The JI
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

The Journal of Immunology, 2007, 179, 7741 -7750
Copyright © 2007 by The American Association of Immunologists, Inc.
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow A correction has been published
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lee, W. Y.
Right arrow Articles by Parkos, C. A.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Lee, W. Y.
Right arrow Articles by Parkos, C. A.

Novel Structural Determinants on SIRP{alpha} that Mediate Binding to CD471

Winston Y. Lee*, Dominique A. Weber*, Oskar Laur{dagger}, Eric A. Severson*, Ingrid McCall*, Rita P. Jen*, Alex C. Chin*, Tao Wu*, Kim M. Gernet{ddagger} and Charles A. Parkos2,*

* Department of Pathology and Laboratory Medicine, Emory University, Atlanta, GA 30322; {dagger} Emory Vaccine Center, Department of Microbiology and Immunology, Emory University, Atlanta, GA 30329; and {ddagger} BioMolecular Computing Resource, Emory University, Atlanta, GA 30322


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Signal regulatory proteins (SIRP-{alpha}, -β, and -{gamma}) are important regulators of several innate immune functions that include leukocyte migration. Membrane distal (D1) domains of SIRP{alpha} and SIRP{gamma}, but not SIRPβ, mediate binding to a cellular ligand termed CD47. Because the extracellular domains of all SIRPs are highly homologous, we hypothesized that some of the 16 residues unique to SIRP{alpha}.D1 mediate binding to CD47. By site-directed mutagenesis, we determined that SIRP{alpha} binding to CD47 is independent of N-glycosylation. We also identified three residues critical for CD47 binding by exchanging residues on SIRP{alpha} with corresponding residues from SIRPβ. Cumulative substitutions of the critical residues into SIRPβ resulted in de novo binding of the mutant protein to CD47. Homology modeling of SIRP{alpha}.D1 revealed topological relationships among critical residues and allowed the identification of critical residues common to SIRP{alpha} and SIRPβ. Mapping these critical residues onto the recently reported crystal structure of SIRP{alpha}.D1 revealed a novel region that is required for CD47 binding and is distinct and lateral to another putative CD47 binding site described on that crystal structure. The importance of this lateral region in mediating SIRP{alpha}.D1 binding to CD47 was confirmed by epitope mapping analyses of anti-SIRP Abs. These observations highlight a complex nature of the ligand binding requirements for SIRP{alpha} that appear to be dependent on two distinct but adjacent regions on the membrane distal Ig loop. A better understanding of the structural basis of SIRP{alpha}/CD47 interactions may provide insights into therapeutics targeting pathologic inflammation.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosures
 References
 
Polymorphonuclear neutrophil (PMN)3 migration across mucosal surfaces, while necessary for host defense, is a central component of disease pathophysiology in many inflammatory conditions. For example, severity of patient symptoms in conditions such as ulcerative colitis and bronchitis/cystitis correlates with the degree of PMN infiltration across mucosal epithelium (1). Previous studies have established that PMN migration across intestinal epithelial monolayers is a complex process involving sequential receptor ligand interactions between PMN and epithelial cells (2). CD47, an ubiquitous transmembrane protein, has been shown to regulate the rate of PMN transmigration across cell monolayer and matrix. The physiological importance of CD47 is apparent from studies using knockout mice which rapidly succumb to Escherichia coli peritonitis due to delayed recruitment of PMN at the site of infection (3). These in vivo observations are consistent with in vitro studies demonstrating a delayed PMN transepithelial migration in the presence of anti-CD47 Abs (4, 5). New insights into the function of CD47 came with the identification of its cellular ligand signal regulatory protein {alpha} (SIRP{alpha}).

SIRP{alpha} is a transmembrane glycoprotein belonging to the Ig superfamily. Other members of the SIRP family include SIRPβ and SIRP{gamma}. The extracellular regions of the SIRPs are highly homologous and consist of three Ig-like loops (6, 7). The most membrane distal loop contains an IgV domain, whereas the two membrane proximal loops contain IgC domains. Interestingly, an isoform of SIRP{alpha} that lacks the two membrane proximal IgC domains has been described in monocytes (8). In contrast to CD47, the expression of SIRP{alpha} is restricted to leukocytes, neurons, and muscle cells (9, 10, 11). Recently, our laboratory has demonstrated that SIRPβ is expressed as a homodimer with a disulfide link between the most membrane proximal Ig loops (12).

In addition to a role in regulating PMN migration, SIRP{alpha}-CD47 interactions also mediate negative regulation of several monocyte/macrophage functions. For example, CD47-deficient RBCs are rapidly phagocytosed by splenic macrophages when infused into normal mice due to the absence of CD47-SIRP{alpha} generated inhibitory signals that serve to dampen phagocytosis (13). Similar mechanisms are presumed to mediate autoimmune hemolytic anemia (14, 15). The SIRP{alpha}-CD47 interaction has also been shown to play a similar role in platelet homeostasis (16, 17). Additional evidence comes from in vitro studies demonstrating that ligation of SIRP{alpha} with CD47 fusion proteins or anti-SIRP{alpha} Abs results in inhibition of macrophage/monocyte phagocytosis and oxidative burst (17, 18, 19, 20). Furthermore, SIRP{alpha}-CD47 interactions have been shown to be crucial in regulating macrophage-mediated clearance of apoptotic cells (21, 22). As highlighted above, several studies have demonstrated that SIRP{alpha} has negative regulatory properties, and such inhibitory signals are mediated by ITIM present in the cytoplasmic domain. In contrast, SIRPβ has been shown to have positive regulatory functions by virtue of the ability to recruit ITAM binding adaptors. SIRP{gamma} is unable to recruit any signaling adaptors and may act as a decoy receptor (6, 7).

A number of studies have provided clues to the structural requirements of SIRP{alpha} binding to CD47. Previously, we and others demonstrated that the membrane distal Ig loop of SIRP{alpha} binds to the Ig-like extracellular loop of CD47 in a specific manner (23, 24, 25). Importantly, the highly homologous SIRPβ does not bind to CD47 (25). Several other reports suggest that glycosylation plays a role in regulation of the affinity of binding of SIRP{alpha} to CD47, however the conclusions have been contradictory (26, 27, 28, 29).

In an attempt to define the CD47 binding site on SIRP{alpha}, we took advantage of the high degree of homology between SIRP{alpha} and SIRPβ. Through mutagenesis of residues in SIRP{alpha} followed by analyses of mutant proteins binding to CD47, structural modeling, and additional mutagenesis, we have defined a region on the membrane distal Ig loop of SIRP{alpha} that mediates binding to CD47. By mapping critical residues identified in this study on the recently reported crystal structure, we delineate a CD47-binding region on the lateral surface of the membrane distal domain of SIRP{alpha} that is distinct from the CD47 binding site recently deduced by Hatherley et al. (30). The importance of this region in mediating binding to CD47 is further supported by epitope mapping analyses of anti-SIRP mAbs. The significance of these findings with respect to the recently reported crystal structure is discussed.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosures
 References
 
