1. Volume 31, Issue 2, Pages 266-277 (July 2008)
Paired Receptor Specificity Explained by Structures of Signal Regulatory Proteins Alone and Complexed with CD47 
Deborah Hatherley, Stephen C. Graham, Jessie Turner, Karl Harlos, David I. Stuart, A. Neil Barclay 
Molecular Cell 
Volume 31, Issue 2, Pages (July 2008)
DOI: /j.molcel
Copyright © 2008 Elsevier Inc. Terms and Conditions

2. Figure 1
Interactions between Macrophages and Other Cells, Including the SIRP Paired Receptor Family
The approximate sizes of the proteins are indicated with ovals representing IgSF domains and rectangles representing the scavenger receptor cysteine-rich domains of CD6. Filled circles denote the approximate positions of potential N-linked glycosylation sites, although typical N-linked sugars are usually larger than indicated in the cartoon. Data are from Barclay et al. (1997) and Wright et al. (2000).
Molecular Cell  , DOI: ( /j.molcel )
Copyright © 2008 Elsevier Inc. Terms and Conditions

3. Figure 2 Crystal Structures of the IgSF Domain of CD47
(A) The secondary structure of CD47 is shown above the sequence of the IgSF domain of human CD47 with arrows and cylinders representing β sheets and α helices, respectively. The N-terminal glutamine residue that cyclizes to form a pyroglutamate is boxed. The cysteine residues that form a conserved disulphide bond between β sheets are highlighted in yellow. C15, which forms a disulphide link with the 5 transmembrane helix C-terminal domain of CD47, is starred. Residues where N-linked glycosylation was observed in at least one monomer of CD47 are colored magenta.
(B) CD47 (shown as ribbons) crystallized as a dimer where the G strands are swapped between pairs of molecules. The structure is color ramped from blue (N terminus) to red (C terminus), and the strands are labeled A–G. The three insets show the N-terminal pyroglutamate, the region between the domains together with the Mg2+ that is assumed to facilitate the strand-swap, and one of the sites of N-linked glycosylation. In all insets, the final refined model is shown in 2FO-FC electron density (1.3 σ) calculated at the end of refinement.
Molecular Cell  , DOI: ( /j.molcel )
Copyright © 2008 Elsevier Inc. Terms and Conditions

4. Figure 3 The Structure of CD47 in Complex with SIRPα
(A) The structure of CD47 (yellow ribbons) complexed with SIRPα d1 (blue ribbons). Sites of N-linked glycosylation are shown as magenta spheres. A schematic representation of the 5 transmembrane helix C-terminal domain of CD47 is shown, including the proposed disulfide bond between C15 (Gly in our structure; yellow sphere) and C245. The inset shows a close-up view of the CD47/SIRPα interaction interface.
(B) The structure of CD47 in complex with SIRPα (yellow) is almost indistinguishable from that of CD47 solved in isolation (orange and gray). Remarkably, the interaction of strand G with the rest of the CD47 monomer is almost identical regardless of whether it is strand swapped or not. The inset shows L101 and T102, which mediate the turn at the tip of the FG loop in the non-strand-swapped CD47 monomer.
(C) The surface of SIRPα as viewed by CD47. The molecular surface of SIRPα d1 is shown, colored by electrostatic potential. Regions of CD47 that interact with SIRPα are shown as sticks. In (C) and (D), areas of particular interest are identified by roman numerals (see [E]) and loop/strand names are shown, with NA denoting the N-terminal 6 residues of CD47.
(D) The surface of CD47 as viewed by SIRPα. The molecular surface of CD47 is shown, colored by electrostatic potential. Loops of SIRPα that interact with CD47 are shown as sticks.
(E) Selected portions of the interface. Residues of CD47 (yellow carbon atoms) and SIRPα (blue carbon atoms) are shown as sticks, and water molecules are shown as red spheres. The molecular surface of SIRPα is shown (white, semitransparent). Hydrogen bonds and salt bridges are shown as orange dashes. The roman numerals of the panels identify the regions of the interaction on the molecular surfaces in (C) and (D).
Molecular Cell  , DOI: ( /j.molcel )
Copyright © 2008 Elsevier Inc. Terms and Conditions

