Download C1qRP Is a Heavily O-Glycosylated Cell Surface Protein Involved in

C1qRP Is a Heavily O-Glycosylated Cell
Surface Protein Involved in the Regulation of
Phagocytic Activity
This information is current as
of August 3, 2017.
Subscription
Permissions
Email Alerts
J Immunol 1999; 162:3583-3589; ;
http://www.jimmunol.org/content/162/6/3583
This article cites 32 articles, 22 of which you can access for free at:
http://www.jimmunol.org/content/162/6/3583.full#ref-list-1
Information about subscribing to The Journal of Immunology is online at:
http://jimmunol.org/subscription
Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by
The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 1999 by The American Association of
Immunologists All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
References
R. R. Nepomuceno, S. Ruiz, M. Park and A. J. Tenner
C1qRP Is a Heavily O-Glycosylated Cell Surface Protein
Involved in the Regulation of Phagocytic Activity1
R. R. Nepomuceno,2 S. Ruiz,3 M. Park, and A. J. Tenner4
I
nteraction of human monocytes and culture-derived macrophages with three structurally similar molecules, the classical
complement pathway recognition molecule C1q, mannosebinding lectin (MBL),5 and pulmonary surfactant protein A (SPA)
(1– 4), results in an enhanced ability of these cells to phagocytose
particles suboptimally opsonized with IgG or complement (1, 2, 4,
5). These multidomain proteins appear to have important roles in
the clearance of immune complexes, microbial pathogens, and cellular debris (6 –9). All three molecules are large, containing multipolypeptide chains, each of which consists of a distinctive collagen-like region in the N-terminal portion contiguous with a
globular noncollagen-like domain. It has been shown that the collagen-like region of C1q is responsible for the enhanced phagocytosis effect since only the pepsin-resistant, but not collagenaseresistant fragments of C1q enhance monocyte phagocytic capacity
(2). The similar collagen-like structure of C1q, MBL, and SPA and
the ability of C1q to inhibit the binding of SPA (10, 11)- and MBL
Department of Molecular Biology and Biochemistry, University of California, Irvine,
CA 92697
Received for publication September 25, 1998. Accepted for publication December
18, 1998.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
1
This work was supported by grants from the Arthritis Foundation and the National
Institutes of Health (AI 41090).
2
Current address: Stanford University, MSLS, 3rd Floor, MC:5492, Stanford, CA
94305.
3
Current address: C. Nal. Farmacobiologia, Instituto de Salud Carlos III, Madrid,
Spain.
4
Address correspondence and reprint requests to Dr. Andrea J. Tenner, 3205 Biological Sciences II, Dept. Molecular Biology and Biochemistry, University of California, Irvine, CA 92697-3900.
5
Abbreviations used in this paper: MBL, mannose-binding lectin; BAG, benzyl
2-acetamido-2-deoxy-a-D-galactopyranoside; CHO, Chinese hamster ovary; E, sheep
erythrocyte; HSA, human serum albumin; PI, phagocytic index; SPA, pulmonary
surfactant protein A.
Copyright © 1999 by The American Association of Immunologists
(12)-coated particles to human monocytes support the hypothesis
that a common motif in the collagen-like region of these molecules
may be responsible for cell activation via a common cell surface
receptor. The fact that mAbs R139 and R3 are able to abrogate the
enhanced phagocytosis by C1q and MBL indicates that C1qRP is
at least a shared, functional component of the receptor that mediates the enhanced phagocytosis induced by these two ligands (4, 13).
Using the monoclonal anti-C1qRP to affinity purify this surface
protein, we obtained 10 peptide sequences, designed oligonucleotide primers and probes based on those sequences, and cloned and
sequenced a cDNA that encodes a novel type 1 membrane protein
(GenBank Accession U94333). However, the amino acid sequence
deduced from this cDNA indicates that the mature protein is composed of 631 amino acids, which is calculated to be 66,495 Da
(14), while previous characterization of C1qRP demonstrated that
it migrates in SDS-PAGE gels with a relative mobility of 100,000,
which shifts upon reduction to 126,000. While it was known that
the molecule recognized by R139 and R3 contained some sialic
acid (15), the basis for the large discrepancy between the predicted
molecular mass and the migration of the protein in SDS-PAGE had
not been discerned.
The ability to up-regulate this powerful effector mechanism of
the innate immune response at early stages of infection before the
development of adaptive responses would be a potentially useful
prophylactic and/or therapeutic approach. Thus, it is critical to
characterize the molecular interaction sites both on the ligand and
the receptor. To initiate these studies, we generated new data to
validate that the molecule recognized by the mAb, R139 and R3,
which we designate C1qRP, plays a critical role in inducing this
enhancement of phagocytic function. The data presented document
that SPA is also able to trigger enhanced phagocytosis through
C1qRP, and that cross-linking of C1qRP via the IgM mAb, R3, was
sufficient to induce this cellular response. Furthermore, the structural features contributing to the migration of the mature C1qRP in
SDS-PAGE were characterized by comparing the electrophoretic
mobility of the purified C1qRP with the recombinantly expressed
0022-1767/99/$02.00
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
C1q, mannose-binding lectin (MBL), and pulmonary surfactant protein A (SPA) interact with human monocytes and macrophages, resulting in the enhancement of phagocytosis of suboptimally opsonized targets. mAbs that recognize a cell surface
molecule of 126,000 Mr, designated C1qRP, have been shown to inhibit C1q- and MBL-mediated enhancement of phagocytosis.