Antibodies

SIRP{alpha} and SIRPβ binding mAbs SE7C2 and B4B6 (BD Biosciences), affinity purified goat polyclonal anti-GST Abs (GE-Amersham), HRP-conjugated goat anti-mouse IgG, HRP-conjugated anti-rabbit Fc, and HRP-conjugated donkey anti-goat IgG (Jackson ImmunoResearch Laboratories) are all commercially available. J10.1 is a murine mAb against Jam-A, and was used as an IgG1 isotype control (31). SAF10.1 is a murine mAb produced in our laboratory against human SIRP{alpha}, crossreacts with other SIRPs and recognizes an epitope in the membrane most distal domain (D1). SAF10.1 (IgG1) was generated by immunizing female BALB/c mice i.p. with an eukaryotically expressed SIRP{alpha}-Fc fusion protein consisting of the extracellular domain and emulsified with complete Freund’s adjuvant (Sigma-Aldrich), and followed by two subsequent boosts of SIRP{alpha}-Fc fusion protein emulsified with IFA (Sigma-Aldrich). The mouse with the highest Ab titer was given a final i.v. boost with SIRP{alpha}-Fc and splenocytes were harvested and fused with P3U1 myeloma cells 4 days later. Fused cells were then cultured in selection medium containing DBA-2 feeder thymocytes as described previously (4). Supernatants of hybridoma colonies were screened by ELISA. Positive colonies were subcloned by limiting dilution and weaned from selection medium. The Ab was purified using protein A-Sepharose (Sigma-Aldrich) from culture supernatant. The protocol used in mouse experiments in this study is approved by Institutional Animal Care and Use Committee (IACUC) at Emory University, Atlanta GA.

Generation and mutagenesis of GST/Fc tagged full-length and truncated SIRP plasmid constructs

The generation of constructs encoding rabbit Fc, GST tagged full-length, and truncated extracellular regions of SIRP has been described previously (25). The cDNA sequence of SIRP{alpha} used in this study corresponds to GenBank entry BC 029662.1, and the cDNA sequence of SIRPβ corresponds to NM 006065. To simplify the fusion of GST tags, we made a modified pcDNA3 (Invitrogen Life Technologies) vector by inserting GST from pGEX4T1 (Invitrogen Life Technologies) into a multicloning site leaving HindIII and BamHI (New England Biolab) restriction sites available for subsequent insertions of DNA fragments (pcDNA3GSTmod). Site-directed mutations were introduced by overlap extension using PCR and complementary primers carrying the mutation (32). Forward and reverse end primers contained HindIII and BamHI restriction sites respectively, so that the resultant products could be inserted into pcDNA3GSTmod. All constructs were fully sequenced.

Production of recombinant SIRPs

COS 7.2 cells (American Type Culture Collection) were cultured in DMEM high glucose supplemented with 10% heat inactivated FBS and supplements in 5% CO2 at 37°C. Transient transfections of COS 7.2 with SIRP plasmids were performed by a DEAE-Dextran (GE-Amersham)-based method (33). In brief, the day before transfection, COS 7.2 cells were seeded onto a 60 mm petri dishes (Costar; 0.4 x 106 cells per dish). The following day, after washing cells in serum-free medium, 2 ml of serum-free DMEM containing 500 µg DEAE-Dextran, 2 µg of DNA, and 0.1 mM chloroquine was added and incubated for 4 h in 5% CO2 at 37°C. Cells were shocked with 2 ml of 10% DMSO (Sigma-Aldrich) in serum-free DMEM for 2 min at room temperature (RT) and further cultured in 4 ml of DMEM 10% FCS for 3–4 days. Protein production was assessed by dot blot and ELISA. For dot blots, 5 µl of cell culture supernatants from each transfection were spotted onto a nitrocellulose membrane and allowed to air dry. The nitrocellulose membrane was blocked with a 5% milk/TBS solution for 1 h, incubated with anti-GST, followed by incubation with donkey anti-goat-IgG HRP. Washed blots were developed using a peroxidase chemiluminescence kit from Roche.

To assess recombinant protein in cell culture supernatants, ELISAs were also performed. Microtiter wells (Immulon II) were coated with goat anti-GST or anti-rabbit Fc, each at 10 µg/ml in PBS overnight at 4°C. After blocking with 1% BSA/TBS and washing in TTBS, supernatants from transfectant cultures were added for 1 h at RT to capture recombinant proteins. After washing, an anti-pan SIRP{alpha} mAb SAF10.1 was added for 1 h, followed by washing and addition of goat anti-mouse IgG-HRP. Color was developed with ABTS (Sigma-Aldrich) and absorbance measured at 405 nm.

Immunoblotting of recombinant proteins

Ten microliters of medium containing recombinant proteins were boiled in 1x sample buffer under reducing conditions and separated using SDS-PAGE followed by transfer to polyvinylidene difluoride membranes (BioRad). Nonspecific binding was blocked in 5% milk in TBS. Rabbit-Fc tagged proteins were detected with HRP conjugated polyclonal anti-rabbit IgG Abs, and GST tagged proteins were detected with affinity purified polyclonal goat anti-GST followed by HRP conjugated polyclonal donkey anti-goat IgG Abs. Blots were developed using a chemiluminescence kit from Roche.

In vitro binding assay of SIRP and CD47

A modified version of a previously described assay (25) was used in this study. In brief, affinity-purified polyclonal goat anti-GST Ab or goat-anti-rabbit-Fc was added to microtiter wells (Immulon II) and incubated overnight at 4°C. After blocking with 1% BSA in PBS and washing with Tween 20 and Tris-buffered saline, cell culture supernatants of COS 7.2 transfectants were added to the plates to allow the capture of recombinant proteins for 1 h at RT. Wells were washed with Tween 20 and Tris-buffered saline and cell culture supernatant containing CD47-AP was added for 1 h. After washing, binding to SIRP proteins was assessed colorimetrically using p-nitrophenyl phosphate (Sigma-Aldrich) at 405 nm.

For Ab inhibition assays, following capture of recombinant SIRP proteins in microtiter wells as above, 10 µg/ml SE7C2 or B4B6 in 1% BSA/PBS was added for 1 h at RT. After washing, CD47-AP was added and binding assessed as described as above.

Homology modeling of SIRP{alpha}

A comparative analysis of the membrane distal domain of SIRP{alpha} to a database of known protein crystal structures was performed using the Reverse Position Specific Basic Local Alignment Search Tool (RPS-BLAST) search engine (34). An Fv fragment of an Ab (protein data base code 1MFA; Ref. 35) identified to be most homologous to SIRP{alpha}.D1 (16% identical residues and 49% conserved residues) was chosen as the template for homology modeling. An alignment of SIRP{alpha}.D1 to 1MFA generated by RPS-BLAST was used by MODELLER (36) to obtain a preliminary homology model. The model was then visually inspected and optimized using Sybyl (Tripos). For each step of optimization, the alignment of the protein sequences was first adjusted followed by generation of a new homology model. Improvement in the modeled structure was assessed through a decrease in Discrete Optimized Protein Energy as evaluated by MODELLER (37). Molecular graphics images were produced using the UCSF Chimera package from the Resource for Biocomputing, Visualization, and Informatics at the University of California, San Francisco (supported by National Institutes of Health P41 RR-01081; Ref. 38).

