| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
RIIb21
Departments of
*
Internal Medicine and
Physiology and Cell Biology, Ohio State University, Columbus, OH 43210;
Institut Curie, Paris, France; and
University of Bielefeld, Bielefeld, Germany
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
receptor (FcR) expression by immunohistology and
immunoblotting. Several pan-FcRII Abs that label the placental
endothelium displayed a distribution pattern that correlated well with
transport functions, being intense in the terminal villus and nil in
the cord. In contrast, the MHC class 1-like IgG transporter, FcRn, and
the classical FcRIIa were not expressed in transport-related
endothelium of the placenta. Our inference, that FcRIIb was the
likely receptor, we confirmed by analyzing purified placental villi,
enriched in endothelium, by immunoblotting with a new Ab specific for
the cytoplasmic tail of FcRIIb. These experiments showed that the
FcRII expressed in villus endothelium was the b2 isoform whose
cytoplasmic tail is known to include a phosphotyrosyl-based motif that
inhibits a variety of immune responses. We suggest that this receptor
is perfectly positioned to transport IgG although as well it may
scavenge immune complexes. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The villus EC layer, transporting materials arriving from the
syncytiotrophoblast, is marked by continuous tight junctions, highly
attenuated cell bodies, and abundant intracellular transport structures
called caveolae (8). These caveolae have a characteristic
flask shape when associated with the cell membrane and a consistent
size ranging from 
50 to 100 nm (9). Caveolae appear to
function in cellular transport pathways in a manner similar to, but
distinct from, clathrin-coated vesicles (10) and likely
form the structural basis for IgG transit across the fetal villus EC.
Early studies using aggregated IgG and IgG-coated RBC
(11, 12) showed that the villus EC expressed receptors
specific for IgG (FcR). With the later development of Abs against
the classical FcR, one of the three groups of FcR, namely
FcRII, was shown to be present (13, 14, 15). This finding
is remarkable because as a rule (with one exception showing FcRII
expression in skin; Ref. 16) endothelium in the body does
not express FcR. Whether FcRII is associated with caveolae has
not been established.
FcRII (CD32) is the most abundant and widely distributed group
within the FcR family, which includes FcRI (CD64) and FcRIII
(CD16). In humans, the FcRII group consists of at least six
different proteins encoded by three distinct genes (A, B, and C)
(17, 18). Of these several proteins there are two that are
likely candidates for expression on the villus EC, namely, FcRIIa
and FcRIIb, major products of genes A and B, respectively. As a rule
these two receptors when expressed elsewhere mediate opposing signals.
FcRIIa initiates such functions as endocytosis, inflammatory
mediator release, gene transcription, and others. FcRIIb displays an
essentially antagonistic character by transducing inhibitory signals
that down-regulate several immune functions (19, 20, 21, 22, 23). The
molecular basis for this dichotomy resides in different
phosphotyrosyl-based sequence motifs in the cytoplasmic tails of these
otherwise very similar receptors. These different motifs serve as
docking sites for different sets of SH2 domain-containing enzymes,
which in turn distinguish the receptor responses (24).
In the mouse there is only a single FcRII gene, and it resembles the
human B gene. As a rule, its products mediate inhibitory responses, as
in the human. However, distinct from the situation with the human
receptors, the capacity of its products to mediate endocytosis has been
studied in detail, but only by means of transfection experiments. Only
one of the two products of the murine FcRII gene, namely,
FcRIIb2, is capable of mediating rapid IgG endocytosis and
transcytosis by means of clathrin-coated vesicles (25).
The other product, FcRIIb1, expresses a 47-aa insert in its
cytoplasmic tail that appears to inhibit receptor inclusion in coated
pits and thus to inhibit endocytic capacity (26, 27). A
comparable insert (19 aa) appears in the human homolog, FcRIIb1,
although similar studies of endocytic capacity have not been performed
in humans. There is no FcRIIa in the mouse.