Similar inhibition of the SPA-mediated enhancement of phagocytosis by these mAbs now suggests that C1qRP is a common
component of a receptor for these macromolecules. Ligation of human monocytes with immobilized R3, a IgM mAb recognizing
C1qRP, also triggers enhanced phagocytic capacity of these cells in the absence of ligand, verifying the direct involvement of this
polypeptide in the regulation of phagocytosis. While the cDNA for C1qRP encodes a 631 amino acid membrane protein, Chinese
hamster ovary cells transfected with the cDNA of the C1qRP coding region express a surface glycoprotein with the identical
126,000 Mr in SDS-PAGE as the native C1qRP. Use of glycosylation inhibitors, cleavage of the mature C1qRP with specific
glycosidases, and in vitro translation of C1qRP cDNA demonstrated that both posttranslational glycosylation and the nature of the
amino acid sequence of the protein contribute to the difference between its predicted m.w. and its migration on SDS-PAGE. These
results verify that the cDNA cloned codes for the mature C1qRP, that C1qRP contains a relatively high degree of O-linked
glycoslyation, and that C1qRP cross-linked directly by monoclonal anti-C1qRP or engaged as a result of cell surface ligation of
SPA, as well as C1q and MBL, enhances phagocytosis. The Journal of Immunology, 1999, 162: 3583–3589.
3584
O-GLYCOSYLATED C1qRP MEDIATES ENHANCED PHAGOCYTOSIS
C1qRP, with glycosidase-treated C1qRP, and with C1qRP synthesized in an in vitro system lacking the ability to posttranslationally
modify the protein.
Materials and Methods
Reagents, Abs, and cell culture
Phagocytosis assay
Opsonized target particles for the phagocytosis assay were sheep erythrocytes (E) bearing either IgG anti-SRBC (EAIgG) or IgM anti-SRBC and
C4b (EAIgMC4b) (prepared as previously described (21)) to assess FcRand CR1-mediated phagocytosis, respectively. Eight-well Lab Tek chambers (Nalgene, Naperville, IL) were coated with varying concentrations of
C1q, SPA, HSA, iron-saturated transferrin, control IgM, or R3 (1). Monocytes (6.25 3 104 cells/well) were added to each chamber, and the cells
were centrifuged at 700 rpm (RT6000; DuPont Sorvall, Newtown, CT) for
3 min and subsequently placed at 37°C in 5% CO2 for 45 min. Targets were
then added (107/100 ml) and the slides were again subjected to centrifugation (700 rpm, 3 min) and incubated for 30 min at 37°C. For CR1mediated phagocytosis, 10 ng/ml phorbol dibutyrate was added with the
opsonized targets, as it is well known that while monocytes can bind to
targets via CR1, this receptor must be activated in these less differentiated
myeloid cells to mediate phagocytosis of complement-opsonized targets
(1). After removing unbound targets by washing, bound, noningested targets were removed by hypotonic lysis (1). Cells were then fixed in 1%
glutaraldehyde and stained with Giemsa. Phagocytosis was quantitated using light microscopy. The number of E targets ingested per 100 effector
cells was defined as the phagocytic index (PI), whereas the percentage of
effector cells ingesting at least one E target was defined as the percent
phagocytosis. Each experiment, performed on separate days with different
donors, used duplicate sample wells per condition. Controls of unopsonized E were not ingested by monocytes or macrophages under any conditions. Statistical analysis was performed using the paired Student’s t test.
rC1qRP expression
Reverse-transcriptase PCR was performed using RNA isolated from U937
cells with the SuperScript Preamplification System for First Strand cDNA
Synthesis Kit, according to manufacturer’s instructions. A 2034-bp cDNA
was amplified that contains the entire C1qRP coding region, using Pfu
polymerase (Stratagene, La Jolla, CA), and 59-GCAGAGGGCCACACA
GAGACCG-39 as the forward primer, and the oligonucleotide 59-GCTCT
GAGGATGGTGGCTGGTG-39 as the reverse primer. After purification of
the C1qRP cDNA using the Geneclean DNA purification kit (Bio101, La
Jolla, CA), according to manufacturer’s instructions, 2.5 U Taq polymerase, 20 mM Tris-HCl, pH 8.4, 50 mM KCl, 1.5 mM MgCl2, and 0.2 mM
dATP were added in a 100 ml reaction and incubated for 1 h at 37°C to add
terminal A’s for cloning into the pGEM-T vector (Promega, Madison, WI).
Inhibition of glycosylation
U937 cells were grown in the absence or presence of 1–5 mM benzyl
2-acetamido-2-deoxy-a-D-galactopyranoside (BAG) or 6, 12, and 18
mg/ml tunicamycin for 24 –72 h in a 37°C, 5% CO2 incubator. The cells
were then washed once with PBS, lysed with extraction buffer above, and
immunoprecipitated using the R139 mAb. After SDS-PAGE under reducing conditions and transfer to nitrocellulose, the receptor was detected by
Western blotting with the polyclonal QR1 Ab, then developed by enhanced
chemiluminescence.
O-Glycosidase digestion
U937 cells (5 3 108) growing in log phase were harvested and lysed with
extraction buffer. The lysate was precleared using packed Sepharose
CL-40, applied to a R3-affinity column, and eluted with 0.1 M glycine, pH
2.4, 0.5 M NaCl, 0.05% Nonidet P-40, and 1 mM PMSF. The eluate was
concentrated by Centricon (Amicon, Beverly, MA) and boiled with 0.5%
SDS for 2 min. After cooling to room temperature, cacodylate buffer was
added (20 mM NaCacodylate, pH 6, 0.5% Nonidet P-40) and boiled again
for 2 min. The sample, cooled to room temperature, was divided into five
equal aliquots and incubated at 37°C for 18 h in presence of O-glycosidase
(1 mU; Boehringer Mannheim, Indianapolis, IN) alone, neuraminidase (2
mU; Calbiochem, La Jolla, CA) alone, O-glycosidase and neuraminidase (1
mU, 2 mU, respectively), or with no enzymes added. A parallel sample
with no enzyme added was left on ice for the duration of digestion to insure
there was no protein modification in absence of enzymes during the incubation at 37°C. After the incubation, all of the samples were analyzed by
SDS-PAGE under reducing conditions, followed by Western blotting, as
described above with the QR1 Ab.
