|
Volume 270,
Number 14,
Issue of April 7, 1995 pp. 8164-8171
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
A Novel Role for
IgG-Fc
TRANSDUCTIONAL POTENTIATION FOR HUMAN HIGH AFFINITY Fc
RECEPTOR (Fc RI) SIGNALING (*)
(Received for publication, December 8, 1994)
Lorraine C.
Pfefferkorn
(1), (§),
Jan G. J.
van de Winkel
(2),
Sharon L.
Swink
(1)From the
(1)Department of Microbiology, Dartmouth
Medical School, Lebanon, New Hampshire 03756 and the
(2)Department of Experimental Immunology, University
Hospital Utrecht, Utrecht, The Netherlands
ABSTRACT
INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
ABSTRACT
Human type 1 Fc receptors (Fc RI) bind with high
affinity (K =
10 M) Fc regions of monomeric IgG1 and
IgG3. As demonstrated in this report, interaction of IgG-Fc with the
ligand binding site on Fc RI alters its capacity for
aggregation-dependent signaling. This Fc-dependence was demonstrated in
normal monocytes and U937-10.6 cells exposed to monomeric IgG and
then to anti-Fc RI F(ab`) that cross-link the receptor.
Using O production to measure cell
signaling, we found that binding by high affinity IgGs of various
species, as well as by murine hybrid IgGs containing only one high
affinity heavy chain, resulted in a marked increase in
Fc RI-mediated signaling. Preaggregated Fc RI/IgG had a ratio
of one. IgG binding after aggregation of unligated Fc RI did not
restore signaling. Dose responses indicated that concentrations of IgG
that saturated Fc RI optimized transductional activity. The
inclusion of unligated with ligated Fc RI in aggregates depressed
activity, indicating a lack of trans-activation of unligated Fc RI.
Significantly, IgG-binding markedly increased aggregation-dependent
tyrosine phosphorylation of Fc RI -chains and the association
of tyrosine phosphorylated Syk. Thus, the consequences of IgG-Fc
binding were increases in aggregation-dependent phosphorylation of
Fc RI -chains, recruitment of pp72Syk to Fc RI, and
signaling of the NADPH oxidase pathway.
INTRODUCTION
Fc receptors (Fc R) ( )bind the Fc region of
immunoglobulin G (IgG) and are expressed on the surface of a variety of
cells. The human high affinity Fc R (Fc RI) is a multichain
receptor comprised of a 70-kDa glycoprotein -chain and homodimeric
largely intracellular -chain subunits(1, 2) .
Surface expression of the receptor on monocytes and U937 cells is
induced by IFN- (3, 4, 5) . Fc RI
binds with high affinity (K =
3 nM) human IgG subclasses IgG1 and IgG3. It has one log
less affinity for IgG4 and no apparent affinity for
IgG2(6, 7) . The receptor binds with high affinity
certain IgGs from other species that possess sequence homologies in the
C 2 Fc regions(6) . Fc RI engagement of complexed IgG
or anti-receptor antibodies(8, 9) aggregates the
receptor(10, 11) , triggering several very early
responses including fluxes of intracellular
Ca (12, 13) , tyrosine
phosphorylations of several proteins (14) and activation of the
NADPH oxidase pathway(12) . Myelocytic signaling functions
of Fc RI, Fc RIIIA, Fc RI, and Fc R (15) are
mediated by tyrosine activation of -chains (10, 11, 13) in motifs recognized by members
of src family(16) . Src kinases Hck and Lyn associate
with Fc RI -chains to constitute a preaggregation multichain
transduction unit(17) . Aggregation of Fc RI induces the
tyrosine phosphorylation of its -chains(15, 18) and the tyrosine phosphorylation of Syk
kinase(18, 19, 20) . Syk, a 72-kDa structural
homologue of the T-cell signal-transducing ZAP70, is recruited to
-chains on activated Fc RI(21) . We have previously
reported that an equimolar mixture of anti-Fc RI monoclonal
antibodies 32 and 22 is a potent combination for triggering
Fc RI-mediated NADPH oxidase activity(8, 9) .
Subsequently, we found that the F(ab`) fragments of these
anti-Fc RI trigger poorly, and that this situation is reversed by
binding high affinity IgG to the receptors(22) . For this
report, we analyzed the consequences of IgG-binding on signaling
through Fc RI. By engaging the ligand binding site on Fc RI
with monomer ligand and then cross-linking the receptors through
anti-Fc RI F(ab`) , we show evidence of a novel
dependence on IgG-Fc for Fc RI transductional activity. We show
that IgG-Fc binding increased aggregation-dependent
O production, tyrosine phosphorylation of
Fc RI -chains, and the association of pp72Syk kinase to
activated Fc RI.
MATERIALS AND METHODS
CellsU937 cells (10.6 subclone) were cultured
for 5-6 days in RPMI 1640 medium containing 10% fetal bovine
serum, gentomycin, and 20 units/ml IFN- (Genentech, S. San
Francisco, CA) (23) . Monocytes were obtained by cytopheresis
and purified on Ficoll-Hypaque as described(24) ; further
purification was through repeated rate-sedimentation separations from
lymphocytes until >90% purity was achieved. Monocytes were cultured
for 2 days in the RPMI medium containing 20 units/ml IFN- .
Antibodies and ImmunoglobulinsPurified
anti-Fc RI mAbs 32 and 22, 32 F(ab`) and 22
F(ab`) , and anti-Fc RII IV.3 Fab were obtained from Dr.
R. Graziano at Medarex, Annendale, NJ. Affinity-purified sheep
F(ab`) anti-murine IgG-F(ab`) and goat
F(ab`) anti-murine IgG were obtained from Organon Technika
(Durham, NC). Covalently cross-linked conjugates of 22Fab with high
unit multiplicity were prepared (25) and provided by Dr. M. W.
Fanger (Dartmouth Medical School, Lebanon, NH). Monomer Igs in this
paper include human IgG1 (hIgG1) purified from myeloma serum as
described previously(8) ; purified human serum IgG subclasses
IgG1, IgG2, IgG3, and IgG4 (The Binding Site LTD, (Birmingham, UK);
purified goat and rabbit IgG (Organon Technika); hybrid murine IgGs
(anti-human IgA1 (2a/2a), anti-human IgA/anti horseradish peroxidase
(2a/2b), and anti-horseradish peroxidase (1/2a) were prepared by Dr. J.
