Follicular dendritic cells control engulfment of apoptotic bodies by secreting Mfge8

Abstract

The secreted phosphatidylserine-binding protein milk fat globule epidermal growth factor 8 (Mfge8) mediates engulfment of apoptotic germinal center B cells by tingible-body macrophages (TBMphis). Impairment of this process can contribute to autoimmunity. We show that Mfge8 is identical to the mouse follicular dendritic cell (FDC) marker FDC-M1. In bone-marrow chimeras between wild-type and Mfge8(-/-) mice, all splenic Mfge8 was derived from FDCs rather than TBMphis. However, Mfge8(-/-) TBMphis acquired and displayed Mfge8 only when embedded in Mfge8(+/+) stroma, or when situated in lymph nodes draining exogenous recombinant Mfge8. These findings indicate a licensing role for FDCs in TBMphi-mediated removal of excess B cells. Lymphotoxin-deficient mice lacked FDCs and splenic Mfge8, and suffer from autoimmunity similar to Mfge8(-/-) mice. Hence, FDCs facilitate TBMphi-mediated corpse removal, and their malfunction may be involved in autoimmunity.

Figure 1.

FDC-M1 and Mfge8 are identical. (A) Two-color immunolabeling of a WT spleen stained with anti-Mfge8 antibody 18A2-G10 (green) and anti–FDC-M1 antibody 4C11 (red), or anti-Mfge8 antibody 2422 (green) and 4C11 (red). Both anti-Mfge8 antibodies showed colocalization with 4C11. (B) Preincubation with 25 μg/ml rMfge8 blocked the labeling of FDCs with 18A2-G10 or 4C11. To visualize FDCs, sections were stained with anti-PrP antibody POM2. For control, sections were preincubated with rEGF or rPrP. Bars, 100 μm. (C) Splenic protein extracts (WT and Mfge8^−/−) were immunoprecipitated with 2422 or 4C11, or to a rat IgG2c isotype control antibody. Control beads were coupled with 2422 or 4C11 but were not exposed to splenic extracts. Western blots were probed with 18A2-G10. Mfge8-specific bands are indicated (* and x). Arrows indicate nonspecific bands. (D) Sensograms indicating binding of each protein after subtraction of their binding to a control protein-coupled surface. Black arrows indicate antibody injections. (insets) Schematic representations of binding and competition events. For control, injections of all proteins were made on two flow cells, with one cell coated with control protein and another coated with the protein of interest.

Figure 2.

Analysis of splenic Mfge8 expression by immunofluorescence and ISH. (A) BM chimeras were stained with anti-Mfge8 antibody (clone 2422; green) and anti-CD68 antibody (red). Mfge8 immunoreactivity of FDCs and CD68⁺ TBMφs was only observed when FDCs were of WT origin. Figures show areas inside follicles. White squares mark the areas shown at a higher magnification. Bars, 20 μm. (B) Splenic Mfge8 expression was assessed by ISH. Consecutive sections were immunolabeled with 4C11 and anti-CD68. Mfge8 expression and 4C11 immunostaining was only found in WT mice irrespective of the BM genotype (top and second from bottom). Mfge8^−/− mice receiving BM from either Mfge8^−/− or WT mice showed no Mfge8-specific signal after ISH and no 4C11 immunostaining. Bars, 100 μm.

Figure 3.

TBMφs bind extracellular Mfge8. Footpads of Mfge8^−/− mice were injected with 10 μg rMfge8. For control, PBS or rPrP was injected into the contralateral footpad of Prnp^o/o mice. 20 h later, popliteal lymph nodes were collected and analyzed by immunofluorescence with anti-Mfge8 and anti-PrP antibodies. TBMφs and FDCs showed strong and weak Mfge8 staining, respectively. rPrP was undetectable in the lymph nodes of Prnp^o/o mice. Bars, 50 μm.

Figure 4.

Impaired engulfment of apoptotic bodies in the absence of stromal Mfge8. (A) Apoptotic cells, TBMφs, and GCs were visualized by TUNEL, anti-CD68, and PNA, respectively, on splenic cryosections 9 wk after BM reconstitution and after immunization. (right) Each datapoint represents the mean number of TUNEL⁺ cells per TBMφ in one individual GC. Mfge8^−/−→Mfge8^−/− and WT→Mfge8^−/− mice showed increased numbers of TUNEL⁺ cells per TBMφ. Horizontal bars represent means. White circles (left) indicate GCs. Bars, 100 μm. (B) Ultrastructural features of TBMφs of aged BM-chimeric mice 41 wk after reconstitution. Apoptotic cells in various degradation stages were observed inside TBMφs of all chimeric mice. (C) Engulfment of apoptotic cells by TBMφs in WT, Mfge8^−/−, Ltbr^−/−, Lta^−/−, and Ltb^−/− mice was analyzed by TUNEL (green) and CD68 (red) staining. WT TBMφs contained copious TUNEL⁺ material. The latter was also observed in Mfge8^−/− mice, but most TUNEL⁺ cells were large and intact. Ltbr^−/−, Lta^−/−, and Ltb^−/− macrophages were small and only contained intact TUNEL⁺ cells. At least three mice per genotype and ≥10 follicles per mouse were analyzed. Bars, 20 μm.

Figure 5.

Mfge8-dependent removal of apoptotic cells from the GC. (A) In acute inflammatory conditions, macrophages secrete Mfge8 (step 1; reference 30), which targets apoptotic bodies (step 2) and allows macrophages to bind them via integrins (step 3). The Mfge8–integrin interaction results in engulfment (step 4). (B) Revised model of splenic Mfge8-dependent engulfment. In this case, Mfge8 is not produced by macrophages but by FDCs. These establish a local Mfge8 gradient within GCs (step 1). TBMφs are therefore licensed for selective engulfment at the sites of apoptosis (step 2). Apoptotic cells are presumably decorated by Mfge8 while in contact with FDCs (step 3).