Structural alignment of SIRPs

The SIRP{alpha} crystal structure (pdb id: 2uv3) and the SIRPβ NMR structure (pdb id: 2d9c) were obtained from the RCSB Protein Data Base (www.pdb.org). The structures were aligned using the MatchMaker function in Chimera (39). Matching was iterated by pruning long atom pairs until no pairs exceeded 1 angstrom. The matching alignment between SIRP{alpha} crystal structure and SIRPβ NMR structure (conformation #0.1) is shown in Fig. 8.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosures
 References
 
The most membrane-distal Ig loop (D1) of SIRP{alpha} binds to CD47

Using an in vitro binding assay, we and others have shown previously that the extracellular region of SIRP{alpha} specifically interacts with the Ig loop of CD47 (23, 24, 25). The extracellular region of SIRP{alpha} consists of three domains with classic Ig folds. In this study, we have modified our in vitro binding assay (see Materials and Methods). The recombinant SIRPs were captured onto a plate coated with anti-GST or anti-Fc Abs instead of coating the microtiter plates with purified recombinant proteins. To validate the modified in vitro binding assay, we verified the previous findings that the most membrane distal loop of SIRP{alpha} mediates the binding to CD47 (23, 24, 40). We produced tagged fusion proteins containing full-length and truncated extracellular regions of SIRP{alpha}. Specifically, we produced soluble rabbit Fc tagged fusion proteins containing only the most membrane-distal loop (SIRP{alpha}.D1-Fc; residues 1 to 133), the two membrane-proximal loops (SIRP{alpha}.D2D3-Fc; residues 1 to 33 fused with residues 131 to 366), or the most membrane-proximal loop (SIRP{alpha}.D3-Fc; residues 1 to 33 fused with residues 227 to 336), as well as the full-length extracellular region of SIRP{alpha} (SIRP{alpha}.D1D2D3-Fc; residues 1 to 366) (Fig. 1A). Using the in vitro binding assay, only SIRP{alpha}.D1D2D3-Fc and SIRP{alpha}.D1-Fc bound to CD47 (Fig. 1C). We also produced the following GST fusion proteins: SIRP{alpha}.D1D2D3-GST and SIRP{alpha}.D1-GST. Given that SIRPβ does not bind to CD47 (24, 25), we generated SIRPβ.D1D2D3-GST and SIRPβ.D1-GST to be used as negative controls (Fig. 1B). Consistent with the results obtained using Fc-tagged proteins, only fusion proteins containing SIRP{alpha}.D1 bound to CD47 (Fig. 1D). As detailed in the Materials and Methods section, all cell culture supernatants were assessed both by dot blots and capture ELISAs to ensure that a lack of CD47 binding was not due the absence of recombinant SIRPs (data not shown). As shown in Fig. 1, our modified assay was able to confirm that the most membrane distal Ig loop of SIRP{alpha} contains the domain that interacts with CD47. Therefore, we conclude that the assay modification does not alter our binding measurements.


Figure 1
View larger version (24K):
[in this window]
[in a new window]

  		FIGURE 1. CD47 binds to SIRP{alpha}.D1 but not SIRPβ. A, Schematic illustration of soluble tagged SIRP{alpha} fusion proteins containing various Ig domains. B, Western blot confirming production of soluble SIRP fusion proteins by transfected COS7.2 cells analyzed with murine monoclonal anti-pan-SIRP Ab SAF10.1 and goat anti-mouse IgG-HRP. C and D, In vitro binding assay assessing binding of SIRP{alpha} fusion proteins with CD47: CD47-AP containing culture supernatant was added to immobilized recombinant protein and binding assessed colorimetrically using p-nitrophenyl phosphate.

 
Comparative analysis of protein sequences of SIRP{alpha}.D1 and SIRPβ.D1

As shown in Fig. 1D, CD47 interacts with SIRP{alpha}.D1 but not with SIRPβ.D1, despite the fact that SIRP{alpha} and SIRPβ are highly homologous (9, 24, 40). In an attempt to characterize the CD47 binding site on SIRP{alpha}, we aligned the protein sequences of SIRP{alpha}.D1 and SIRPβ.D1 using CLUSTALW to investigate the differences between the two molecules (41). As illustrated in Fig. 2, the two sequences are highly homologous IgV loops that share 89% identity and differ only at 16 positions, of which 12 are conservative substitutions and 4 are nonconservative substitutions. As defined by the N-X-S/T motifs, there are two potential glycosylation sites at positions N73 and N80 in SIRPβ.D1, whereas there is only one potential glycosylation site at N80 in SIRP{alpha}.D1 (Fig. 2). We reasoned that the presence of a carbohydrate moiety at N73 in SIRPβ may introduce a spatially bulky domain that might interfere with the binding to CD47. We substituted N73D in SIRPβ to eliminate the putative carbohydrate moiety at that position. As shown in Fig. 3, the resultant SIRPβ mutant did not bind to CD47. Conversely, the lack of the carbohydrate moiety at D73 in SIRP{alpha} could have allowed binding to CD47. We substituted D73N to introduce a potential neoglyclosylation site in SIRP{alpha} and observed that the SIRP{alpha} mutant still bound to CD47 (Fig. 3). It has been reported that SIRP{alpha} can be differentially glycosylated in tissues and the degree of glycosylation may affect its affinity to CD47 suggesting that the sugar moiety at N80 in SIRP{alpha} may also play a role in the interaction (26, 29). Therefore, we generated a SIRP{alpha} mutant (N80A) and a SIRPβ double mutant (N73D/N80A) to eliminate all potential N-linked glycosylation sites. As shown in Fig. 3, the SIRP{alpha} mutant did not lose binding to CD47 and the SIRPβ mutant did not gain the ability to bind CD47. Taken together, these data suggest that N-linked glycosylation of SIRP{alpha} is not required for binding to CD47.


Figure 2
View larger version (36K):
[in this window]
[in a new window]

  		FIGURE 2. Alignment of protein sequences demonstrating that SIRP{alpha}.D1 and SIRPβ.D1 are highly homologous. The circled residues are potential N-glycosylation sites in SIRP{alpha} or SIRPβ. Boxed residues represent candidate amino acids in this study that may be important in the interaction with CD47. Residues boxed in solid lines are disparate between SIRP{alpha} and SIRPβ; residues boxed in dashed lines are identified as potential critical residues through homology modeling. Substitutions of SIRP{alpha} residues with either corresponding SIRPβ residues or other nonconserved residues are indicated on top of the boxed amino acids; the corresponding reverse substitutions into SIRPβ are labeled with italicized font under the respective boxed residues.

 

Figure 3
View larger version (13K):
[in this window]
[in a new window]

  		FIGURE 3. N-glycosylation does not play a role in the interaction of SIRP{alpha} with CD47 nor does it prevent binding of SIRPβ to CD47. Binding of CD47-AP to SIRP glycosylation mutants was measured as described in Materials and Methods. SIRP{alpha}.D1-WT contains 1 N-glycosylation motif at N80. The mutant D73N contains 2 motifs, and the mutant N80A contains none. In contrast, SIRPβ.D1-WT contains 2 potential sites (N73/N80). The mutant N73D contains one site, and the double mutant N73D/N80A does not contain any glycosylation sites.

 
Analyses of the role of disparate residues between SIRP{alpha}.D1 and SIRPβ.D1 in mediating binding to CD47

Because we did not observe a role of glycosylation in binding of SIRP{alpha}.D1 to CD47, we reasoned that residues of SIRP{alpha}.D1 not present in SIRPβ.D1 must be important for binding to CD47. To test this hypothesis, we swapped the disparate residues in SIPR{alpha}.D1 with the corresponding residues in SIRPβ.D1 one at a time through site-directed mutagenesis. These SIRP residues are boxed with solid lines in Fig. 2. Mutations: V27M, Q37M, and P44A/A45G were first introduced into the SIRP{alpha}.D1.Fc construct; and later the mutations: S66L, E70N, M72L, and +D101 (insertion of Asp at position 101) were introduced into SIRP{alpha}.D1.GST (Fig. 2). Among the SIRP{alpha}-Fc mutants, substitutions at V27M and Q37M completely abolished binding to CD47; whereas the substitutions at P44A/A45G had no effect (Fig. 4A). Among the SIRP{alpha}-GST mutants, substitutions at S66L eliminated binding to CD47; whereas at M72L retained binding to CD47. Insertion of Asp at position 101 (+D101) in SIRP{alpha}.D1-GST had no effect (Fig. 4B). CD47 binding experiments using full-length wild type and mutant SIRP{alpha} proteins expressed on the surface of transfected CHO cells yielded similar results (data not shown). These data suggest that residues V27, Q37, and S66 play important roles in binding SIRP{alpha} to CD47 and will be referred to as "critical residues." Conversely, we will refer to amino acids that can be substituted without altering binding as "noncritical residues."