5. Figure 4
The Structures of SIRP Family Members Reveal Structural Flexibility at the Site of CD47/SIRPα Binding and Polymorphism Away from This Interface
(A) Sequence alignment of SIRP family members (accession numbers: SIRPα, CAA71403; SIRPα (sequence 2), NP_542970; SIRPβ, NP_006056; SIRPβ(2), CAI21700; SIRPγ, NP_061026). The sequence of SIRPα is shown in black text as are residues of SIRPα(2), SIRPβ, SIRPβ(2) and SIRPγ that differ from SIRPα. All other residues are shown in pale gray. Residues of SIRPα that interact with CD47 are shown in bold face and highlighted in light blue. The cysteine residues that form a conserved disulphide bond between β sheets are highlighted in yellow. Secondary structure is shown above the sequence with arrows and cylinders representing β sheets and α helices, respectively.
(B) Backbone traces are shown for the structures of SIRPβ (magenta lines, 2 copies), SIRPβ(2) (violet lines, 2 copies), SIRPγ (green lines, 1 copy), SIRPα in isolation (thin blue lines, 2 copies; PDB code 2uv3; Hatherley et al., 2007), and SIRPα in complex with CD47 (thick blue lines, 4 copies). The superposition of all SIRP molecules (residues 4–114; SIRPα numbering) was performed using THESEUS (Theobald and Wuttke, 2006). The molecular surface of CD47, oriented relative to SIRPα as in the CD47/SIRPα complex, is shown for reference (yellow semitransparent surface and yellow ribbons). Two orthogonal views are shown, the first corresponding to the view of the complex presented in Figure 3A.
(C) Mapping onto the structure of SIRPα in complex with CD47 reveals that polymorphisms in SIRPα segregate from the CD47-interaction interface. The positions of polymorphic side chains are indicated by red spheres. Three orthogonal views are shown.
Molecular Cell  , DOI: ( /j.molcel )
Copyright © 2008 Elsevier Inc. Terms and Conditions

6. Figure 5
The Inability of SIRPβ and SIRPβ(2) to Bind CD47 Results from Disruption of Specific CD47/SIRPα Interactions
(A) Left panel: the structure of SIRPα (blue carbon atoms) in complex with CD47 (yellow carbon atoms). Right panel: the structure of SIRPβ (magenta carbon atoms). Replacement of V27 with Met causes a rearrangement of I36 side chain to a conformation incompatible with the position of CD47 T102 (semitransparent sticks, yellow carbon atoms) as seen in the CD47/SIRPα complex.
(B) Left panel: the structure of SIRPα (blue carbon atoms) in complex with CD47 (yellow carbon atoms). Right panel: the structure of SIRPβ(2) (violet carbon atoms). Replacement of D100 with His prevents it from forming a salt bridge with SIRPβ(2) K96. The extended conformation of K96 is inconsistent with CD47 binding, as it would clash with the side chain of CD47 Y37. For (A) and (B), final refined models are shown in 2FO-FC electron density (1.3 σ) calculated at the end of refinement.
(C) Surface plasmon resonance shows that replacement of SIRPβ(2) H100 with Asp confers SIRPα-like binding affinity. The bar indicates the duration of injection of CD47 over the SIRPs with strong binding to SIRPα, SIRPβ(2) H100D, and no binding to SIRPβ(2) or the negative control CD4d3+4. The right panel shows equilibrium binding for CD47 to SIRPα, SIRPβ(2) H100D, and SIRPβ(2).
Molecular Cell  , DOI: ( /j.molcel )
Copyright © 2008 Elsevier Inc. Terms and Conditions