G. J. van de Winkel as described(26) ); and murine mAbs of the
IgG1, IgG2a, and IgG2b subclasses, murine mAbs of the IgM, IgA, and IgE
classes, and rat mAbs of the IgG1, IgG2a, and IgG2b subclasses. mAbs
used as ligands were obtained from hybridoma cultures. mAbs are
identified by their hybridoma ATCC designations or mAb names in the
order of the Fig. 2B list: m2a-1, -2, -3 are murine
IgG2a mAbs w632, HB63, HB32; m2a+2b represents murine IgG2a plus
mAb IgG2b BBMI; r2b-1,-2 are rat IgG2b mAbs TIB128, TIB129; m2b-1, -2,
-3 are murine IgG2b mAbs OKMI, IV.3, AML 2-23; m1-1, -2,
-3, -4, -5, -6, -7, -8 are murine IgG1 mAbs 3G8, P3, TIB191, TIB139,
HB205, HB195, HB202, HB203; r1-1 is rat IgG1 mAb TIB168; r2a-1 is
a rat IgG2a mAb TIB166; mM-1, -2, -3 are murine IgM mAbs PM81,
PMN6, TIB200; mA-1 is murine IgA mAb TIB196; mE-1 is murine IgE mAb
TIB141.
Figure 2:
Effects of Igs of various species, class,
and subclass on Fc RI signaling. Cells were preincubated in various
Igs and, after a brief wash, were analyzed for the percent occupancy of
Fc RI-ligand binding sites and for Fc RI-mediated
O production. A, preincubations
contained hIgG1 (H), goat IgG (G), rabbit IgG (R) (10 µg/ml), or medium (none). B,
preincubation media were hybridoma supernatants (200 µl) containing
murine (m) or rat (r) mAbs of IgG subclasses 2a (2a), 1 (1), and 2b (2b) or of the IgM (M), A (A), or E (E) classes. Identities of
mAbs are listed under ``Materials and Methods''; those with
reactivity toward monocyte surface antigens are indicated (*). Assays
of mAb-preincubated cells included control cells preincubated with
hIgG1 or medium. In B, the O
production by cells receiving 32F(ab`) and 22F(ab`) but no IgG was subtracted. Therefore, data in B represent the enhancement of O production due to ligand. (Negative values are slight depressions
in response by unligated receptors.)
Assay for O
ProductionO production was
measured by luminol-enhanced chemiluminescence(27) . Cells were
incubated at 22 °C for 20 min with primary antibody or ligands (10
µg/ml or in hybridoma supernatants) after which they were suspended
in O assay medium (15) at a
concentration of 2-5 10 /ml. Cells were
stimulated by mixing with an equal volume of 10 M luminol (5-amino-2,3-dihydrophthalazine-1,4-dione)
containing anti-Fc RI F(ab`) , 22Fab 22Fab
conjugates, or polyclonal anti-murine IgG antibodies as described
previously(15) . Luminometry on a BioOrbit luminometer 1250 was
used to measure the instantaneous velocity of O production at 25 °C; data was recorded in mV/s. Measurements
were recorded every 0.5-1.5 min until the respiratory burst
subsided to 10% of peak activity (about 10-15 min). For
relative totals of O in bursts, velocity
data were integrated and are expressed in mV. Chemiluminescence of
cells lacking stimulants was subtracted.
FACS Analysis of Cell-associated Antibodies and
LigandsCells were incubated at 22 °C for 20 min with
primary antibody or ligand (10 µg/ml or in hybridoma supernatants).
After washing 3 times with PBS containing 1 mg/ml BSA (PBS/BSA), cell
samples were incubated at 4 °C for 90 min with PBS/BSA containing
one of the following FITC-conjugated second antibodies or FITC-hIgG1:
FITC-conjugated F(ab`) goat anti-murine IgG (FITC-GAM)
(CALTAG, San Francisco, CA), FITC-conjugated hIgG1 (FITC-hIgG1)
prepared by conjugating purified hIgG1 with fluoroisothiocyanate, and
FITC-conjugated rat IgG2a mAb against the murine Ig -chain
(FITC-anti- ) (Zymed, San Francisco, CA). After washing 3 times,
the cells were fixed in 1% paraformaldehyde in PBS. Cell-associated
fluorescence was analyzed using a FACScan flow cytometer (Becton
Dickinson)(9) ; measurements of the mean fluorescence
intensity/cell were made on a 4-log scale. Data were converted to
FITC/cell using a standard curve generated from microspheres of known
FITC content (Flow Cytometry Research Corporation, San Juan, Puerto
Rico) as described previously(15) . For assays using
FITC-anti-IgG, data are expressed as FITC-second antibody binding
sites/cell. For assays using FITC-hIgG1, cells were exposed to various
unlabeled IgGs, and their binding was determined by the extent to which
subsequent FITC-hIgG1 (10 µg/ml) was blocked. Thus, 100% binding by
IgG means 100% blocking of FITC-hIgG1 after subtraction of
nonspecifically associated FITC-hIgG1 (which was measured in the
presence of 10 mg/ml hIgG). For assays using FITC-anti -chain mAb,
calculations for the number of -chains/cell were attained by
correcting FITC/cell values for the FITC/second antibody ratio.
Aggregation and InternalizationFc RI
aggregation was assessed by measuring cross-linking-induced receptor
internalization. Internalization was measured by fluorescent labeling
and FACS analysis. Cells were preincubated for 20 min at 22 °C in
O assay medium containing 32F(ab`) with or without ligand. Fc RI were cross-linked by adding 5
µg/ml FITC-22F(ab`) and incubating at 37 °C for 25
min. Time course assays established that induced internalization did
not continue after 25 min. The cells were washed with cold PBS/BSA and
incubated again at 4 °C for 60 min with PBS/BSA (100 µl) or
PBS/BSA containing 100 µg/ml rabbit anti-FITC antibodies (Molecular
Probes, Eugene, OR) to block fluorescence of cell surface FITC. To
assess the efficiency of this blocking, the procedure using
32F(ab`) and ligand binding was the same, but exposure to
FITC-22F(ab`) was done at 4 °C for 60 min. Subsequent
washes and incubation with and without anti-FITC Abs was the same as
before. After a final wash, fluorescence was measured by FACScan
analysis and data converted to FITC second antibody binding sites/cell.