In vitro expression
TNT Coupled Reticulocyte Lysate System (Promega) was used according
to manufacturer’s instructions. The 1956-bp C1qRP cDNA subcloned into
the pcDNA3.1(1) plasmid and the cDNA encoding the mature C1qRP protein excluding the 21 amino acid signal sequence, which was amplified
from the C1qRP-containing pcDNA3.1(1) plasmid as described above using Pfu polymerase and cloned into the pGEM-T vector, were used as DNA
templates. One microgram of DNA template was added to the rabbit reticulocyte reaction mixture containing 20 mM amino acid mixture minus
cysteine, 10 U RNA polymerase, [35S]cysteine (Amersham, Arlington
Heights, IL), and 40 U RNasin ribonuclease inhibitor (Promega), and incubated at 30°C for 60 min. The reaction mixture was then analyzed by
SDS-PAGE under reducing conditions, followed by exposure to Fuji RX
Medical x-ray film (Stamford, CT) overnight (12–16 h) at 270°C.
Results
mAb R139 inhibits SPA-mediated enhancement of phagocytosis
Previous experiments had shown that the anti-C1qRP mAb, R139,
was able to inhibit both C1q- and MBL-mediated but not fibronectin-mediated enhancement of phagocytosis (4, 15). Since SPA also
enhances FcR- and CR1-mediated phagocytosis (3), the effect of
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
The RPMI 1640 medium, F12 Nutrient Mixture (HAM), SuperScript Preamplification System for First Strand cDNA Synthesis Kit, Lipofectin Reagent, G418, and Taq DNA polymerase were purchased from Life Technologies (Grand Island, NY). C1q was isolated from plasma-derived
human serum by the method of Tenner et al. (16) and modified as described
(17). The preparations used were active, as determined by hemolytic titration, and homogeneous, as assessed by SDS-PAGE. Protein concentration
was determined using an extinction coefficient (E1%) at 280 nm of 6.82 for
C1q (18). Human SPA isolated from alveolar proteinosis patients (5) was
a generous gift from Dr. Jo Rae Wright, Duke University (Durham, NC).
Except where noted otherwise, all other reagents were purchased in the
highest quality from Sigma (St. Louis, MO).
Anti-C1qRP mAbs R139 (IgG2b) and R3 (IgM), generated using C1qbinding proteins as the immunogen (13), were purified before use, as previously described (14). The rabbit polyclonal anti-C1qRP Ab, QR1, was
generated using R3-purified C1qRP, as described (14).
COS-1 cells and the human histiocytic cell line, U937, were grown in
RPMI 1640 medium containing 10% supplemented bovine calf serum (HyClone, Logan, UT) and 10 mM HEPES, pH 7.4. CHO-K1 cells were grown
in F12 Nutrient Mixture containing 10% FCS (HyClone). Human peripheral blood monocytes were isolated from blood units collected into CPDA1
blood collection bags (Baxter, Deerfield, IL) at the UCI Medical Plaza
(Irvine, CA). Monocytes were isolated by counterflow elutriation using a
modification of the technique of Lionetti et al. (19), as described (20).
Greater than 95% of the cells in each preparation were monocytes according to size analysis on a Coulter Channelyzer (Hialeah, FL). Previous analysis has substantiated that such populations are nonspecific esterase positive and .98% viable (20).
The PCR product was then gel purified and cloned into pGEM-T. Insertcontaining plasmids were screened for orientation by restriction digest
mapping and were subcloned into pcDNA3.1(1) (Invitrogen, Carlsbad,
CA). The plasmid insert was sequenced at both ends to ensure it contained
the full-length coding region and for proper orientation for expression.
For transient expression of the receptor, plasmid constructs were transfected into COS-1 cells grown in six-well culture plates (Corning Costar,
Encinitas, CA) using 2 mg of DNA per well and the liposome formulation,
Lipofectin Reagent, according to manufacturer’s instructions. After 48 –72
h, the transfected cells were washed with PBS, then lysed in 1 ml/well of
extraction buffer (10 mM triethanolamine, pH 7.4, 1 mM CaCl2, 1 mM
MgCl2, 0.15 M NaCl, 1% Nonidet P-40, 2 mg/ml pepstatin, 10 mg/ml
leupeptin, 10 mg/ml aprotinin, and 1 mM PMSF). C1qRP immunoprecipitation with the R139 mAb, SDS-PAGE (7.5%), and Western blotting with
biotinylated R3 mAb were performed as described previously (14), and
developed by enhanced chemiluminescence with the HRPL kit (National
Diagnostics, Atlanta, GA).
For stable transfections, constructs were transfected into CHO-K1 cells
grown in 100-mm culture dishes (Corning Costar) using 5.5 mg of DNA
per dish, as described above. After 72 h, the transfected cells were split into
medium containing 400 mg/ml G418. Individual colonies were selected and
tested for C1qRP surface expression by FACS analysis, as described (13).
For additional analysis of the rC1qRP, 1 3 107 cells were washed with
PBS, then lysed, immunoprecipitated, and detected, as described above.
The Journal of Immunology
preincubation of isolated monocytes with anti-C1qRP mAb on the
response triggered by SPA was tested. Fig. 1 presents data from
four experiments in which monocytes pretreated with buffer or
mAb were added to wells that had been coated with HSA, transferrin, SPA, or C1q, and phagocytosis assessed. Anti-C1qRP inhibited both the C1q and SPA enhancement of CR1-mediated
phagocytosis by human monocytes in these experiments and others, with an average percent inhibition of PI of 50 6 17% for C1q
(n 5 5, p , 0.02) and 55 6 14% for SPA (n 5 5, p , 0.02) when
coated at 8 mg/ml, and 77 6 10% for C1q and 48 6 22% for SPA
when the C1q/SPA concentration used for coating was 4 mg/ml
(n 5 3). Inhibition of C1q- or SPA-enhanced phagocytosis was not
observed in parallel samples incubated with control mouse IgG2b.
Immobilized mAb R3 mimics the C1q-, MBL-, or SPA-mediated
enhancement of phagocytosis
Pretreatment of human monocytes with R3, a IgM mAb that recognizes C1qRP, blocked the C1q-mediated enhancement of phago-
FIGURE 2. mAb R3 (anti-C1qRP) triggers enhanced phagocytosis. Human monocytes purified by countercurrent elutriation were added to Lab
Tek chambers that had been precoated with 5 (open bars), 25 (horizonal
stripes), or 50 (cross-hatched) mg/ml R3; 25 (horizonal stripes), or 50
(cross-hatched) mg/ml control mouse IgM; or 8 mg/ml C1q (solid bars).