In the studies reported here we have used a variety of approaches to
determine that FcRIIb2 is the predominant receptor expressed by the
terminal villus EC. These results suggest that villus FcRIIb2 may be
functioning as an IgG transporter as well as perhaps an immune complex
scavenging receptor.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following anti-FcR mAbs were used for immunolabeling;
KB61 (IgG1; D. Mason, Radcliffe Hospital, Oxford, U.K.), KU79 (IgG2b;
T. Mohanakumar, St. Louis, MO), 41H16 (IgG2a; T. Zipf, Houston, TX),
CIKM5 (IgG1; G. Pilkington, Melbourne, Australia), 2E1 (IgG2a;
Immunotech, Westbrook, ME), IV3 (IgG2b; Medarex, Lebanon, NH), 32.2
(IgG1; Medarex), and 3G8 (IgG1; Medarex) (28). All were
applied at a working concentration of 10 µg/ml as defined below.
Anti-FcRn Abs in these studies were directed against the cytoplasmic
tail (anti-CT) and the extracellular region (anti-H2) of the
molecule (29); both were used at 20 µg/ml. Additional
control Abs were anti-cytokeratin, anti-placental alkaline
phosphatase (prediluted mAb and rabbit sera, respectively; both
obtained from Zymed, San Francisco, CA), and anti-CD31 (10 µg/ml;
Sigma, St. Louis, MO). Isotype controls included IgG fractions of
myeloma proteins MOPC-21, MOPC-141, and HOPC-1 (10 µg/ml; Sigma).
For immunoblotting studies, isoform-specific rabbit polyclonal
antiserum Ab 260 (30) was used to detect FcRIIa while a
recently developed rabbit polyclonal antiserum (Ab 163.96) directed
against GST fused to the cytoplasmic portion of FcRIIb1 at its N
terminus (GST-ALPGY; produced by J.-L.T.) was used to detect both b1
and b2. Both antisera were used at a 1:2000 dilution. In additional
studies anti-FcRII mAbs II8D2 (1 µg/ml) and II1A5 (supernatant
at 1:1000) (31) were used to confirm isoform-specific
reactions in control cells and placental samples.
Secondary Abs for immunohistology were FITC-conjugated goat F(ab')2 anti-mouse IgG and FITC-conjugated goat F(ab')2 anti-rabbit IgG used at 1:50 dilution (Caltag, South San Francisco, CA). For immunoblotting, HRP-conjugated sheep anti-mouse and anti-rabbit IgG Abs were used at a 1:5000 dilution (Amersham Pharmacia Biotech, Piscataway, NJ).
Tissue procurement and handling
Placental tissue samples were supplied by the regional tissue
procurement facility of The Ohio State University Medical Center. These
were randomly selected from normal full-term Caesarian deliveries and
were processed within 30–45 min of delivery. Individual lobes were
dissected out and washed with fresh cold PBS (0.15 M PBS, pH 7.2).
Tissue pieces 1 cm3 were excised, rinsed by
immersion in cold PBS, and fixed in 4% paraformaldehyde/PBS at 4°C
for 12 h. Samples were washed with four changes of PBS (10x
volume) for a total of 16 h at 4°C. Tissues were then suffused
by immersion in 2.5 M sucrose/PBS for 8–16 h (until sample sank
following saturation). Next, the samples were rinsed with PBS and
placed in TBS freeze medium (Triangle Biomedical Sciences, Durham, NC)
within CMS Tissue Path molds (Fisher Scientific, Pittsburgh, PA) and
snap-frozen by immersion in liquid nitrogen. Resultant blocks were
stored at -70°C until sectioning.
Immunohistology
Tissue sample blocks were warmed to -20°C for 2 h before 5-µm sections were cut. These were air-dried onto Superfrost microscope slides (Fisher Scientific) and stored at -20°C.
To label, slides with sections were thawed 5 min at room temperature,
submerged in PBS for 5 min to solubilize the TBS medium, and washed in
PBS for 5 min. Sections were then blocked with 5% goat serum in PBS
for 1 h at room temperature in a humidified chamber. Blocking
solution was replaced with 50 µl of primary Abs in 5% goat
serum/PBS, and sections were incubated 2 h at room temperature
(anti-FcR) or 16 h at 4°C (anti-FcRn) in a humid
chamber. Next, slides were washed with three changes of PBS for 15 min
each. Sections were then incubated with 50 µl of appropriate
FITC-conjugated secondary Abs in 5% goat serum/PBS for 1 h.
Finally, these were washed again in PBS three times for 15 min, and
coverslips were mounted with Gelmount (Fisher Scientific).
