Plexin-B2 negatively regulates macrophage motility, Rac, and Cdc42 activation

Abstract

Plexins are cell surface receptors widely studied in the nervous system, where they mediate migration and morphogenesis though the Rho family of small GTPases. More recently, plexins have been implicated in immune processes including cell-cell interaction, immune activation, migration, and cytokine production. Plexin-B2 facilitates ligand induced cell guidance and migration in the nervous system, and induces cytoskeletal changes in overexpression assays through RhoGTPase. The function of Plexin-B2 in the immune system is unknown. This report shows that Plexin-B2 is highly expressed on cells of the innate immune system in the mouse, including macrophages, conventional dendritic cells, and plasmacytoid dendritic cells. However, Plexin-B2 does not appear to regulate the production of proinflammatory cytokines, phagocytosis of a variety of targets, or directional migration towards chemoattractants or extracellular matrix in mouse macrophages. Instead, Plxnb2(-/-) macrophages have greater cellular motility than wild type in the unstimulated state that is accompanied by more active, GTP-bound Rac and Cdc42. Additionally, Plxnb2(-/-) macrophages demonstrate faster in vitro wound closure activity. Studies have shown that a closely related family member, Plexin-B1, binds to active Rac and sequesters it from downstream signaling. The interaction of Plexin-B2 with Rac has only been previously confirmed in yeast and bacterial overexpression assays. The data presented here show that Plexin-B2 functions in mouse macrophages as a negative regulator of the GTPases Rac and Cdc42 and as a negative regulator of basal cell motility and wound healing.

Figure 1. Plexin-B2 is expressed in the human and mouse immune system.

cDNA expression level of Plxnb2 in A) mouse and B) human immune cells from BioGPS database . C) Plexin-B2 protein expression by flow cytometry of ex-vivo mouse spleen and bone marrow cells. pDCs were gated on B220⁺ PDCA-1⁺ cells. cDCs were gated on CD11c⁺ B220⁻ cells. Macrophages were gated on F4/80⁺ CD11c⁻ cells. B cells were gated on B220⁺ cells. T cells were gated on TCR⁺ cells. NK cells were gated on NK1.1⁺ TCR⁻ cells, and NK T cells on NK1.1⁺TCR⁺ cells. Wild type fetal liver reconstituted mice stain positive with anti-Plexin-B2 antibody in cells that express Plexin-B2 and are indicated by unfilled histograms. Plxnb2^−/− reconstituted mice do not stain positive for Plexin-B2 and are indicated by filled histograms. Results are representative of at least four experiments, n = 8 per group.

Figure 2. Schematic of Fetal Liver Transplant.

A Plxnb2^+/− male is crossed to a Plxnb2^+/− female. Fetal liver cells are harvested from Plxnb2^+/+ or Plxnb2^−/− littermate fetal livers at day 14 of gestation. 2×10⁶ fetal liver cells are injected i.v. into congenic, lethally irradiated CD45.1 recipient mice. Mice are utilized 6–8 weeks post-transplant.

Figure 3. Immune system reconstitution with Plxnb2^−/− fetal livers is similar to wild type.

A) Flow cytometry of reconstitution of congenic CD45.1⁺ mice with wild type and Plxnb2^−/− fetal livers (CD45.2⁺) in the splenic and bone marrow compartments six weeks post-transplant. cDCs are designated as B220−/CD11c+ cells, pDCs as CD11c⁺ B220⁺, and B cells as B220⁺ CD11c⁻. pDCs were further defined as CD11c^low and PDCA-1⁺. Macrophages were defined as F4/80⁺ cells. T cells were first gated on TCR⁺ cells, and then divided into subpopulations by the CD4⁺ and CD8⁺ markers. NK cells were defined as NK1.1+/CD4− and NKT cells as NK1.1+CD4+. B) Total donor reconstitution in the spleen and bone marrow were similar between wild type and Plxnb2^−/− mice. CD45.2 marked the donor, Ly5.1⁺ cells. CD45.1 marked the residual host, Ly5.2⁺ cells. Figures are representative plots of four separate experiments. n = 5 mice per group.

Figure 4. Plexin-B2 does not affect cytokine secretion.