After 45-min incubation at 37°C, EAIgMC4b targets were added and phagocytosis was assessed (similar results were obtained when EAIgG targets
were used). Results show data from one of four separate experiments.
Values presented are the mean 6 SD of duplicate samples. A, PI (the
number of targets ingested by 100 monocytes). B, The percentage of monocytes ingesting at least one EAIgMC4b target.
cytosis (13) triggered by the adherence of monocytes to C1qcoated wells, similar to the inhibition by R139. However, since R3,
as a IgM mAb, has the potential for cross-linking multiple C1q
receptors without engaging Fc receptors, the ability of surfacebound R3 alone to modulate monocyte phagocytic capacity was
tested. Monocytes were added to wells that had been coated with
5–50 mg/ml of R3 or a control irrelevant mouse IgM or 8 mg/ml of
C1q, and phagocytosis of EAIgMC4b targets was assessed. Fig. 2
shows results from a representative experiment in which R3 mediates enhancement of phagocytosis in a concentration-dependent
manner. When data from four separate experiments were averaged,
the PI for cells adhered to wells coated with 25 and 50 mg/ml R3
was 69 6 20 and 132 6 34, respectively, while the average PI for
cells adhered to wells coated with 25 and 50 mg/ml control mouse
IgM was 25 6 21 and 52 6 19, respectively (n 5 4). While the
absolute values varied with the donors tested, the differences between R3 and IgM are significant ( p , 0.003) using the Student’s
paired t test for analysis. The same results were observed when
EAIgG targets were used (data not shown). Again, no such effect
was seen with several distinct irrelevant mouse IgM run in parallel,
demonstrating the specificity of this response for anti-C1qRP.
These data suggest that immobilized R3 via the ligation of C1qRP
mimics the C1q-/MBL-/SPA-mediated enhancement of phagocytosis.
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
FIGURE 1. Anti-C1qRP inhibits SPA-enhanced phagocytosis. Human
monocytes were preincubated with buffer (open bars), or 10 mg/ml (final
concentration) R139 (anti-C1qRP) (horizontal stripes), or control mouse
IgG2b (cross-hatched bars) before addition to wells precoated with 8 mg/ml
HSA, transferrin, C1q, or SPA. After 45-min adherence, EAIgMC4b targets
were added along with 10 ng/ml phorbol dibutyrate, and phagocytosis was
assayed, as described in Materials and Methods. The data presented are the
mean 6 SD (SD) of four separate experiments performed in duplicate. A,
PI, the number of targets ingested per 100 monocytes. B, % Phagocytosis,
the percentage of monocytes ingesting at least one target. (p, p , 0.006
compared with no Ab; pp, p , 0.02 compared with no Ab.)
3585
3586
O-GLYCOSYLATED C1qRP MEDIATES ENHANCED PHAGOCYTOSIS
FIGURE 3. Western blot of rC1qRP. C1qRP was immunoprecipitated
with 5 mg R139 from U937 cells (lane 1) or COS-1 cells (A) transiently
transfected with the C1qRP cDNA insert cloned in the forward orientation
for expression (lane 2), in the reverse orientation (lane 3), or mock transfected (lane 4). Alternatively (B and C), lysates of U937 cells (lanes 1 and
2), untransfected CHO cells (lanes 3 and 4), or CHO cells transfected with
C1qRP (lanes 5 and 6) or vector only (lanes 7 and 8) were immunoprecipitated with either control IgG2b (lanes 1, 3, 5, and 7) or R139 (lanes 2,
4, 6, and 8). The precipitated protein was separated by SDS-PAGE (7.5%)
under reducing (A, C) or nonreducing (B) conditions and transferred to
nitrocellulose. Nonreduced C1qRP was detected with biotinylated R3 mAb
and streptavidin-horseradish peroxidase (B). Reduced C1qRP was detected
with the QR1 anti-C1qRP polyclonal Ab and peroxidase-labeled anti-rabbit
IgG (A, C). All blots were developed by enhanced chemoluminescence.
Positions of the m.w. standards are indicated to the right of each blot.
Recombinant expression of C1qRP in mammalian cells
The 631 deduced amino acid coding region of the mature C1qRP
predicts a polypeptide with a molecular mass of 66,495 Da, which
is substantially smaller than the 126,000 relative mobility of the
reduced native protein as assessed by SDS-PAGE. Thus, it was
necessary to determine whether the expressed product of the
cloned cDNA for C1qRP would migrate similarly to verify that the
entire coding region has been determined and that no errors were
made during the DNA sequencing that would result in a false,
premature stop codon. Flanking primers and Pfu polymerase were
used to amplify the coding region by reverse-transcriptase PCR
from U937 total RNA. The resulting cDNA was cloned into the
mammalian expression vector, pcDNA3.1(1). The plasmid construct was transiently transfected into COS cells, and the transfected cells were lysed with detergent. R139 mAb was added for
immunoprecipitation, and the resulting protein complexes were
separated by SDS-PAGE under reducing conditions, and transferred to nitrocellulose. C1qRP was detected with the anti-C1qRP
polyclonal Ab, QR1. As shown in Fig. 3A, cells that were transfected with the C1qRP construct in the correct orientation express
a protein that is indistinguishable in size from native C1qRP immunoprecipitated from control U937 cells.