Densitometry
After labeling and mounting, sections were examined within 24–48 h with a Nikon Optiphot Epifluorescent microscope and photographed on TMax 400 film rated at EI 1600. The resultant negatives were then digitally scanned with a Polaroid SprintScan 35 slide scanner. Full frame grayscale scans were made of each negative at 1024 DPI resolution. These files were imported into Sigma-Image analysis software (Jandel Scientific, Corte Madera, CA), and density measurements performed. Three linear measurements of pixel intensity were taken at 60° angles to each other across selected vessels. In any given image, vessels with clear cross-sections were measured, whereas tangentially cut vessels were excluded. In addition, at least three background measurements were also collected in each villus cross-section, which included the syncytiotrophoblast layer and avascular stroma. All observations were transferred to Excel 97 (Microsoft, Seattle, WA), and calculations were performed to subtract average background values from observed intensities. Vessels were grouped according to the morphologic nature of the villi into terminal, intermediate, or stem villi and cord vessels (CV). Then the average areas under the peak were calculated and plotted for each grouping (n > 15 independent observations for each grouping).
Immunoblotting
Tissue lysates were prepared by selective dissection and isolation of villus tips (a modification of methods by Kacemi et al., Ref. 32) from samples collected as described above. In this procedure, following extensive washing with PBS, fresh 2 cm3 blocks of tissue were placed in cold PBS and progressively diced by cross-cutting with paired razor blades, which released large numbers of very small villus pieces into suspension. After dissection, the suspended fragments were transferred to a 50-ml conical centrifuge tube, and large pieces were allowed to settle. Chorionic villus tips remained suspended and were removed to a 15-ml centrifuge tube after 5 min and were pelleted at 1000 RCF in a Beckman model GPR centrifuge. This produced samples enriched for terminal villus tips (TVT) from the suspension as well as mixed villus (MV) samples from the initial sunken material. Relative sample contents were confirmed by phase contrast microscopic examination.
Immediately following harvest, 0.1 g of pelleted material was
resuspended (working volumes 30 µl), and 100 µl of lysis buffer
(25 mM HEPES, 20 mM
Na4P2O7.10
H2O, 100 mM NaF, 4 mM EDTA, 2 mM
Na3VO4, 1% Triton X-100,
0.34 mg/ml PMSF, 0.01 mg/ml aprotinin, and 0.01 mg/ml leupeptin) was
added. Following a 30-min incubation on ice, debris was pelleted in a
refrigerated Eppendorf 5415 microfuge at 16,000 relative centrifugal
force (RCF), and the lysate was stored at -20°C (typically for
18 h). Control cells (U937, Raji and Daudi; American Type Culture
Collection (ATCC), Manassas, VA) were grown to an approximate density
of 1–2 x 106 cells/ml in RPMI 1640 medium
(Life Technologies, Rockville, MD) containing 15% FBS (HyClone, Logan,
UT), 100 U/ml of penicillin, and 100 µg/ml streptomycin (Life
Technologies). These were then pelleted at 1000 RCF, washed three times
in cold PBS, and resuspended to 1 x 107
cells/tube. These were then pelleted and subjected to lysis with 500
µl of buffer as above. Additional isoform controls in these
experiments included transfected COS-7 fibroblast cells (ATCC). In each
case, pCEXV-3 FcRIIa, IIb1, IIb2 constructs (provided by Dr. J.
Ravetch, Rockefeller University, New York, NY) or empty vectors were
transfected into cells using LipofectAMINE reagent (Life Technologies)
as previously described (33).
Samples were next thawed, and 250 µl of each was subjected to
immunoprecipitation with 50 µl of anti-FcRII mAb [KB61 (1
µl ascites), AT10 (1 µl ascites), and IV3 (1 µg IgG)]-coated
F(ab')2 goat anti-mouse IgG (Pierce,
Rockford, IL)-conjugated Sepharose beads (Amersham Pharmacia Biotech).
Resulting samples were then subjected to SDS-PAGE on 10% gels as
described (29). Proteins were transferred to
nitrocellulose membranes (Hybond ECL; Amersham Pharmacia Biotech),
blocked, and probed with test Abs in 5% low fat milk in 10 mM
TBS/0.1% Tween 20. The signal was then detected by addition of ECL
substrate reagent (Amersham Pharmacia Biotech) and visualized on Kodak
X-OMAT AR film (Eastman Kodak, Rochester, NY).
