) was activated by dilution to 1 mg/ml in PBS containing 10 mM cysteine and 5 mM EDTA at pH 8.8 and incubating in a 37°C water bath for 30 min. Activated papain was added to IgG at 1% (w/w), and the digest mixture was maintained at 37°C. The digest was monitored by SDS-PAGE and quenched with 20 mM iodoacetamide upon completion. Fab was purified from Fc and fragments by protein A affinity chromatography.
Construction of human/mouse PECAM-1 chimeric constructs
A total of six human/mouse chimeric mutants are described in this work. The constructs are named by assuming a huPECAM-1 backbone followed by a listing of any regions of the molecule that contain mouse sequences. Thus, huPECAM-1(D1mu) represents huPECAM-1 construct in which Ig-like domain 1 has been replaced by muPECAM-1, while huPECAM-1(D1/1–25mu) designates a huPECAM-1 mutant in which aa 1–25 of Ig-like domain 1 have been replaced by the homologous muPECAM-1 sequence.
huPECAM-1, muPECAM-1, huPECAM-1(D1mu), and huPECAM-1(D2mu)
The preparation of these constructs has been previously described (17).
huPECAM-1(D1/1–25mu), huPECAM-1(D1/1–45mu), huPECAM-1(D1/46–103mu), and huPECAM-1(D1/71–103mu)
To generate mutant cDNAs of huPECAM-1 in which
sequences of domain 1 were replaced by the corresponding sequences from
domain 1 of muPECAM-1 (see Fig. 4
), the cDNA of
huPECAM-1 in the PBK.CMV expression vector (Stratagene, La
Jolla, CA) was cut with two unique restriction endonucleases,
ApaI and BstEII. ApaI cuts early in
the signal sequence at bp 159, while BstEII cuts at bp 819,
a site between the second and third Ig-like domains. Replacement
sequences were then engineered for each construct in a two-step process
in which two PCR-generated sequences (5' fragment and 3' fragment) were
joined together by means of sequence overlap extension
(20). The gel-purified 5' and 3' fragments were then used
as templates and joined together to create inserts using the forward
primer for the 5' fragment and the reverse primer for the 3' fragment.
The replacement sequences were then digested with ApaI and
BstEII and ligated into the previously digested
huPECAM-1/PBK.CMV construct. (The primers and templates used
to generate these constructs are available upon request.) All of the
mutant constructs were sequenced in the regions of change, and protein
expression was confirmed in COS cells by immunofluorescence staining. A
mesothelioma line, REN, was subsequently transfected with these
constructs using previously described procedures (21), and
protein expression was confirmed by FACS analysis and
immunoprecipitation. Unless otherwise stated, each clone studied
expressed PECAM-1 on its surface on greater than 90% of the
cells.
|
The aggregation assay used in these studies has been described
in detail previously (17). Briefly, stable L cell
transfectants, which had been plated (8–10 x
106 cells/75-cm2 flask) and
grown overnight, were removed with trypsin for huPECAM-1
constructs and nonenzymatically for muPECAM-1 constructs. The
cells were washed twice with 10 mM EDTA in PBS, pH 7.2, and twice with
HBSS without divalent cations. Cells were finally resuspended to a
concentration of 
0.8–1 x 106/ml in HBSS
with or without 1 mM calcium. After the cells had been dispersed to a
single cell suspension, 1-ml aliquots with and without Ab
(concentration 100 µg/ml) were transferred to wells in a 24-well
nontissue culture plastic tray (Costar, Cambridge, MA) that had been
previously incubated with 2% BSA in HBSS for at least 1 h and
washed thoroughly with HBSS immediately before use to prevent
nonspecific binding to the plastic of the tray. The trays containing
the suspended L cells were rotated on a gyratory platform (100 rpm) at
37°C for 30 min. Aggregation was quantified by examining
representative aliquots from each sample on a hemocytometer grid using
phase-contrast optics. The number of single cells (cells in aggregates
3) remaining vs those present in aggregates of greater than three
cells were counted from four 1-mm squares. At least 400 cells were
counted from each sample. The percentage of total cells present in
aggregates was used as the measure of aggregation. For comparison, the
aggregation data for each mAb were subsequently expressed as the
percentage of control aggregation, which was calculated by dividing the
aggregation in the presence of the mAb by the mean control aggregation
in the absence of mAb.
Binding of PECAM-1/IgG chimera to cellular transfectants
The generation of the PECAM-1/IgG chimera has been previously described (18). A total of 5 x 105 cells in 100 µl of HBSS containing 10% FBS with or without Ab (100 µg/ml) was mixed with 5 µg of PECAM-1/IgG in 25 µl of HBSS and allowed to incubate for 90 min at 22°C. Cells were then washed by centrifugation with 1 ml of HBSS and then resuspended in 100 µl HBSS containing 0.8 ml of FITC-labeled goat anti-human IgG chain (Jackson ImmunoResearch, West Grove, PA). Following a 30-min incubation at 4°C, the cells were washed once more in HBSS, resuspended in 0.6 ml of ice-cold HBSS, and subjected to flow-cytometric analysis using FACScan (Becton Dickinson, San Jose, CA) (18). For Ab studies, Ab concentration was 10 µg/ml.
