FLIP: a novel regulator of macrophage differentiation and granulocyte homeostasis

Abstract

FLIP is a well-established suppressor of death receptor-mediated apoptosis. To define its essential in vivo role in myeloid cells, we generated and characterized mice with Flip conditionally deleted in the myeloid lineage. Myeloid specific Flip-deficient mice exhibited growth retardation, premature death, and splenomegaly with altered architecture and extramedullary hematopoiesis. They also displayed a dramatic increase of circulating neutrophils and multiorgan neutrophil infiltration. In contrast, although circulating inflammatory monocytes were also significantly increased, macrophages in the spleen, lymph nodes, and the peritoneal cavity were reduced. In ex vivo cultures, bone marrow progenitor cells failed to differentiate into macrophages when Flip was deleted. Mixed bone marrow chimera experiments using cells from Flip-deficient and wild-type mice did not demonstrate an inflammatory phenotype. These observations demonstrate that FLIP is necessary for macrophage differentiation and the homeostatic regulation of granulopoiesis.

Figure 1

Deletion of Flip in myeloid linage results in postnatal growth retardation and premature death. (A) Schematic of the Flip targeting strategy. A 9.5-kb Flip genomic DNA segment (exon 2 to 4 from a C57BL/6J mouse BAC library) was used to engineer the Flip conditional mutagenesis construct. Exons are indicated by rectangular boxes and E and B represent the restriction cleavage sites of EcoRI and BamHI. Three loxP sites were inserted as indicated by triangles. The dashed lines represent the flanking region for Southern blot probes. The location of 3 PCR primers (a, b, c) and orientation are indicated by arrows. Amplification using primers a and b generates a 220-bp fragment for wild-type (Flip⁺) and a 270-bp fragment for the floxed (Flip^f) allele. Cre-induced recombination generates a 150-bp fragment for depleted Flip allele (Flip^d) using primers a and c. All recombinant DNA and animal procedures were approved by the Office of Research Safety and the Institutional Animal Care and Use Committee of Northwestern University. (B) PCR genotyping of the LysM-cre–induced cell type-specific deletion of Flip^f. The representative Flip^f/+, LysM^c/+ littermate control and Flip^f/d, LysM^c/+ knockout (KO) mice were genotyped from tail biopsy, and the different cell types were isolated from bone marrow (BM), and peripheral blood were genotyped by 3 PCR primers indicated in panel A. Granulocytes (Gran) were 11b⁺/Gr1⁺ F4/80⁻; peripheral blood monocytes or bone marrow monocyte precursors (Mono) were 11b⁺, F4/80⁺; B cells were 11b⁻/CD19⁺, and T cells were 11b⁻CD3⁺. (C) Body weight of Flip^f/d, LysM^c/+ (n = 53) and littermate controls including: Flip^+/+, Flip^f/+ or Flip^f/f, LysM^+/+ (n = 9); Flip^f/d or Flip^+/d, LysM^+/+ (n = 10); Flip^+/+, LysM^c/+ (n = 5); Flip^f/+or Flip^+/d, LysM^c/+ (n = 24); and Flip^+/+, Flip^f/+or Flip^+/d, LysM^c/c (n = 12). All mice are between 6 to 24 weeks of age. ***P < .001, compared with indicated groups. (D) Postnatal viability of Flip^f/d, LysM^c/+ (n = 101) and the Flip^f/d, LysM^c/c (n = 17) and littermate controls (n = 475), which included: Flip^+/+, Flip^f/+ or Flip^f/f, LysM^+/+ (n = 112), Flip^f/d or Flip^+/d, LysM^+/+ (n = 112), Flip^+/+, LysM^c/+ (n = 47), Flip^f/+or Flip^+/d, LysM^c/+ (n = 127), and Flip^+/+, Flip^f/+or Flip^+/d, LysM^c/c (n = 77) mice.

Figure 2

Flip deletion in myeloid linage results in leukocytosis. (A) Peripheral blood from Flip^f/d, LysM^c/+ mice (n = 56) and their littermate controls (mixed genotypes, n = 50) were examined for completely blood count, and neutrophils and monocytes are presented. ***P < .001 compared with the controls. (B) Representative blood smears from a Flip^f/d, LysM^c/+ mice and controls stained with Hema-3. Data are representative of smears from 3-4 mice for each group. (C) Representative flow cytometric analysis of circulating monocytes and neutrophils of Flip^f/d, LysM^c/+ and sex-matched littermate controls. Cells are gated by side scatter (SSC) and the expression of CD115, CD62L, and Gr1. (D) Absolute cell count of resident and inflammatory monocytes calculated from the total number of monocytes and the percentage of each subset using 5 Flip^f/d, LysM^c/+ and sex-matched littermate controls.

