CRP and SAP bind to FcαRI in solution. (A) The binding between serial dilutions of CRP, SAP, or PTX3 in micromolar concentrations (u) and immobilized FcαRI on a CM5 sensor chip. (B) The D1 (green) and D2 (blue) domains of FcγRIIa (3D5O), FcεRI (1F2Q), FcαRI (1OW0), and KIR2DL2 (2DL2) are shown in respective D2 orientations. The CRP binding sites on FcγRIIa and the IgA binding site on FcαRI are shown as surface patches.

CRP binds to FcαRI on transfected RBL cells. (A and B) Anti-FcαRI (MIP8a) (gray) or isotype control (black) staining of RBL cells (A) or FcαRI-transfected RBL (G248) cells (B). (C) CRP (150 μg/mL) bound to G248 cells (gray) better than to RBL cells (black). The CRP binding to G248 cells was blocked by MIP8a (heavy dashed line) but not by its isotype control (thin dashed line). Filled areas represent unstained cells. Horizontal axes show fluorescence intensities. (D) Dose-dependent CRP binding to G248 or RBL cells detected using FITC-2C10. Data are representative of at least three experiments.

CRP induces neutrophil FcαRI surface expression, phagocytosis, and TNF-α secretion. (A and B) Confocal images of neutrophils stained with anti-FcαRI (A59) and AF488 goat anti-mouse (green). Neutrophils were incubated for 30 min at room temperature with PKH26 (red)-labeled (A) or CRP-opsonized (B) PnC-SRBC. (C) Uptake of CRP-opsonized FITC-S. pneumoniae by neutrophils, expressed as phagocytic index (bacteria per 100 neutrophils), with or without inhibitors. Data are mean ± SEM of four experiments. ***P < 0.001; **P < 0.01. (D) CRP or IgA (200 μg/mL) cross-linking induced TNF-α release in human neutrophils. The TNF-α secretion upon either CRP or IgA treatment was inhibited by the Fab fragment of MIP8a.

CRP and IgA bind at nonidentical sites on FcαRI. (A) The solution binding response for CRP (2.9 μM), recombinant FcαRI (4.7 μM), or their combination onto an IgA immobilized CM5 sensor chip. CRP alone did not bind IgA. (B) The binding of CRP alone (red) in the presence of IgA (1,000 μg/mL) (green), or mAb A59 (blue) to G248 cells (solid lines) and RBL cells (dashed lines). Unstained G248 cells (gray line) and RBL cells (shaded gray) are shown. Horizontal axes of the histograms show fluorescence intensities. (C) Cy3-IgA binding to G248 cells alone (green) or in the presence of 150 μg/mL CRP (red), blocking MIP8a (blue shaded), and nonblocking anti-FcαRI mAb A59 (blue line). Unstained G248 cells (black line) and Cy3-IgA–stained RBL cells (shaded gray) are shown. (D) Bar graph shows the geometric mean channel fluorescence for Cy3–IgA binding.

FcαRI recognizes the CRP effector face. (A) Competitive solution binding between C1q and immobilized FcαRI or FcγRIIa to CRP using BIAcore. Recombinant FcαRI and FcγRIIa were immobilized individually on CM5 chips. The analytes were CRP (2.7 μM) in the presence or absence of 0.25 or 1 μM C1q. C1q alone resulted in close to zero response at 0.25 μM and negative responses at 1 μM, probably because of higher binding to the dextran sulfate surface in the reference cell. (B) The equilibrium binding responses of wild-type and mutant CRP to immobilized FcαRI. (Sensorgrams are shown in Fig. S2.) (C) The structure of the SAP–FcγRIIa complex (3D5O) and the docked CRP–FcαRI model in two orthogonal views. The mutation sites used in B are shown by green sticks. The putative interface residues in the CRP–FcαRI model are listed in Table S1.

CRP induces ERK phosphorylation, degranulation, and cytokine secretion in FcαRI-transfected RBL cells. (A) FcαRI-transfected 9.4 cells or untransfected RBL cells were preincubated with CRP or IgA (200 μg/mL), then incubated with 2C10, anti-IgA, or MIP8a at time point 0. Cells were lysed at time point 0 and at 5 min, and phospho-ERK was detected by Western blot. Blots were stripped and reprobed for total ERK. Results are representative of three experiments. (B) RBL 9.4 cells (open bars) or untransfected RBL cells (solid bars) were treated as in A, and ERK phosphorylation was determined at 5, 10, and 15 min by flow cytometry. (C) FcαRI-transfected 9.4 cells (solid lines) or RBL cells (dashed lines) were preincubated with CRP (circle) or IgA (square), then cross-linked with 2C10 or anti-IgA. No significant β-hexosaminidase release was detected in 9.4 cells treated with 2C10 alone (diamond). β-Hexosaminidase release was measured and expressed as the percentage of total activity. Mean ± SEM of triplicate wells from one experiment are shown. (D) β-Hexosaminidase release was measured as in C except on FcαRI-transfected G248 cells with streptavidin cross-linked biotin-CRP. (E) IL-4 secretion in G248 cells (open bars) or untransfected RBL cells (solid bars) treated with streptavidin cross-linked with CRP or without piceatannol.