CR1 with Phagocytosis, Transfer, and Associations Receptor 1

Home , C3b, Complement receptor 1, Glycophorin A

Processing of C3b-Opsonized Immune Complexes Bound to Non-Complement Receptor 1 (CR1) Sites on Red Cells: Phagocytosis, Transfer, and Associations with This information is current as CR1 of September 24, 2021. Maria L. Craig, John N. Waitumbi and Ronald P. Taylor J Immunol 2005; 174:3059-3066; ; doi: 10.4049/jimmunol.174.5.3059 http://www.jimmunol.org/content/174/5/3059 Downloaded from

References This article cites 48 articles, 21 of which you can access for free at:

http://www.jimmunol.org/content/174/5/3059.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on September 24, 2021

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts 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 © 2005 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Processing of C3b-Opsonized Immune Complexes Bound to Non-Complement Receptor 1 (CR1) Sites on Red Cells: Phagocytosis, Transfer, and Associations with CR11

Maria L. Craig,* John N. Waitumbi,† and Ronald P. Taylor2*

Severe anemia is a lethal complication of Plasmodium falciparum malaria, particularly in children. Recent studies in children with severe P. falciparum anemia have demonstrated elevated levels of E-bound Abs, reduced E-associated complement receptor 1 (CR1) and decay-accelerating factor (DAF), and pronounced splenic enlargement, suggesting a mechanism for E loss involving Abs, complement, and phagocytosis. Motivated by these reports, we have developed an in vitro model in which human E with Abs and complement bound to CR1, DAF, or glycophorin A are incubated with model human macrophages (the THP-1 cell line). Previous work has demonstrated that immune complex (IC) substrates bound to E CR1, either by an Ab or via C3b, are Downloaded from transferred to macrophages with loss of CR1. In this study, we report that IC bound to DAF or glycophorin A by an Ab linkage are also transferred to macrophages. DAF is lost from the E during the transfer of DAF-bound IC, but the transfer of CR1-bound IC does not lead to a signiﬁcant loss of DAF. Using glycophorin A-bound IC, we observe competition between transfer of IC and phagocytosis of the E: a fraction (<15%) of the E was phagocytosed, while the remaining E were stripped of IC. We also examined the organization of CR1 and DAF in the presence of E-bound Ab/complement. We ﬁnd that CR1, but not DAF, colocalizes with

IgM mAb-C3b and IC-C3b substrates attached to glycophorin A. We observe that the binding of the IgM mAb-C3b to glycophorin http://www.jimmunol.org/ A induces a novel unclustering of CR1. The Journal of Immunology, 2005, 174: 3059–3066.

alaria is caused by a protozoan parasite, Plasmodium factor (DAF or CD55), and CD59 (membrane inhibitor of reactive falciparum, which in the pathogenic stage of its life lysis) (18). CR1, organized in clusters on primate E (19–21), ap- M cycle forms merozoites and invades E. Severe anemia pears to be a privileged site in comparison with other E proteins in children Ͻ5 years of age is one of the most serious complica- such as glycophorin A, the Rho (D) Ag, or band 3, in that the tions of P. falciparum malaria, and this and other complications binding of IgG and/or IC to CR1 leads to little in vitro erythroph- cause the death of Ͼ1 million children each year in sub-Saharan agocytosis (21, 22). Similarly, in vivo, binding of IC to E via CR1

Africa (1). Only a small fraction of a child’s E becomes infected does not induce E destruction (23, 24), while formation of IC at by guest on September 24, 2021 (1, 2), and therefore, direct lysis of the infected E cannot explain other sites promotes E lysis and/or clearance (11). In the ﬁrst case, the severe anemia (3). The presence, in the circulation of malaria- E are spared because IC bound to CR1, either via the activated infected animals and patients (4–9), of anti-malaria Abs, malarial large cleavage fragment of the third component of complement Ags, immune complexes (IC),3 and complement degradation prod- (C3b) or through the action of an anti-CR1 Ab, are taken up by ucts, suggests that E loss may occur by a mechanism involving the macrophages in the liver and spleen in a process known as the interaction of Abs, IC, and complement with the E. transfer reaction (25). Studies in both in vitro and in vivo models Binding of these substrates to E in vivo may lead to one of delineate a mechanism for this transfer in which engagement of several possible fates for the E: hemolysis by complement attack CR1-bound IC by macrophage Fc receptors leads to cleavage of (10), erythrophagocytosis by ﬁxed tissue macrophages in the liver CR1, resulting in loss of CR1 and the associated IC from the E (22, and spleen (10–15), or removal of E-bound Abs or IC by transfer 25–28). Thus, the transfer reaction is likely to explain, at least in to macrophages (16, 17). Under normal circumstances, E are pro- part, the reduced CR1 levels in patients with a variety of diseases tected from hemolysis by the presence of complement control pro- involving circulating IC (29), particularly autoimmune diseases teins, including complement receptor 1 (CR1), decay-accelerating (29, 30) and AIDS (31). In the second case, E that have Abs and complement bound to other sites are removed and destroyed by

*Department of Biochemistry and Molecular Genetics, University of Virginia, Char- fixed tissue macrophages (10–15). lottesville, VA 22908; and †U.S. Army Medical Research Unit and Kenya Medical Recently, studies in children with severe malarial anemia have Research Institute, Nairobi, Kenya suggested that E-bound Abs or IC may play important roles in the Received for publication September 16, 2004. Accepted for publication December pathogenesis of malarial anemia (1, 9). As in autoimmune diseases 6, 2004. and in AIDS, CR1 levels on the E of these children are reduced; in The costs of publication of this article were defrayed in part by the payment of page addition, E levels of a second complement-regulatory protein, charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. DAF, are also reduced. Other evidence suggestive of involvement 1 This work was supported by National Institutes of Health Grant AR43307 and the of E-bound Abs or IC in the pathogenesis of malarial anemia in- Ellison Medical Foundation. cludes increased levels of circulating IC and E-bound IgG, in- 2 Address correspondence and reprint requests to Dr. Ronald P. Taylor, Department creased susceptibility of the E to in vitro phagocytosis, and splenic of Biochemistry and Molecular Genetics, P.O. Box 800733, University of Virginia, enlargement (1, 9). Abs to malarial proteins displayed on the sur- Charlottesville, VA 22908. E-mail address: [email protected] face of infected E have been identified (32, 33), but the mode of 3 Abbreviations used in this paper: IC, immune complex; Al, Alexa; CMFDA, 5-chlo- romethylfluorescein diacetate; CR, complement receptor; DAF, decay-accelerating Ab binding to uninfected E is not well understood. Glycophorin A, factor; FL, fluorescence. B, and C, being among the most abundant E surface proteins, are