Flow cytometry
The relative blotting efficiency of the isoform-specific
FcRII antisera (Ab 260 and Ab 163.96) was quantified in the
following manner. We compared FcRII expression on cultured U937
cells (expressing virtually only the FcRIIa isoform) and Raji cells
(expressing only the FcRIIb isoforms) by flow cytometry using
pan-FcRII mAb KB61 as previously described (34). We
found the two cells to express roughly equal numbers of FcRII.
Simultaneously, we assessed the FcRII band density of immunoblots
labeled with Ab 260 and Ab 163.96 of whole cell lysates of varying
numbers of U937 and Raji cells. For an equal number of FcRII from
the two cell types, we found the Ab 163.96 signal to be 2.5 times
brighter than the Ab 260 signal.
Endoglycosidase assay
To evaluate the glycosylation of placental FcRII proteins,
FcRII was immunoadsorbed from detergent lysates as above, then
submitted to N-glycosidase F (Boehringer Mannheim,
Indianapolis, IN) treatment as described (35). Briefly,
after three washings with PBS, adsorbed samples were denatured with 0.1
M 2-ME/0.1% SDS and heated to 90°C for 5 min. Then 10 µl of each
sample was combined with 3 µl of 0.5 M Tris-Cl, pH 8.6; 5 µl of
H2O; 2 µl of 10% Triton X-100, and 5 µl of
250 mU/ml N-glycosidase F or 5 µl of 0.5 M Tris-Cl, and
these were incubated for 16 h at 37°C. These samples were
analyzed by SDS-PAGE and immunoblotting as above.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
RII mAb labeling of placental villus vasculature
Using a library of mAbs, we evaluated the distribution of FcRII
by indirect immunofluorescence of paraformaldehyde-fixed frozen
sections of villus tissue from six randomly selected normal placentas.
Table I
presents an overview of
observations made throughout the placental vasculature. It can be seen
from these data that three patterns of cell-specific reactivity were
detected with this anti-FcRII mAb library. Pattern 1 was seen
with the mAbs KB61, 41H16, and Ku79. With these three Abs, the
endothelium of all capillaries within terminal villi, as well as those
of the intermediate villus peripheral capillary network, displayed an
intense positive fluorescence signal. Larger vessels within the
placental vascular tree (intermediate and stem villi) showed a
decreasing signal as distance from the capillaries increased. In the
umbilical CV, little or no signal was seen in EC. Non-EC within the
stroma of larger villi and cord tissue also labeled with a positive
signal. Based on distribution and overall morphology, these positive
stromal cells were identified as fetal tissue macrophages (Hofbauer
cells).
|
, A and B,
illustrates pattern 1 reactivity seen with KB61. Several types of
chorionic villi from terminal (marked with "T") to intermediate
(marked with "I") are shown in A. All terminal villus
capillaries (TVC) are labeled with anti-FcRII Ab, as are
peripheral capillaries of intermediate villi (arrows). In B,
an intermediate villus (center) shows both peripheral capillaries
(arrows) and larger vessels (arrowhead) labeled with positive
fluorescence signal. As indicated in Table I
, mAbs 41H16 and Ku79
produced virtually identical labeling throughout the vasculature as
that seen with mAb KB61.
|
, these mAbs showed a low level of EC
labeling, whereas stromal cells presented a strong positive signal. The
third type of reactivity (pattern 3) was observed with mAb IV3 and was
characterized by a strong positive signal in the stromal Hofbauer cells
with very little or no endothelial labeling. Fig. 1, C and
D, shows chorionic villi labeled with mAb IV3. In many of
the larger villi positive stromal cells (arrows) can be seen. Based on
location and relative distribution within the tissue, these cells were
considered to represent the Hofbauer cells commonly reported in larger
villi.
In addition to anti-FcRII mAbs, we also examined placental
sections with Abs to FcRI and FcRIII (mAbs 32.2 and 3G8,
respectively). In both cases, stromal cells were positive but no
endothelial or trophoblast signal was detected (Table I). Although both
of these Abs presented similar signals (pattern 4), it should be noted
that differences in signal distribution were observed between each of
these mAbs and those signals seen with mAb IV3 or CIKM5/2E1 (above).
