Granulocyte macrophage-colony stimulating factor: A key modulator of renal mononuclear phagocyte plasticity

Abstract

Macrophage-colony stimulating factor (M-CSF) and granulocyte macrophage-colony stimulating factor (GM-CSF) play key roles in the differentiation of macrophages and dendritic cells (DCs). We examined the effect of treatment with M-CSF-containing macrophage medium or GM-CSF-containing DC medium upon the phenotype of murine bone marrow-derived macrophages and DCs. Culture of macrophages for 5 days in DC medium reduced F4/80 expression and increased CD11c expression with cells effectively stimulating T cell proliferation in a mixed lymphocyte reaction. DC medium treatment of macrophages significantly reduced phagocytosis of both apoptotic cells and latex beads and strongly induced the expression of the chemokine receptor CCR7 known to be involved in DC trafficking to lymph nodes. Lysates of obstructed murine kidneys expressed both M-CSF and GM-CSF though M-CSF expression was dominant (M-CSF:GM-CSF ratio ∼30:1). However, combination treatment with both M-CSF and GM-CSF (ratio 30:1) indicated that small amounts of GM-CSF skewed macrophages towards a DC-like phenotype. To determine whether macrophage phenotype might be modulated in vivo we tracked CD45.1⁺ bone marrow-derived macrophages intravenously administered to CD45.2⁺ mice with unilateral ureteric obstruction. Flow cytometry of enzyme dissociated kidneys harvested 3 days later indicated CD11c and MHC Class II upregulation by adoptively transferred CD45.1⁺ cells with CD45.1⁺ cells evident in draining renal lymph nodes. Our data suggests that GM-CSF modulates mononuclear phagocyte plasticity, which likely promotes resolution of injury and healing in the injured kidney.

Keywords: DC; GM-CSF; Inflammation; M-CSF; Macrophage; Mononuclear phagocyte.

Fig. 1

Macrophages can be induced to adopt properties characteristic of dendritic cells and vice versa. Bone marrow cells were grown in macrophage (Mφ) or dendritic cell (DC) media for 7 days prior to assessment of cell surface expression of F4/80 (a), CD11c (b) and MHC Class II (c) by flow cytometry. The level of expression is expressed as mean fluorescent intensity (MFI). Day 7 bone marrow-derived Mφ and DCs were then washed and incubated for a further 5 days in DC medium or Mφ medium respectively. Control Mφ and DCs underwent further culture in Mφ or DC medium respectively. At day 12, cells were recovered and the expression of F4/80 (d), CD11c (e) and MHC Class II (f) was determined. Again, day 7 bone marrow-derived Mφ and DCs were either switched to DC or Mφ medium for a further 5 days. Now cells were recovered at days 8, 10 and 12 and real-time PCR undertaken for CCR7 mRNA expression (g). Day 12 cells from control Mφ or DC cultures or ‘medium switched’ cultures from C57BL/6 mice were replated with splenocytes from BALB/c animals for 48 h. 1 mCi of [3 H]-TdR was added to each well and plates incubated overnight before harvesting and counting using a liquid scintillation counter to measure T cell proliferation. Data expressed as counts per minute (cpm) (h). Results are representative of multiple experiments (>3). * p < 0.05, ** p < 0.01, ***p < 0.

Fig. 2

Culture in DC medium reduces macrophage phagocytosis of apoptotic cells and beads. Day 7 bone marrow-derived Mφ were incubated for 5 days in macrophage (Mφ) or dendritic cell (DC) medium (n = 3 mice). Cells were then incubated with CellTracker Green labeled apoptotic murine thymocytes or 3 μm Fluoresbrite YG microsphere beads for 60 min. Cells were vigorously washed and detached from the plates for flow cytometric assessment of the proportion of cells that had phagocytosed apoptotic cells or beads. Representative flow cytometry dot plots of Mφ cultured in Mφ medium (cells alone (a), following incubation with beads (c) or apoptotic cells (e)) or DC medium (cells alone (b), following incubation with beads (d) or apoptotic cells (f). Quantification of the proportion of cells exhibiting phagocytosis demonstrates that Mφ cultured in Mφ medium (Mφ/Mφ) remain highly phagocytic but exhibit reduced phagocytic capacity following incubation in DC medium (Mφ/DC) (g & h). ** p < 0.01, ***p < 0.001 (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article).