Copyright © 2005 by The American Association of Immunologists, Inc. 0022-1767/05/$02.00 3060 PROCESSING OF IC BOUND TO NON-CR1 SITES ON RED CELLS some of the receptors used by P. falciparum during entry into E 10 min at 37°C, washed, and then incubated for 10 min at 37°C with 9 (34–37). These intriguing findings have motivated us to initiate in ␮g/ml unlabeled, Cy5-labeled, or Texas Red-labeled rabbit IgG anti-mouse vitro studies aimed at understanding what happens when Abs and IgM. E were then allowed to activate complement by treatment with serum, as above. Control samples were prepared in which serum contained 10 mM complement are bound at non-CR1 sites on the E, in particular EDTA to block complement activation. DAF and glycophorin A. Our study has yielded several novel results regarding the transfer reaction and the organization of CR1 on the E surface. Notably, Transfer/phagocytosis reactions we find that DAF- or glycophorin A-bound IC (Fig. 1A) consisting E(1ϫ 108/ml in BSA-PBS) and THP-1 cells (2 ϫ 107 cells/ml in culture of multiple Abs, with or without complement, are transferred from medium, but lacking FBS and gentamicin) were combined at varying ra- E to model macrophages. The transfer of IC bound to DAF leads tios, mixed, and pelleted by centrifugation (1 min, 2000 ϫ g, 4°C). After to the loss of DAF from the E, and this loss of DAF appears to incubating at 37°C, without shaking, for 30 min or 1 h, cells were respectively analyzed by flow cytometry or fluorescence microscopy. In some occur independently of the loss of CR1. When the IC are bound to flow cytometry experiments (Fig. 2), E were separated from THP-1 cells by glycophorin A, we observe, in agreement with Reinagel et al. (21), layering over 60% Percoll (Amersham Biosciences) in PBS, followed by that a fraction of the E is phagocytosed; however, we find that the centrifugation (4°C, 1250 ϫ g, 20 min). In other experiments (Fig. 3), the ϫ E that escape phagocytosis are almost completely stripped of IC by cell pellets were resuspended in BSA-PBS to a THP-1 concentration of 1 107 cells/ml, and, to selectively stain the THP-1 cells, PE anti-CD45 mAb the macrophages. Fluorescence microscopy of E with IC bound to was bound (1/20 final dilution) by incubating for 10 min at 37°C. In certain glycophorin A shows that CR1, but not DAF, colocalizes with the cases (Fig. 3D), E that had not been internalized were removed by hypo- IC when complement is activated and C3b deposits on the E-bound tonic lysis in 1.5 ml of 0.1% (w/v) NaCl for 1 min (21). For fluorescence

IC during opsonization. Moreover, when IgM Abs are used to microscopy (Fig. 4), the reactions were stopped by placing the tubes on ice Downloaded from target glycophorin A and promote C3b deposition, we observe an and resuspending, using repetitive pipetting, in 1 ml of cold BSA-PBS containing 2 mg/ml rabbit or human IgG. As a negative control (Fig. 4A), unusual binding pattern characterized by rearrangement and un- E and THP-1 cells were added directly to 1 ml of ice-cold BSA-PBS clustering of CR1. containing 2 mg/ml rabbit or human IgG. The cells were pelleted (2 min, 2000 ϫ g, 4°C) and ﬁxed by suspending in 500 ␮l of 1% paraformaldehyde Materials and Methods in PBS.

Antibodies http://www.jimmunol.org/ The anti-CR1 mAbs (38) 1B4, HB8592, and 9H3, and the anti-C3b Abs Probing of naive and opsonized E with ﬂuorescently labeled Abs 7C12, 9F9, and 10H5 (39, 40) were produced at the University of Virginia and streptavidin Lymphocyte Culture Center and were labeled with Alexa (Al) 488 dye (kit ϳ ϫ 7 from Molecular Probes) or EZ-Link N-hydroxysuccimide-long chain-bi- Eat 5 10 cells/ml were separately incubated with one or more probes at the following concentrations for 10 min at 37°C and then washed with otin (Pierce). Anti-CD55 mAbs IA10 (biotinylated; BD Pharmingen) and ␮ BRIC216 (BioSource International) have been described previously (41). BSA-PBS: 70 g/ml Al488-rabbit anti-mouse IgG; a 2.5-fold dilution of biotinylated IA10 (anti-DAF); 50 ␮g/ml Al488-streptavidin; 40 ␮g/ml bi- Mouse IgM mAb to glycophorin A (A63-B/C2) and PE-labeled mouse ␮ IgG1 mAb to leukocyte common Ag (CD45, mAb HI30) were obtained otinylated 9H3 (anti-CR1); and 50 g/ml Al488-7C12 (anti-C3b). For the from Ancell and Caltag Laboratories, respectively. Commercially available ﬂuorescence microscopy studies (Figs. 5 and 6, below) of IC/receptor organization, E (9 ϫ 106 cells/ml) were incubated for 10 min at 37°C or 1 h secondary Abs were as follows: Al488- and Al594-labeled rabbit anti- by guest on September 24, 2021 at 4°C with the following IgG mAb probes, all at a concentration of 9 mouse IgG, Al594-labeled goat anti-rabbit IgG, and Al488- and Al594- ␮ labeled streptavidin were purchased from Molecular Probes; unlabeled and g/ml, unless noted (targeted molecule in parentheses): Al488-7C12 Cy5- and Texas Red-labeled rabbit IgG anti-mouse IgM were purchased (C3b); Al488-9F9 (C3b); Al488-10H5 (C3b); biotinylated 9H3 (CR1); biotinylated HB8592 (CR1); biotinylated IA10 (50-fold dilution, DAF); and from Jackson ImmunoResearch Laboratories; and unlabeled rabbit IgG anti- ␮ mouse IgM was also obtained from Cortex Pharmaceuticals. Al594-goat anti-rabbit IgG (50 g/ml, polyclonal). Biotinylated IA10 (DAF) and Al594-streptavidin, respectively, were at a 50-fold dilution and Cells 50 ␮g/ml. For all probes except anti-mouse IgG, excess mouse IgG (Ն2 mg/ml) was added to prevent binding of the probes to anti-mouse IgG/IgM E were obtained from blood samples of normal human donors and were already incorporated into IC on the E. washed thoroughly with BSA-PBS (1% BSA in PBS) before use. THP-1 cells were cultured and treated with all-trans retinoic acid, as described previously (17). Fluorescence microscopy