FACS analysis
The various cell lines used in this study were treated with various anti-human or anti-murine PECAM-1 mAbs for 1 h at 4°C. The primary Ab was then removed, the cells were washed with PBS, and a 1/200 dilution of FITC-labeled goat anti-mouse or anti-rat secondary Ab (ICN Pharmaceuticals, Cappell, Aurora, OH) was added for 30 min at 4°C. After washing in PBS, flow cytometry was performed using an Ortho Cytofluorograph 50H cell sorter equipped with a 2150 data handling system (Ortho Instruments, Westwood, MA).
Skin transplantation
The protocols for skin transplantation have been described previously in detail (13, 22). Briefly, SCID mice were obtained from a colony maintained at the Wistar Institute Animal Facility. At 4–6 wk of age, the plasma of each mouse was tested for IgM production and only fully immunodeficient mice were used. Neonatal foreskins from elective circumcisions obtained using sterile techniques were cut into two halves and trimmed to a diameter of 1–1.5 cm. Full-thickness human skin grafts from the same donor were then transplanted into full-thickness, size-matched wound beds prepared on each flank of a SCID mouse. Mice were used for experiments only within the second month following human skin transplantation to ensure engraftment and stabilization. Only those mice whose grafts grossly showed no signs of inflammation or rejection were used.
Immunoperoxidase staining
Immunoperoxidase staining was performed as previously described (13, 22). Counterstaining was not used to enhance the detection of positive cells.
Quantification of leukocyte infiltration
To identify murine leukocytes, frozen sections of skin grafts were stained as described above with the anti-mouse Mac-1 mAb, MI/70.15, which does not cross-react with human Mac-1 (13). Four to six sections were cut from the center of each skin biopsy, as marked by colloidal carbon (see below), to obtain representative samples of the injection site. The number of leukocytes per 3–6 microscope fields (x200) within close proximity of the colloidal carbon in each section were directly counted to give an estimate of the number of leukocytes. Because some grafts demonstrated a cuff of leukocytes at the murine-human skin junctional regions, the areas counted were in the immediate subepidermal regions not adjacent to murine skin.
Effect of Abs on TNF-induced migration of murine leukocytes
The protocols for this model of TNF-induced neutrophil
recruitment have been described previously (13, 22). A
skin graft on one side of each mouse was injected intradermally with 50
µl of endotoxin-free saline containing 2 mg/ml of BSA (Sigma, St.
Louis, MO) as a protein carrier plus colloidal carbon (Sigma) to mark
the site of injection (control side). The skin graft on the
contralateral side was injected with 6000 U of human rTNF-
(Boehringer Mannheim), diluted in 50 µl of endotoxin-free saline
containing 2 mg/ml BSA with colloidal carbon (experimental side). Abs,
as Fab fragments, were injected i.v. via tail vein in a final volume of
100 µl with endotoxin-free saline containing 2 mg/ml BSA immediately
before the intradermal skin injections. Mice were sacrificed 6 h
after saline or TNF- intradermal injection, and the skin grafts were
carefully dissected from the animals. The grafts were cut into two
halves through the center of the injection site marked by the colloidal
carbon. Each half was oriented, placed in OCT compound, and snap frozen
for immunohistochemical analysis, as described above.
mAb biotinylation
Protein A-purified anti-PECAM-1 mAbs were biotinylated by reaction with normal human serum/biotin according to manufacturer’s specification (Boehringer Mannheim). Following dialysis, biotinylated anti-PECAM-1 mAbs were tested for their ability to bind ELISA wells that had previously been coated with 5 µg/ml of 1,2PECAM-IgG4 in bicarbonate buffer. The assays were developed by incubation first with avidin-HRP and then the chromogenic substrate o-phenylenediamine (Sigma). For each biotinylated mAb, the concentration corresponding to the upper linear portion of the titration curve was determined and was selected as the labeled mAb concentration used in subsequent competitive binding assays. In most cases, this standard concentration was about 1 µg/ml.
Competitive Ab-binding studies
Anti-PECAM-1 mAbs with related epitope specificity were identified by pairwise competitive binding studies, as follows. ELISA wells were coated with 5 µg/ml of 1,2PECAM-IgG4 in bicarbonate buffer washed and blocked with TBS containing 0.05% Tween 20. Each of the unlabeled mAbs was titrated with the standard, nonsaturating amount of the individual biotinylated mAbs determined above. Ab pairs that could reciprocally and completely inhibit binding of each other were considered to recognize overlapping, spatially related epitopes.
Determination of species cross-reactivity
Primary and endothelial cell lines form the indicated species were purchased from Cell Systems (Kirkland, WA) and grown to confluence in 96-well tissue culture plates. Media were removed, and murine anti-PECAM-1 Abs (20 µg/ml) were added and incubated for 1 h at 37°C. Cells were washed twice with media, and 1 µg/ml of radiolabeled 125I-anti-mouse secondary Ab (goat anti-mouse IgG F(ab')2; Jackson ImmunoResearch) with a sp. act. of 1 mCi/mg was added and incubated for 30 min at 37°C. Unbound Ab was removed with three washes, and the radioactivity in each well was counted.