Internalization was calculated according to the equation A - ((T - A)F) = I, where T = total sites/cell; A = total sites/cell of anti-FITC blocked cells; (T - A) = surface sites blocked by anti-FITC; F = correction factor for percent blocking efficiency
of anti-FITC, obtained from the percent unblocked/percent blocked
fluorescence of the 4 °C control procedure; I =
internalized sites/cell.
Immunoprecipitation, Immunoblotting, and
AutoradiographyCells (2 10 /ml) were
preincubated for 20 min at 22 °C in O assay medium with or without hIgG1 (10 µg/ml) and then for 5
min at 37 °C after the addition of an equal volume of luminol
containing 22F(ab`) and/or 32F(ab`) (5
µg/ml each). Cells were rapidly chilled and washed in cold PBS
containing 200 uM Na VO . They were
solubilized for 30 s in 4 °C lysis buffer containing 0.5% Nonidet
P-40 and 0.5% digitonin, and then for another 10 min after the addition
of an equal volume of lysis buffer containing 1% digitonin. Both
buffers contained aprotinin (0.23 units/ml), 2 mM phenylmethylsulfonyl fluoride, 5 mM Na EDTA,
10 mM NaF, and 200 uM Na VO in
PBS, pH 7.4. After a 10-min centrifugation at 16,000 g, supernatants were agitated with goat anti-mouse Ig
-chain (Pierce, Rockford IL) conjugated to Biosupport beads AB1
according to instructions of the manufacturer. Beads were washed 3
times with lysis buffer, and bead-precipitated proteins were separated
by nonreducing SDS-polyacrylamide gel electrophoresis on 16% acrylamide
gels followed by transfer to nitrocellulose membranes. Membranes were
blocked in 5% BSA, and immunoblotted with rabbit anti-phosphotyrosine
antibody (anti-PY) (a gift of Gustav Leinhard, Dartmouth Medical
School) or rabbit anti- -chain antibody (anti- ) (a gift of
J.-P. Kinet, NIH, NIAID). For immunoblotting with rabbit anti-Syk
antibody (UBI, Lake Placid, NY), membranes were blocked and blotted in
5% nonfat milk. Immunoblotting and membrane stripping were performed as
described previously(15) . Horseradish peroxidase-conjugated
anti-rabbit antibodies and the ECL system (Amersham, Arlington Heights,
IL) were used to develop blots. Positions corresponding to nonreduced
and phospho- dimers have been previously
established(15) .
RESULTS
IgG-Fc Binding Increases Fc RI-mediated
SignalingTo determine whether signaling through Fc RI is
altered by IgG-binding to its ligand binding site, we preincubated 10.6
cells with or without hIgG1 and reacted the cells with anti-Fc RI
mAb 32 and 22(9, 28) F(ab`) fragments.
These anti-Fc RI bind to nonoverlapping epitopes outside of the
Fc RI ligand binding site and trigger through receptor
cross-linking(8, 9) . The effect on cell signaling was
analyzed by measuring the production of
O . As shown in Fig. 1A,
cross-linking elicited a substantially greater respiratory burst from
cells that had been preincubated with hIgG1. A second assay performed
in the presence of Fc RII-blocking IV.3Fab gave similar results,
indicating that the effect IgG1 had on oxidase activity did not involve
Fc RII.
Figure 1:
The effect of hIgG on
Fc RI-mediated O production. A, U937 10.6 cells, differentiated by IFN- for 6 days,
were preincubated for 10 min in O assay
medium with and without hIgG1 (G1) (10 µg/ml). IV.3Fab (25
µg/ml) was added to block Fc RII. The preincubated cells were
added to luminol containing 32F(ab`) and 22F(ab`) (F) (5 µg/ml each), or 22F(ab`) alone as
control. O production was measured by
luminometry and is expressed in mV ± S.D. Controls were
subtracted. B, human monocytes cultured with IFN- for 2
days were preincubated with or without hIgG1 for 5 min and stimulated
by the addition of cells to luminol containing 5 µg/ml intact mAbs
32 and 22 (W) or 32F(ab`) and 22F(ab`) (F). C, U937 10.6 cells were preincubated with
human IgG subclasses 1, 2, 3, and 4, or no IgG (none) for 10 min and then washed.
Duplicate cell samples were stimulated by the addition of
32F(ab`) and 22F(ab`) for an
O assay. Another set of duplicates were
incubated with FITC-hIgG1 and analyzed by FACS to measure hIgG subclass
binding. O production and IgG-binding to
Fc R are expressed as a percent of the hIgG1 control. There was no
detectable binding by hIgG2. Data represent the mean of three
experiments.
A similar assay was performed with monocytes that were
treated with IFN- to induce the surface expression of newly
synthesized Fc RI. Treatment was done in culture medium lacking IgG
so that induced receptors would be occupied by high affinity ligands.
When anti-Fc RI F(ab`) triggered receptors in the
absence of IgG, there was almost no activity (Fig. 1B),
but monocytes that had been preincubated with IgG produced a
respiratory burst that was comparable with that triggered by intact
anti-Fc RI. We conclude that IgG binding profoundly affected
Fc RI signaling in normal human receptors. To determine whether
other human subclasses exert a similar effect, we preincubated 10.6
cells with hIgG subclasses 1, 2, 3, and 4; washed the cells briefly;
and then cross-linked their Fc RI. As shown in Fig. 1C, hIgG1- and hIgG3-preincubated cells had
enhanced Fc RI signaling, whereas hIgG4 preincubation produced an
intermediate response. FACS analysis showed that bound IgG-binding to
the cells (Fig. 1C) correlated with enhanced
O production (Fig. 1C).