Figure 4
View larger version (18K):
[in this window]
[in a new window]

  		FIGURE 4. Identification of critical residues mediating SIRP{alpha} binding to CD47 through substitution of disparate SIRP{alpha} residues with corresponding amino acids on SIRPβ. The substitution of a nonconserved SIRP{alpha} residue with a corresponding SIRPβ residue or vice versa was achieved using site-directed mutagenesis. Using the CD47-AP in vitro binding assay, we analyzed binding to: A, SIRP{alpha}.D1 Fc mutants; B, SIRP{alpha}.D1-GST mutants; and C, SIRPβ.D1-GST mutants carrying single or cumulative substitutions of SIRP{alpha} residues.

 
The data in Fig. 4 demonstrate that substitution of a single critical residue in SIRP{alpha} can abolish binding to CD47. In a complementary fashion, SIRPβ mutants carrying incomplete sets of reverse substitutions of SIRP{alpha} critical residues could be viewed as equivalent to a SIRP{alpha} mutant carrying one or more critical mutations. We reasoned that in order for SIRPβ to gain the ability to bind to CD47, it would be necessary to replace cumulatively all residues that correspond to the critical residues in SIRP{alpha}. Indeed as shown in Fig. 4C, when these mutations were introduced singly and in combinations, no binding to CD47 was observed except for the recombinant protein carrying all the critical mutations. Thus, these data serve to identify all nonconserved residues that are necessary for SIRP{alpha} binding to CD47.

Homology modeling of SIPRa.D1 reveals the topological relationships of critical residues and identifies a region critical for CD47 binding

Changes in critical residues that abrogated the interaction with CD47 could result from two mechanisms. First, such residues could participate in direct binding interactions with CD47. Second, some of these residues could be necessary for maintaining structural integrity of SIRPs by affecting tertiary structures. In this latter possibility, amino acid substitutions would indirectly alter SIRP binding with CD47. To explore these possibilities, we performed detailed structural modeling studies to gain insight into the localization of critical residues in the tertiary structure of SIRP{alpha}. At the time of this study, the crystal structure had not been solved, and we proceeded to construct a structural model of SIRP{alpha}.D1 using homology modeling techniques. Using the RPS-BLAST search engine (34), the structure of SIRP{alpha}.D1 was identified to be homologous to a preclustered IgV family (CD00099; PSSMID 28982). Members of this family that share homology to SIRP{alpha}.D1 can be broadly divided into immunoglobulins (1MFA, 1ADQ, 1NFD, and 1C5D) and T cell receptors (1J8H, 1FYT, 1AO7, and 1BWM). Among these structures, 1MFA was identified to be most similar to SIRP{alpha}.D1. 1MFA is a crystal structure of the Fv fragment of an Ab in a complex with a carbohydrate Ag (35). The alignment of SIRP{alpha}.D1 and 1MFA by MODELLER demonstrates that 16% of the residues are identical and 49% are conserved. A model was constructed using MODELLER and visually optimized using SYBYL for aberrant folding or unusual residue interactions.

The modeling studies predict that the SIRP{alpha}.D1 IgV fold begins at residue E1 and ends at P118 with a disulfide linkage occurring between C25 and C91. Interestingly, the model predicts that critical residue Q37 is located within the hydrophobic core adjacent to the disulfide bond, whereas the other two critical residues V27 and S66 are positioned near each other on a protein surface opposite to the plane containing Q37 (Fig. 5A). The model also predicts that noncritical residues P44/A45 and +D101 are distant to the critical residues, while residues M72 and E70 are positioned close to the critical residues. Taken together, these mutagenesis and modeling data suggest that critical residues V27 and S66 form part of a region that may play an important role in the interaction with CD47.


Figure 5
View larger version (42K):
[in this window]
[in a new window]

  		FIGURE 5. A computational homology model of SIRP{alpha}.D1 reveals additional critical residues common to SIRP{alpha} and SIRPβ that mediate binding to CD47. Critical residues V27 and S66 are located in close proximity, whereas critical residue Q37 is away from both V27 and S66 and is near the hydrophobic core and disulfide bond (yellow). A, The left panel is a ribbon diagram with sidechains depicting a front view of the SIRP{alpha}.D1 model. The right panel is the space-filled configuration of the model. Based on the assumption that V27 and S66 are involved in binding of SIRP{alpha} to CD47, the model predicts other residues (labeled with *) surrounding V27 and S66 that may also be important in binding to CD47. Critical residues (Q8, T26, V27, Q37, S66, R69, S75, and M72) are labeled in red, and noncritical residues (H24, E70, and +D101) are labeled in green. Yellow residues represent cysteines that form a disulfide bond. B, Results of CD47-AP binding to SIRP{alpha}.D1 mutants with nonconserved mutations of candidate residues predicted by the homology model.

 
Identification of critical residues that are common to SIRP{alpha} and SIRPβ

Because our model predicted that several critical residues were clustered into a distinct region, we tested the possibility that additional residues in the region containing V27 and S66 may be involved in the interaction with CD47. The selected residues (Q8, H24, T26, S75, S66, R69, E70, and V27) are marked with * on our homology model (Fig. 5A), and are also highlighted in the primary sequence using boxes with dashed and solid borders (Fig. 2). Among these residues, we have already shown that V27 and S66 are critical residues (colored in red in Figure 5A), and E70 is a noncritical residue (colored in green in Fig. 5A). Because the substitution of M72L is highly conserved, the replacement of M72 with nonconserved residues, such as arginine and alanine, might help define the role of M72. Neither M72L (Fig. 4B) nor M72R (Fig. 5B) abrogated the binding to CD47. In contrast, the substitution of M72 with alanine abolished the binding to CD47 (Fig. 5B). These data suggest that the physical presence of a medium sized sidechain at residue 72 has an important role in the interaction with CD47.

We applied a similar approach to test whether other candidate residues are important for binding by replacing them with residues not conserved in size, polarity or charge. Using SIRP{alpha}.D1.GST as a template, the following mutants were made: Q8L, H24N, H24L, T26A, R69A, R69E, and S75L (Fig. 5B). Nonconserved substitutions Q8L, T26A, R69A, and S75L completely abrogated binding to CD47. Conversely, mutations H24L or H24N had no effect on SIRP{alpha} binding to CD47. These results lend strong support to the idea that the region on SIRP{alpha}.D1 containing Q8, T26, V27, S66, R69, M72, and S75 plays an important role in binding CD47.