Creation of a molecular model of domains 1 and 2 of huPECAM-1
Molecular modeling was done using Sybyl 6.2 on an SGI Crimson
R44000-150 with VGXT graphics. Molecular alignment was done with
MegAlign from DNAstar (Madison, WI). The sequence of VCAM-1 was aligned
with PECAM-1. Using the crystal structure of the first two
domains of VCAM-1 (23) and the sequence alignment, amino
acids in VCAM-1 were electronically mutated to the PECAM-1
sequence. Each change was visually examined for steric and electronic
interactions. Where unfavorable interactions existed, the side chains
of the relevant amino acids were automatically repositioned using
Sybyl. There were four regions of PECAM-1 that differed in
amino acid number from VCAM-1, two requiring the addition of amino
acids and two requiring deletions. For each region, a
computer-generated loop search was done using anchoring residues 5
amino acids from either end of the altered regions. The results of each
loop search were individually electronically grafted into the
PECAM-1 model, and the resulting structures were examined for
steric and electronic compatibility. Most structures were eliminated
based on backbone-backbone or backbone-side chain interactions. In all
instances, the remaining structures were similar. Where severe side
chain-side chain interactions were present, side chains were
automatically repositioned. The most energetically favored structures
were selected and combined in the PECAM-1 model. The crystal
structure of VCAM-1 does not contain amino acids corresponding to the
first seven amino acids of PECAM-1. To fill in this region, a
BLAST search was done using PECAM-1[1–16]. Several similar
structures were identified. The sequence 37–52 of Neutral Protease
(E.C.3.4.24.27, PDB1NPC.ENT) was selected and grafted on the N terminus
of the PECAM-1 model using a least square fit of the
carbons of the sequence NSV (PECAM-1[8–10]) for
positioning. Essential hydrogens and charges were added to the
structure. The model was subjected to 250 cycles of minimization
(steepest descent, NB cutoff 8 Å, dielectric 10). The resulting
structure was solvated with water and subjected to an additional 100
cycles of minimization (steepest descent). This was followed by 50 fs
of dynamics simulation (300°K, NB cutoff 8 Å, dielectric 10, random
velocity), 100 cycles of minimization (steepest descent), and 1000
cycles of conjugate gradient minimization.
Statistical analysis
Differences among groups were analyzed using one-way ANOVA. When statistically significant differences were found (p < 0.05), individual comparisons were made using the Bonferroni/Dunn test.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A panel of murine mAbs was generated using standard fusion
procedures following immunization with recombinant protein composed of
Ig-like domains 1 and 2 of huPECAM-1 fused to an Fc region of
human IgG
4 isotype (1,2PECAM-1-IgG). Using cellular transfectants
expressing mutant forms of huPECAM-1, six of these Abs (mAbs
24, 35, 37, 40, 51, and 62) were found to bind only to the Ig-like
domain 1 (Fig. 1) and were subsequently
studied in more detail. All Abs were of the IgG1 isotype except for mAb
62, which was IgG2a.
|
The Abs were first investigated for their effects on
PECAM-1-dependent adhesion. Initial adhesion studies were done with
an aggregation assay involving L cell fibroblasts transfected with
PECAM-1 (17, 24). Previous studies with this
system have demonstrated that L cell transfectants expressing the
full-length huPECAM-1 mediate cell-cell aggregation that
involves, on opposing cells, the interaction of huPECAM-1
with an unidentified non-PECAM-1 ligand (possibly a
heparin-containing proteoglycan) (heterophilic adhesion). In contrast,
a mutant missing exon 14 (huPECAM-1
14) exclusively
forms cellular aggregates that are the result of
PECAM-1/PECAM-1-binding interactions (homophilic
adhesion). Given this, the Abs were therefore tested for their effect
on PECAM-1-dependent L cell aggregation mediated by either
huPECAM-1 or huPECAM-114 (Fig. 2). Five of the six Abs (100 µg/ml)
inhibited PECAM-1-dependent L cell aggregation: three of
these Abs, mAbs 24, 40, and 62, blocked L cell aggregation triggered by
huPECAM-1 and huPECAM-114 (homophilic and
heterophilic aggregation), while the other two Abs, mAbs 37 and 51,
only inhibited aggregation mediated by huPECAM-1
(heterophilic aggregation). Unlike the other Abs, mAb 35 increased the
aggregation seen with transfectants expressing either of the
PECAM-1 constructs.
|
). All of the Abs blocked the
binding of the PECAM/IgG chimera except for mAb 35, which appeared to
augment adhesion (by 145.5 ± 6.5% compared with control;
n = 2). (The effects of the Abs in these two adhesion
assays are summarized in Fig. 4
.)
|
Domain 1 anti-PECAM-1 Abs that inhibit PECAM-1-dependent homophilic L cell aggregation also inhibit in vivo neutrophil accumulation
A model of the human vasculature has been developed in which human
skin (neonatal foreskin) is transplanted onto SCID mice (13, 22). The vessels of the grafts remain principally human for
several months, while the circulating murine leukocytes (principally
neutrophils) retain the ability to interact with the human vasculature
and thus are able to participate in an inflammatory reaction. In this
model, two halves of the same foreskin are transplanted onto the back
of each SCID mouse. Intradermal injection of TNF- induces a
2–4-fold increase in the number of leukocytes accumulating in the
dermis compared with grafts injected with saline. Using this model, we
have demonstrated that Ab against huPECAM-1 blocks
TNF--induced recruitment of neutrophils into the skin grafts
(9, 13). To our knowledge, this model represents the only
system available to look at the response of the human vasculature to an
inflammatory stimulus in an in vivo context, although this assay is
limited by the requirement that the human molecules must be able to
interact with the homologous mouse counter-receptor.