HIgG2 did not bind and had no effect on signaling over the IgG-free
control. Certain other nonhuman IgGs bind human Fc RI. To
analyze their effect specifically on Fc RI signaling, we tested
cells after exposure to several murine and rat mAbs of various Ig
classes and subclasses as well as rabbit and goat polyclonal IgGs. As
shown in Fig. 2A, rabbit IgG, which binds with high
affinity to Fc RI, supported signaling to the same extent as did
hIgG1, whereas goat IgG neither bound nor affected response above the
control. As shown in Fig. 2B, murine IgG2a and rat
IgG2b mAbs are the subclasses that bind to Fc RI, whereas murine
mAbs of the IgA, IgE, and IgM classes, murine IgG1 and IgG2b mAbs, and
rat IgG2a mAbs bind poorly or not at all. As shown, only the mAbs that
bound enhanced O production through
Fc RI. Some mAbs had antigen specificities for non-Fc RI
monocyte epitopes, but these had no apparent affect on
O production attributable to antigen (Fig. 2A, asterisks). Some exceptions were
mIgG1 mAbs that recognize leukocyte function-associated antigens and an
IgE mAb. These mAbs reduced Fc RI-binding sites without affecting
O production. Overall, the results
suggest that ligand occupancy by high affinity IgG enhanced
Fc RI-signaling for the oxidative burst.
Ligand-bound StructureHybrid IgGs contain only
one IgG heavy chain capable of ligating Fc RI(26) . To
determine whether enhanced O production
requires both of the heavy chains in normal high affinity IgGs, hybrids
with one or two IgG2a heavy chains were tested. Binding by the 2a/2a
hybrid resulted in 100% of the enhancement of the hIgG1 control (Fig. 3). The 2a/1 and 2a/2b hybrids were less effective,
perhaps due to an insufficient time for saturation binding, but were
nevertheless capable of enhancing Fc RI signaling. As the hybrid
low affinity chain can interact with Fc RII(29) , very
short preincubation times were used, and duplicate assays were
conducted in the presence of IV.3 Fab. As shown, IV.3 blocking of
Fc RII did not interfer with hybrid-enhanced Fc RI signaling (Fig. 3), indicating that Fc RII had not been involved.
Figure 3:
Hybrid Abs enhance Fc RI-mediated
signaling. Cells were preincubated for 10 min with hIgG1 or hybrid
murine mAbs comprising two IgG2a chains (2a/2a), one IgG2a and
one IgG2b chain (2a/2b), or one IgG2a and one IgG1 chain (2a/1) (7 µg/ml each). Anti-Fc RI F(ab`) were added as in Fig. 2, and Fc RI-mediated
O production was measured. The
preincubations and assay were repeated in the presence of 25 µg/ml
IV.3Fab. Data are expressed as a percent of the hIgG1
control.
The effect of hybrids suggested a ligated structure involving one
receptor/IgG as opposed to a trimer composed of two receptors each
binding to one heavy chain on IgG. For a more direct approach in
establishing this ratio, we measured cell-associated 32, 32Fab, and
32F(ab`) , to establish the number of Fc RI on cells; we
measured cell-associated IgG2a, to establish the number of ligand
binding sites/cell. The results of a FACS analysis (Table 1) show
that IgG-binding sites equal the number of Fc RI, indicating that
the ligated structure must have a ratio of one Fc RI/bound IgG.
IgG-binding Does Not Prime Cells or Increase Fc RI
ExpressionThe results above suggested that high affinity IgG
increased the transmission of an excitatory signal. A plausible
alternative was that IgG binding induced cell priming, a phenomenon
whereby exposure to one reagent increases the oxidase response to a
subsequent unrelated stimulant(30, 31) . To test this,
we triggered hIgG1-preincubated cells through the receptor for fMLP
with the result that fMLP-signaling was found unchanged by Fc RI
ligation (Fig. 4).
Figure 4:
Effect of IgG on the fMLP-triggered
respiratory burst. Cells were preincubated with medium (med),
hIgG1, 32F(ab`) , or hIgG1 plus 32F(ab`) . They
received 22F(ab`) to trigger Fc RI or fMLP (5
10 M) to stimulate chemotactic receptors.
O production is expressed in
mV.
Measurements by FACS analysis of the
number of surface Fc RI following a typical preincubation of cells
with IgG revealed no significant increase in Fc RI numbers (Table 2). Thus, neither cell-priming nor increased numbers of
surface Fc RI account for the increase in signaling.
Enhanced Fc RI Signaling Is Not Due to a Specific
Mode of Anti-Fc RI BindingAs signal-enhancement may have
been due to a specific mode of anti-Fc RI binding, we questioned
whether these Abs engaged ligated differently than unligated receptors
or could desensitize the receptors. First, we examined the possibility
that anti-Fc RI might be capable of a ligand-blockable receptor
desensitization. For this, O production
was measured after binding one anti-Fc RI followed by ligand, or
the reverse order, and cross-linking by the other anti-Fc RI. As
shown in Table 3, either order of binding elicited similar
oxidative responses. That anti-Fc RI did not form more monovalent
complexes with unligated Fc RI is also implied by this result.
Furthermore, we speculated that ligation might involve a rearrangement
involving rebinding of the first anti-Fc RI. However, the addition
of 32Fab to block rebinding by 32F(ab`) did not block
ligand-increased signaling (Table 3), suggesting that
ligand-induced productive rebinding did not occur.
Second, the
possibility that ligation might increase signaling through an increased
rate of anti-Fc RI binding was tested. Rates of binding depend on
concentrations of the anti-Fc RI
antibodies(8, 9) . Since there is a linear
relationship between the rates of binding of cross-linking antibody and
O generation(8, 9) ,
times of peak rates of response correspond to peak rates of binding.
Therefore, peak times could be used as an index for the rate of
binding. As shown in Table 4, peak times were changed by antibody
concentration but not by IgG-ligation.
Third, Fc RI were
aggregated through anti-Fc RI F(ab`) plus sheep or goat
anti-murine antibodies, or through 22Fab covalently conjugated
multimers that cross-link Fc RI and would have a different gross
physical structure compared with 22F(ab`) . In each case,
ligand-increased signaling was found (Table 5). These data show
that enhancement was not due to a unique mode of bridging by
22F(ab`) . Collectively, the results indicate that
enhancement was not due to a specific mode of anti-Fc RI binding.
Cross-linking of Unligated Fc RI Induces Receptor
Aggregation and InternalizationThe possibility that suboptimal
signaling occurred because unligated Fc RI could not be aggregated
was assessed by measuring aggregation through induced internalization.