These data suggest the possible presence of different FcR expression
patterns within villus Hofbauer cells at term, although further pursuit
of this question is beyond the scope of this report.
Labeling with anti-cytokeratin mAb was used in these studies to
identify the trophoblast layer (Fig. 1E, arrows). None of
the anti-FcR mAbs used in these studies labeled normal
trophoblast. Myeloma proteins MOPC-21, MOPC-141, and HOPC-1 were all
applied to test for nonspecific FcR binding of mouse Ab classes
IgG1, IgG2a, and IgG2b, respectively. Fig. 1F, showing
MOPC-21 labeling, illustrates that no endothelial or other villus
labeling was detected with these reagents.
Distribution of FcRII in placental vascular tree
Our observations with the anti-FcRII mAbs that label the
villus endothelium most intensely (pattern 1 summarized in Table I)
suggested that the fluorescence signal in the placental vascular tree
was brightest in TVC, virtually absent in vessels of the umbilical
cord, and that the change in signal intensity between these two
extremes along the vascular branch was gradual. This point is
illustrated microscopically in Fig. 2, where the three panels show the entire intensity spectrum of
endothelial labeling. Panel A shows a TVT labeled with
anti-FcRII mAb KB61 in which all capillary endothelium is
intensely bright (arrows). Panel B shows intermediate villi
where the large vessel is only weakly fluorescent (arrowhead), whereas
peripheral capillaries are intensely positive (arrows). Panel
C shows the lack of endothelial labeling of umbilical
CV.
|
RII distribution
To validate our qualitative impressions of the anti-FcRII
signal gradation illustrated in Fig. 2, we measured pixel intensities
of representative digitized micrographs of fluorescence-labeled vessels
along the length of the vascular tree. Fig. 3, A and B,
illustrates this approach. Following digitization of the micrograph,
pixel intensity measurements were taken of lines drawn at 60° angles
to each other across randomly selected vessels of each type. These
lines record both individual pixel intensities and length of the
measurement. For simplicity, Panel A shows a single line of
the trio drawn across each of two vessels, one a capillary and one a
larger vessel. Plots of the pixel intensity vs distance for these two
vessels (B) indicate that in this case the thickness of the
endothelial layers are fairly equal (5.2 µm) despite the marked
difference in vessel diameters (ratio of 4:1) and that the smaller
vessel is 3 times brighter than the larger vessel (area beneath the
curve).
|
RII signals. The
resultant data clearly documented our qualitative impression that
FcRII signal intensity decreased as the size of the vessel increased
and as location moved away from the capillaries and toward the cord. In
distinct contrast, the distribution of FcRn in endothelium displayed a
reciprocal pattern of expression, with brighter signal being seen in
the cord and minimal signal in the TVC (D). Overall, it can
be seen from these data that the FcRII signal detected in terminal
capillaries has an 4.8-fold higher intensity than that detected in
CV. In contrast, FcRn can be seen to have an inverse signal
differential of 5.1-fold from CV to capillaries. It should be noted
that with both anti-FcRII and anti-FcRn, very low background
pixel intensities were recorded within the cord and capillaries,
respectively. This background represents nonspecific autofluorescence,
which is commonly seen in placental tissue samples. In these studies,
such background is considered to be negative for specific Ab
binding.
The predominant FcRII isoform in placental EC is b2
In light of published reports of specificity of the
anti-FcRII mAbs (36, 37, 38, 39, 40), the cell-specific
reactivity patterns presented in Table I suggest that villus
endothelium may express predominantly the FcRIIb isoform. However,
none of these mAbs is known to label unequivocally only FcRIIa or
FcRIIb. Therefore, a different approach was taken to define which
isoform is actually expressed in the TVC. In these studies, immunoblot
analysis using isoform-specific rabbit anti-sera
(anti-FcRIIa Ab 260 and anti-FcRIIb1 Ab 163.96) was
combined with a specific tissue dissection technique to evaluate the
relative distribution of FcRIIa and FcRIIb within the villus
branch. This was accomplished by fine dissection followed by isolation
of villus branches based on relative size and resulted in two samples,
one enriched for TVT and the other for MV containing both terminal and
intermediate villi. Based on villus morphology (41),
samples containing primarily small terminal villi (TVT) are greatly
enriched for EC relative to the core tissue samples (MV). FcRII was
immunoadsorbed with anti-FcRII mAbs from detergent lysates of
the two villus samples, separated by SDS-PAGE, and identified by
immunoblotting with anti-FcRIIa Ab 260 or anti-FcRIIb Ab
163.96. Cell lines were included as controls for FcRIIa (U937) and
FcRIIb (Raji and Daudi), as were COS 7 cells transfected with
expression vectors containing FcRIIa, FcRIIb1, and
FcRIIb2 cDNA.