Therefore, to evaluate the effect of these Abs on in vivo inflammation,
we studied their effect on the accumulation of neutrophils in
TNF--stimulated skin grafts (Fig. 5).
As these Abs do not recognize muPECAM-1 (Fig. 1), in human/SCID mice chimeras they bind
only to huPECAM-1 on the human vasculature of the graft. As
previously observed in animals given BSA i.v., the intradermal
injection of TNF-, but not saline, resulted in significant
accumulation of leukocytes within the dermis (9, 13).
However, i.v. injection of mAb 24, 40, 35, and 62 Fab fragments (100
µg Ab/mouse) significantly blocked the TNF--induced extravasation
of leukocytes into the dermis (p < 0.05). In
contrast, mAb 37 and 51 Fab fragments (100 µg Ab/mouse) did not
significantly impair neutrophil recruitment in response to intradermal
TNF-. Fab fragments were used for these experiments, and thus the
effects of the blocking Abs were unlikely to be due to Fc interactions.
The Abs, as Fab fragments, did not induce neutropenia. As mAbs 24, 40,
35, and 62 blocked or augmented PECAM-1-dependent homophilic
L cell aggregation while mAbs 37 and 51 did not, these data suggest
that the ability to disrupt PECAM-1-mediated homophilic
adhesion as defined by the L cell assay may correlate with the ability
of the domain 1 Abs to block in vivo inflammation in the human
skin/SCID mouse model.
|
To better understand the basis of the differential effects of these domain 1 Abs, their binding epitopes were localized. To accomplish this, we employed a sequential strategy that involved: 1) competitive Ab inhibition; 2) analysis of Ab binding to human/mouse PECAM-1 chimeric mutants; and 3) Ab species cross-reactivity and comparisons of the amino acid sequences of human, rat, bovine, and PECAM-1.
Four pairs of domain 1 anti-PECAM-1 Abs are identified by competitive Ab-binding studies
We first studied the ability of unlabeled Abs to compete for the binding of biotinylated Abs to 1,2-PECAM-1-IgG recombinant protein. From these studies, four pairs of cross-blocking Abs (24 and 40; 37 and 51; 35 and 62; and 51 and 62) were defined (Fig. 6). These data suggested that the Abs in each pair bound to the same epitope or to epitopes that were spatially close to each other.
The binding epitopes for the domain 1 anti-PECAM-1 Abs are clustered in two distinct linear sequences
To locate the linear sequences within domain 1 that contained the
binding epitopes, we took advantage of the fact that these Abs do not
recognize muPECAM-1 and determined their binding to chimeric
receptors made by interchanging homologous regions of domain 1 from
mouse and human PECAM-1. This approach is particularly
appealing for epitope analysis, as the generation of chimeric
human/mouse PECAM-1 receptors results in mutants in which the
tertiary structure is presumably preserved. Analysis of the pattern of
Ab binding to these chimeric constructs indicated that the binding
epitopes for the Abs were located in two distinct but separated
sequences (aa 26–45 and aa 71–103) (Fig. 7). Specifically, mAbs 37, 51, and 62
bound to constructs containing the human sequence for aa 26–45, while
mAbs 24, 40, and 35 bound to mutants that carried the human sequence
for aa 71–103.
|
The Abs were tested for their ability to bind to endothelial cells
derived from human, rat, bovine, and murine sources (Fig. 8). All of the Abs recognized
PECAM-1 on human but not murine endothelial cells (see also
Table I). mAb 37 also bound rat endothelial cells, while 62 also bound
to endothelial cells of rat and bovine sources. The species reactivity
of these Abs coupled with the data described above allowed us to map
the presumed binding epitopes by comparing the amino sequences for the
different species.
|
A shows the amino acid
sequence in domain 1 (aa 26–45) that contains the epitopes for mAbs
37, 51, and 62. As mAb 62 binds to human, rat, and bovine but not
murine endothelial cells, it must recognize a sequence that is common
to human, rat, and bovine but not murine PECAM-1. The
sequence, LTLQC, best fulfills this requirement. mAb 37 binds to
endothelial cells derived from human and rat sources, but not to those
from bovine or murine sources. It therefore must bind to a sequence
that is present in human and rat but not bovine or murine
PECAM-1. The sequence most consistent with this is PQHQ. mAb
51 recognizes only huPECAM-1, but its epitope must be
spatially close to mAb 37, as the two competitively inhibit the binding
of each other. Given this, a review of the sequences suggests that the
epitope for mAb 51 includes HVKP.