Cells were exposed to the usual reagents, but cross-linking was
performed using FITC-22F(ab`) . FACS analysis for
internalized fluorescence demonstrated induced uptake of unligated
Fc RI as well as of ligand-occupied receptors (Table 6). The
difference in uptake ( 0.15-fold) was minor compared with the
effect of IgG on Fc RI signaling (usually 3-5-fold),
suggesting an indirect influence on uptake, possibly through
signal-dependent changes in cytoplasmic milieu.
IgG Predisposes Fc RI for Aggregation-dependent
SignalingFc RI signaling for O production is transient (9) . Following cross-linking,
receptors enter a surface accessible postsignaling stage(8) .
Unligated Fc RI, being aggregable, are also likely to enter a
postsignaling stage following cross-linking. If IgG-binding predisposes
Fc RI to signal, it should have no effect on postcross-linked
unligated receptors. Results consistant with this are shown in Fig. 5. In the control assay (Fig. 5A),
cross-linking antibody was added at various times to discrete samples
of cells to demonstrate their capacity to respond. In a second assay (Fig. 5B), the cells were exposed to IgG before or
after the zero time addition of cross-linking anti-Fc RI. As shown
in Fig. 5B, IgG given to cells 3 min after Fc RI
were cross-linked produced a small reaction, but IgG given at later
times gave responses identical those receiving no IgG. This implies
that aggregated unligated receptors reached a postsignaling stage
beyond that affected by IgG-binding. It confirms that
aggregation-dependent signaling requires preoccupancy of the receptors.
Figure 5:
IgG-binding prior to
Fc R-cross-linking produces the critical change in transductional
activity. A, cells were preincubated with hIgG1 and
22F(ab`) , and individual samples of these cells were given
32F(ab`) at 0, 3, 6, 9, or 12 min to stimulate
O production. One sample received no
32F(ab`) (none). B, cells were
preincubated in 22F(ab`) for 10 min, and stimulated by the
addition of 32F(ab`) . HIgG1 was added to individual samples
of these cells at 0, 3, 6, 9, or 12 min, or no hIgG1 was added (none). Results are presented as rate of
O production, measured in mV/s and
expressed as a percent of the optimum.
Dose Response of Fc RI/IgG in AggregatesWe
examined the question of whether each (Fc RI/IgG) is capable of
delivering a unitary excitation. Cells were incubated with saturating
and subsaturating concentrations of IgG. The amount of ligand-bound to
cells and their capacity to trigger respiratory bursts through mixtures
of unligated and ligated-Fc RI were measured. Data in Fig. 6on the dose dependence indicated that 2 µg of hIgG1/ml
was both saturating for Fc RI (Fig. 6B) and optimal
for the Fc RI-mediated respiratory burst (Fig. 6A).
At the lower concentrations where ligated Fc RI were diluted in
aggregates by unligated receptors, O production was not proportional to ligand-occupied receptors (Fig. 6C). Rather, production decreased exponentially
with increasing numbers of unligated receptors. This indicated that the
presence of unligated Fc RI depressed responses to aggregated
ligated receptors, and suggested that there was no trans-activation of
unligated Fc RI by ligated receptors, at least for
O signaling.
Figure 6:
Effect of IgG concentration on
aggregation-dependent Fc RI signaling. In sequential experiments,
cells were preincubated in O assay medium
containing 0, 0.06, 0.125, 0.25, 0.5, 1, 2, and 4 µg/ml hIgG1 (opencircles) or 0.7, 7, 23, 70, 230, and 7000
µg/ml hIgG1 (filledcircles). Samples in each
were assayed for Fc R-mediated respiratory bursts (A) and
hIgG1-binding (B). B was measured by FACS analysis
following an incubation with FITC-conjugated anti-hIgG antibody; data
is expressed as the mean fluorescence intensity of cells. C,
data were calculated from results in A and B to
determine O production as a function of
hIgG1-binding. On a log plot, data falls on a straight line indicating
that O production/ligated receptor
decreases by a simple exponential function (not shown). The saturation (SAT'N) point for IgG1-binding is indicated by an arrow. Data represent the mean ± the half-range of
seven assays.
Phosphorylation of Fc RI -Chains and Association
of SykSince aggregation controls transmembrane signaling by
clustering Fc RI -chains inducing their tyrosine
phosphorylation (15, 18) , we examined the effect of
IgG on this activity. Following cross-linking, the receptors were
precipitated, and their co-precipitating -chains were analyzed by
anti-phosphotyrosine and anti- -chain immunoblot. As shown in Fig. 7, increased numbers of phosphorylated -chains were
identified in aggregates of Fc RI/IgG compared with the unligated
clusters. Receptor/subunit recovery was assessed through anti-
immunoblot, which demonstrated that ligated receptors, aggregated or
not, precipitated less efficiently. Experiments with similar recoveries
revealed even larger differences in phospho- (not shown). As
shown, phospho- was not detected in ligated, unaggregated
Fc RI. Interestingly, a 72-kDa tyrosine phosphorylated protein (Fig. 7A) also associating more extensively with
Fc RI/IgG aggregates was found through immunoblotting (Fig. 7B) to contain pp72Syk. ( )These
results show that IgG directly affected the receptor by increasing
phosphorylation of -chains and the association of tyrosine
phosphorylated pp72Syk.
Figure 7:
Effect of IgG on -chain
phosphorylation and the co-precipitation of pp72Syk. A, cells
preincubated with or without hIgG1 were incubated in
O assay medium containing 22F(ab`) or 22F(ab`) and 32F(ab`) . After 5 min at
37 °C, the cells were centrifuged through cold PBS/100 uM Na VO (1 ml) and solubilized in digitonin
lysis buffer. The receptors were precipitated by GAMk-AB1 beads,
separated by nonreducing SDS-polyacrylamide gel electrophoresis,
transferred, and analyzed by immunoblot using anti-phosphotyrosine
(anti-PY). The membrane was stripped and reblotted with
anti- -chain (anti- ) antibody. Blots were developed using
horseradish peroxidase-conjugated second antibody and ECL.
Phospho- 2 is indicated by brackets. The anti- -reactive
18-22-kDa proteins beneath phospho- are unphosphorylated
-chain dimers. Arrow indicates pp72. B, anti-PY
and anti-Syk immunoblots of 72-kDa proteins that co-precipitated with
cross-linked Fc RI.