5 6 Cell labeling Cells (10 -10 per slide) were placed on poly(L-lysine)-coated slides (Erie Scientific) either by centrifugation at 1000 rpm for 3 min in a Cytospin 4 ϫ 9 ␮ E(1 10 cells/ml) were mixed with 6 M PKH67 dye in Diluent C (both centrifuge (Thermo Shandon) or by adherence of 10 ␮l of a suspension. from Sigma-Aldrich) for 5 min at room temperature. An equal volume of Prolong antifade reagent (Molecular Probes) was used to maintain fluo- BSA-PBS was added to stop the reaction and, after centrifugation and rescent signals. Cells were covered with glass coverslips (No. 1, 18 ϫ 18 removal of the supernatant, E were washed twice more with BSA-PBS. mm) and sealed with nail varnish, and slides were examined using a BX40 ϫ 8 ␮ Alternatively, E (5 10 cells/ml) were combined with 24 M 5-chlo- fluorescent microscope (Olympus), equipped with a Magnafire digital romethylfluorescein diacetate (CMFDA; Molecular Probes) in PBS for 15 camera. min at 37°C, and, after centrifugation and removal of the supernatant, were diluted to 2 ϫ 107 cells/ml in PBS, held at 37°C for 30 min, and washed thoroughly with BSA-PBS. Flow cytometry Binding of IC and complement to DAF and CR1 E, THP-1 cells, and E/THP-1 mixtures were analyzed using CellQuest software on a FACSCalibur flow cytometer (BD Biosciences). Fluores- E(5ϫ 107 cells/ml) were incubated for 10 min at 37°C with anti-DAF (50 cence intensities were converted to molecules of equivalent soluble fluo- ␮g/ml BRIC 216 or a 1.5-fold dilution of biotinylated IA10) or with anti- rochrome using calibrated standard beads (Spherotech). CR1 (90 ␮g/ml 9H3 or 1B4) mAbs, washed, and then further opsonized for To determine the levels of the anti-glycophorin A substrates (FL4 pos- 10 min at 37°C with 0.1 mg/ml unlabeled or Al633-labeled rabbit anti- itive) bound to E and THP-1 cells in the transfer/phagocytosis experiments, mouse IgG. E were washed again, and, in some cases, cell pellets were then we separated signals on E from those on THP-1 cells using a gating scheme suspended at 2 ϫ 108 cells/ml in neat human autologous serum, held for 10 based upon fluorescence (FL)1 (for PKH67-labeled E), FL2 (for THP-1 min at 37°C, and thoroughly washed. cells, labeled with PE anti-CD45), and side scatter signals. Inclusion and Binding of IgM Abs, IC, and complement to glycophorin A exclusion gates were used simultaneously to insure that THP-1 cells that had internalized E did not contribute to the FL4 signal of the E. However, E(1ϫ 107 cells/ml) were mixed with subsaturating amounts (4.5 ␮g/ml) our gating scheme did not exclude THP-1 cells with internalized E from of either unlabeled or an Al647-labeled IgM anti-glycophorin A mAb for contributing to the total signal of the THP-1 cells. The Journal of Immunology 3061

Results DAF-bound IC on E are transferred to THP-1 model macrophages To examine possible mechanisms for the reduction in E-associated CR1 and DAF in malaria (1), we created model IC (Fig. 1) bound to either CR1 or DAF by first binding anti-CR1 or anti-DAF mAbs to E and subsequently adding Al633-labeled rabbit anti-mouse IgG. In some cases, these IC were then allowed to activate complement to induce C3b deposition. We incubated E containing these IC either with retinoic acid-treated THP-1 cells or in buffer alone, and, after separation of the E from the macrophages, used Al488-labeled Abs to determine levels of mouse IgG, CR1, or DAF on the E. Strikingly, the processing of DAF IC closely resembles processing of CR1 IC (Fig. 2). In particular, two key results are evident: 1) the DAF-bound IC, like the CR1-bound IC, are removed from the E in the presence of the macrophages, and 2) this loss of DAF- and CR1-bound IC from the E is accompanied by respective loss Downloaded from of the IC-bearing receptor (Fig. 2A). After binding of C3b, we observed similar trends, except that the loss of DAF from the E was not as pronounced (Fig. 2B). In addition, the majority of the C3b was also lost from the E (Fig. 2B). FIGURE 2. Transfer of IC bound to CR1 or DAF occurs with loss of the engaged receptor. IC (A) or C3b-IC (B) consisting of anti-CR1 (9H3 or Glycophorin A-bound IC are transferred to THP-1 model 1B4) or anti-DAF (BRIC 216 or biotinylated IA10) mAbs and Al633- macrophages labeled rabbit anti-mouse IgG were lost from the E during a 30-min trans- http://www.jimmunol.org/ fer reaction with THP-1 cells at 37°C. Levels of the Al633 anti-mouse IgG We also examined the handling of Abs and IC bound at an addi- and of the following fluorescently labeled probes of anti-CR1/DAF, CR1, tional E site, glycophorin A. We chose this site for several reasons: DAF, or C3b, respectively, were measured by flow cytometry: Al488-rab- first, previous studies of the phagocytosis of E with glycophorin bit anti-mouse IgG (which was able to bind in the presence of the Al633- A-bound IC created a framework for this work (21); second, gly- rabbit anti-mouse IgG), biotinylated 1B4 or biotinylated 9H3, biotinylated cophorin A is abundant on the E; and, third, glycophorin A has IA10 or FITC-BRIC 216, or Al488 7C12, plus Al488-SA when necessary Ϯ been implicated as a site of P. falciparum entry into E (36). The to measure cell-bound biotin. The mean SD are shown for three (A)or two (B) independent experiments. abundance of glycophorin A on E necessitated the use of subsatu-

rating amounts of anti-glycophorin A mAb to minimize cross-link- by guest on September 24, 2021 ing of the E. We compared the handling of four different types of substrates bound to glycophorin A, as illustrated in the Fig. 1 schematic (figure derived from Ref. 42). These substrates were prepared with an IgM anti-glycophorin A mAb, rabbit IgG anti-mouse IgM Abs, and serum as a complement source (for C3b opsonization). Levels of binding of the IgM anti-glycophorin A mAb to E were similar for three of the four substrates; however, binding under similar conditions was 80% lower for the IC-C3b substrate, possibly due to complement-induced disruption of the IC (43). To examine both phagocytosis and transfer simultaneously, we designed a three-color system in which E, THP-1 cells, and substrates were labeled with, consecutively, PKH67 (green), PE anti- CD45, and either Al647-IgM anti-glycophorin A mAb or Cy5- rabbit anti-mouse IgM. The phagocytosis of E by THP-1 cells produced a population of green/PE double-positive cells that could be measured by flow cytometry, and therefore, we measured phagocytosis by enumerating these green/PE double-positive cells after an E lysis step that selectively removed E that had not been internalized. Transfer was evaluated using flow cytometry to fol- low loss from the E and uptake by the THP-1 cells of the Al647- IgM anti-glycophorin A or Cy5-rabbit anti-mouse IgM. FIGURE 1. Schematic representation of model IC substrates bound to E We found that both phagocytosis and transfer occurred, with surface proteins CR1, DAF, or glycophorin A. A, Four different types of phagocytosis of up to 15% of E. E that escaped phagocytosis were E-bound substrates were studied: 1) mAbs alone, 2) complement-opso- stripped of most of the bound IC and IC-C3b (Fig. 3A), and uptake nized mAbs, 3) IC consisting of mAbs and rabbit IgG secondary Abs, and of both substrates by the THP-1 cells was apparent (Fig. 3B). For 4) complement-opsonized IC. Anti-CR1 and anti-DAF mAbs were of the IgG isotype, and the anti-glycophorin A mAbs were of the IgM isotype. reasons we do not understand, more loss from the E of the non-IC Substrate 2 requires an IgM mAb, because IgG does not efficiently capture mAb and mAb-C3b substrates was observed in a mock control C3b unless it is part of a large IC. B, Two types of processes were exam- (lacking THP-1 cells) than in the presence of THP-1 cells (data not ined: 1) transfer of the substrates from E to model macrophages, or 2) shown), and there was little uptake of these two substrates by the phagocytosis of the substrate-bearing E by the macrophages. THP-1 cells (Fig. 3B). 3062 PROCESSING OF IC BOUND TO NON-CR1 SITES ON RED CELLS

FIGURE 3. The binding of IC to glycophorin A may re- sult in either transfer of the IC to or phagocytosis of the E by THP-1 cells. IC consisting of IgM anti-glycophorin mAbs and rabbit anti-mouse IgM Ϯ C3b are lost from the E in a transfer reaction analogous to that shown in Fig. 2. Loss from E(A) and uptake by THP-1 cells (B) of the Al647-labeled anti-glycophorin mAb portion of the IC (detected in the FL4 channel by ﬂow cytometry) are shown as a function of the ratio of E:THP-1 cells. The mean Ϯ SD for two independent experiments is shown, and these results agree qualitatively with two additional experiments in which a Cy5-labeled rabbit anti-mouse IgM was used (data not shown). C, Flow cytometry dot plots of mixtures of PKH67-labeled E (FL1 positive) and PE anti-CD45-labeled THP-1 cells (FL2 positive) demonstrate that IC-coated E (bottom), but not naive E (top) are taken up by THP-1 cells to yield double-positive events Downloaded from (3.8%). D, The percentage of THP-1 cells taking up E (% phagocytosis) is shown as a function of the E substrates described in Fig. 1. For both C and D, the THP-1 cell to E ratio was 5:1. http://www.jimmunol.org/

We determined the percentage of the THP-1 cells internalizing green tracking dye. The E were then used in the transfer/phago- E (percentage of phagocytosis) based on the ratio of green/PE cytosis reaction, as described above. Before the reaction (Fig. 4A), double-positive cells to total PE-positive cells (Fig. 3C), and these the IC were observed as clusters of red ﬂuorescence on the green- values are reported in Fig. 3D. The IC substrate consistently in- dyed E. After the reaction (Fig. 4B), the E were devoid of the duced the most phagocytosis, with the IgM mAb-C3b and IC-C3b IC-associated red color, which had been transferred to the THP-1 by guest on September 24, 2021 substrates also inducing more phagocytosis than the IgM mAbs cells. After an E lysis step, instances of phagocytosis were ob- alone (Fig. 3D). served (Fig. 4C). Fig. 4C clearly demonstrates the dual transfer/ Fluorescence microscopy studies further conﬁrmed the dual phagocytosis phenomenon, showing an E that has been phagocy- transfer/phagocytosis phenomenon. IC labeled with Al647 anti- tosed by a THP-1 cell (arrow) adjacent to a THP-1 cell that has glycophorin A (red) were bound to E labeled with CMFDA, a taken up IC in the transfer reaction (asterisk).

FIGURE 4. Fluorescence microscopy images illus- trating the dual transfer/phagocytosis phenomenon. IC (red) consisting of unlabeled or Al647-labeled mouse IgM anti-glycophorin A mAb and, respectively, Texas Red-labeled or unlabeled rabbit anti-mouse IgM were bound to E labeled with CMFDA (green). Transfer/ Phagocytosis reactions were performed for1hat37°C. Images are shown in sets of four: a, phase; b, green; c, red; and d, red ϩ green composite. A, Results of a control in which the transfer/phagocytosis reaction is blocked by the addition of excess mouse IgG and in- cubation on ice demonstrate the red IC, organized in clusters, on the E before the reaction. B, After the reaction, the E have been stripped of IC, which are now associated with the THP-1 cells. C, After an E lysis step, only E that have been phagocytosed by THP-1 cells remain (see arrow). A THP-1 cell that has phagocytosed a green E is located next to one that has taken Results shown are .(ء up IC in the transfer reaction (see representative of two independent experiments. The Journal of Immunology 3063

FIGURE 5. CR1 organization on E reﬂects the positioning of glycophorin A-bound C3b. E with substrates (labels to the left of the ﬁgure) bound to glycophorin A were probed at 4°C for CR1 (red) using a mixture of biotinylated mAbs (9H3 and HB8592) and Al594-SA and for C3b (green) using a mixture of Al488-labeled mAbs (7C12, 9F9, and 10H5). When an IgM mAb is bound (top row), CR1 retains its normal clustered, punctate distribution. However, when the IgM-coated E are incubated with serum to deposit C3b (second row), CR1 reorganizes, taking on a distribution like that of the C3b. When the cells are opsonized with

IC and C3b (third row), CR1 and C3b colocalize in Downloaded from punctate clusters. Enlargement (fourth row) of an area from the third row clearly demonstrates the colocalization (note yellow colors in zoom overlay) between CR1 and C3b. Results are representative of two to four similar experiments. http://www.jimmunol.org/ by guest on September 24, 2021

FIGURE 6. DAF organization on E is independent of the positioning of glycophorin A-bound C3b. Exper- imental procedures were the same as Fig. 5, except that E were probed for DAF (red), instead of CR1, using biotinylated mAbs (IA10) and Al594-SA. DAF is organized in punctate clusters when the IgM mAb is bound (top row). Unlike CR1, DAF does not reorganize when C3b is deposited on the IgM-coated E (second row). Furthermore, DAF clusters do not colocalize with C3b clusters when the substrate is IC-C3b (third and fourth rows). Note the preponderance of independent green and red spots. Results are representative of two to four similar experiments. 3064 PROCESSING OF IC BOUND TO NON-CR1 SITES ON RED CELLS