|
B shows the amino acid sequence in domain 1 that
contains the epitopes for mAbs 24, 35, and 40 (aa 71–103). Comparison
of the human and murine sequences indicates that from aa 77–96 there
is a high degree of homology, while there are two regions (EVRIYD, aa
71–76 and QLLVE, aa 97–101) in which the two sequences differ
significantly. This suggests that the binding epitopes for these three
Abs are located in these latter sequences. Polymorphism of
PECAM-1 has been observed in which a frequent
CTG
GTG substitution has been found at codon
125, resulting in alleles that code either leucine or valine at aa 92
in wild-type huPECAM-1 (26). We noted that mAbs
24 and 40 bound 1.5 times more strongly to cells that were homozygous
for valine than those that were homozygous for leucine at this site,
suggesting that the epitopes for 24 and 40 were located at QLLVE. Given
that mAb 35 did not competitively inhibit the binding of mAbs 24 and
40, this suggested that epitope for mAb 35 was located in the sequence
EVRIYD. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-induced intradermal
accumulation of neutrophils in a model of human skin transplanted onto
SCID mice. Localization of the epitopes of these Abs indicated that
these function-blocking Abs mapped to specific regions of domain 1,
regions that may contain or be close to the binding sites for
PECAM-1-dependent homophilic adhesion, and thus may represent
potential targets for the development of antiinflammatory reagents. The involvement of PECAM-1 in acute inflammation (8) has been established by studies that have shown that inhibition of PECAM-1 with Abs or soluble PECAM-1 constructs blocks neutrophil (and monocyte) recruitment in models of peritonitis (9, 10, 11, 13, 14, 15), acute lung injury (9), myocardial ischemia-reperfusion injury (11, 12), and cutaneous inflammation (9, 13). Recently, Liao et al. (14) have demonstrated that a chimeric protein composed of the first Ig-like domain PECAM-1 fused to an IgG Fc region, which presumably binds only homophilically, blocks leukocyte recruitment in chemical-induced peritonitis. Our finding that Abs against domain 1 that block homophilic adhesion in the L cell aggregation assay are also active in vivo is therefore consistent with their observations. Interestingly, neutrophil extravasation into extravascular sites was associated with a loss of PECAM-1 from the surface of extravasated leukocytes (13, 27). Given these observations, understanding the mechanism(s) and the ligand-binding interactions that are involved in these processes has been a subject of recent investigation.
Based on its concentration at intercellular junctions, ability to transduce intracellular signals, and its multiple potential ligand-binding interactions, two distinct levels of involvement in leukocyte recruitment have been proposed for PECAM-1. First, PECAM-1/PECAM-1 interactions between the neutrophil and the endothelium facilitate movement of the leukocyte across the endothelium. These homophilic interactions could activate integrin-mediated adhesion to the endothelium or make endothelial intercellular junctions permeable to neutrophils by increasing intercellular calcium. Second, following passage through endothelial cell-cell junctions, the binding of neutrophil PECAM-1 to one of its yet to be identified non-PECAM-1 ligands triggers the movement of neutrophils through the basement membrane possibly by stimulating integrin-dependent migration or protease release. Evidence supporting this proposal has come from in vitro studies of neutrophil transendothelial migration that demonstrated that Abs that blocked PECAM-1 homophilic adhesion inhibited movement across the endothelium, while Abs that inhibited heparin-inhibitable PECAM-1 heterophilic adhesion blocked passage through extracellular matrix (16).
A critical element of this model is the presumed ability of PECAM-1 to bind to itself (homophilic adhesion) or other non-PECAM-1 molecules (heterophilic adhesion). Data from a number of assay systems have convincingly established that PECAM-1 is able to bind homophilically (reviewed in Ref. 6). Evidence indicates that PECAM-1-dependent homophilic adhesion involves the direct interaction of the first two extracellular Ig-like domains in an antiparallel manner, in which domain 1 of one PECAM-1 molecule binds to domain 1 (19) or domain 2 of the counter molecule of PECAM-1 (17, 18).
PECAM-1 has also been reported to mediate heterophilic
binding to a number of non-PECAM-1 ligands. These include a
heparin-containing proteoglycan (25, 28, 29, 30), the integrin
vß3 (29, 31), and CD38 (32). The binding sites in
PECAM-1 responsible for interactions with two of these
ligands, the putative proteoglycan and
vß3, were reported to
be localized to regions in the second Ig-like domain (25, 29). Recently, however, the validity of some of these presumed
ligand interactions has been challenged (18).
Specifically, in assays that studied the binding of PECAM-1/IgG
chimeric proteins, or purified PECAM-1 in phospholipid
vesicles, to PECAM-1-expressing cells,
PECAM-1/vß3
and PECAM-1/heparin interactions were not confirmed
(18). Furthermore, in a subsequent detailed analysis,
PECAM-1 was not found to bind heparin (33).