DISCUSSION
In this report, we demonstrate that signaling resulting in
activation of the NADPH oxidase pathway and involving the tyrosine
phosphorylation of -chains on Fc RI is markedly increased by
IgG-Fc interaction with the receptor ligand binding site. To show the
effect on Fc RI-mediated NADPH oxidase activity, we triggered
through Fc RI on monocytic cells exposed to a variety of IgGs and
found that all IgGs that bound to IFN- -treated U937-10.6
monocytic cells, including rabbit, human subclasses 1 and 3, murine
subclass 2a, and rat subclass 2b, increased the Fc RI-mediated
production of O . Low affinity IgGs and
IgA, IgM, and IgE had no effect. Binding to some leukocyte integrins
caused slight Fc RI blocking but did not increase Fc RI
signaling. Significantly, binding by IgG had a direct effect on the
receptor in that it increased the aggregation-dependent tyrosine
phosphorylation of Fc RI -chains(15, 18) .
This was accompanied by the binding to activated Fc RI of a 72-kDa
tyrosine phosphoprotein, identified as pp72Syk. This result is
consistent with recent studies showing that Syk binds to tyrosine
phosphorylated -chains in cells stimulated through
Fc RI(18) , to -chains in activated Fc RI
complexes(21, 32) , and to homologous
tyrosine-phosphorylated motifs in activated B-cell receptor
complexes(33) . We propose that IgG-binding creates a change in
availability of Fc RI -chains that permits
aggregation-dependent tyrosine phosphorylation of -chain motifs
and recruitment of pp72Syk. Another possibility is that ligand binding
influences the affinity of the -subunits for associated Src
kinases (17) , perhaps modifying the kinetics of reactions in
aggregates. This study also shows a correlation between
-phosphorylation, pp72Syk association, and
O production, suggesting a link between
Fc RI-associated tyrosine kinases and the activation of the NADPH
oxidase pathway. Experiments to elucide possible effects of IgG
other than on signaling through Fc RI demonstrated that enhancement
was not due to global cell-priming, because triggering through fMLP was
unaffected. It was not due to an increase in the numbers of surface
Fc RI nor to specific modes of cross-linking by anti-Fc RI
antibodies. It was also not due to a lack of aggregability of unligated
Fc RI. These negative results help to support the proposed
hypothesis that aggregation without ligand is transductionally
insufficient. The IFN- -induced receptors on U937-10.6
cells appeared somewhat less IgG-Fc-dependent than those on monocytes.
However, induced receptors on premonocytic U937 cells were not tested
after a comparable expression time because several days of additional
culture are required to develop an oxidative capacity in the tumor
cells(23) . Furthermore, greater numbers of Fc RI are
expressed through induction on the U937-10.6 subclone. Thus, age
or number (or some other factor) could have converted a minority of
receptors to what may be an easily achieved signal-efficient
configuration. Regardless of the reason, characterizing newly expressed
normal monocyte Fc RI as IgG-Fc-dependent strengthens our
conclusions on the importance of Fc to Fc RI signaling. The
nature of the improved signaling is not yet understood. However,
several observations were made that partially characterize the
ligand/receptor relationship. (i) IgG that was bound to cells was not
itself cross-linked, making it unlikely that a mechanical strain
altered the receptor and more likely that the interaction with IgG-Fc perse brought about relevant changes. (ii)
Concentrations of IgG optimizing transductional activity were identical
to those saturating Fc RI, supporting the biochemical linkage and
the specificity of the reaction. (iii) This interaction in
unaggregated, signal-quiescent Fc RI did not increase
phosphorylation of . Therefore, potentiation was not accompanied
by an obvious increase in activity of protein-tyrosine kinases
associating with -chains(17) . However, binding could have
increased protein-tyrosine kinase affinity, a point presently under
investigation. (iv) When ligand was bound after receptor cross-linking,
there was no return to oxidase activation. This outcome is consistant
with the view that, with or without ligand, aggregation provides only a
transient opportunity (interval/receptor) for transductional activity (8, 9) and that the ligand is effecting a change
within this window. (v) At various subsaturating IgG concentrations,
anti-Fc RI cross-linking must have produced mixed aggregates of
ligated and unligated Fc RI, and in these cases,
O production decreased exponentially with
the decrease in ligand-occupied receptors (Fig. 6C).
Evidently, the unligated receptors were not only refractory as triggers
but seemed obstructive to efficient signaling. The surprising
implication is that the alteration in ligated Fc RI seems not to
have been transmitted to proximate unligated Fc RI. If the
alteration were a matter of affinity or availability of Src kinases on
-chains, it is possible that the kinase concentration was diluted
below a critical level for a trans-activation of the unligated
Fc RI. In any case, this evidence indicates that ligated and
unligated Fc RI are markedly different in ways that are critical
and specific to transductional activity. A further characterization
using hybrid antibodies indicated that signal-enhancement was achieved
through the binding of IgG containing only one high affinity heavy
chain capable of binding Fc RI. This evidence alone would not have
excluded the possibility of hybrid low affinity chain engagement of a
second Fc RI or an Fc RII. However, we made direct measurements
that show that the potentiated Fc RI structure has a ratio of one
receptor/hIgG1 ligand. This indicated that the other heavy chain on
hIgG1 does not engage with high affinity the ligand-binding site on a
second Fc RI. Additionally, measured with IV.3Fab as a block, the
second heavy chain does not occupy the ligand binding site on
Fc RII. Therefore, as predicted by the hybrids, only one of the two
heavy chains of IgG binds in a manner sufficient for potentiation.