Placental TVT and MV samples probed with Ab 163.96 (Fig. 4, top) show an intense band
at 35 kDa characteristic of FcRIIb2 (lower
arrow). This band comigrates with the major band observed in the
FcRIIb2-transfected COS 7 cells and moves faster than the major band
(38 kDa) in FcRIIb1 transfectant or Raji/Daudi control lanes
(upper arrow). No such bands were seen in samples of
FcRIIa transfectants or of U937 cells, which express abundant
FcRIIa. In addition to these major bands, Ab 163.96 also detected a
somewhat weaker and slower moving band in the MV sample at 50 kDa.
Similar bands are also seen in Raji (55 kDa) and both the
FcRIIb-transfected COS 7 cells (55 and 50 kDa). Although there
are minor differences in the position of these slower moving bands, the
degree of similarity and distribution within all positive controls
suggests that these are different forms of FcRIIb, perhaps resulting
from differential posttranslational modifications. In additional
experiments immunoadsorbing with Ab 163.96 and blotting with the
anti-FcRIIb mAb II8D2 or pan-FcRII mAb II1A5, both the 35-kDa
placental bands and slower moving Raji/Daudi bands were observed (data
not shown), which supports the interpretation of these bands as
FcRIIb.
|
, lower panel) failed to
detect any FcRIIa signal in the placental TVT sample in this
experiment. The MV sample did show a minor band at 43 kDa, which
corresponded with the major FcRIIa band detected in U937 cells as
well as a specific band in the FcRIIa-transfected COS 7 sample.
Thus, very little FcRIIa was seen in our placental samples and none
in the TVT.
This experiment was repeated on eight separate placental samples. In
all eight experiments, we saw robust FcRIIb2 bands in both TVT and
MV lanes. Moreover, in six of eight, there was little or no FcRIIa
signal in the TVT and either a weak or no signal in MV. However, in two
of the eight experiments, the FcRIIa signal was present in both TVT
and MV lanes. Therefore, we quantified the efficiency by which each of
these Abs recognizes its target, determining stoichiometric
relationships by flow cytometry, as described in Materials and
Methods. We found that in both of these experiments the number of
FcRIIb found in TVT was at least 2.5-fold the number of FcRIIa
detected. Our experiments, in sum, indicate that FcRIIb2 is the
predominant isoform in placental TVT and, by inference, in the terminal
villus endothelium.
FcRII glycosylation
We next examined the glycosylation of placental FcRII to assess
the size of the core protein. TVT and MV along with U937 and Raji cell
lysates were immunoadsorbed on beads conjugated with anti-FcRII
mAbs, and the immunoadsorbed receptor was incubated with and without
N-glycosidase F. Both glycosylated and deglycosylated
samples were then immunoblotted with Ab 163.96. Fig. 5 shows that the 35-kDa FcRIIb
band, detected by Ab 163.96 in both TVT- and MV-glycosylated samples
(upper arrow), was replaced by a major band at 29 kDa
after deglycosylation. Raji cell lysates showed the same pattern
(arrowheads), although the respective MWs were 3 kDa higher in both
cases than the placental bands (at 38 and 31 kDa). The
deglycosylated sizes of these proteins correspond to the expected core
protein sizes of FcRIIb2 and FcRIIb1, the relative difference
corresponding to the 19-aa cytoplasmic tail insert of FcRIIb1. No
FcRIIb signal was detectable in U937.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
RII
expressed in the endothelium of placental villi is predominantly the b2
isoform. Our strongest and most direct evidence comes from
immunoblotting experiments using a new rabbit polyclonal Ab directed
toward the entire cytoplasmic tail of FcRIIb1. The data in Fig. 4
show that this Ab recognizes the b1 isoform produced by transfected
fibroblasts and human B cell lines (Daudi and Raji) as well as the b2
isoform (lacking the 19-aa exonic insert) produced in transfected
cells, but it does not cross-react with FcRIIa present in COS 7
transfectants or in the human monocyte line U937. So, we judge that it
is specific for the two forms of FcRIIb. Parenthetically, we would
add that recent studies indicate that U937 cells express low levels of
FcRIIb (S. Tridandapani, K. Siefken, J.-L. Teilland, J. E.
Carter, M. D. Wewers, and C. L. Anderson, manuscript in
preparation).
For the immunoblotting experiments with this Ab, we partially purified
villus tips from placenta such that the cells of our tissue lysate were
predominantly EC and syncytiotrophoblast, whereas Hofbauer cells were
rarely present. Thus, the major FcRII-expressing cell in this
preparation, according to our immunofluorescence studies (Table I), was
endothelium. Lysates of these villus tips, in the majority of our
experiments, expressed only the IIb form of FcRII. This conclusion
was supported by additional experiments (data not shown) in which
FcRII was immunoadsorbed from villus tips with Ab 163.96 or
anti-FcRII mAbs and identified by immunoblotting with the
anti-FcRIIb-specific mAb II8D2 or pan-FcRII mAb II1A5.
Furthermore, the mobility in sizing gels of the FcRIIb band from
villus tips was identical with that expressed by the FcRIIb2
transfectant and faster than the band derived from the FcRIIb1
transfectant and from both human B cell lines. Although B cells express
both b1 and b2 isoforms (42), b1 appears to be the
predominant form (31). The mobilities of the
deglycosylated receptors also confirmed these observations (Fig. 5).
Thus, we conclude that villus endothelium predominantly expresses the
FcRIIb2 isoform.
The occasional presence of FcRIIa, seen in two of eight
immunoblotting experiments (where it represented <30% of total
FcRII molecules), we suspect was due to villus tip preparations with
greater numbers of blood leukocytes or Hofbauer cells, which are known
to express high concentrations of the FcRIIa isoform (Table I)
(34). This conclusion is supported by immunofluorescence
studies (Table I), which indicated that mAb IV3, known to be relatively
specific for FcRIIa, bound to only occasional EC in the same
placental sections where Hofbauer cells were positive, whereas three
pan-anti-FcRII mAbs bound well to endothelium. The possibility
that a product of the C gene of FcRII might be expressed is
eliminated by the negative immunoblotting results with Ab 260 directed
toward the cytoplasmic tail of FcRIIa, which is identical with the
cytoplasmic tail of FcRIIc (43).
Whether the b2 isoform of FcRII participates in the movement of IgG
across the villus endothelium is currently a matter of conjecture. It
would appear that the large amount of IgG that traverses the
endothelium moves in transport vesicles, possibly in the abundant
caveolae (44, 45). A transport function for the receptor
is definitely compatible with our observation that expression varies
with location in the placental vascular tree. Expression is highest in
the villus tips and nil in the cord, correlating with transport
functions in general. As well, the murine form of this receptor has
been studied by transfection in Madin-Darby canine kidney cells and has
been shown to transport Fab of anti-FcRII Ab from one pole of
the cell to the other, suggesting its capacity to mediate IgG transport
(27). Were such a postulated transport process to be
active, the receptor would require more than a single affinity state,
which has not been described. However, recent studies of the crystal
structures of both FcRIIa and FcRIIb have suggested models in
which the receptors bind ligand in a complex consisting of two receptor
molecules and a single IgG ligand (46, 47). Although the
details of these two models differ, it seems reasonable to postulate
that in such trimolecular structures the receptor may be capable of two
different affinities for ligand, a high and a low affinity. A high
affinity would be seen when two receptors bind a single ligand, and a
low affinity would result from a single receptor binding a single
ligand. Certainly, it is conceivable that the switch from one
configuration to the other might be mediated by phosphorylation of the
tyrosine-based signal motif in the cytoplasmic tail of FcRIIb2.
Furthermore, it is possible that an association with specialized
membrane microdomains such as caveolae could facilitate both the
ligand/receptor interaction and the postulated shift in the
phosphorylation/affinity state of the receptor.
Whether endothelial FcRIIb2 serves to scavenge immune complexes is
also a consideration, although the endothelium is not generally thought
of as a scavenging cell type, and other cells, namely Hofbauer cells,
perform this function in the villus. However, FcRIIb2, when
expressed by transfection in fibroblasts and B cells, has been shown to
mediate the endocytosis of immune complexes via clathrin-coated pits
and vesicles (25, 26, 27). Resolution of the precise function
of endothelial FcRIIb will await additional experiments.
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Clark L. Anderson, 473 West 12th Avenue, Room 415, Columbus, OH 43210.
3 Abbreviations used in this paper: EC, endothelial cell(s): TVT, terminal villus tips; MV, mixed villus/villi; FcR, Fc receptor for IgG; RCF, relative centrifugal force; TVC, terminal villus capillary/ies; CV, cord vessel(s).
Received for publication October 4, 2000. Accepted for publication January 10, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
receptors in the human placenta. Placenta 2:355.[Medline]
receptor heterogeneity in the human placenta. Scand. J. Immunol. 32:561.[Medline]
receptors and membrane immunoglobulin inhibits anti-immunoglobulin-induced blastogenesis. J. Immunol. 132:627.[Abstract]
RII-deficient mice. Nature 379:346.[Medline]
receptor-mediated
phagocytosis by a nonphagocytic Fc receptor. Blood
91:1762.
RII (CD32) and FcRIII (CD16): characterization and use in various assays. Hybridoma 16:519.[Medline]
receptor-mediated signal transduction in myeloid cells. J. Biol. Chem. 275:20480. receptor II (CD32) isoforms produced in insect cells. Biochemistry 38:8469.[Medline]
RIIA, FcRIIB, and FcRIIC in hematopoietic cells: analysis of transcripts. Mol. Immunol. 30:451.[Medline]
-receptor IIb: a new member of the immunoglobulin superfamily at 1.7 Å resolution. EMBO J. 18:1095.[Medline]
This article has been cited by other articles:
|
|
 |
|
  J. Kim, S. Mohanty, L. P. Ganesan, K. Hua, D. Jarjoura, W. L. Hayton, J. M. Robinson, and C. L. Anderson FcRn in the Yolk Sac Endoderm of Mouse Is Required for IgG Transport to Fetus J. Immunol., March 1, 2009; 182(5): 2583 - 2589. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Takizawa, C. L. Anderson, and J. M. Robinson A Novel Fc{gamma}R-Defined, IgG-Containing Organelle in Placental Endothelium J. Immunol., August 15, 2005; 175(4): 2331 - 2339. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Liu, E. Masuda, M. C. Blank, K. A. Kirou, X. Gao, M.-S. Park, and L. Pricop Cytokine-mediated regulation of activating and inhibitory Fc{gamma} receptors in human monocytes J. Leukoc. Biol., May 1, 2005; 77(5): 767 - 776. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. L. Kepley, S. Taghavi, G. Mackay, D. Zhu, P. A. Morel, K. Zhang, J. J. Ryan, L. S. Satin, M. Zhang, P. P. Pandolfi, et al. Co-aggregation of Fc{gamma}RII with Fc{epsilon}RI on Human Mast Cells Inhibits Antigen-induced Secretion and Involves SHIP-Grb2-Dok Complexes J. Biol. Chem., August 20, 2004; 279(34): 35139 - 35149. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Tridandapani, R. Wardrop, C. P. Baran, Y. Wang, J. M. Opalek, M. A. Caligiuri, and C. B. Marsh TGF-{beta}1 Suppresses Myeloid Fc{gamma} Receptor Function by Regulating the Expression and Function of the Common {gamma}-Subunit J. Immunol., May 1, 2003; 170(9): 4572 - 4577. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. G. Kolonin, R. Pasqualini, and W. Arap Teratogenicity induced by targeting a placental immunoglobulin transporter PNAS, October 1, 2002; 99(20): 13055 - 13060. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Tridandapani, K. Siefker, J.-L. Teillaud, J. E. Carter, M. D. Wewers, and C. L. Anderson Regulated Expression and Inhibitory Function of Fcgamma RIIb in Human Monocytic Cells J. Biol. Chem., February 8, 2002; 277(7): 5082 - 5089. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