It is clear that some (but by no means all) of the differences may be due to the adhesion assay system used (cell-cell aggregation vs the binding of PECAM-1 constructs to cells). Each assay is admittedly artificial, with limitations, and each may in fact be looking at a different aspect of PECAM-1-dependent adhesion. We would note, however, for one animal model of neutrophil recruitment, the L cell aggregation assay appeared to correlate the best with in vivo observations. Specifically, for adhesion mediated by the PECAM/Ig chimera, all of the Abs (excluding mAb 35) blocked in vitro (homophilic) adhesion, yet two of the five Abs did not inhibit cutaneous inflammation, while none of the Abs that were active during in vivo inflammation blocked the aggregation PECAM-1 L1/2 transfectants. In contrast, the L cell aggregation assay identified two groups of Abs: one that was capable of blocking PECAM-1 homophilic aggregation and a second that exclusively blocked heterophilic aggregation. Only those Abs that were able to block homophilic aggregation inhibited cutaneous inflammation. Supporting this are the findings that mAb 1.3 (epitope includes residues 37–42 of domain 1; P. J. Newman and C. Paddock, unpublished observations), which does not effect PECAM-1-dependent L cell aggregation (34), and mAb 4G6 (epitope in domain 6; see Ref. 35), which blocks only heterophilic aggregation (34), also do not inhibit neutrophil accumulation in the human skin/SCID mouse chimera model (13 and H. DeLisser, unpublished observations). Currently, the one exception to the above is mAb Hec 7. This Ab, which has a complex epitope involving both domains 1 and 2 (16, 35), has been previously shown to inhibit cutaneous inflammation (13), but not PECAM-1-dependent L cell aggregation (34). mAb 35 differs from the other Abs in that although it promotes in vitro adhesion, it blocks cutaneous inflammation. The reasons for the inhibition observed in vivo are not clear, but may represent antagonism of PECAM-1 function due to inappropriate or excessive adhesion.
It has been previously noted that L cell transfectants expressing muPECAM-1 or huPECAM-1 when mixed together do not form mixed aggregates with each other, indicating that in the L cell system, muPECAM-1 and huPECAM-1 do not interact with each other (17). Thus, although the results of the L cell aggregation assay correlated the best with SCID mouse/human skin model of inflammation, these previous data suggest a possible limitation of L cell system. There is, however, evidence from immunostaining of human skin grafts on SCID mice that for murine vessels that have fused with human vessels, mouse and human PECAM-1 on the endothelium do in fact interact with each other (36 , H. DeLisser, unpublished observations).
An important question, raised by the model that proposes that PECAM-1-dependent heterophilic adhesion is involved in in vivo leukocyte recruitment, is why heterophilic blocking Abs (mAbs 37, 51, and 4G6) do not also block neutrophil accumulation in the human skin/SCID mouse model? The lack of effect of these Abs on cutaneous inflammation is not surprising, however, if the species reactivity is considered. As these Abs do not bind muPECAM-1, they cannot bind murine neutrophils, and thus they would not be expected to impact on this step of the process that is proposed to involve only neutrophil PECAM-1.
Newton and associates (19) have recently identified two
clusters of acidic residues (D11 and K89 and D33, D51 and K50) that map
to opposite faces of domain 1 and that appear to mediate
PECAM-1-dependent homophilic binding. Given these findings and the
results of the L cell aggregation studies, the location of the binding
epitopes for our domain 1 Abs was compared with the position of these
sequences using a computer-generated model of the structure of the
first Ig-like domain of PECAM-1 (see the methods for the
strategies employed in producing this model). From this model, a number
of observations can be made about the Abs we studied (Fig. 10). The binding epitopes for our
domain 1 Abs do not include any of these five residues. This is not
unexpected given that our anti-human PECAM-1 Abs were
generated in mice and thus only nonhomologous regions would be
immunogenic. However, the epitopes of mAbs 62 and 24 and 40 are
positioned such that Ab binding would be anticipated to shield two of
these residues (D11 and D33), although their effects could still be
mediated by distortions of the conformation of the molecule. In
contrast, mAbs 37 and 51 bind to sequences that are remote from any of
these five acidic residues and thus, not surprisingly, Ab binding did
not inhibit L cell-dependent homophilic aggregation. Ab 35 is
intriguing in that its binding epitope is close to residues D51 and
K50, but its binding in fact augments rather than inhibits adhesion.
This suggests that this Ab may alter the conformation of the molecule
in a way that makes these residues more accessible for binding to other
PECAM-1 molecules. Finally, the Ab cross-blocking studies
indicated that the epitopes for mAbs 35 and 62 were spatially close to
one another, but our model proposes that the epitopes are widely
separated. This points to a potential limitation of using homologies
with VCAM-1 to derive the structure of PECAM-1 and suggests
that determination of the actual crystal structure of the first Ig-like
domain of PECAM-1 will reveal that the epitopes for these two
Abs are located more closely to each other than is suggested by our
computer-generated model.
|
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Horace M. DeLisser, 806, BRB II/III, 421 Curie Boulevard, Philadelphia, PA 19104-6160. E-mail address:
3 Abbreviations used in this paper: PECAM-1, platelet endothelial cell adhesion molecule-1; huPECAM-1, human PECAM-1; 125I-anti-mouse, 125I-labeled anti-mouse; MFI, mean fluorescence intensity; muPECAM-1, murine PECAM-1.