The CHO-K1 cell line was then transfected with the C1qRPcontaining plasmid, and individual clones stably expressing the
receptor were identified by FACS analysis with the anti-C1qRP
mAbs. Surface expression of mAb-reactive epitopes (Fig. 4) demonstrates that successful expression and transport to the plasma
membrane had occurred. Cells were then lysed with detergent, and
the expressed protein was immunoprecipitated with the R139
mAb. The protein complexes were separated by SDS-PAGE under
reducing and nonreducing conditions, and transferred to nitrocellulose. C1qRP was detected with biotinylated R3 mAb for the nonreduced protein, and with QR1 for the reduced protein. As shown
in Fig. 3, B and C, only cells that were transfected with the C1qRP
construct express a protein that is indistinguishable in size from
native C1qRP immunoprecipitated from control U937 cells. Untransfected cells, as well as cells transfected with the pcDNA vector only, do not express anti-C1qRP Ab-reactive proteins (Fig. 3, B
and C, lanes 3, 4 and 7, 8). The fact that the recombinant receptor
is recognized by both the R139 and R3 mAbs (Fig. 3B) whose
epitopes are conformation dependent demonstrates that at least
some proper folding of the expressed receptor occurs and the correct disulfide bonds are formed. Additionally, the recombinantly
expressed protein migrates in SDS-PAGE gels with the same characteristic shift as the native receptor, from the 100,000 Mr nonreduced form (Fig. 3B) to the 126,000 Mr reduced form (Fig. 3C).
These results indicate that the cloned cDNA encodes the polypeptide sequence sufficient for complete expression and posttranslational modifications of the mature C1qRP protein.
Inhibition of C1qRP glycosylation
To begin to investigate the possible posttranslational modifications
that may contribute to the altered mobility of the C1qRP polypeptide in SDS-PAGE, U937 cells were cultured in the presence of the
protein glycosylation inhibitors tunicamycin (for N-linked glycosylation) or BAG (for O-linked glycosylation). C1qRP immunoprecipitated from cells grown for 24 h in up to 18 mg/ml of tunicamycin did not show a detectable shift in mobility in SDS-PAGE
gels compared with receptor from untreated cells (data not shown).
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
FIGURE 4. FACS analysis for detection of rC1qRP surface expression.
U937 cells (A), untransfected CHO cells (B), or CHO cells transfected with
C1qRP (C) or vector only (D) were incubated with 5 mg of the R139
anti-C1qRP mAb (thick lines) or an isotype-matched IgG2b control Ab
(thin lines). Bound Abs were detected with FITC-labeled anti-mouse IgG.
Cell number is shown on the vertical axis, while relative fluorescence intensity is shown on the horizontal axis.
The Journal of Immunology
3587
FIGURE 7. O-Glycosidase treatment alters the relative mobility of
C1qRP. C1qRP was affinity purified on a R3 column from U937 cell lysate
and incubated with neuraminidase alone (2 mU, lane 1), O-glycosidase
alone (1 mU, lane 2), neuraminidase and O-glycosidase (2 and 1 mU,
respectively, lane 3), or in the absence of enzymes (lane 4) at 37°C for
18 h. Another sample of C1qRP with no enzymes added was left on ice for
the duration of the incubation (lane 5). After SDS-PAGE, the proteins were
transferred to nitrocellulose and probed with anti C1qRP, QR1. The result
shown here is a typical representation of four separate experiments.
However, since the predicted amino acid sequence of the receptor
indicates only one Asn-X-Ser/Thr N-linked glycosylation site (14),
it is possible that preventing glycosylation at this single site would
not result in a significantly different migration pattern of the protein. Use of the O-Glyci.bas program, developed by Elhammer et
al. for predicting O-glycosylation sites based on amino acid sequences surrounding potential serine and threonine acceptor sites
(22), suggested a relatively high number of sites favorable for
O-linked glycosylation in C1qRP (Fig. 5). As predicted, treatment
of U937 cells with BAG did result in a marked change in the
mobility of C1qRP. As shown in Fig. 6, C1qRP immunoprecipitated from cells grown in 3 mM BAG had a significant difference
in mobility, 113,000 6 1,400 Mr (n 5 5), relative to the native
C1qRP (126 –128,000 Mr). Additional experiments with cultures
grown in 5 mM BAG for up to 72 h (not shown) showed no further
increase in mobility.
As an alternative approach to determine the extent to which
O-linked glycosylation affects the migration of C1qRP on SDSPAGE, affinity-purified C1qRP from U937 cell lysate was subjected to O-glycosidase digestion. O-Glycosidase hydrolyzes
FIGURE 6. Western blot of C1qRP immunoprecipitated from BAGtreated U937 cells. C1qRP was immunoprecipitated with the R139 mAb
from U937 cells grown for 48 h in the absence (2BAG) or presence of 3
mM BAG (1BAG). After separation of the precipitated protein by SDSPAGE (7.5%) under reducing conditions and transfer to nitrocellulose,
C1qRP was detected with the QR1 anti-C1qRP polyclonal Ab and peroxidase-labeled anti-rabbit IgG using the enhanced chemoluminescence technique for development. Positions of the m.w. standards are indicated at the
left.
O-linked glycans bound to serine and threonine. However, since
the presence of terminal sialic acid on the oligosaccharide prevents
this hydrolysis, the sialic acid must be removed by treatment with
neuraminidase to permit efficient O-glycosidase activity. No shift
on mobility on SDS-PAGE was detected in C1qRP treated with
O-glycosidase alone compared with untreated C1qRP. However,
when treated with neuraminidase alone, C1qRP mobility shifted to
109,100 6 5,000 Mr, which was further shifted to 107,100 6 4,500
Mr when treated simultaneously with neuraminidase and O-glycosidase (n 5 4) (Fig. 7). Importantly, while these experiments indicate that glycosylation does not account for the entire discrepancy between the predicted and observed size of the receptor, they
do indicate significant glycosylation and exclude the possibility
that the 126,000 Mr protein is a dimer of the 66,495-kDa polypeptide.
In vitro translation of C1qRP cDNA
C1qRP cDNA was transcribed and translated in an in vitro system,
employing rabbit reticulocyte components that lack the capacity
for posttranslational modification. The major polypeptide produced in this in vitro translation reaction migrates in SDS-PAGE
as 85,500 6 3,900 Mr (n 5 6). A slightly larger product is detected, as expected, when the coding region, including the signal
peptide, is used as a template (Fig. 8, lane 2 versus lane 1). It is
expected that the signal sequence would be cleaved in the mature
protein, and thus it is the smaller ;85,500-Mr band that represents
the peptide without other posttranslational modifications. Both the
major 85,500-Mr product and the ;65,000 minor product seen in
FIGURE 8. In vitro transcription and translation of C1qRP cDNA.