Figure 3

Flip deletion in myeloid linage results in multiorgan neutrophil infiltration. Hematoxylin and eosin staining of representative sections of tissues from the indicated organs from Flip^f/d, LysM^c/+ and control mice. The area in the box is enlarged in the panel below. Data are representative of sections from 3-4 mice for each group.

Figure 4

Normal apoptosis and function in Flip-deficient circulating neutrophils. (A) Time-dependent loss of mitochondrial transmembrane potential (Δ Ψ_m) was assessed by decreased Rh123 fluorescence (x-axis) and the loss of membrane integrity assessed by uptake of DAPI (y-axis). (B) The percentage of live cells is identified as Rh123⁺, DAPI⁻ (n = 4 for each group); *P < .05 and **P < .01 between groups. (C) Representative myeloperoxidase staining of peripheral blood smears, observed by light microscopy (400×). Data are representative of 3 Flip^f/d, LysM^c/+ and sex-matched littermate controls. (D) The ability to oxidize nonfluoresent dihydrorhodamine 123 was accessed as increased mean florescence intensity after PMA activation. (E) Neutrophil degranulation was determined by increased mean florescence intensity of cell surface CD11b after PMA stimulation. The observations were obtained from 5 Flip^f/d, LysM^c/+ and sex-matched littermate controls.

Figure 5

Decreased macrophages and FLIP expression in the spleen and lymph node. Spleen size (n = 31; A) and total number of spleen cells (n = 16; B) in Flip^f/d, LysM^c/+ and littermate controls. A representative flow histogram of forward scatter (FSC) from ungated splenocytes of Flip^f/d, LysM^c/+ and littermate control mice is presented in the inset of panel B. Immunofluorescence microscopy of spleen was performed to identify red pulp macrophages (anti-F4/80; C) or marginal zone macrophages (anti-CD169; D), and of lymph nodes with anti-F4/80 antibodies (E). The data are representative of sections from 3-4 mice of each group. The area in the box is enlarged in the bottom panel. Randomly selected spleens (F) and lymph nodes (G) from littermate controls and Flip^f/d, LysM^c/+ mice were used to examine the expression of FLIP determined by immunoblot analysis. The data are representative of > 4 mice for each group.

Figure 6

Flip deletion results in increased neutrophils and decreased macrophages in the peritoneum. Resident (A-B) or thioglycollate elicited (E-F) cells from peritoneal cavities were analyzed by flow cytometry, gaiting on CD11b⁺ cells. The macrophages (MΦ) were identified as F4/80⁺, Gr1⁻, whereas granulocytes (Gran) were identified as F4/80⁻, Gr1⁺. (B) The total number of cells, resident macrophages and granulocytes from controls (n = 20) and Flip^f/d, LysM^c/+ mice (n = 26) are summarized. (C) There was an inverse relationship between the percentage of peritoneal macrophages and granulocytes in the wild-type and Flip^f/d, LysM^c/+ mice. (D) PCR genotyping of granulocytes and macrophages isolated from peritoneal cavities were performed using the 3 PCR primers indicated in Figure 1. (F) The total number of thioglycollate elicited macrophages and granulocytes at 72 hours (n = 7 for each group) are summarized. *P < .05, **P < .01, and ***P < .001 between groups.

Figure 7

Deletion of Flip suppresses macrophage differentiation. cKit⁺ hematopoietic stem cells from Flip^f/+ and Flip^f/d were isolated from bone marrow and seeded at 2.5 × 10⁵ cells/well followed by infection of retroviral vectors expressing GFP alone (A-B) or GFP and Cre (C-D). After the infection, the cells were differentiated in vitro to macrophages in 20% L929 medium. The total number of GFP⁺ cells (A,C) and the GFP⁺, CD11b⁻, F4/80⁻, the GFP⁺, CD11b⁺, F4/80⁻, and the GFP⁺, CD11b⁺, F4/80⁺ cells (B,D) were determined after 1, 2, and 5 days of differentiation. Data in panels B,D represents the mean ± SEM of 4 independent experiments.