Tissue-Resident Macrophages Promote Renal Cystic Disease

Abstract

Background: Mutations affecting cilia proteins have an established role in renal cyst formation. In mice, the rate of cystogenesis is influenced by the age at which cilia dysfunction occurs and whether the kidney has been injured. Disruption of cilia function before postnatal day 12-14 results in rapid cyst formation; however, cyst formation is slower when cilia dysfunction is induced after postnatal day 14. Rapid cyst formation can also be induced in conditional adult cilia mutant mice by introducing renal injury. Previous studies indicate that macrophages are involved in cyst formation, however the specific role and type of macrophages responsible has not been clarified.

Methods: We analyzed resident macrophage number and subtypes during postnatal renal maturation and after renal injury in control and conditional Ift88 cilia mutant mice. We also used a pharmacological inhibitor of resident macrophage proliferation and accumulation to determine the importance of these cells during rapid cyst formation.

Results: Our data show that renal resident macrophages undergo a phenotypic switch from R2b (CD11c^lo) to R2a (CD11c^hi) during postnatal renal maturation. The timing of this switch correlates with the period in which cyst formation transitions from rapid to slow following induction of cilia dysfunction. Renal injury induces the reaccumulation of juvenile-like R2b resident macrophages in cilia mutant mice and restores rapid cystogenesis. Loss of primary cilia in injured conditional Ift88 mice results in enhanced epithelial production of membrane-bound CSF1, a cytokine that promotes resident macrophage proliferation. Inhibiting CSF1/CSF1-receptor signaling with a CSF1R kinase inhibitor reduces resident macrophage proliferation, R2b resident macrophage accumulation, and renal cyst formation in two mouse models of cystic disease.

Conclusions: These data uncover an important pathogenic role for resident macrophages during rapid cyst progression.

Keywords: CSF1R; cilia; cystic kidney; macrophages; renal injury.

Graphical abstract

Figure 1.

Resident macrophages from wild-type mouse kidneys undergo a rapid phenotypic switch during postnatal development and have a unique gene expression profile. (A) Representative flow cytometry plot indicting the gating strategy used to identify R1 bone marrow–derived infiltrating (CD11b^hi F4/80^lo) and R2 resident macrophages (CD11b^lo F/480^hi). Resident macrophages were further gated based on expression of CD11c and subgrouped into R2a (CD11c^hi) and R2b (CD11c^lo). (B) Representative flow cytometry plot showing CD11c expression in resident macrophages at different points after birth in a wild-type mouse kidney. Quantification of R2a and R2b as a percentage of total cells (all kidney cells) is shown as the mean±SEM. n=3–4 mice, two replicates. (C) Heat map from RNA-seq of R2a and R2b macrophages isolated from 8-week-old C57Bl/6 mice. n=5 mice per population.

Figure 2.

R2b macrophages accumulate in kidneys of P7-induced, P28-harvested and injured adult–induced cilia mutant mice. (A) Representative flow cytometry plot showing CD11c expression in resident macrophages isolated from P7-induced, P28-harvested cilia wild-type (WT) and cilia mutant mice. The number of R2a and R2b macrophages as a percentage of total kidney cells are quantified for each time point and shown as mean±SEM. n=4–5, two replicates. **P<0.01, t test. (B) Quantification of the number of R2a and R2b macrophages as a percentage of total kidney cells is shown as the mean±SEM. Two-way ANOVA using sex as a covariate. n=8–13, three to four replicates. *P<0.05, **P<0.01. (C) Representative confocal image showing F4/80 (red), CD11c (white), CD206 (green), and Hoescht (blue) staining in control and cilia mutant mice harvested 35 days postinjury; n=3. Original magnification, ×400; objective, ×40; scale bar, 50 µm. IR, ischemia reperfusion.

Figure 3.

Cilia mutant kidneys have increased expression of macrophage chemoattractants, proinflammatory, and profibrotic genes after IRI compared with injured wild-type (WT) kidneys. RNA from whole kidney lysates of control and cilia mutant mice was isolated using TRIzol and levels of (A) macrophage chemoattractants and activators (Ccl2, total Csf-1, and membrane-bound Csf-1), (B) proinflammatory cytokines (Il1β, Tnfα, Arginase 1 [Arg1], and Nos2), and (C) profibrotic genes (inhibinß_A, fibronectin 1 [Fn1], col1a2, and col3a1) determined by qRT-PCR. Values were normalized to hypoxanthine-guanine phosphoribosyltransferase (Hprt). For the qRT-PCR analysis, values represent the mean±SEM. n=2–6 mice, two replicates. Two-way ANOVA *P<0.05. RQ, relative quantification; IR, ischemia reperfusion.