CR1, but not DAF, colocalizes with mAb-C3b and IC-C3b and B; Fig. 4, A and B), which is structurally different (46), are also bound to glycophorin A on E transferred to macrophages. Transfer can occur in the absence of ␥ To determine whether CR1 and DAF interact with Ab/complement complement (Fig. 2A), suggesting an Fc R-based mechanism like substrates bound to non-CR1, non-DAF sites on E, we used fluo- that described for CR1 and CR2. The DAF transfer reaction further rescence microscopy to examine the same four anti-glycophorin A resembles transfer of CR1 and CR2 because DAF is lost from E substrates studied above. In this experiment, unlabeled substrates (Fig. 2) when DAF-bound mAb are removed. We were unable to were bound to E and followed by fluorescently labeled mAb investigate glycophorin A loss due to its abundance on E; use of probes specific for CR1, DAF, or C3b (Figs. 5 and 6). Interest- saturating amounts of anti-glycophorin A caused E cross-linking, ingly, we found that CR1 signals increased when the complement- and we could not measure a reduction in glycophorin A when less opsonized substrates, but not the noncomplement-opsonized sub- Ab was used. Thus, we cannot address all details of the mechanism strates, bound to E (2- to 4-fold for mAb-C3b and 2- to 3-fold for of release of glycophorin A-bound IC from E. IC-C3b based on molecules of equivalent soluble fluorochrome We tested two hypotheses that could explain reduced levels of E Ϯ values; data not shown). Results of probing for 10 min at 37°C DAF in children with malarial anemia: 1) DAF-bound IC ( com- agreed well with results of probing for1hat4°C, indicating that plement) may be transferred to macrophages in a process resem- reorganization of molecules on the E surface was not due to probe bling the CR1 transfer reaction and therefore involving loss of binding. In addition, we confirmed that C3b deposited on the E DAF from E; and 2) an interaction between CR1 (or CR1-bound colocalized with the IgM anti-glycophorin A mAb (data not C3b) and DAF may promote cotransfer of DAF and CR1. Al- shown). though we could find no reports of anti-DAF Abs in malaria, our We addressed the issue of association between CR1 or DAF and results (Fig. 2A) are consistent with hypothesis 1, showing that Downloaded from the glycophorin A-bound C3b. CR1 is organized in clusters on E binding of IC to DAF leads to loss of IC and DAF from E. Thus, (19–21), but glycophorin A is distributed more evenly over the E we speculate that Abs may bind to DAF in malaria, either directly surface (21). The IgM anti-glycophorin mAb used in these studies or perhaps by way of DAF-bound P. falciparum proteins. In fact, binds evenly across the entire E surface (data not shown) without several noncomplement proteins can bind to DAF during invasion altering the clustered organization of CR1 (Fig. 5, top row). In by other pathogens (47). We found little evidence to support hy- contrast, binding of C3b to E opsonized with this mAb induces loss pothesis 2. In the absence of C3b, CR1 or DAF loss was restricted http://www.jimmunol.org/ of CR1 clustering (Fig. 5, second row). Conversion of these mAbs to the IC-bound receptor (Fig. 2A). However, in the presence of to IC by the addition of IgG anti-IgM mAbs induces a clustering C3b, a small loss of CR1 or DAF was observed when IC were and condensation of the mAbs (Fig. 4A), and, when C3b is depos- bound, respectively, to DAF or CR1 (Fig. 2B), and thus, we cannot ited on these IC clusters, we observe substantial colocalization of exclude the second hypothesis on the basis of these data. Finally, CR1 clusters with the deposited C3b clusters (Fig. 5, third and C3b binding to the receptor-specific IC leads to less DAF loss fourth rows). A comparable experiment using the anti-glycophorin (from 49 to 32%; Fig. 2), and this might be due to partial blockade opsonization paradigm, but in which DAF is examined (Fig. 6), of the Fc regions of the anti-DAF mAb by C3b (28). demonstrates that the distribution of DAF on the E is not affected Transfer vs phagocytosis by guest on September 24, 2021 by these opsonization procedures and that DAF does not colocalize with C3b. Our results indicate that phagocytosis and transfer are competing processes; the former promotes E destruction, while the latter may Discussion protect E. Under our conditions, transfer is the preferred process. However, the balance between transfer vs phagocytosis may de- General approach pend upon a number of factors, including the number of Fc regions Modifications of the E surface, including Ab binding, C3 fragment attached to the E surface and the organization of IC-bound recep- deposition, and reductions in CR1 and DAF, have been observed tors. Reinagel et al. (21) proposed that the clustered organization in malaria-infected children with severe anemia (1, 7, 9). When of CR1 protects E from phagocytosis when IC are bound to CR1; taken in combination with reports of malarial proteins on the E however, the anti-glycophorin substrate that induced the most surface (44) and of low parasitemia in monkey models of severe phagocytosis (Fig. 3D) was organized in clusters (Fig. 4A). There- malarial anemia (2), this evidence suggests that Abs and comple- fore, the number and spacing of clusters on each E or the number ment may bind to malarial proteins adhered to the surface of both of IgG Fc moieties per cluster may be more important factors infected and uninfected E, possibly causing loss of CR1 and DAF influencing the balance between transfer and phagocytosis. The from E, and, ultimately, inducing E destruction. With the goal of IgM mAb-C3b substrate contains no IgG Fc, and little transfer of understanding possible mechanisms for E loss, we developed an in this substrate was observed; however, this substrate induced sig- vitro model in which Abs and complement are bound to select E nificant E phagocytosis. Yan et al. (15) have indeed demonstrated sites. Our investigations reveal novel insights into the E-macroph- in mice that, via CR3, macrophages promote clearance of IgM- age interactions and E surface rearrangements resulting from bind- opsonized E containing deposited C3b and iC3b. An understanding ing of complex substrates to E. of the determining factors in the decision between phagocytosis vs transfer could have implications in future therapeutic approaches Transfer of IC substrates from E to macrophages for immune-mediated anemias. C3b-opsonized IC bound to E CR1 and B cell CR2 are transferred to macrophages in a process known as the transfer reaction. This Association of CR1 with C3b-opsonized substrates bound to process does not require complement, but occurs by a mechanism glycophorin A in which macrophage Fc␥R bind IC-associated Fc regions to form We report that the organization on the E surface of CR1, but not E (or B cell)-macrophage linkages that are released by the removal of DAF, can change in response to binding of C3b-opsonized Abs of CR1 (CR2) and bound IC from the cells, possibly by proteolytic and IC to a non-CR1 site, glycophorin A. E CR1 is normally or- cleavage of the receptors (17, 26, 28, 45). Our investigation reveals ganized in clusters (19–21), characterized by a punctate fluores- that IC bound at two other E sites, DAF (Fig. 2), which is struc- cence when visualized using Ab probes. Glycophorin A, however, turally similar to CR1 and CR2 (18), and glycophorin A (Fig. 3, A is distributed more evenly over the E surface (21). We demonstrate The Journal of Immunology 3065 in this study that C3b opsonization of IgM mAbs bound to glyco- 6. Adam, C., M. Geniteau, M. Gougerot-Pocidalo, P. Verroust, J. Lebras, C. Gibert, phorin A induces marked changes in CR1 organization: CR1 be- and L. Morel-Maroger. 1981. Cryoglobulins, circulating immune complexes, and complement activation in cerebral malaria. Infect. Immun. 31:530. comes unclustered, taking on the distribution of IgM anti- 7. Abdalla, S., and D. J. Weatherall. 1982. The direct antiglobulin test in P. falci- glycophorin A and C3b (Fig. 5). Natural instances of CR1 parum malaria. Br. J. Haematol. 51:415. unclustering may occur when autoantibodies bind to E in autoim- 8. Kawamoto, Y., K. Kojima, Y. Hitsumoto, H. Okada, M. Holers, and A. Miyama. 1997. The serum resistance of malaria-infected erythrocytes. Immunology 91:7. mune hemolytic anemias (10) or when Abs bind to foreign proteins 9. Stoute, J. A., A. O. Odindo, B. O. Owuor, E. K. Mibei, M. O. Opollo, and adhered to E during infections by agents such as P. falciparum (32, J. N. Waitumbi. 2003. Loss of red blood cell-complement regulatory proteins and 33). In addition to colocalization of CR1 with C3b during the increased levels of circulating immune complexes are associated with severe malarial anemia. J. Infect. Dis. 187:522. unclustering events, we also observe CR1-C3b colocalization for 10. Gehrs, B. C., and R. C. Friedberg. 2002. Autoimmune hemolytic anemia. the clustered IC-C3b substrate (Fig. 5). Am. J. Hematol. 69:258. Based on these observations, we propose that CR1 can diffuse 11. Schreiber, A. D., and M. M. Frank. 1972. Role of antibody and complement in the immune clearance and destruction of erythrocytes. I. In vivo effects of IgG through the E membrane, interacting with deposited C3b. Previous and IgM complement-fixing sites. J. Clin. Invest. 51:575. studies indicate that CR1 is mobile on neutrophils (48) and lym- 12. Frank, M. M. 1989. The role of macrophages in bloodstream clearance. In Human phocytes (49), but the mobility of CR1 in the E membrane has not, Monocytes. M. Zembala and G. L. Asherson, eds. Academic, London, p. 337. 13. Malaise, M. G., P. Franchimont, and P. R. Mahieu. 1989. The ability of normal to our knowledge, been reported. If CR1 can diffuse through the E human monocytes to phagocytose IgG-coated red blood cells is related to the membrane, then E must house a mechanism for maintaining CR1 number of accessible galactosyl and mannosyl residues in the Fc domain of the clusters. This mechanism may involve self-association of CR1; anti-red blood cell IgG antibody molecules. J. Immunol. Methods 119:231. 14. Indik, Z. K., J. G. Park, S. Hunter, and A. D. Schreiber. 1995. The molecular Fearon (50) previously reported that purified CR1 was organized dissection of FcgRIII receptor mediated phagocytosis. Blood 86:4389. in complexes of four to six molecules. In addition, Ghiran et al. 15. Yan, J., V. Vetvicka, Y. Xia, M. Hanikyrova, T. N. Mayadas, and G. D. Ross. Downloaded from (51) found that an interaction between the CR1 cytoplasmic do- 2000. Critical role of Kupffer cell CR3 (CD11b/CD18) in the clearance of IgM- opsonized erythrocytes or soluble ␤-glucan. Immunopharmacology 46:39. main and the fifth PDZ domain (protein-protein interaction module 16. Lindorfer, M. A., C. S. Hahn, P. L. Foley, and R. P. Taylor. 2001. Heteropoly- that binds to C-terminal motifs in the cytoplasmic tails of trans- mer-mediated clearance of immune complexes via erythrocyte CR1: mechanisms membrane proteins) of Fas-associated phosphatase-1 may contrib- and applications. Immunol. Rev. 183:10. 17. Craig, M. L., A. J. Bankovich, and R. P. Taylor. 2002. Visualization of the ute to CR1 clustering on E. In any case, interactions that maintain transfer reaction: tracking immune complexes from erythrocyte complement re-