Fig. 2

Culture in DC medium reduces macrophage phagocytosis of apoptotic cells and beads. Day 7 bone marrow-derived Mφ were incubated for 5 days in macrophage (Mφ) or dendritic cell (DC) medium (n = 3 mice). Cells were then incubated with CellTracker Green labeled apoptotic murine thymocytes or 3 μm Fluoresbrite YG microsphere beads for 60 min. Cells were vigorously washed and detached from the plates for flow cytometric assessment of the proportion of cells that had phagocytosed apoptotic cells or beads. Representative flow cytometry dot plots of Mφ cultured in Mφ medium (cells alone (a), following incubation with beads (c) or apoptotic cells (e)) or DC medium (cells alone (b), following incubation with beads (d) or apoptotic cells (f). Quantification of the proportion of cells exhibiting phagocytosis demonstrates that Mφ cultured in Mφ medium (Mφ/Mφ) remain highly phagocytic but exhibit reduced phagocytic capacity following incubation in DC medium (Mφ/DC) (g & h). ** p < 0.01, ***p < 0.001 (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article).

Fig. 3

The phenotypic shifts of bone marrow-derived macrophages and dendritic cells are due to whole cell population changes and not to an outgrowth of progenitor cells. Day 7 bone marrow derived macrophages (BMMφ) or dendritic cells (BMDC) were labeled with PKH green. Flow cytometry was carried out to confirm that all cells were labeled with PKH dye (a). The labeled Mφ and DCs were then replated and cultured in either the same medium (control Mφ/Mφ and DC/DC) or switched to the opposite medium (Mφ/DC and DC/Mφ) for a further 5 days. At day 12 cells were analysed by flow cytometry to assess the expression of F4/80 and CD11c (y-axis) and the presence of PKH green (x-axis) (b).

Fig. 4

Exposure to a 50:50 mix of macrophage and dendritic cell medium modulates the phenotype of macrophages and dendritic cells. Bone marrow cells were grown in either macrophage (Mφ) medium, DC medium or a 50:50 mix of Mφ and DC media (Mφ, DC or mix) for 7 days before prior to analysis of cell expression of F4/80 (a), CD11c (b) and MHC Class II expression (c) by flow cytometry. The level of expression is expressed as mean fluorescent intensity (MFI). In further experiments, day 7 bone marrow-derived Mφ were washed and incubated for a further 5 days in either Mφ medium (Mφ/Mφ), DC medium (Mφ/DC) or the 50:50 mix (Mφ/mix). Similarly, day 7 bone marrow-derived DCs were washed and incubated for a further 5 days in either DC medium (DC/DC), Mφ medium (DC/Mφ) or the 50:50 mix (DC/mix). At day 12, cells were recovered and the expression of F4/80 (d, g), CD11c (e, h) and MHC Class II (f, i) was determined by flow cytometry whilst real-time PCR was used to determine CCR7 mRNA expression (j, k). Results are representative of multiple experiments (>3). * p < 0.05, ** p < 0.01, ***p < 0.001.

Fig. 5

The obstructed kidney expresses both M-CSF and GM-CSF with modulation of macrophage and dendritic cell phenotype by recombinant M-CSF and GM-CSF. C57BL/6 mice underwent unilateral ureteric obstruction and kidneys were removed at days 3, 5 and 7 for protein extraction. The concentrations of M-CSF (a) and GM-CSF (b) were determined by ELISA. Bone marrow cells were cultured for 7 days in either recombinant M-CSF (20 ng/ml), GM-CSF (667 pg/ml) or a 30:1 mix (both) of M-CSF and GM-CSF (20 ng/ml M-CSF: 667 pg/ml GM-CSF). At day 7, the expression of F4/80 (c), CD11c (d) and MHC Class II (e) was determined by flow cytometry whilst real-time PCR was used to determine CCR7 mRNA expression (f). The level of expression is expressed as mean fluorescent intensity (MFI). Results are representative of multiple experiments. * p < 0.05, ** p < 0.01, ***p < 0.001.