The lateral surface of SIRP{alpha}.D1 is important in binding CD47

Recently Hatherley et al. (30) reported the solution of a SIRP{alpha}.D1 crystal structure and the identification of a CD47 binding site deduced through homology comparison, site directed mutagenesis, and Ab studies. Because only two of our critical mutations overlap with their findings (S66 and R69), we proceeded to map our critical and noncritical residues onto their reported crystal structure (Fig. 6). In this figure, we highlighted the CD47 binding site deduced by Hatherley et al. in cyan. Interestingly, our critical residues: Q8, T26, V27, S66, M72, and S75 (colored in red) are clustered in a region on the lateral surface of the crystal structure that is distinct from the binding site deduced by Hatherley et al. Thus, our mutagenesis results in Figs. 3 and 4 indicate that a lateral surface of SIRP{alpha} plays an important role in binding CD47. The lateral region we describe overlaps greatly with a putative "transdimerization region" proposed by Hatherley et al. However, there is no published evidence that transdimerization of SIRP{alpha} occurs under physiological conditions. Thus, in this report, we refer to this region as the "lateral surface" of the crystal structure.


Figure 6
View larger version (54K):
[in this window]
[in a new window]

  		FIGURE 6. The lateral surface of SIRP{alpha}.D1 mediates binding to CD47. Critical residues (red) and noncritical residues (green) are mapped onto the recently reported SIRP{alpha}.D1 crystal structure (30 ). Residues in the putative CD47 binding site deduced by Hatherley et al. (cyan) are completely conserved between SIRP{alpha} and SIRPβ. Two critical residues labeled with * were identified by us and Hatherley et al. Critical residues identified by us are clustered at a surface (circled in red) that is distinct and lateral from the binding site proposed by Hatherley (circled in cyan). The main figure represents an en face view of the lateral surface of SIRP{alpha}.D1 whereas the inset represents the structure rotated counterclockwise 90 degrees.

 
A mAb that binds to a residue adjacent to the lateral surface of SIRP{alpha}.D1 blocks binding of CD47

We performed an Ab epitope mapping study to investigate the hypothesis that the lateral surface of SIRP{alpha}.D1 is involved in binding to CD47. It has been previously published that mAb SE7C2 binds specifically to SIRP{alpha} and blocks CD47 binding, whereas mAb B4B6 binds specifically to SIRPβ (12, 24, 25). We reasoned that identifying epitopes of such SIRP-specific Abs would provide additional insight into the nature SIRP{alpha}/CD47 interactions. We screened these two Abs for the ability to bind to our SIRP{alpha}.D1 mutants. As shown in Fig. 7A, SE7C2 bound to all mutants of SIRP{alpha}.D1 except SIRP{alpha}.D1-E70N. Therefore, SE7C2 binding to SIRP{alpha} seemed to be dependent on the presence of E at position 70, suggesting that the epitope of SE7C2 includes E70 on SIRP{alpha}. Furthermore SE7C2 was shown to consistently inhibit CD47 binding to wild-type SIRP{alpha}.D1 as well as SIRP{alpha}.D1-P44A/A45G (Fig. 7, A and B). In the SIRP{alpha}.D1 crystal structure, residue E70 is immediately adjacent to the cluster of critical residues that we have identified on the lateral surface (Fig. 7C).


Figure 7
View larger version (35K):
[in this window]
[in a new window]

  		FIGURE 7. CD47 binding is inhibited by a mAb that maps to the SIRPa.D1 lateral surface. A, mAb SE7C2, specific for human SIRP{alpha}, but not SIRPβ, and B, mAb B4B6, specific for human SIRPβ, but not SIRP{alpha}, were screened for binding to SIRP{alpha} mutants by ELISA as described in the Materials and Methods (10 µg/ml mAb in 1%BSA/PBS buffer). C, The protein structure shows the epitope of SE7C2 (circled in orange) contains E100, which is at the upper edge of the lateral surface. The graph shows that CD47-AP binding to SIRP{alpha}.D1 is blocked by SE7C2 in an in vitro binding assay but not by control mAb J10.4. D, The protein structure shows the neoepitope of B4B6 (circled in orange) introduced into SIRP{alpha}.D1 with the noncritical mutation P44A/A45G is distant from the lateral surface. The graph shows no inhibition of CD47-AP binding to SIRP{alpha}.D1.P44A/A45G in the presence of B4B6.

 
In contrast, B4B6, a mAb specific for SIRPβ, did not bind to any of the SIRP{alpha}.D1 mutants except to SIRP{alpha}.D1-P44A/A45G (Fig. 7B). This result suggests that the epitope of B4B6 contains A44/G45 on SIRPβ. As shown in Fig. 4A, the substitutions P44A/A45G in SIRP{alpha} create a SIRP{alpha} mutant that not only is still capable of binding to CD47 but also contains a neoepitope of B4B6. On the crystal structure of SIRP{alpha}.D1, this epitope is positioned in a region distant from both the CD47 binding site deduced by Hatherley et al. and from the lateral surface (Fig. 7D). Indeed, consistent with our results and those reported by Hatherley, B4B6 did not inhibit the mutant protein SIRP{alpha}.D1-P44A/A45G binding to CD47 (Fig. 7D).


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosures
 References
 
Given that SIRP{alpha}.D1 but not SIRPβ.D1 is a ligand for CD47 (23, 24, 25), our comparative analysis of the primary sequences of the two proteins revealed several key differences that may be important in defining the interaction of SIRP{alpha} and CD47. One of the first important distinctions was the different patterns of potential N-linked glycosylation sites. It has been reported that SIRP{alpha} in different tissues exhibit differential galactosylation, which can affect its binding specificity to cellular targets (26). However, a recent report demonstrated that recombinant SIRP{alpha} proteins produced from tunicamycin treated cells could bind to CD47 as well as the glycosylated SIRP{alpha} (27). These conflicting reports prompted us to explore the role of glycosylation in defining the ligand specificity of SIRPs through the use of site-directed mutagenesis. By eliminating the glycosylation concensus motifs in the SIRP proteins, we could address the question unequivocally. We found that glycosylation neither plays a role in facilitating binding of CD47 to SIRP{alpha}, nor in preventing binding of CD47 to SIRPβ.

The second major consideration is the presence of a relatively small number of disparate residues between SIRP{alpha} and SIRPβ. Substitutions of seven candidate residues in SIRP{alpha} with the corresponding residues in SIRPβ revealed that three residues (V27, Q37, and S66) are crucial in the ability of SIRP{alpha} to bind CD47. The distribution of the residues in the primary sequence provided no direct clues as to the nature of the binding site. Furthermore, when this study was performed, the crystal structure was not available. Therefore, we constructed a computational model of SIRP{alpha}.D1 to gain insights into the role of these critical residues in binding CD47. The model revealed that V27 and S66 are located on the flexible loops at the edge of a β sheet, whereas residue Q37 is on the opposing β sheet in the interior, proximal to the disulfide bond and a hydrophobic tryptophan that are present in the conserved hydrophobic structural core of IgVs. Because the region containing V27 and S66 is more surface accessible than the region containing Q37, we hypothesized that other residues adjacent to V27 and S66 in that region might also have major effects on ligand binding. Along this line of reasoning, our model predicted that residues Q8, H24, T26, R69, M72, and S75 should influence ligand binding. Among these residues, Q8, H24, R69, and S75 are conserved between SIRP{alpha} and SIRPβ. When we replaced these residues in SIRP{alpha}.D1 with leucine, a medium sized nonpolar amino acid, both Q8L and S75L mutants lost their ability to bind CD47. However, after replacing H24, a large positively charged residue, with leucine or a medium sized polar residue (N), there was no inhibitory effect. In contrast, when we exchanged a small polar residue such as T26 for a small nonpolar residue (A), or a medium sized nonpolar residue (L), the CD47 binding capacity of the mutant was abrogated. It is interesting to note that T26 is conserved in all SIRPs except for SIRPβ and is identified here as being critical for CD47 binding. When the other residue (R69) that is common between SIRP{alpha} and SIRPβ was mutagenized into a small nonpolar residue (A), or a large negatively charged residue (E), binding was obliterated. These results lead us to classify Q8, T26, R69, and S75 as critical residues. Finally, when we exchanged M72 in SIRP{alpha}.D1 for the corresponding leucine of SIRPβ.D1, the mutant M72L retained binding to CD47, although the level of binding was consistently lower than wild type. Because M72L is a fairly conservative substitution, we replaced M72 with nonconserved residues consisting of either a small hydrophobic residue (A), or a large positively charged residue (R). Interestingly, only the mutant M72A lost its ability to bind CD47, suggesting that the physical presence of a sidechain at residue 72 in SIRP{alpha} is important in binding to CD47. From these results, it is clear that M72 is also a critical residue in mediating binding interactions with CD47.