C1qRP cDNA excluding (lane 1) or including (lane 2) the 21 amino acid
signal sequence transcribed and translated in a rabbit reticulocyte in vitro
system. As a positive control for the transcription and translation reaction,
luciferase control DNA was used (lane 3). A reaction with no DNA template added was conducted as a negative control for the reaction (lane 4).
The autoradiogram shown is a representative of six similar experiments.
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
FIGURE 5. Potential O-glycosylation sites in C1qRP. The deduced
amino acid sequence of C1qRP was analyzed using the O-Glyci program
provided by Elhammer et al. (22). The N terminus of the mature protein is
designated as amino acid “1.”
3588
O-GLYCOSYLATED C1qRP MEDIATES ENHANCED PHAGOCYTOSIS
Fig. 8, lane 1, are seen when the in vitro translation reaction is
immunoprecipitated with polyclonal antiC1qRP, suggesting that
the lower band represents a truncated translation product. Since the
molecular mass deduced from the amino acid sequence of the mature protein is 66,495 Da, these data demonstrate that the composition of the amino acid sequence of the protein itself dictates
altered mobility of SDS-PAGE (23). In addition, although significantly reduced from Mr of 126,000 mature protein, there remains
a discrepancy between the neuraminidase/O-glycosidase-treated
protein (107,900 Mr) and the 85,500 Mr of the unmodified nascent
polypeptide chain, suggesting that some posttranslational modification other than O-glycosylation on serine and threonine may
contribute to the mature protein.
Discussion
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
The data presented in this work demonstrate that C1qRP is a highly
glycosylated cell surface protein that plays a critical role in the
enhancement of phagocytic function by multiple ligands. The enhancement of phagocytosis by SPA is inhibited by the anti-C1qRP
mAb R139 to the same extent as was previously seen for MBL and
C1q (4, 15), providing more compelling evidence that the recently
cloned glycoprotein, C1qRP, is a common receptor (or a common
component of a receptor) shared by C1q, SPA, and MBL. (It is
important to note that this mAb does not inhibit phagocytosis or
the enhancement of phagocytosis in general, as baseline or fibronectin-mediated enhancement of phagocytosis is not inhibited
by R139 (15).) The ability of multiple ligands to effectively engage
this receptor, resulting in enhanced phagocytic function, suggests
that C1qRP is similar to a pattern recognition molecule with a
potentially significant role in innate immunity. While similar in
general to other recognition receptors such as CD14 and macrophage scavenger receptors (24), C1qRP differs in that it binds host
defense collagens (25), namely C1q, MBL, and SPA, once they
have bound pathogenic material displaying a variety of specificities, including microbes or cellular debris. This is conceptually
similar to FcR-Ab interactions, although with a much more limited
repertoire of recognition than that supplied by the large diversity of
Ab specificity.
The degree of inhibition by the anti-C1qRP mAb of ligand-induced enhancement of phagocytic function rarely is greater than
80% (Fig. 1), similar to that seen in previous studies with C1q and
MBL (4, 13, 15). This would be consistent with a requirement for
continuous inhibition of receptor-ligand interactions, rather than an
all or none signaling mechanism. In addition, it is possible that the
avidity of soluble Ab for C1qRP is lower than that of immobilized
ligand for the C1qRP or receptor complex. Previous studies reported that anti-C1qRP R139 does not inhibit C1q binding to cells,
and R3, a IgM anti-C1qRP, only partially inhibits the C1q binding
to monocytes, to the monocytic-like cell line U937, and to neutrophils (13). Thus, the mechanisms by which the Abs inhibit the
ligand-induced response are not known at this time. One possibility is that the Abs block a conformational alteration critical for
signaling. Studies using the recombinantly expressed C1qRP will
define whether the component of the phagocytosis-modulating receptor that is recognized by the mAbs R139 and R3 actually binds
the ligand or if it is involved in the signaling function of the receptor or both.
Previous studies have shown that multivalent presentation of
C1q or a conformational change induced by aggregation or immobilization of the C1q monomer seems to be required for the generation of a functional response (reviewed in Refs. 5, 26, and 27).
The mAb R3 is a IgM pentamer with the ability to cross-link
multiple receptors, and thus to potentially mimic the response trig-
gered by multivalent C1q. The data presented in this work clearly
show that cross-linking of C1qRP by immobilized R3 increases the
monocyte phagocytic capacity in a dose-dependent manner, identifying C1qRP, the polypeptide recognized by the mAb, as a critical component of this important cellular response, and confirming
the requirement for a multivalent presentation to activate C1q receptors at least in monocytes.
Characterization of the structural features of this novel cell surface molecule is critical to understand the ligand-receptor interactions and subsequent signal transduction mechanisms involved in
this host response. While the mAb that inhibited the enhancement
of phagocytosis consistently immunoprecipitated a molecule of
126,000 Mr (C1qRP) from monocytes, neutrophils, U937 cells
(15), and platelets (28), the primary amino acid sequence deduced
from the cloned cDNA predicted a protein with a molecular mass
of 66,495 Da. Translation in an in vitro system (Fig. 8) that lacks
translational modification demonstrated that the amino acid sequence itself results in an aberrant migration in SDS-PAGE, similar to that reported for other molecules with a high percentage of
alanine, proline, and charged amino acids in their primary sequence (23, 29). The results of the experiments investigating glycosylation of this surface molecule are consistent with multiple
O-linked glycosylation sites. Indeed, the extracellular portion of
the molecule proximal to the transmembrane domain has an abnormally high percentage of serine, threonine, and proline residues, providing high probability for O-glycosylation (14) (Fig. 5).
While some of the O-linked oligosaccharides contain N-acetyl
galactosamine, as indicated by the inhibitory effect of BAG, which
acts as an acceptor for UDP-gal:galNAc-b1,3 galactosyltransferase (30), the smaller apparent size of neuraminidase/O-glycosidase-treated C1qRP (107,000 Mr) suggests that either there are
other O-linked sugars that are not inhibited by BAG, or that the
inhibitor is not completely effective. In addition, as mentioned
above, there remains a discrepancy between the in vitro translated
product of 85,500 Mr and the glycosidase-treated protein (107,000
Mr), suggesting additional posttranslational modifications, such as
other forms of O-glycosylation, phosphorylation, and/or sulfation.