Unlike the ligand-induced dimerization characteristic of several
cytokine receptor family members(34) , monomer ligand does not
dimerize Fc RI. The results of this study do not alter the
general view that aggregation, not a ligand-induced physical change in
FcR, is the key event in transmembrane signaling(35) . But in
addition to that, IgG-Fc binding must alter Fc RI such that those
events normally involved in aggregation-dependent signal transduction
can occur. It is our speculation on this point that Fc binding
stabilizes an activatable configuration of the receptor, one that may
be present to varying extents in the nonphysiologic conditions of
ligand-free cell culture. Fc region involvement other than in passive
mediation of myelocytic antigen recognition has been suggested by Brown
and Koshland(36, 37) , and some distinctions in
FcR-mediated cytotoxicity due to hIgG-Fc in subclasses 1 and 3 have
also been described(38, 39) . Physicochemical studies
have shown that coupling of IgG to Fc RI and IgE to Fc RI
induces conformational alterations in the ligands (40) and
possibly in Fc RI during an affinity shift(41) . Physical
distances between FcR in aggregates (42) and interactions with
cytoskeletal proteins may also be important(43) . Whether
ligation changes the configuration of Fc RI and how binding is
communicated to cytoplasmic components will require further study. In summary, the results in this report provide evidence that
Fc RI is altered by IgG-Fc interaction with the Fc RI
ligand-binding site and that this alteration permits efficient
signaling of the oxidative inflammatory response. Phosphorylation of
associated -chains and association by pp72Syk kinase were also
increased by ligation of the receptors.
FOOTNOTES
- *
- This work
was supported by Grant AI29455 from the National Institutes of Health
(to L. C. P.) and by a grant from the Hitchcock Foundation, Hanover,
NH. Flow cytometry was supported in part by the Core Grant of the
Norris Cotton Cancer Center (CA23108). The costs of publication of this
article were defrayed in part by the payment of page charges. This
article must therefore by hereby marked
``advertisement'' in accordance with 18 U.S.C.
Section 1734 solely to indicate this fact.
- §
- To whom correspondence should be addressed.
- (
) - The abbreviations used are: Fc
R, Fc
receptors; Fc RI, high affinity Type I Fc receptor; Fc RI,
high affinity Fc receptor; IgG, immunoglobulin G;
O , superoxide; IFN- , interferon
; FACS, fluorescence-activated cell sorting; PBS,
phosphate-buffered saline; BSA, bovine serum albumin; FITC, fluorescein
isothiocyanate.
- (
) - L. C. Pfefferkorn and S. L.
Swink, manuscript in preparation.
ACKNOWLEDGEMENTS
We thank Dr. Clark L. Anderson, Ohio State University
College of Medicine for many stimulating discussions.
REFERENCES
- Ernst, L. K., Duchemin, A.-M., and Anderson, C. L. (1993) Proc. Natl. Acad. Sci. U. S. A. 90, 6023-6027
[Abstract/Free Full Text]
- Wang, A. V. T., Scholl, P. R., and Geha, R. S. (1994) J. Exp. Med. 180, 1165-1170
[Abstract/Free Full Text]
- Guyre, P. M., Morganelli, P. M., and Miller, R. (1983) J. Clin. Invest. 72, 393-397
- Allen, J. M., and Seed, B. (1989) Science 243, 378-381
[Abstract/Free Full Text]
- Pan, L., Mendel, D. B., Zurlo, J., and Guyre, P. M. (1990) J. Immunol. 145, 267-275
[Abstract]
- Woof, J. M., Partridge, L. J., Jeffris, R., and Burton, D. R. (1986) Mol. Immunol. 23, 319-330
[CrossRef][Medline]
[Order article via Infotrieve]
- Anderson, C. L., and Looney, R. J. (1986) Immunol. Today 7, 264-266
[CrossRef]
- Pfefferkorn, L. C., and Fanger, M. W. (1989) J. Immunol. 143, 2640-2649
[Abstract]
- Pfefferkorn, L. C., and Fanger, M. W. (1989) J. Biol. Chem. 264, 14112-14120
[Abstract/Free Full Text]
- Ravetch, J. V., and Kinet, J.-P. (1991) Annu. Rev. Immunol. 9, 457-492
[Medline]
[Order article via Infotrieve]
- Ravetch, J. V. (1994) Cell 78, 553-560
[CrossRef][Medline]
[Order article via Infotrieve]
- van de Winkel, J. G. J., and Anderson, C. L. (1991) J. Leukocyte Biol. 49, 511-524
[Medline]
[Order article via Infotrieve]
- van de Winkel, J. G. J., and Capel, P. J. A. (1993) Immunol. Today 14, 215-221
[CrossRef][Medline]
[Order article via Infotrieve]
- Scholl, P. R., Ahern, D., and Geha, R. S. (1992) J. Immunol. 149, 1751-1757
[Abstract]
- Pfefferkorn, L. C., and Yeaman, G. R. (1994) J. Immunol. 153, 3228-3236
[Abstract]
- Songyang, Z., Shoelson, S. E., Chaudhuri, M., Gish, G., Pawson, T., Haser, W. G., King, F., Roberts, T., Ratnofsky, S., Lechleider, R. J., Neel, B. G., Birge, R. B., Fajardo, J. E., Chou, M. M., Hanafusa, H., Schaffhausen, B., and Cantley, L. C. (1993) Cell 72, 767-778
[CrossRef][Medline]
[Order article via Infotrieve]
- Wang, A. V. T., Scholl, P. R., and Geha, R. S. (1994) J. Exp. Med. 180, 1165-1170
- Duchemin, A.-M., Ernst, L. K., and Anderson, C. L. (1994) J. Biol. Chem. 269, 12111-12117
[Abstract/Free Full Text]
- Kiener, P. A., Rankin, B. M., Burkhardt, A. L., Schieven, G. L., Gilliland, L. K., Rowley, R. B., Bolen, J. B., and Ledbetter, J. A. (1993) J. Biol. Chem. 268, 24442-24448
[Abstract/Free Full Text]
- Agarwal, A., Salem, P., and Robbins, K. C. (1993) J. Biol. Chem. 268, 15900-15905
[Abstract/Free Full Text]