Our mutational studies identifying M72 as a critical residue revealed that our initial comparison of primary sequences between SIRP{alpha} and SIRPβ was only useful in determining critical residues that have become sufficiently divergent. There is a widely held view that ITAM/ITIM protein pairs, such as SIRP{alpha} and SIRPβ, sharing a highly homologous extracellular region, probably arose from gene duplication and other mechanisms (42, 43). It is conceivable that, as SIRPβ evolved from the duplication of SIRP{alpha}, binding to CD47 was lost. However, this does not imply that all residues important in binding to CD47 have become divergent. As demonstrated by our study, one mutation at a critical residue in SIRP{alpha} is all that is necessary to abolish binding to CD47. Consequently, the comparative analyses of the primary sequences only reveal nonconserved changes in residues and do not provide any information regarding the significance of conserved residues. Therefore, homology modeling was useful in predicting additional critical candidate residues. As is further discussed below, our model identified residues that, when mapped onto the recently published crystal structure, confirm the importance of a novel region involved in binding of SIRP{alpha} to CD47.

Recently, another group adopted a similar approach comparing various members of SIRPs to identify residues important in interaction with CD47 (40). They concluded that V27 and Q37 were critical residues necessary for interaction, whereas M72 "assisted" the interaction. Although the observation that V27 and Q37 are critical residues is consistent with our findings, no data exploring the role of conserved residues in SIRP{alpha} binding interactions with CD47 were presented. Furthermore, two mutually exclusive homology models proposed in this study are inconsistent with the recently reported crystal structure.

Recently Hatherley et al. (30) reported the solution of a SIRP{alpha}.D1 crystal structure. These authors described two major domains on the crystal structure consisting of transdimerization and putative CD47 binding sites. The proposed transdimerization site consists of a region of close contact between two SIRP{alpha}.D1 molecules in trans at loop DE (residues 60–78), strand B1B2 (residues 20–27), and strand A1 (residues 6 and 7). Despite this observation in the crystal structure, it was acknowledged that no biochemical or functional evidence exists indicating that SIRP{alpha}.D1 forms transdimers under physiological conditions (30). Because of the lack of evidence for transdimerization, we refer to this region on SIRP{alpha} as the "lateral surface."

Through homology comparison, site-directed mutagenesis, and an Ab epitope study, a CD47 binding region was deduced by Hatherley et al. consisting of residues I31, V33, R69, K96, and D100. As shown in that study, the epitope of the SIRP{alpha} and SIRPβ-binding mAb SE5A5 appears to be conformationally dependent as it spans two different loops (K96 on loop G1F and V33 and I31 on loop B2C) on the top of SIRP{alpha}. Interestingly, this epitope also overlays much of the region on SIRP{alpha}.D1 proposed by these authors as a CD47 binding site on SIRP{alpha}.D1. However, because SE5A5 also binds to SIRPβ, it can be inferred that the conformation of Hatherley’s proposed binding site is highly similar to the corresponding region in SIRPβ. This inference is further corroborated by the alignment of the SIRP{alpha} crystal structure to the recently available NMR structure of SIRPβ (Fig. 8). The structurally based alignment demonstrates that the three dimensional conformations of SIRP{alpha}.D1 and SIRPβ.D1 are highly conserved at the five critical residues that are reported by Hatherley to constitute the CD47 binding domain. Given these observations and the fact that the critical residues reported by these authors as well as the amino acids adjacent to these critical residues are identical between SIRP{alpha} and SIRPβ, it is hard to envision how this region can alone be sufficient to mediate binding to CD47.


Figure 8
View larger version (46K):
[in this window]
[in a new window]

  		FIGURE 8. SIRP{alpha}D1 and SIRPβD1 exhibit highly similar structural conformations. Alignment of SIRP{alpha}D1 crystal structure (2uv3; white) with SIRPβD1 NMR structure (2d9c; magenta) reveals a high degree of similarity between the two structures in the region of the putative CD47 binding site (cyan) proposed by Hatherley et al. Note, the residues in this region are identical between both molecules, thus it is unclear how this domain by itself can confer specificity of binding to CD47. Critical residues reported by us are highlighted in red and cluster on the lateral surface (noncritical residues are in green). Although the conformations of main peptide chains in the lateral surfaces of both proteins are similar between SIRP{alpha}D1 and SIRPβD1, four critical residues (labeled with *) in the lateral surface are unique to SIRP{alpha} and presumably confer binding specificity through sidechain interactions.

 
When we mapped the critical residues identified in our study onto the crystal structure of SIRP{alpha}, most of them (Q8, T26, V27, S66, R69, M72, and S75) are located in a cluster on the lateral surface of SIRP{alpha}.D1. As highlighted in Fig. 6, this region on the lateral surface is distinct from Hatherley’s binding site. Interestingly, these authors excluded the possibility that the lateral surface is a CD47 binding site although it was reported that S66D and R69E were critical mutations, whereas E70K and M72R were not (30). However, these authors only examined residues in the DE loop, but none were examined in the B1B2 and A1 strands, which are also parts of the lateral surface. Residue R69 is actually positioned on the edge of the lateral surface, but it was considered part of their reported CD47 binding site on the top of the molecule. Because of its distance from the putative CD47 binding site, it was also indicated that residue S66 is important in maintaining the structural integrity perhaps through the formation of a potential intramolecular hydrogen bond, that is absent in SIRPβ, to the main peptide chain. However, the high degree of similarity between the three-dimensional conformations of SIRP{alpha}.D1 and SIRPβ.D1 would argue that an intramolecular hydrogen bond at S66 is not essential in maintaining the structural integrity. From their results, Hatherley et al. proposed that the lateral surface does not have a role in binding CD47. In this study, we probed the lateral region of SIRP{alpha}.D1 in a more comprehensive fashion based on our structural model. We identified six additional critical residues on the lateral surface (Q8, T26, V27, S66, and S75) including M72 since M72A, but not M72L or M72R, abolished binding to CD47. These critical residues identified in our study are present in the B1B2 and A1 strands as well as the DE loop. In further support of involvement of this region in binding to CD47, we demonstrated de novo binding of SIRPβ to CD47 by exchanging four SIRPβ specific residues with corresponding SIRP{alpha} unique residues (A26T, M27V, M37Q, and L66S). Among these residues, three of them are located on the lateral surface. Lastly, involvement of the lateral surface of SIRP{alpha}.D1 in binding to CD47 is also supported by our epitope mapping experiments with the SIRP{alpha}-specific mAb SE7C2 that inhibits binding to CD47. The epitope of SE7C2 contains E70, a residue that is immediately adjacent to the critical domain on the lateral surface.