While future experiments will be necessary to further define the
functional significance of this glycosylation, earlier experiments
suggested that at least sialic acid was not critical for the enhancement of phagocytosis (15). It has been suggested that O-linked
glycosylation produces an elongated polypeptide structure (31). In
some surface proteins, a short glycosylated domain separates the
functional portion of the protein from the transmembrane domain,
possibly serving to extend the functional domain of the molecule
well beyond the cell surface for interaction with extracellular matrix proteins or other cell or particle surfaces. This may be the
function of this serine/threonine-rich region in C1qRP.
Another cellular response triggered by C1q is the production of
superoxide by neutrophils and eosinophils. However, several lines of
evidence suggest that this response may be mediated by a receptor,
C1qRO22, distinct from C1qRP. First, none of the anti-C1qRP mAbs
inhibit (13) or mimic (unpublished data) C1q-mediated oxidative
burst in neutrophils. In addition, neither immobilized SPA nor MBL
triggers superoxide production in neutrophils (32), suggesting that the
interaction motif on C1q required for the receptor that mediates superoxide generation differs from that required for enhancing phagocytosis. Finally, the region of the C1q molecule required to stimulate
superoxide production (just above the kink in the collagen-like region)
has little homology (beyond the GXY collagen motif) with any region
of SPA, and only minimal homology with MBL (33). These data
indicate that the neutrophil C1qR/receptor complex differs in some
way from C1qRP. Similarly, a recent report has provided evidence
supporting earlier suggestions that there is an additional SPA receptor
The Journal of Immunology
Acknowledgments
We thank Dr. J. R. Wright for pulmonary surfactant protein A, and Lisa
Salazar-Murphy for initial experiments investigating the glycosylation of
C1qRP.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
References
1. Bobak, D. A., M. M. Frank, and A. J. Tenner. 1988. C1q acts synergistically with
phorbol dibutyrate to activate CR1-mediated phagocytosis by human mononuclear phagocytes. Eur. J. Immunol. 18:2001.
2. Bobak, D. A., T. G. Gaither, M. M. Frank, and A. J. Tenner. 1987. Modulation
of FcR function by complement: subcomponent C1q enhances the phagocytosis
of IgG-opsonized targets by human monocytes and culture-derived macrophages.
J. Immunol. 138:1150.
3. Tenner, A. J., S. L. Robinson, J. Borchelt, and J. R. Wright. 1989. Human pulmonary surfactant protein (SP-A), a protein structurally homologous to C1q, can
enhance FcR- and CR1-mediated phagocytosis. J. Biol. Chem. 264:13923.
4. Tenner, A. J., S. L. Robinson, and R. A. B. Ezekowitz. 1995. Mannose binding
protein (MBP) enhances mononuclear phagocyte function via a receptor that
contains the 126,000 Mr component of the C1q receptor. Immunity 3:485.
5. Wright, J. R., R. E. Wager, S. Hawgood, L. Dobbs, and J. A. Clements. 1987.
Surfactant apoprotein Mr 5 26,000 –26,000 enhances uptake of liposomes by
type II cells. J. Biol. Chem. 262:2888.
6. Manz-Keinke, H., H. Plattner, and J. Schlepper-Schafer. 1992. Lung surfactant
protein A (SP-A) enhances serum-independent phagocytosis of bacteria by alveolar macrophages. Eur. J. Cell Biol. 57:95.
7. Bobak, D. A., R. G. Washburn, and M. M. Frank. 1988. C1q enhances the phagocytosis of Cryptococcus neoformans blastospores by human monocytes. J. Immunol. 141:592.
8. Tino, M. J., and J. R. Wright. 1996. Surfactant protein A stimulates phagocytosis
of specific pulmonary pathogens by alveolar macrophages. Am. J. Physiol. 270:
L677.
9. Schweinle, J. E., R. A. B. Ezekowitz, A. J. Tenner, M. Kuhlman, and
K. A. Joiner. 1989. Human mannose-binding protein activates the alternative
6
N. Jasinskiene, R. Rochford, S. Ruiz, and A. J. Tenner. Complement component
C1q selectively enhances phagocytosis, but not the production of proinflammatory
cytokines. Submitted for publication.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
complement pathway and enhances serum bactericidal activity on a mannose-rich
isolate of Salmonella. J. Clin. Invest. 84:1821.
Geertsma, M. F., W. L. Teeuw, P. H. Nibbering, and R. Van Furth. 1994. Pulmonary surfactant inhibits activation of human monocytes by recombinant interferon-g. Immunology 82:450.
Geertsma, M. F., P. H. Nibbering, H. P. Haagsman, M. R. Daha, and
R. Van Furth. 1994. Binding of surfactant protein A to C1q receptors mediates
phagocytosis of Staphylococcus aureus by monocytes. Am. J. Physiol. 267:L578.
Soell, M., E. Lett, F. Holveck, M. Scholler, D. Wachsmann, and J.-P. Klein.
1995. Activation of human monocytes by streptococcal rhamnose glucose polymers is mediated by CD14 antigen, and mannan binding protein inhibits TNF-a
release. J. Immunol. 154:851.
Guan, E., S. L. Robinson, E. B. Goodman, and A. J. Tenner. 1994. Cell surface
protein identified on phagocytic cells modulates the C1q-mediated enhancement
of phagocytosis. J. Immunol. 152:4005.
Nepomuceno, R. R., A. H. Henschen-Edman, W. H. Burgess, and A. J. Tenner.
1997. cDNA cloning and primary structure analysis of C1qRP, the human C1q/
MBL/SPA receptor that mediates enhanced phagocytosis in vitro. Immunity
6:119.
Guan, E., W. H. Burgess, S. L. Robinson, E. B. Goodman, K. J. McTigue, and
A. J. Tenner. 1991. Phagocytic cell molecules that bind the collagen-like region
of C1q: involvement in the C1q-mediated enhancement of phagocytosis. J. Biol.
Chem. 266:20345.