Figure 4.

Resident macrophage accumulation in cilia mutant mice is independent of peripheral blood monocytes. (A) Schematic of how wild-type (WT) and cilia mutant CD45.2 mice were parabiotically joined to CD45.1 mice in this experiment. A timeline for key events in the experiment is shown. (B) Percentage chimerism, calculated as percentage nonhost cells (CD45.1 positive cells harvested in CD45.2 mice), is shown for each macrophage population. Values are shown as the mean±SEM. n=5–6 mice, four replicates. One way ANOVA, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001. IR, ischemia reperfusion.

Figure 5.

Resident macrophages from cilia mutant mice have increased proliferation compared with injured wild-type (WT) resident macrophages. (A) Representative flow cytometry plots showing Ki67 versus F4/80 staining in resident (R2), R2a (CD11c^hi), and R2b (CD11c^lo) macrophages 3 days post injury. (B) Quantification of the percentage of each cell population that is positive for Ki67 in R2, R2a, and R2b macrophage populations is shown in the associated graph as the mean±SEM. n=8–13, three to four replicates. Two-way ANOVA, *P<0.05, **P<0.01, ****P<0.001. IR, ischemia reperfusion.

Figure 6.

Membrane-bound Csf-1 is increased in LTA+ proximal tubule epithelium from cilia mutant mice. (A–C) mRNA levels of (A) membrane-bound Csf-1, (B) full-length Csf-1, and (C) Csf-1R from FACS-sorted kidney cells 3 days post-IRI as determined by qRT-PCR. Values represent the mean±SEM. n=4–5 mice, two to three replicates. *P<0.05, t test. RQ, relative quantification; HPRT, hypoxanthine-guanine phosphoribosyltransferase; IR, ischemia reperfusion; WT, wild type.

Figure 7.

CSF-1R inhibition reduces resident macrophage proliferation and prevents R2b resident macrophage accumulation. (A) Quantification of Ki67+ cells as a percentage of each cell population is shown for injured cilia mutant mice harvested 3 days after daily treatment with vehicle or GW2580. Values represent the mean±SEM. n=4–5, three replicates. (B) Quantification of the number of R2a and R2b macrophages as a percentage of total kidney cells from cilia mutant mice treated with vehicle or GW2580 for 3 days after injury. Values represent the mean±SEM. n=4–5 replicates. **P<0.01. (C) Confocal images of kidney sections harvested 35 days post injury stained with F4/80 (green), Ki67 (red), and LTA (white). F4/80+ area/nuclear area is quantified in the associated graph as the mean±SEM. n=3, one replicate. *P<0.05, **P<0.01, ***P<0.001, t test. Original magnification, ×400; objective, ×40; scale bar, 100 µm.

Figure 8.

GW2580 reduces cystic index and number in two mouse models of cystic disease. (A) Representative hematoxylin and eosin image from male mice treated with vehicle or GW2580. For each group, the number of cysts/total area as well as the cystic index (cystic area/total kidney area) is shown as the mean±SEM. n=7–8, three replicates. **P<0.01, t test. Original magnification, ×10; ×1 objective; scale bar, 1000 µm. (B) Representative image of kidneys that were isolated from cilia mutant mice treated with vehicle only or GW2580 at 35 days postinjury and stained with picrosirius red. Quantification of the percent area that was picrosirius red positive/kidney area was obtained using ImageJ. Values represent the mean±SEM. n=4–5, two replicates. *P<0.05. Original magnification, ×400; ×40 objective; scale bar, 50 µm. (C) Representative pretreatment (P11) and post-treatment (P21) MRI images of Cys1^cpk/cpk males treated with GW2580 or vehicle. (D) Graph showing the increase in TKV and TKV/body length (BL) ratio (equivalent to height-adjusted TKV used in humans) in control and Cys1^cpk/cpk mice. The effect of GW2580 on body and kidney indices were tested with analyses of covariance with pretreatment scores and gender serving as covariates. (E) Representative confocal images showing Cys1^cpk/cpk mice with and without GW2580 treatment that were stained with F4/80 (red) and Hoescht (blue). Quantification of macrophage area/nuclear area is shown. n=2–3. Original magnification, ×400; objective, ×40; scale bar, 50 µm.