- Jouvin, M.-H. E., Adamczewski, M., Numerof, R., Letourneur, O., Valle, A., and Kinet, J.-P. (1994) J. Biol. Chem. 269, 5918-5925
[Abstract/Free Full Text]
- Pfefferkorn, L. C. (1994) FASEB J. 8, A749
- Pfefferkorn, L. C., Guyre, P. M., and Fanger, M. W. (1990) Mol. Immunol. 27, 263-272
[CrossRef][Medline]
[Order article via Infotrieve]
- Shen, L., Lasser, R., and Fanger, M. W. (1989) J. Immunol. 143, 4117-4122
[Abstract]
- Shen, L., Guyre, P. M., Anderson, C. L., and Fanger, M. W. (1986) J. Immunol. 137, 3378-3382
[Abstract]
- Koolwijk, P., Spierenburg, G. T., Frasa, H., Boot, J. H. A., van de Winkel, J. G. J., and Bast, E. J. E. G. (1989) J. Immunol. 143, 1656-1662
[Abstract]
- Hallett, M. B., and Campbell, A. K. (1983) Biochem. J. 216, 459-465
[Medline]
[Order article via Infotrieve]
- Anderson, C. L., Guyre, P. M., Whitin, J. C., Ryan, D. H., Looney, R. J., and Fanger, M. W. (1986) J. Biol. Chem. 261, 12856-12864
[Abstract/Free Full Text]
- Koolwijk, P., van de Winkel, J. G. J., Pfefferkorn, L. C., Jacobs, C. W. M., Otten, I., Spierenburg, G. T., and Bast, B. J. E. G. (1991) J. Immunol. 147, 595-602
[Abstract]
- Bender, J. G., McPhail, L. C., and Van Epps, D. E. (1983) J. Immunol. 130, 2316-2323
[Abstract]
- Bass, D. A., McPhail, L. C., Schmitt, J. D., Morris-Natschke, S., McCall, C. E., and Wykle, R. L. (1989) J. Biol. Chem. 264, 19610-19617
- Benhamou, M., Ryba, N. J., Kihara, H., Rishikata, H., and Siraganian, R. P. (1993) J. Biol. Chem. 268, 23318-23324
[Abstract/Free Full Text]
- Law, D. A., Sidorenko, P., Chandran, A., Draves, K. E., Chan, A. C., Weiss, A., Edelhoff, S., Disteche, M., and Clark, E. A. (1994) J. Biol. Chem. 269, 12310-12319
[Abstract/Free Full Text]
- Longmore, G. D., Watowich, S. S., Hilton, D. J., and Lodish, H. F. (1993) J. Cell Biol. 123, 1305-1308
[Free Full Text]
- Metzger, H. (1992) J. Immunol. 149, 1477-1487
[Medline]
[Order article via Infotrieve]
- Brown, J. C., and Koshland, M. E. (1977) Proc. Natl. Acad. Sci. U. S. A. 74, 5682-5686
[Abstract/Free Full Text]
- Brown, J. C., and Koshland, D. E. (1975) Proc. Natl. Acad. Sci. U. S. A. 72, 5111-5115
[Abstract/Free Full Text]
- Rozsnyay, Z., Sarmay, G., Walker, M., Maslanka, K., Valasek, Z., Jefferis, R., and Gergely, J. (1989) Immunology 66, 491-498
[Medline]
[Order article via Infotrieve]
- Gergely, J., and Sarmay, G. (1988) Immunol. Lett. 20, 1-4
- Zheng, Y., Shopes, B., Holowka, D., and Baird, B. (1992) Biochemistry 31, 7446-7456
[CrossRef][Medline]
[Order article via Infotrieve]
- Ortega, E., Schweitzer-Stenner, R., and Pecht, I. (1991) Biochemistry 30, 3473-3483
[CrossRef][Medline]
[Order article via Infotrieve]
- Kane, P. M., Holowka, D., and Baird, B. (1988) J. Cell Biol. 107, 969-980
[Abstract/Free Full Text]
- Ohta, Y., Stossel, T. P., and Hartwig, J. H. (1991) Cell 67, 275-282
[CrossRef][Medline]
[Order article via Infotrieve]
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.

CiteULike Complore Connotea Del.icio.us Digg Reddit Technorati What's this?
This article has been cited by other articles:

|
 |

|
 |
 
C. M. Sayes, R. Wahi, P. A. Kurian, Y. Liu, J. L. West, K. D. Ausman, D. B. Warheit, and V. L. Colvin
Correlating Nanoscale Titania Structure with Toxicity: A Cytotoxicity and Inflammatory Response Study with Human Dermal Fibroblasts and Human Lung Epithelial Cells
Toxicol. Sci.,
July 1, 2006;
92(1):
174 - 185.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
D. Pilling, N. M. Tucker, and R. H. Gomer
Aggregated IgG inhibits the differentiation of human fibrocytes
J. Leukoc. Biol.,
June 1, 2006;
79(6):
1242 - 1251.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
J. C. Edberg, H. Qin, A. W. Gibson, A. M. F. Yee, P. B. Redecha, Z. K. Indik, A. D. Schreiber, and R. P. Kimberly
The CY Domain of the Fcgamma RIa alpha -Chain (CD64) Alters gamma -Chain Tyrosine-based Signaling and Phagocytosis
J. Biol. Chem.,
October 18, 2002;
277(43):
41287 - 41293.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
C. A. Guyre, T. Keler, S. L. Swink, L. A. Vitale, R. F. Graziano, and M. W. Fanger
Receptor Modulation by Fc{gamma}RI-Specific Fusion Proteins Is Dependent on Receptor Number and Modified by IgG
J. Immunol.,
December 1, 2001;
167(11):
6303 - 6311.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
J. C. Edberg, A. M. F. Yee, D. S. Rakshit, D. J. Chang, J. A. Gokhale, Z. K. Indik, A. D. Schreiber, and R. P. Kimberly
The Cytoplasmic Domain of Human Fcgamma RIa Alters the Functional Properties of the Fcgamma RI{middle dot}gamma -Chain Receptor Complex
J. Biol. Chem.,
October 15, 1999;
274(42):
30328 - 30333.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
D. J. Hackam, O. D. Rotstein, A. Schreiber, W.-j. Zhang, and S. Grinstein
Rho is Required for the Initiation of Calcium Signaling and Phagocytosis by Fcgamma Receptors in Macrophages
J. Exp. Med.,
September 15, 1997;
186(6):
955 - 966.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
R. K. Park, Y. Liu, and D. L. Durden
A Role for Shc, Grb2, and Raf-1 in Fcgamma RI Signal Relay
J. Biol. Chem.,
June 7, 1996;
271(23):
13342 - 13348.
[Abstract]
[Full Text]
[PDF]
|
 |
|
Copyright © 1995 by the American Society for Biochemistry and Molecular Biology.
|
Advertisement
Advertisement
|