Fig. 6

The phenotype of bone marrow-derived macrophages is modified following adoptive transfer to mice with ureteric obstruction. 15 × 10⁶ mature day 7 bone marrow-derived macrophages (BMMφ) generated from CD45.1⁺ C57BL/6 mice were intravenously administered to CD45.2⁺ C57BL/6 mice (WT C57BL/6) in 3 separate injections 3–4 days following unilateral ureteric obstruction (UUO). Whole kidneys were recovered and enzymatically dissociated for flow cytometric analysis at day 7 following UUO. Flow cytometry for F4/80 (a), CD11c (b) or MHC Class II (c) was undertaken on day 7 BMMφ on the day of injection and the CD45.1 cells from the digested kidneys. Representative histograms are shown (continuous line - day 7 injected BMMφ, dashed line - CD45.1 cells from obstructed kidney) whilst the level of expression is expressed as mean fluorescent intensity (MFI). ** p < 0.01.

Fig. 7

The phenotype of adoptively transferred bone marrow-derived macrophages localizing to multiple organs. Day 7 mature bone marrow-derived macrophages (BMMφ) generated from CD45.1⁺ C57Bl/6 mice were intravenously administered to CD45.2⁺ animals (WT C57BL/6) 3–4 days after unilateral ureteric obstruction (UUO) had been performed (total cells injected ∼15 × 10⁶) in order to track the fate of the injected BMMφ. The obstructed kidneys, liver, spleen and draining lymph node were recovered at day 7 and enzymatically dissociated for flow cytometric analysis of CD45.1+ cells for expression of F4/80, CD11c, MHC Class II and CD86 (the level of expression is quantified as mean fluorescent intensity [MFI]). CD45.1+ cells that localized to the obstructed kidney exhibited a reduction in F4/80, increased CD11c and a striking increase in MHC Class II expression with no change in CD86 expression (a). This phenotype was not evident in CD45.1+ cells that localized to the spleen (↓F4/80, CD11c & CD86) or liver (↓F4/80 & CD86, ↑ MHC Class II) (b). ** p < 0.01, ***p < 0.001.

Fig. 8

Adoptively transferred bone marrow-derived macrophages localize to the draining renal lymph node in mice with ureteric obstruction. 15 × 10⁶ mature day 7 bone marrow-derived macrophages (BMMφ) generated from CD45.1⁺ C57BL/6 mice were intravenously administered to CD45.2⁺ C57BL/6 mice (WT C57BL/6) in 3 separate injections 3–4 days following unilateral ureteric obstruction (UUO). The lymph nodes draining the left kidney following UUO or sham surgery were removed and CD45.1 expression detected by immunofluorescent microscopy (CD45.1 - Green, DAPI - blue) (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article).

Fig. 9

Schematic diagram of the potential effects of M-CSF and GM-CSF upon macrophages in the obstructed kidney. M-CSF contributes to the switch of M1 inflammatory Mφ to a reparative M2 phenotype (a). GM-CSF, despite being expressed at lower levels compared to M-CSF, also contributes to the M1 to M2 Mφ phenotypic shift (a). GM-CSF causes a reduction of CSF1-R expression (b) possibly through cleavage of the CSF1-R receptor (c).⁴⁷ GM-CSF induces the upregulation of the chemokine receptor CCR7 in Mφ (d) thereby promoting chemotaxis of Mφ to the chemokines CCL19 and CCL21 expressed by lymphatic vessels (e). Thus may, GM-CSF promote the egress of CCR7⁺ cells from the inflamed kidney to draining lymph nodes.