CR1 clusters are apparently weaker than those between CR1 and ceptor 1 to macrophages. Clin. Immunol. 105:36. http://www.jimmunol.org/ multimeric C3b, because interaction with the IC-C3b promotes 18. Morgan, B. P., and C. L. Harris. 1999. Complement Regulatory Proteins. Aca- demic, San Diego. disruption of the clusters. 19. Paccaud, J. P., J. L. Carpentier, and J. A. Schifferli. 1988. Direct evidence of the We found no evidence for DAF colocalization with the com- clustered nature of complement receptors type 1 on the erythrocyte membrane. plexed CR1-C3b-mAb or CR1-C3b-IC (Fig. 6). This is consistent J. Immunol. 141:3889. 20. Chevalier, J., and M. D. Kazatchkine. 1989. Distribution in clusters of comple- with the lack of cooperation between CR1 and DAF in the transfer ment receptor type one (CR1) on human erythrocytes. J. Immunol. 142:2031. reactions (Fig. 2B), and both results are perhaps expected because 21. Reinagel, M. L., M. Gezen, P. J. Ferguson, E. N. Martin, and R. P. Taylor. 1997. DAF binds to C3 convertases rather than to C3b. Thus, in this The primate erythrocyte complement receptor (CR1) as a privileged site: binding of immunoglobulin G to erythrocyte CR1 does not target erythrocytes for phago- study, we cannot provide a mechanism for DAF loss in the chil- cytosis. Blood 89:1068. by guest on September 24, 2021 dren with severe malarial anemia. 22. Kuhn, S. E., A. Nardin, P. E. Klebba, and R. P. Taylor. 1998. Escherichia coli Based on observations in children with severe P. falciparum bound to the primate erythrocyte complement receptor via bispecific monoclonal antibodies are transferred to and phagocytosed by human monocytes in an in vitro anemia, we have developed an in vitro model to study IC- and model. J. Immunol. 98:5088. complement-based mechanisms of E destruction. We have dem- 23. Cornacoff, J. B., L. A. Hebert, D. J. Birmingham, and F. I. Waxman. 1984. onstrated that IC substrates bound to non-CR1 sites on E, specif- Factors influencing the binding of large immune complexes to the primate erythrocyte CR1 receptor 1. Clin. Immunol. Immunopathol. 30:255. ically DAF and glycophorin A, are transferred from E to macro- 24. Hebert, L. A., and F. G. Cosio. 1987. The erythrocyte-immune complex-glomer- phages, and that DAF is lost from E during this transfer process. In ulonephritis connection in man. Kidney Int. 31:877. addition, we have found that binding of these substrates to glyco- 25. Taylor, R. P., E. N. Martin, M. L. Reinagel, A. Nardin, M. Craig, Q. Choice, R. Schlimgen, S. Greenbaum, N. L. Incardona, and H. D. Ochs. 1997. Bispecific phorin A induces either transfer of these substrates to macrophages monoclonal antibody complexes facilitate erythrocyte binding and liver clearance or phagocytosis of E by macrophages. Finally, we have observed of a prototype particulate pathogen in a monkey model. J. Immunol. 97:4035. that CR1 colocalizes with C3b on the E surface, even becoming 26. Craig, M. L., A. J. Bankovich, J. L. McElhenny, and R. P. Taylor. 2000. Clear- ance of anti-double-stranded DNA antibodies: the natural immune complex clear- unclustered when C3b is uniformly distributed over the E surface. ance mechanism. Arthritis Rheum. 43:2265. These studies may provide the foundation for analysis of blood and 27. Taylor, R. P., P. J. Ferguson, E. N. Martin, J. Cooke, K. L. Greene, K. Grinspun, E from children at diverse phases of malaria infections. M. Guttman, and S. Kuhn. 1997. Immune complexes bound to primate erythrocyte complement receptor (CR1) via anti-CR1 mAbs are cleared simultaneously with loss of CR1 in a concerted reaction in a rhesus monkey model. Clin. Im- Disclosures munol. Immunopathol. 82:49. 28. Reinagel, M. L., and R. P. Taylor. 2000. Transfer of immune complexes from The authors have no financial conflict of interest. erythrocyte CR1 to mouse macrophages. J. Immunol. 164:1977. 29. Ross, G. D., W. J. Yount, M. J. Walport, J. B. Winfield, C. J. Parker, C. R. Fuller, R. P. Taylor, B. L. Myones, and P. J. Lachmann. 1985. Disease-associated loss References of erythrocyte complement receptors (CR1, C3b receptors) in patients with sys- 1. Waitumbi, J. N., O. O. Malachi, R. O. Muga, A. O. Misore, and J. A. Stoute. temic lupus erythematosus and other diseases involving autoantibodies and/or 2000. Red cell surface changes and erythrophagocytosis in children with severe complement activation. J. Immunol. 135:2005. Plasmodium falciparum anemia. Blood 95:1481. 30. Iida, K., R. Mornaghi, and V. Nussenzweig. 1982. Complement receptor (CR1) 2. Egan, A. F., M. E. Fabucci, A. Saul, D. C. Kaslow, and L. H. Miller. 2002. Aotus deficiency in erythrocytes from patients with systemic lupus erythematosus. New World monkeys: model for studying malaria-induced anemia. Blood J. Exp. Med. 155:1427. 99:3863. 31. Tausk, F. A., A. McCutchan, P. Spechko, R. D. Schreiber, and I. Gigli. 1986. 3. Price, R. N., J. A. Simpson, F. Nosten, C. Luxemburger, L. Hkirjaroen, Altered erythrocyte C3b receptor expression, immune complexes, and comple- F. ter Kuile, T. Chongsuphajaisiddhi, and N. J. White. 2001. Factors contributing ment activation in homosexual men: varying risk groups for acquired immune to anemia after uncomplicated falciparum malaria. Am. J. Trop. Med. Hyg. deficiency syndrome. J. Clin. Invest. 78:977. 65:614. 32. Bull, P. C., B. S. Lowe, M. Kortok, and K. Marsh. 1999. Antibody recognition 4. Seitz, H. M. 1972. Demonstration of malarial antigens in the sera of Plasmodium of Plasmodium falciparum erythrocyte surface antigens in Kenya: evidence for berghei-infected mice. J. Parasitol. 58:179. rare and prevalent variants. Infect. Immun. 67:733. 5. June, C. H., C. E. Contreras, L. H. Perrin, P. H. Lambert, and P. A. Miescher. 33. Siddique, A. B., N. Ahlborg, M. Warsame, P. Perlmann, and K. Berzins. 1999. 1979. Circulating and tissue-bound immune complex formation in murine ma- Antibodies to a non-repeat region of Plasmodium falciparum antigen Pf155/ laria. J. Immunol. 122:2154. RESA in individuals from malaria-endemic areas. Clin. Exp. Immunol. 116:493. 3066 PROCESSING OF IC BOUND TO NON-CR1 SITES ON RED CELLS

34. Dolan, S. A., J. L. Proctor, D. W. Alling, Y. Okubo, T. E. Wellems, and ment receptor: development of a heterologous mouse model. J. Immunol. Meth- L. H. Miller. 1994. Glycophorin B as an EBA-175 independent Plasmodium ods 270:183. falciparum receptor of human erythrocytes. Mol. Biochem. Parasitol. 64:55. 43. Schifferli, J. A., S. R. Bartolotti, and D. K. Peters. 1980. Inhibition of immune 35. Jakobsen, P. H., P. M. Heegaard, C. Koch, K. Wasniowska, M. M. Lemnge, precipitation by complement. Clin. Exp. Immunol. 42:387. J. B. Jensen, and B. K. Sim. 1998. Identiﬁcation of an erythrocyte binding peptide 44. Taraschi, T. F., M. O’Donnell, S. Martinez, T. Schneider, D. Trelka, from the erythrocyte binding antigen, EBA-175, which blocks parasite multipli- V. M. Fowler, L. Tilley, and Y. Moriyama. 2003. Generation of an erythrocyte cation and induces peptide-blocking antibodies. Infect. Immun. 66:4203. vesicle transport system by Plasmodium falciparum malaria parasites. Blood 36. Duraisingh, M. T., A. G. Maier, T. Triglia, and A. F. Cowman. 2003. Erythro- 102:3420. cyte-binding antigen 175 mediates invasion in Plasmodium falciparum utilizing 45. Lindorfer, M. A., H. B. Jinivizian, P. L. Foley, A. D. Kennedy, M. D. Solga, and sialic acid-dependent and -independent pathways. Proc. Natl. Acad. Sci. USA R. P. Taylor. 2003. B cell complement receptor 2 transfer reaction. J. Immunol. 100:4796. 170:3671. 37. Pasvol, G. 2003. How many pathways for invasion of the red blood cell by the 46. Telen, M. J. 1993. The mature erythrocyte. In Wintrobe’s Clinical Hematology, malaria parasite? Trends Parasitol. 19:430. 9th Ed. G. R. Lee, T. C. Bithell, J. Foerster, J. W. Athens, and J. N. Lukens, eds. Lea & Febiger, Philadelphia, p. 108. 38. Nickells, M., R. Hauhart, M. Krych, V. B. Subramanian, K. Geoghegan-Barek, 47. Lea, S. 2002. Interactions of CD55 with non-complement ligands. Biochem. Soc. H. C. Marsh, Jr., and J. P. Atkinson. 1998. Mapping epitopes for 20 monoclonal Trans. 30:1014. antibodies to CR1. Clin. Exp. Immunol. 112:27. 48. Rasmusson, B. J., J. L. Carpentier, J. P. Paccaud, and K. E. Magnusson. 1996. 39. Tosic, L., W. M. Sutherland, J. Kurek, J. C. Edberg, and R. P. Taylor. 1989. The N-formyl methionyl peptide, formyl-methionyl-leucyl phenylalanine Preparation of monoclonal antibodies to C3b by immunization with C3b(i)- (fMLF), increases the lateral diffusion of complement receptor 1 (CR1/CD35) in Sepharose. J. Immunol. Methods 120:241. human neutrophils; a causative role for oxidative metabolites? Biosci. Rep. 40. Kennedy, A. D., P. V. Beum, M. D. Solga, D. J. DiLillo, M. A. Lindorfer, 16:391. C. E. Hess, J. J. Densmore, M. E. Williams, and R. P. Taylor. 2004. Rituximab 49. Kucik, D. F., M. L. Dustin, J. M. Miller, and E. J. Brown. 1996. Adhesion- infusion promotes rapid complement depletion and acute CD20 loss in chronic activating phorbol ester increases the mobility of leukocyte integrin LFA-1 in lymphocytic leukemia. J. Immunol. 172:3280. cultured lymphocytes. J. Clin. Invest. 97:2139. 41. Coyne, K. E., S. E. Hall, S. Thompson, M. A. Arce, T. Kinoshita, T. Fujita, 50. Fearon, D. T. 1979. Regulation of the ampliﬁcation C3 convertase of human D. J. Anstee, W. Rosse, and D. M. Lublin. 1992. Mapping of epitopes, glyco- complement by an inhibitory protein isolated from human erythrocyte membrane. Downloaded from sylation sites, and complement regulatory domains in human decay accelerating Proc. Natl. Acad. Sci. USA 76:5867. factor. J. Immunol. 149:2906. 51. Ghiran, I., L. Klickstein, and A. Nicholson-Weller. 2004. The cytoplasmic tail of 42. Henderson, A., M. Lindorfer, A. Kennedy, P. Foley, and R. Taylor. 2002. Con- human erythrocyte complement receptor 1 (CR1) binds FAP-1, a scaffolding certed clearance of immune complexes bound to the human erythrocyte comple- protein. Mol. Immunol. 41:201. http://www.jimmunol.org/ by guest on September 24, 2021