Wild type and Plxnb2^−/− macrophages were plated in complete DMEM and treated for 16 hours with TLR ligands Poly(I∶C) (10 ug/ml) (TLR 3 agonist), LPS (1 ug/ml) (TLR 4 agonist), and R837 (4 ug/ml) (TLR 7 agonist). Poly(I∶C), LPS, and R837 were purchased from InvivoGen. Supernatants were collected at the 16 hour time point and assessed for secretion of A) TNF and B) IL-6 by ELISA. Graphs are representative of three independent experiments, n = 3 mice per group.

Figure 5. Plxnb2^−/− cells have higher motility than wild type macrophages.

A) Cells were plated in complete DMEM, rested overnight, and treated with sham (warmed DMEM) or warmed DMEM and 50 ug/ml M-CSF. Representative cell movement tracks over a 2.5 hour time period on sham or M-CSF treated wild type or Plxnb2^−/− macrophages. B) Combined velocity scores of wild type and Plxnb2^−/− bone marrow macrophages in sham or treated (50 ng/ml M-CSF) groups in 5A. Cell velocity was scored in 45 cells per group per experiment. The experiment was repeated three separate times using cells from different mice in each experiment. C) Macrophage transwell migration towards CSF, CXCL12, or FCS. Cells were placed in the upper chamber in plain DMEM and chemokines in plain DMEM in the lower chamber of migration plates. Cells were allowed to migrate for four hours towards different concentrations of cytokines (M-CSF, 0–500 ng/ml or CXCL12, 0–1000 ng/ml) or overnight towards serum (0–20%). Cells were quantified and normalized to a standard curve of each cell type. Graphs are representative of at least three independent experiments. n = 4 mice per group.

Figure 6. Plexin-B2 inhibits active Rac and Cdc42.

WT and Plxnb2^−/− macrophages were plated, rested overnight, and treated with sham (warmed DMEM) or M-CSF (50 ng/ml of M-CSF in warmed DMEM) and assessed for A) RacGTP, Cdc42GTP, total Rac and total Cdc42 using a GST-PBD pulldown assay followed by western blot. B) Western blot of ERK activation as assessed by phospho-ERK level in cells treated as above. Total protein level was determined by western blot of GAPDH. Results are representative of four separate experiments. n = 4 mice per group.

Figure 7. Plexin-B2 does not significantly affect phagocytosis.

Wild type and Plxnb2^−/− bone marrow derived macrophages were incubated with A) GFP-E. coli, B) latex beads, or C) opsonized thymocytes (1∶10 ratio E. coli, bead, or T cells: macrophage) for 30, 60, or 90 minutes. Uptake was measured by fluorescence and flow cytometry. Graphs are representative or four independent experiments. n = 5 mice per group.

Figure 8. Plxnb2^−/− cells demonstrate faster wound healing than wild type macrophages.

A) F-actin distribution in sham and M-CSF treated macrophages. Wild type and Plxnb2^−/− macrophages were plated on glass coverslips, rested overnight, and fixed following sham (plain DMEM) treatment or treatment with 50 ng/ml M-CSF in DMEM for 2 or 10 minutes. Cells were imaged on an Olympus IX-70 (Center Valley, PA). Pictures are representatives of two different experiments with 10 pictures per treatment group, with cells from four mice per group. B) Migration of macrophages into extracellular matrix protein. Wild type and Plxnb2^−/− macrophages were seeded into the top chamber of a fibronectin coated transwell apparatus in plain DMEM. Cells were allowed to migrate towards plain DMEM, DMEM with 100 ng/ml M-CSF, and DMEM with 10% FBS for 20 hours and quantified by optical density. Graph is representative of two separate experiments performed with cells from four mice per group. C) Adhesion to extracellular matrix protein. Wild type and Plxnb2^−/− macrophages were plated on fibronectin (20 ug/ml) coated plates and allowed to adhere for 20 minutes. Plates were washed and adherent cells were quantified by ToxiLight Bioassay Kit (Lonza, Rockland, ME). Graph is representative of two separate experiments with cells from four mice per group. D) Wound healing assay. Wild type and Plxnb2^−/− macrophages (1×10⁶) were plated on 12 well plates. An artificial wound was introduced using a 200 ul pipette tip, and cells were imaged closing the wound every 10 minutes at four points per scratch for 24–48 hours using an Olympus-IX-70 inverted microscope with live cell imaging capability. The rate of change of the area of the scratch was measured over time using the NIH ImageJ program. The calculated area is in terms of pixels, and the rate of movement is calculated as pixels/sec and converted to micrometers/sec. Experiments were repeated two times with 6 wells per group, using cells from four different mice per group.