MEK inhibition exerts temporal and myeloid cell-specific effects in the pathogenesis of neurofibromatosis type 1 arteriopathy

Abstract

Mutations in the NF1 tumor suppressor gene are linked to arteriopathy. Nf1 heterozygosity (Nf1+/-) results in robust neointima formation, similar to humans, and myeloid-restricted Nf1+/- recapitulates this phenotype via MEK-ERK activation. Here we define the contribution of myeloid subpopulations to NF1 arteriopathy. Neutrophils from WT and Nf1+/- mice were functionally assessed in the presence of MEK and farnesylation inhibitors in vitro and neutrophil recruitment to lipopolysaccharide was assessed in WT and Nf1+/- mice. Littermate 12-15 week-old male wildtype and Nf1+/- mice were subjected to carotid artery ligation and provided either a neutrophil depleting antibody (1A8), liposomal clodronate to deplete monocytes/macrophages, or PD0325901 and neointima size was assessed 28 days after injury. Bone marrow transplant experiments assessed monocyte/macrophage mobilization during neointima formation. Nf1+/- neutrophils exhibit enhanced proliferation, migration, and adhesion via p21^Ras activation of MEK in vitro and in vivo. Neutrophil depletion suppresses circulating Ly6C^low monocytes and enhances neointima size, while monocyte/macrophage depletion and deletion of CCR2 in bone marrow cells abolish neointima formation in Nf1+/- mice. Taken together, these findings suggest that neurofibromin-MEK-ERK activation in circulating neutrophils and monocytes during arterial remodeling is nuanced and points to important cross-talk between these populations in the pathogenesis of NF1 arteriopathy.

Figure 1

Neurofibromin controls neutrophil function in vitro and in vivo. (A–C) Quantification of WT (white bars) and Nf1+/– (black bars) neutrophil proliferation  (A), adhesion (B), and migration (C) in response to G-CSF. (A) Data represent thymidine incorporation reported as mean counts per minute (cpm) ± SEM, n = 4–5. *P < 0.01 for WT versus Nf1+/– neutrophils. **P < 0.001 for WT and Nf1+/– neutrophils without growth factor versus WT and Nf1+/– neutrophils incubated with G-CSF.  (B) Data represent mean optical density (600 nm) ± SEM, n = 4–5. *P < 0.001 for WT versus Nf1+/– neutrophils at each time point. **P < 0.001 for WT and Nf1+/– neutrophils at 15 min versus WT and Nf1+/– neutrophils at 60 min. (C) Data represent mean number of migrated cells ± SEM, n = 4–5. *P < 0.01 for WT versus Nf1+/– neutrophils. (D and E) Quantification of percent (D) and absolute number (E) of GR1 + /CD11b + neutrophils in peritoneal lavage fluid 48 h after LPS injection. (D) Data represent mean percent ± SEM, n = 4–5. *P < 0.01 for WT versus Nf1+/– mice. (E) Data represent mean number of cells per mL ± SEM, n = 4–5. *P < 0.001 for WT versus Nf1+/– mice. Analysis by 2-way ANOVA and Student’s t test.

Figure 2

MEK inhibition impairs neurofibromin-deficient neutrophil function. (A–D). Quantification of WT (white bars) and Nf1+/– (black bars) neutrophil proliferation (A and C) and adhesion (B and D) in response to G-CSF in the presence of PD901 (A and B) or rosuvastatin (C and D). (A) Data represent thymidine incorporation reported as mean counts per minute (cpm) ± SEM, n = 4–5. *P < 0.01 for WT versus Nf1+/– neutrophils. **P < 0.001 for WT and Nf1+/– neutrophils versus WT and Nf1+/– neutrophils incubated with indicated concentration of PD901. #P < 0.01 for WT and Nf1+/– neutrophils incubated with 10 nM and 50 nM PD901 versus WT and Nf1+/– neutrophils incubated with 100 nM PD901. (B) Data represent mean optical density (600 nm) ± SEM, n = 4–5. *P < 0.001 for WT versus Nf1+/– neutrophils. #P < 0.001 for WT and Nf1+/– neutrophils versus WT and Nf1+/– neutrophils incubated with 50 nM PD901. (C) Data represent thymidine incorporation (cpm) ± SEM, n = 4–5. *P < 0.01 for WT versus Nf1+/– neutrophils. **P < 0.01 for WT and Nf1+/– neutrophils versus WT and Nf1+/– neutrophils incubated with indicated concentration of rosuvastatin. #P < 0.01 for WT and Nf1+/– neutrophils incubated with 2 μM rosuvastatin versus WT and Nf1+/– neutrophils incubated with 8 μM rosuvastatin. (D) Data represent mean optical density ± SEM, n = 4–5. *P < 0.001 for WT versus Nf1+/– neutrophils. #P < 0.01 for WT and Nf1+/– neutrophils versus WT and Nf1+/– neutrophils incubated with 10 μM rosuvastatin. (E) Quantification of absolute number of GR1 + /CD11b + neutrophils in peritoneal lavage fluid 48 h after LPS injection with/without PD901. *P < 0.001 for WT versus Nf1+/– mice. **P < 0.001 for WT and Nf1+/– mice without LPS injection versus WT and Nf1+/– mice provided LPS. #P < 0.001 for WT and Nf1+/– mice provided LPS versus WT and Nf1+/– mice provide LPS with daily PD901. Analysis by 2-way ANOVA.