Received for publication December 22, 1998. Accepted for publication October 12, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
vß3 integrin involved in adhesion of leukocytes to endothelium. J. Cell Biol. 130:451.This article has been cited by other articles:
|
|
 |
|
  G. Cao, M. L. Fehrenbach, J. T. Williams, J. M. Finklestein, J.-X. Zhu, and H. M. DeLisser Angiogenesis in Platelet Endothelial Cell Adhesion Molecule-1-Null Mice Am. J. Pathol., August 1, 2009; 175(2): 903 - 915. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Garnacho, V. Shuvaev, A. Thomas, L. McKenna, J. Sun, M. Koval, S. Albelda, V. Muzykantov, and S. Muro RhoA activation and actin reorganization involved in endothelial CAM-mediated endocytosis of anti-PECAM carriers: critical role for tyrosine 686 in the cytoplasmic tail of PECAM-1 Blood, March 15, 2008; 111(6): 3024 - 3033. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. M. DeLisser, B. P. Helmke, G. Cao, P. M. Egan, D. Taichman, M. Fehrenbach, A. Zaman, Z. Cui, G. S. Mohan, H. S. Baldwin, et al. Loss of PECAM-1 Function Impairs Alveolarization J. Biol. Chem., March 31, 2006; 281(13): 8724 - 8731. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Wu, P. Stabach, M. Michaud, and J. A. Madri Neutrophils Lacking Platelet-Endothelial Cell Adhesion Molecule-1 Exhibit Loss of Directionality and Motility in CXCR2-Mediated Chemotaxis J. Immunol., September 15, 2005; 175(6): 3484 - 3491. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. D. O'Brien, G. Cao, A. Makrigiannakis, and H. M. DeLisser Role of immunoreceptor tyrosine-based inhibitory motifs of PECAM-1 in PECAM-1-dependent cell migration Am J Physiol Cell Physiol, October 1, 2004; 287(4): C1103 - C1113. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. M. Albelda, K. C. Lau, P. Chien, Z.-Y. Huang, E. Arguiris, A. Bohen, J. Sun, J. A. Billet, M. Christofidou-Solomidou, Z. K. Indik, et al. Role for Platelet-Endothelial Cell Adhesion Molecule-1 in Macrophage Fc{gamma} Receptor Function Am. J. Respir. Cell Mol. Biol., August 1, 2004; 31(2): 246 - 255. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. van Buul and P. L. Hordijk Signaling in Leukocyte Transendothelial Migration Arterioscler Thromb Vasc Biol, May 1, 2004; 24(5): 824 - 833. [Abstract] [Full Text]
|
|
|
|
|
|
B. N. Stein, J. R. Gamble, S. M. Pitson, M. A. Vadas, and Y. Khew-Goodall Activation of Endothelial Extracellular Signal-Regulated Kinase Is Essential for Neutrophil Transmigration: Potential Involvement of a Soluble Neutrophil Factor in Endothelial Activation J. Immunol., December 1, 2003; 171(11): 6097 - 6104. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Christofidou-Solomidou, A. Scherpereel, R. Wiewrodt, K. Ng, T. Sweitzer, E. Arguiri, V. Shuvaev, C. C. Solomides, S. M. Albelda, and V. R. Muzykantov PECAM-directed delivery of catalase to endothelium protects against pulmonary vascular oxidative stress Am J Physiol Lung Cell Mol Physiol, August 1, 2003; 285(2): L283 - L292. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. L. Borjesson, S. I. Simon, E. Hodzic, H. E. V. DeCock, C. M. Ballantyne, and S. W. Barthold Roles of neutrophil beta 2 integrins in kinetics of bacteremia, extravasation, and tick acquisition of Anaplasma phagocytophila in mice Blood, April 15, 2003; 101(8): 3257 - 3264. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Righi, S. Deaglio, C. Pecchioni, A. Gregorini, A. L. Horenstein, G. Bussolati, A. Sapino, and F. Malavasi Role of CD31/Platelet Endothelial Cell Adhesion Molecule-1 Expression in in Vitro and in Vivo Growth and Differentiation of Human Breast Cancer Cells Am. J. Pathol., April 1, 2003; 162(4): 1163 - 1174. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. D. O'Brien, P. Lim, J. Sun, and S. M. Albelda PECAM-1-dependent neutrophil transmigration is independent of monolayer PECAM-1 signaling or localization Blood, April 1, 2003; 101(7): 2816 - 2825. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Tasaka, L. Qin, A. Saijo, S. M. Albelda, H. M. DeLisser, and C. M. Doerschuk Platelet Endothelial Cell Adhesion Molecule-1 in Neutrophil Emigration during Acute Bacterial Pneumonia in Mice and Rats Am. J. Respir. Crit. Care Med., January 15, 2003; 167(2): 164 - 170. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W.-H. Su, H.-i. Chen, and C. J. Jen Differential movements of VE-cadherin and PECAM-1 during transmigration of polymorphonuclear leukocytes through human umbilical vein endothelium Blood, November 15, 2002; 100(10): 3597 - 3603. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Dangerfield, K. Y. Larbi, M.-T. Huang, A. Dewar, and S. Nourshargh PECAM-1 (CD31) Homophilic Interaction Up-Regulates {alpha}6{beta}1 on Transmigrated Neutrophils In Vivo and Plays a Functional Role in the Ability of {alpha}6 Integrins to Mediate Leukocyte Migration through the Perivascular Basement Membrane J. Exp. Med., November 4, 2002; 196(9): 1201 - 1212. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M.-X. Wong, D. Roberts, P. A. Bartley, and D. E. Jackson Absence of Platelet Endothelial Cell Adhesion Molecule-1 (CD31) Leads to Increased Severity of Local and Systemic IgE-Mediated Anaphylaxis and Modulation of Mast Cell Activation J. Immunol., June 15, 2002; 168(12): 6455 - 6462. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Ji, C. D. O'Brien, M. Feldman, Y. Manevich, P. Lim, J. Sun, S. M. Albelda, and M. I. Kotlikoff PECAM-1 (CD31) regulates a hydrogen peroxide-activated nonselective cation channel in endothelial cells J. Cell Biol., April 1, 2002; 157(1): 173 - 184. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Scherpereel, J. J. Rome, R. Wiewrodt, S. C. Watkins, D. W. Harshaw, S. Alder, M. Christofidou-Solomidou, E. Haut, J.-C. Murciano, M. Nakada, et al. Platelet-Endothelial Cell Adhesion Molecule-1-Directed Immunotargeting to Cardiopulmonary Vasculature J. Pharmacol. Exp. Ther., March 1, 2002; 300(3): 777 - 786. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Christofidou-Solomidou, S. Kennel, A. Scherpereel, R. Wiewrodt, C. C. Solomides, G. G. Pietra, J.-C. Murciano, S. A. Shah, H. Ischiropoulos, S. M. Albelda, et al. Vascular Immunotargeting of Glucose Oxidase to the Endothelial Antigens Induces Distinct Forms of Oxidant Acute Lung Injury : Targeting to Thrombomodulin, But Not to PECAM-1, Causes Pulmonary Thrombosis and Neutrophil Transmigration Am. J. Pathol., March 1, 2002; 160(3): 1155 - 1169. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. RayChaudhury, M. Elkins, D. Kozien, and M. T. Nakada Regulation of PECAM-1 in Endothelial Cells during Cell Growth and Migration Experimental Biology and Medicine, July 1, 2001; 226(7): 686 - 691. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. M. Danilov, V. D. Gavrilyuk, F. E. Franke, K. Pauls, D. W. Harshaw, T. D. McDonald, D. J. Miletich, and V. R. Muzykantov Lung uptake of antibodies to endothelial antigens: key determinants of vascular immunotargeting Am J Physiol Lung Cell Mol Physiol, June 1, 2001; 280(6): L1335 - L1347. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Wada, D. E. Reese, and D. M. Bader Bves: prototype of a new class of cell adhesion molecules expressed during coronary artery development Development, June 1, 2001; 128(11): 2085 - 2093. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. T. Kaifi, E. Diaconu, and E. Pearlman Distinct Roles for PECAM-1, ICAM-1, and VCAM-1 in Recruitment of Neutrophils and Eosinophils to the Cornea in Ocular Onchocerciasis (River Blindness) J. Immunol., June 1, 2001; 166(11): 6795 - 6801. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. D. Thompson, K. E. Noble, K. Y. Larbi, A. Dewar, G. S. Duncan, T. W. Mak, and S. Nourshargh Platelet-endothelial cell adhesion molecule-1 (PECAM-1)-deficient mice demonstrate a transient and cytokine-specific role for PECAM-1 in leukocyte migration through the perivascular basement membrane Blood, March 15, 2001; 97(6): 1854 - 1860. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Zhao and P. J. Newman Integrin Activation by Regulated Dimerization and Oligomerization of Platelet Endothelial Cell Adhesion Molecule (Pecam)-1 from within the Cell J. Cell Biol., January 8, 2001; 152(1): 65 - 74. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. D. Thompson, M. W. Wakelin, K. Y. Larbi, A. Dewar, G. Asimakopoulos, M. A. Horton, M. T. Nakada, and S. Nourshargh Divergent Effects of Platelet-Endothelial Cell Adhesion Molecule-1 and {beta}3 Integrin Blockade on Leukocyte Transmigration In Vivo J. Immunol., July 1, 2000; 165(1): 426 - 434. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Ji, C. D. O'Brien, M. Feldman, Y. Manevich, P. Lim, J. Sun, S. M. Albelda, and M. I. Kotlikoff PECAM-1 (CD31) regulates a hydrogen peroxide-activated nonselective cation channel in endothelial cells J. Cell Biol., April 1, 2002; 157(1): 173 - 184. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Cao, C. D. O'Brien, Z. Zhou, S. M. Sanders, J. N. Greenbaum, A. Makrigiannakis, and H. M. DeLisser Involvement of human PECAM-1 in angiogenesis and in vitro endothelial cell migration Am J Physiol Cell Physiol, May 1, 2002; 282(5): C1181 - C1190. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
= inhibition of adhesion; 
= augmentation of adhesion; NE = no effect). Five of
the six Abs inhibited PECAM-1-dependent L cell aggregation:
three (mAbs 24, 40, and 62) blocked both homophilic and heterophilic
interactions, while two (mAb 37, 51) inhibited only heterophilic
interactions. mAb 35 increased PECAM-1-dependent L cell
aggregation. All of the Abs completely blocked the binding of the
PECAM/IgG chimera to PECAM-1-expressing transfectants, except
for mAb 35, which appeared to augment adhesion.