Our site-directed mutagenesis findings coupled with recent crystal structure data strongly suggest that the lateral surface of SIRP{alpha}.D1 plays a key role in binding to CD47. We believe that functional effects of the critical residues in the lateral surface of SIRP{alpha}.D1 are predominantly mediated through sidechains. As shown in Fig. 8, the conformations of the main peptide chains of the lateral surface of SIRP{alpha}.D1 and the corresponding region in SIRPβ.D1 are highly similar. However, the lateral surface of SIRP{alpha}.D1 contains four unique critical residues (T26, V27, S66 and M72). Therefore, it is reasonable to assume that differences in sidechains of these unique critical residues must dictate the specificity of interaction between SIRP proteins and CD47. It is thus possible that the lateral surface of SIRP{alpha}.D1 represents a CD47 binding site. Alternatively, residues in the lateral surface could be necessary for maintenance of the conformational integrity of the binding site proposed by Hatherley, thus indirectly modulating binding interactions with CD47. To distinguish between these possibilities, cocrystallization of SIRP{alpha}.D1 bound to CD47 will be necessary.


   Disclosures
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
References
 
The authors have no financial conflict of interest.


   Footnotes
 
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 National Institutes of Health Grant R01 DK79392. The work at BimCore was supported by Emory School of Medicine, Graduate Division of Biological and Biomedical Sciences and Emory Woodruff Fund. Back

2 Address correspondence and reprint requests to Dr. Charles A. Parkos, Division of Gastrointestinal Pathology, Emory University, Whitehead Biomedical Building, Room 105B, 615 Michael Street, Atlanta, GA 30322. E-mail address: cparkos{at}emory.edu Back

3 Abbreviations used in this paper: PMN, polymorphonuclear neutrophil; SIRP, signal regulatory proteins; RT, room temperature; RPS-BLAST, Reverse Position Specific Basic Local Alignment Search Tool. Back

Received for publication March 26, 2007. Accepted for publication August 30, 2007.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosures
 References
 