Tenner, A. J., P. H. Lesavre, and N. R. Cooper. 1981. Purification and radiolabeling of human C1q. J. Immunol. 127:648.
Young, K. R., J. L. Ambrus, Jr., A. Malbran, A. S. Fauci, and A. J. Tenner. 1991.
Complement subcomponent C1q stimulates immunoglobulin production by human B lymphocytes. J. Immunol. 146:3356.
Reid, K. B. M., D. M. Lowe, and R. R. Porter. 1972. Isolation and characterization of C1q, a subcomponent of the first component of complement, from human
and rabbit sera. Biochem. J. 130:749.
Lionetti, F. J., S. M. Hunt, and C. R. Valeri. 1980. Methods of Cell Separation.
Plenum Publishing Corp., New York, p. 141.
Bobak, D. A., M. M. Frank, and A. J. Tenner. 1986. Characterization of C1q
receptor expression on human phagocytic cells: effects of PDBu and FMLP.
J. Immunol. 136:4604.
Bohnsack, J. F., H. K. Kleinman, T. Takahashi, J. J. O’Shea, and E. J. Brown.
1985. Connective tissue proteins and phagocytic cell function laminin enhances
complement and Fc-mediated phagocytosis by cultured human macrophages.
J. Exp. Med. 161:912.
Elhammer, A. P., R. A. Poorman, E. Brown, L. L. Maggiora, J. G. Hoogerheide,
and F. J. Kezdy. 1993. The specificity of UDP-GalNAc:polypeptide N-acetylgalactosaminyltransferase as inferred from a database of in vivo substrates and
from the in vitro glycosylation of proteins and peptides. J. Biol. Chem. 268:
10029.
Guest, J. R., H. M. Lewis, L. D. Graham, L. C. Packman, and R. N. Perham.
1985. Genetic reconstruction and functional analysis of the repeating lipoyl domains in the pyruvate dehydrogenase multienzyme complex of Escherichia coli.
J. Mol. Biol. 185:743.
Ezekowitz, R. A. B., and J. A. Hoffmann. 1996. Innate immunity. Curr. Opin.
Immunol. 8:1.
Krieger, M., S. Acton, J. Ashkenas, A. Pearson, M. Penman, and D. Resnick.
1993. Molecular flypaper, host defense, and atherosclerosis. J. Biol. Chem.
7:4569.
Tenner, A. J. 1998. C1q receptors: regulating specific functions of phagocytic
cells. Immunobiology 199:250.
Tenner, A. J. 1997. C1q receptors: opportunities for selectively regulating protective and detrimental responses. Clinical Immunology Newsletter 17:173.
Nepomuceno, R. R., and A. J. Tenner. 1998. C1qRP, the C1q receptor that enhances phagocytosis, is detected specifically in human cells of myeloid lineage,
endothelial cells and platelets. J. Immunol. 160:1929.
Sanders, S., M. Jalkanen, S. O’Farrell, and M. Bernfield. 1989. Molecular cloning
of syndecan, an integral membrane proteoglycan. J. Cell Biol. 108:1547.
Kuan, S.-F., J. C. Byrd, C. Basbaum, and Y. S. Kim. 1989. Inhibition of mucin
glycosylation by aryl-N-acetyl-a-galactosaminides in human colon cancer cells.
J. Biol. Chem. 264:19271.
Wilson, I. B. H., Y. Gavel, and G. V. Heijne. 1991. Amino acid distribution
around O-linked glycosylation sites. Biochemistry 275:529.
Goodman, E. B., and A. J. Tenner. 1992. Signal transduction mechanisms of
C1q-mediated superoxide production: evidence for the involvement of temporally
distinct staurosporine insensitive and sensitive pathways. J. Immunol. 148:3920.
Ruiz, S., A. H. Henschen-Edman, and A. J. Tenner. 1995. Localization of the site
on the complement component C1q required for the stimulation of neutrophil
superoxide production. J. Biol. Chem. 270:30627.
Chroneos, Z. C., R. Abdolrasulnia, J. A. Whitsett, W. R. Rice, and
V. L. Shepherd. 1996. Purification of a cell-surface receptor for surfactant protein
A. J. Biol. Chem. 271:16375.
McIntosh, J. C., S. Mervin-Blake, E. Conner, and J. R. Wright. 1996. Surfactant
protein-A protects growing cells and reduces tumor necrosis factor-a activity
from lipopolysaccharide-stimulated macrophages. Am. J. Physiol. 271:L310.
Downloaded from http://www.jimmunol.org/ by guest on August 3, 2017
distinct from C1qRP that modulates phospholipid secretion (34). This
apparent multiplicity of receptors for these multidomain proteins may
permit the primitive orchestration of more than one type of function
in response to a variety of stimuli. It should be noted that Tino and
Wright (8) demonstrated that SPA stimulation of phagocytosis of specific pathogens is inhibited in monocytes adhered to surface-bound
C1q, but not in alveolar macrophages similarly treated. Thus, the differentiation state of the macrophage may dictate the surface distribution of C1qRP and/or may induce additional functional receptors for
these ligands.
The stimulation of phagocytic function may be particularly beneficial as a prophylactic treatment for individuals at risk for infection, such as individuals with genetic immunodeficiencies or infected with HIV, patients undergoing cancer chemotherapy, or
patients undergoing high risk surgery. Other agents that influence
the activation state of phagocytes often trigger the production of
proinflammatory cytokines that can directly cause unwanted side
effects, including, but not limited to, the up-regulation of HIV
production. In contrast, recent studies by Jasinskiene and coworkers6 demonstrate that C1q and SPA do not trigger the release of
proinflammatory cytokines by human monocytes. The ability to
stimulate clearance of cellular debris, pathogens, or immune complexes without triggering the production of cytokines (35)6 that in
some instances could be detrimental for the host, makes C1qRP an
ideal candidate for therapeutic manipulation. Knowledge of the
specific ligand-receptor interactions involved in mediating particular functions should facilitate selective modulation of desired responses (enhanced phagocytic capacity via C1qRP) without the
induction of proinflammatory cytokines or generation of toxic oxygen radicals.
3589