Figure 3

Neutrophil depletion increases neointima size in WT and Nf1+/– mice. (A) quantification of neutrophils after administration of 1A8 to WT (gray) and Nf1+/– (black) mice. *P < 0.01 for each comparison. (B). Van Gieson-stained images of injured carotid arteries from WT and Nf1+/– mice provided IgG or 1A8. Black arrows indicate neointima boundary. Scale bars: 100 μm. (C and D) Data represent mean intima area (μm²) (C) and I/M ratio (D) ± SEM for uninjured and injured arteries from WT and Nf1+/– mice provided IgG or 1A8. n = 8–11 per group. *P < 0.01 for each comparison. **P < 0.01 for Nf1+/– mice receiving IgG versus Nf1+/– mice receiving monoclonal Ly6G antibody. Analysis by 2-way ANOVA.

Figure 4

Early neutrophil depletion suppresses circulating Ly6C^low monocyte frequency. (A) Representative gating strategy to identify Ly6C^hi and Ly6C^low monocytes. (B–D) Quantitation of circulating Ly6C^hi (B), Ly6C^low (C) and Ly6C^hi/Ly6C^low ratio (D) in WT (gray) and Nf1+/– (black) mice provided non-specific IgG or monoclonal Ly6G antibody. Boxes represent the 25th and 75th percentile with median value designated by horizontal line within each box. Whiskers represent the minimum and maximum values. *P < 0.05 and **P < 0.001 for each comparison. #P < 0.01 for WT and Nf1+/– mice provided non-specific IgG versus WT and Nf1+/– mice provided monoclonal Ly6G antibody. Analysis by 2-way ANOVA.

Figure 5

Monocyte/macrophage depletion decreases neointima size in Nf1+/– mice. (A) Representative gating strategy demonstrating depletion of CD115+ /CD11+ /CD11b+ monocytes. (B) Van Gieson-stained images of injured carotid arteries from WT and Nf1+/– mice provided liposomal clodronate or empty liposomes. Black arrows indicate neointima boundary. Scale bars: 100 μm. (C and D). Data represent mean intima area (μm²) (C) and I/M ratio (D) ± SEM, n = 8–11 per group. *P < 0.05 for WT versus Nf1+/– mice. #P < 0.05 for Nf1+/– mice receiving clodronate versus Nf1+/– mice receiving empty liposomes. Analysis by 1-way ANOVA.

Figure 6

Monocytes/Macrophages drive neointima formation via CCR2. (A) Van Gieson-stained images of injured carotid arteries from WT reconstituted with WT bone marrow and Nf1+/– mice reconstituted with either Nf1+/– or Nf1+/– ;CCR2−/− bone marrow. Black arrows indicate neointima boundary. Black boxes identify area of injured artery that is magnified below. Scale bars: 100 μm. (B and C) Data represent mean intima area (μm²) (B) and I/M ratio (C) ± SEM, n = 9–10 per group. *P < 0.05 for WT mice reconstituted with WT bone marrow versus Nf1+/– mice reconstituted with Nf1+/– bone marrow. #P < 0.05 for Nf1+/– mice reconstituted with Nf1+/– bone marrow versus Nf1+/– mice reconstituted with Nf1+/– ;CCR2−/− bone marrow. Analysis by 1-way ANOVA. (D). Representative images of Mac-3 staining in injured carotid arteries from Nf1+/– mice reconstituted with either Nf1+/– or Nf1+/– ;CCR2−/− bone marrow. Black arrows indicate neointima boundary. Black boxes identify area of injured artery that is magnified to the right. Scale bars: 100 μm.

Figure 7

Mature, but not early neointima formation is sensitive to MEK inhibition (A–F) Representative photomicrographs (A and D) and quantification of intima area (B and E) and I/M ratio (C and F) of injured carotid arteries from WT and Nf1+/– mice treated with either placebo or PD901 for 7 days (A–C) or 14 days (D–F) Data represent the mean intima area (B and E) or I/M ratio (C and F) of 3 arterial cross Sects. (400, 800, and 1200 μm proximal to the ligation) ± SEM, n = 8–10 per group. *P < 0.05 for WT injured with placebo treatment versus Nf1+/– injured with placebo treatment. #P < 0.01 for injured with placebo treatment versus WT and Nf1+/– injured with 5 mg/kg/day PD901 treatment. Analysis by 2-way ANOVA.