  1. Chin, A. C., C. A. Parkos. 2006. Neutrophil transepithelial migration and epithelial barrier function in IBD: potential targets for inhibiting neutrophil trafficking. Ann. NY Acad. Sci. 1072: 276-287. [Medline]
  2. Liu, Y., S. K. Shaw, S. Ma, L. Yang, F. W. Luscinskas, C. A. Parkos. 2004. Regulation of leukocyte transmigration: cell surface interactions and signaling events. J. Immunol. 172: 7-13. [Free Full Text]
  3. Lindberg, F. P., D. C. Bullard, T. E. Caver, H. D. Gresham, A. L. Beaudet, E. J. Brown. 1996. Decreased resistance to bacterial infection and granulocyte defects in IAP-deficient mice. Science 274: 795-798. [Abstract/Free Full Text]
  4. Parkos, C. A., S. P. Colgan, T. W. Liang, A. Nusrat, A. E. Bacarra, D. K. Carnes, J. L. Madara. 1996. CD47 mediates post-adhesive events required for neutrophil migration across polarized intestinal epithelia. J. Cell Biol. 132: 437-450. [Abstract/Free Full Text]
  5. Liu, Y., D. Merlin, S. L. Burst, M. Pochet, J. L. Madara, C. A. Parkos. 2001. The role of CD47 in neutrophil transmigration. Increased rate of migration correlates with increased cell surface expression of CD47. J. Biol. Chem. 276: 40156-40166. [Abstract/Free Full Text]
  6. van Beek, E. M., F. Cochrane, A. N. Barclay, T. K. van den Berg. 2005. Signal regulatory proteins in the immune system. J. Immunol. 175: 7781-7787. [Abstract/Free Full Text]
  7. Barclay, A. N., M. H. Brown. 2006. The SIRP family of receptors and immune regulation. Nat. Rev. Immunol. 6: 457-464. [Medline]
  8. Veillette, A., E. Thibaudeau, S. Latour. 1998. High expression of inhibitory receptor SHPS-1 and its association with protein-tyrosine phosphatase SHP-1 in macrophages. J. Biol. Chem. 273: 22719-22728. [Abstract/Free Full Text]
  9. Kharitonenkov, A., Z. Chen, I. Sures, H. Wang, J. Schilling, A. Ullrich. 1997. A family of proteins that inhibit signalling through tyrosine kinase receptors. Nature 386: 181-186. [Medline]
  10. Maile, L. A., J. Badley-Clarke, D. R. Clemmons. 2003. The association between integrin-associated protein and SHPS-1 regulates insulin-like growth factor-I receptor signaling in vascular smooth muscle cells. Mol. Biol. Cell 14: 3519-3528. [Abstract/Free Full Text]
  11. Jiang, P., C. F. Lagenaur, V. Narayanan. 1999. Integrin-associated protein is a ligand for the P84 neural adhesion molecule. J. Biol. Chem. 274: 559-562. [Abstract/Free Full Text]
  12. Liu, Y., I. Soto, Q. Tong, A. Chin, H. J. Buhring, T. Wu, K. Zen, C. A. Parkos. 2005. SIRPβ1 is expressed as a disulfide-linked homodimer in leukocytes and positively regulates neutrophil transepithelial migration. J. Biol. Chem. 280: 36132-36140. [Abstract/Free Full Text]
  13. Oldenborg, P. A., A. Zheleznyak, Y. F. Fang, C. F. Lagenaur, H. D. Gresham, F. P. Lindberg. 2000. Role of CD47 as a marker of self on red blood cells. Science 288: 2051-2054. [Abstract/Free Full Text]
  14. Oldenborg, P. A.. 2004. Role of CD47 in erythroid cells and in autoimmunity. Leukemia Lymphoma 45: 1319-1327. [Medline]
  15. Olsson, M., A. Nilsson, P. A. Oldenborg. 2006. Target cell CD47 regulates macrophage activation and erythrophagocytosis. Transfus. Clin. Biol. 13: 39-43. [Medline]
  16. Olsson, M., P. Bruhns, W. A. Frazier, J. V. Ravetch, P. A. Oldenborg. 2005. Platelet homeostasis is regulated by platelet expression of CD47 under normal conditions and in passive immune thrombocytopenia. Blood 105: 3577-3582. [Abstract/Free Full Text]
  17. Yamao, T., T. Noguchi, O. Takeuchi, U. Nishiyama, H. Morita, T. Hagiwara, H. Akahori, T. Kato, K. Inagaki, H. Okazawa, et al 2002. Negative regulation of platelet clearance and of the macrophage phagocytic response by the transmembrane glycoprotein SHPS-1. J. Biol. Chem. 277: 39833-39839. [Abstract/Free Full Text]
  18. Okazawa, H., S. Motegi, N. Ohyama, H. Ohnishi, T. Tomizawa, Y. Kaneko, P. A. Oldenborg, O. Ishikawa, T. Matozaki. 2005. Negative regulation of phagocytosis in macrophages by the CD47-SHPS-1 system. J. Immunol. 174: 2004-2011. [Abstract/Free Full Text]
  19. Oldenborg, P. A., H. D. Gresham, F. P. Lindberg. 2001. CD47-signal regulatory protein {alpha} (SIRP{alpha}) regulates Fc{gamma} and complement receptor-mediated phagocytosis. J. Exp. Med. 193: 855-862. [Abstract/Free Full Text]
  20. Wang, H., J. VerHalen, M. L. Madariaga, S. Xiang, S. Wang, P. Lan, P. A. Oldenborg, M. Sykes, Y. G. Yang. 2007. Attenuation of phagocytosis of xenogeneic cells by manipulating CD47. Blood 109: 836-842. [Abstract/Free Full Text]
  21. Gardai, S. J., Y. Q. Xiao, M. Dickinson, J. A. Nick, D. R. Voelker, K. E. Greene, P. M. Henson. 2003. By binding SIRP{alpha} or calreticulin/CD91, lung collectins act as dual function surveillance molecules to suppress or enhance inflammation. Cell 115: 13-23. [Medline]
  22. Gardai, S. J., K. A. McPhillips, S. C. Frasch, W. J. Janssen, A. Starefeldt, J. E. Murphy-Ullrich, D. L. Bratton, P. A. Oldenborg, M. Michalak, P. M. Henson. 2005. Cell-surface calreticulin initiates clearance of viable or apoptotic cells through trans-activation of LRP on the phagocyte. Cell 123: 321-334. [Medline]
  23. Vernon-Wilson, E. F., W. J. Kee, A. C. Willis, A. N. Barclay, D. L. Simmons, M. H. Brown. 2000. CD47 is a ligand for rat macrophage membrane signal regulatory protein SIRP (OX41) and human SIRP{alpha} 1. Eur. J. Immunol. 30: 2130-2137. [Medline]
  24. Seiffert, M., P. Brossart, C. Cant, M. Cella, M. Colonna, W. Brugger, L. Kanz, A. Ullrich, H. J. Buhring. 2001. Signal-regulatory protein {alpha} (SIRP{alpha}) but not SIRPβ is involved in T-cell activation, binds to CD47 with high affinity, and is expressed on immature CD34+CD38– hematopoietic cells. Blood 97: 2741-2749.
  25. Liu, Y., H. J. Buhring, K. Zen, S. L. Burst, F. J. Schnell, I. R. Williams, C. A. Parkos. 2002. Signal regulatory protein (SIRP{alpha}), a cellular ligand for CD47, regulates neutrophil transmigration. J. Biol. Chem. 277: 10028-10036. [Abstract/Free Full Text]
  26. van den Nieuwenhof, I. M., C. Renardel de Lavalette, N. Diaz, I. van Die, T. K. van den Berg. 2001. Differential galactosylation of neuronal and haematopoietic signal regulatory protein-{alpha} determines its cellular binding-specificity. J. Cell Sci. 114: 1321-1329. [Abstract]
  27. Subramanian, S., E. T. Boder, D. E. Discher. 2007. Phylogenetic divergence of CD47 interactions with human signal regulatory protein {alpha} reveals locus of species specificity: implications for the binding site. J. Biol. Chem. 282: 1805-1818. [Abstract/Free Full Text]
  28. Subramanian, S., R. Parthasarathy, S. Sen, E. T. Boder, D. E. Discher. 2006. Species- and cell type-specific interactions between CD47 and human SIRP{alpha}. Blood 107: 2548-2556. [Abstract/Free Full Text]
  29. Ogura, T., T. Noguchi, R. Murai-Takebe, T. Hosooka, N. Honma, M. Kasuga. 2004. Resistance of B16 melanoma cells to CD47-induced negative regulation of motility as a result of aberrant N-glycosylation of SHPS-1. J. Biol. Chem. 279: 13711-13720. [Abstract/Free Full Text]
  30. Hatherley, D., K. Harlos, D. C. Dunlop, D. I. Stuart, A. N. Barclay. 2007. The structure of the macrophage signal regulatory protein {alpha} (SIRP{alpha}) inhibitory receptor reveals a binding face reminiscent of that used by T cell receptors. J. Biol. Chem. 282: 14567-14575. [Abstract/Free Full Text]
  31. Liu, Y., A. Nusrat, F. J. Schnell, T. A. Reaves, S. Walsh, M. Pochet, and C. A. Parkos. 2000. Human junction adhesion molecule regulates tight junction resealing in epithelia. 2363–2374.
  32. Horton, R. M., H. D. Hunt, S. N. Ho, J. K. Pullen, L. R. Pease. 1989. Engineering hybrid genes without the use of restriction enzymes: gene splicing by overlap extension. Gene 77: 61-68. [Medline]
  33. Weber, D. A., C. T. Dao, J. Jun, J. L. Wigal, P. E. Jensen. 2001. Transmembrane domain-mediated colocalization of HLA-DM and HLA-DR is required for optimal HLA-DM catalytic activity. J. Immunol. 167: 5167-5174. [Abstract/Free Full Text]
  34. Marchler-Bauer, A., J. B. Anderson, P. F. Cherukuri, C. DeWeese-Scott, L. Y. Geer, M. Gwadz, S. He, D. I. Hurwitz, J. D. Jackson, Z. Ke, et al 2005. CDD: a conserved domain database for protein classification. Nucleic Acids Res. 33: D192-D196. [Abstract/Free Full Text]
  35. Zdanov, A., Y. Li, D. R. Bundle, S. Deng, C. R. MacKenzie, S. A. Narang, N. M. Young, M. Cygler. 1994. Structure of a single-chain antibody variable domain (Fv) fragment complexed with a carbohydrate antigen at 1.7-A resolution. Proc. Natl. Acad. Sci. USA 91: 6423-6427. [Abstract/Free Full Text]
  36. Fiser, A., A. Sali. 2003. Modeller: generation and refinement of homology-based protein structure models. Methods Enzymol. 374: 461-491. [Medline]
  37. Shen, M. Y., A. Sali. 2006. Statistical potential for assessment and prediction of protein structures. Protein Sci. 15: 2507-2524. [Medline]
  38. Pettersen, E. F., T. D. Goddard, C. C. Huang, G. S. Couch, D. M. Greenblatt, E. C. Meng, T. E. Ferrin. 2004. UCSF chimera: a visualization system for exploratory research and analysis. J. Comput. Chem. 25: 1605-1612. [Medline]
  39. Meng, E. C., E. F. Pettersen, G. S. Couch, C. C. Huang, T. E. Ferrin. 2006. Tools for integrated sequence-structure analysis with UCSF chimera. BMC Bioinformatics 7: 339[Medline]
  40. Liu, Y., Q. Tong, Y. Zhou, H. W. Lee, J. J. Yang, H. J. Buhring, Y. T. Chen, B. Ha, C. X. Chen, Y. Yang, K. Zen. 2007. Functional elements on SIRP{alpha} IgV domain mediate cell surface binding to CD47. J. Mol. Biol. 365: 680-693. [Medline]
  41. Thompson, J. D., D. G. Higgins, T. J. Gibson. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22: 4673-4680. [Abstract/Free Full Text]
  42. Viertlboeck, B. C., R. Schmitt, T. W. Gobel. 2006. The chicken immunoregulatory receptor families SIRP, TREM, and CMRF35/CD300L. Immunogenetics 58: 180-190. [Medline]
  43. van den Berg, T. K., J. A. Yoder, G. W. Litman. 2004. On the origins of adaptive immunity: innate immune receptors join the tale. Trends Immunol. 25: 11-16. [Medline]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow A correction has been published
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Lee, W. Y.
Right arrow Articles by Parkos, C. A.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Lee, W. Y.
Right arrow Articles by Parkos, C. A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS