Tissue-Resident Macrophages Promote Renal Cystic Disease

Significance Statement

Disruption of cilia function before postnatal day 12–14 in mice or renal injury in adult mice with cilia dysfunction results in accelerated renal cyst formation. Macrophages have been implicated in promoting cyst formation; however, it is unclear whether infiltrating bone marrow-derived or kidney resident macrophages are responsible. The authors show that a specific population of juvenile-like resident macrophages are present during periods of accelerated cyst formation. Inhibition of juvenile-like resident macrophage accumulation using a colony-stimulating factor 1 receptor kinase inhibitor reduced the severity of cystic disease in two different animal models of cystic disease. These results suggest resident renal macrophages contribute to cystic disease.

Visual Abstract

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.

Cystic kidney disease is a commonly inherited genetic disease resulting from mutations in proteins that localize to primary cilia (e.g., polycystin 1 and polycystin 2) or are required for cilia formation (e.g., intraflagellar transport 88 [IFT88]).^1–4 However, the mechanism connecting cilia dysfunction and renal cyst formation is unknown. 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 (P12–14, critical switch period) leads to rapid cyst formation throughout the kidney within 2–3 weeks.^2,3 In contrast, induction of cilia dysfunction after P14 (adult induced) results in slow focal renal cyst formation.⁵ The slow rate of cyst formation in adult-induced Ift88-, polycystin 1– (Pkd1-), or Pkd2-deficient mice is accelerated by ischemia-reperfusion injury (IRI) with cysts becoming evident by 14 days post injury.^2,5,6 The mechanism driving the rapid rate of cyst formation when cilia function is disrupted in juvenile mice and after injury in adult-induced mice with cilia dysfunction is unknown, but raises the possibility that the environmental conditions are similar during both periods of rapid cyst formation.

A complex role of macrophages and other innate immune factors in the pathogenesis of renal cystic disease was proposed by Mrug et al.⁷ based on genome-wide expression studies of kidneys from the Cys1^cpk mouse model. This concept is consistent with the increased number of macrophages observed in kidneys from patients with autosomal dominant polycystic kidney disease and orthologous mouse models.^8–10 Importantly, depletion of phagocytic macrophages in cystic mice using liposomal clodronate reduced cyst severity and improved renal function^9,10; however, inhibition of macrophage recruitment from the bone marrow using bindarit—a pharmacological inhibitor of C-C motif chemokine ligand 2 (CCL2), CCL7, and CCL8—did not reduce renal cyst formation.^11,12 In contrast to this study, recent findings indicate that disruption of Pkd1 or liver kinase B1 (Lkb1) leads to increased accumulation of bone marrow–derived infiltrating macrophages and accelerated cystogenessis.^13,14 In these studies, genetic inhibition of macrophage recruitment using Ccl2-deficient mice reduced the severity of cystic disease. Our previous work also indicated that infiltrating macrophages promote hepatic fibrosis in Ift88^Orpk mice.¹⁵ The fact that both inhibition of infiltrating macrophage recruitment and depletion of all macrophages in the kidney (both infiltrating and resident) reduces cystic severity suggests that multiple macrophage subtypes are likely important in cystic disease.

In mice, macrophages arise from two distinct origins.^16,17 In contrast to bone marrow–derived infiltrating macrophages (F4/80^lo, CD11b^hi), tissue-resident macrophages (F4/80^hi, CD11b^lo) are derived from embryonic progenitors, migrate into tissues during organogenesis, are independent of the bone marrow and the transcription factor myb,¹⁶ and persist within the tissue through local self-renewal.¹⁸ Resident macrophage self-renewal occurs in part through paracrine signaling involving colony-stimulating factor 1 (CSF1) as evidenced by an accumulation of CSF1 receptor (CSF1R)–enhanced green fluorescent protein–tagged macrophages in regions surrounding epithelial-specific, membrane-bound CSF1.^19–22 Furthermore, inhibition of CSF1R activity reduces resident macrophage proliferation and number and alters the renal injury response.^22,23

In this study, we tested the hypothesis that resident macrophages are involved in the pathogenesis of renal cyst formation. Our data indicate that CD11c^lo macrophages (referred to as R2b macrophages) that are present in the kidney during juvenile periods reappear after renal injury in cilia mutant mice. Further, we show that resident macrophage accumulation in cilia mutant mice after injury is independent of peripheral blood monocyte recruitment and that inhibition of CSF1–mediated resident macrophage proliferation markedly reduces renal cysts in two mouse models of cystic disease. Importantly, recent findings from our laboratory indicate that a resident macrophage-like population of cells exists in multiple species including humans²⁴ and may be a viable therapeutic target for patients with cystic kidney disease.

Methods

Mice

Eight-week-old CAGG-Cre/Esr1/5Amc/J Ift88^f/f male and female mice (referred to as cilia mutant mice) on a C57BL/6J background were bred in-house. The C57BL/6J-Cys1^cpk/cpk mice were purchased from the Jackson Laboratory (Bar Harbor, ME). Animals were maintained in Association for Assessment and Accreditation of Laboratory Animal Care International–accredited facilities in accordance with Institutional Animal Care and Use Committee (IACUC) regulations at the University of Alabama at Birmingham (UAB) (approval numbers: 10130 and 21072).

Induction of Cilia Loss

Eight- to ten-week-old CAGG-Cre^ERT2 Ift88^f/f animals were given an intraperitoneal injection of tamoxifen at 6 mg/40 g body wt once daily for 3 consecutive days. Deletion of Ift88 was confirmed by PCR, quantitative RT-PCR (qRT-PCR), and Western blot (Supplemental Figure 1). The following primers were used to detect Ift88 by PCR: 5′-GCCTCCTGTTTCTTGACAACAGTG, GGTCCTAACAAGTAAGCCCAGTGTT-3′ and 5′-CTGCACCAGCCATTTCCTCTAAGTCATGTA-3′. Cilia deletion was confirmed by immunofluorescence microscopy using the cilia marker acetylated α-tubulin (catalog number 66200–1; Proteintech) (Supplemental Figure 1).

Renal IRI

Three weeks after tamoxifen induction, mice were subjected to unilateral IRI. Mice were weighed and injected with 25 mg/kg body wt tribromoethanol via intraperitoneal injection and maintained under 1% isoflurane during the course of the procedure. A small incision was made on the back and the kidney was removed from the abdominal cavity. Unilateral IRI was performed by clamping the left renal pedicle for 30 minutes followed by clamp release to allow for reperfusion. Mice were injected with buprenorphine (0.05 mg/kg) for pain relief and allowed to recover on a 37°C heating pad with free access to food and water. After the desired number of days, animals were anesthetized with tribromoethanol and perfused with 1% PBS. The kidney receiving IRI was harvested for further processing.

Parabiosis Protocol

The parabiosis protocol was modified from the procedure published by Kamran et al.²⁵ For this procedure, incisions were made through the skin and muscle layer, starting from the elbow joint and extending down the flank to the knee joint. Nonabsorbable 3–0 interrupted sutures were placed around the knee and elbow joints to prevent strain along the suture lines, with care taken to avoid obstructing blood flow to the distal extremities. Analgesia was maintained on all mice according to IACUC guidelines.

Small Animal Magnetic Resonance Imaging

The kidneys of Cys1^cpk/cpk mice were visualized using a 9.4T magnetic resonance imaging (MRI) system BioSpec (Bruker BioSpin, Billerica, MA) with surface coil. A low-resolution image to confirm the location and orientation of the kidneys was followed by the acquisition of axial images covering the entire kidney region with a T2-weighted fast spin echo sequence (rapid acquisition with relaxation enhancement) with a slice thickness of 0.5 mm. Kidney volumes were estimated based on measurements obtained with ImageJ, version 1.47 software (National Institutes of Health, Bethesda, MD).

CSF-1R Inhibitor Treatment

GW2580 (catalog number G-5903; LC Laboratories) was resuspended in 0.5% hydroxypropyl methylcellulose and 0.1% Tween 80 using multiple strokes of a Teflon-glass homogenizer. The resuspended inhibitor was given to mice at a dose of 160 mg/kg through oral gavage once daily throughout the RI. For the Cys1^cpk/cpk inhibitor studies, mice were randomized into vehicle- and drug-treated groups 11 days postbirth based on MRI imaging.

Fixation and Tissue Processing

Following harvest, mouse kidneys were immersion fixed in 4% (wt/vol) paraformaldehyde (PFA) overnight at 4°C, switched to 70% ethanol overnight, embedded in paraffin, sectioned at 5 µm, and stained using hematoxylin and eosin. Picrosirius red staining was performed by the UAB Comparative Pathology Laboratory.

Immunofluorescence Microscopy

Following fixation and overnight cryopreservation in 30% sucrose, 8-µm thick, Optimal Cutting Temperature (OCT)–embedded kidney tissue was fixed with 4% PFA for 10 minutes, permeabilized with 1% Triton X-100 for 8 minutes, and incubated in blocking solution (PBS with 1% BSA, 0.3% Triton X-100, 2% (vol/vol) donkey serum, and 0.02% sodium azide) for 30 minutes at room temperature. Sections were incubated in primary antibody overnight at 4°C according to the manufacturer’s recommendation, washed with PBS, and incubated with the appropriate secondary antibodies in blocking solution for 1 hour at room temperature. Primary antibodies included F4/80 (catalog number14–4801, clone BM8; eBioscience), FITC-conjugated Lotus tetragonolobus agglutinin (LTA) (catalog number FL-1321; Vector Laboratories), FITC-conjugated CD206 (catalog number 141704; Biolegend), CD11c (catalog number 14–0114–82; eBioscience) and phycoerythrin-conjugated Ki67 (catalog number 12–5698–82; eBioscience). Secondary antibodies included the following: Alexa Fluor 647–conjugated anti-hamster (Armenian) (catalog number 405510; Biolegend) and Alexa Fluor 647–conjugated anti-rat (catalog number 712–606–153; Jackson ImmunoResearch). After the addition of secondary antibody, nuclei were stained by Hoechst nuclear stain (Sigma-Aldrich) and samples mounted using IMMU-MOUNT (Thermo). All fluorescence images were captured on a PerkinElmer ERS 6FE spinning disk confocal microscope and images were processed and analyzed in Volocity version 6.1.1 software (PerkinElmer).

Flow Cytometry

After perfusion of the mouse with PBS, kidneys were minced and digested in 1 ml of RPMI 1640 (Gibco, Grand Island, NY) containing 1 mg/ml collagenase type I (Sigma-Aldrich) and 100 U/ml DNase I (Sigma-Aldrich) for 30 minutes at 37°C with agitation. Kidney fragments were passed through a 70-µm mesh (Falcon; BD Biosciences) yielding single-cell suspensions. Red blood cells were lysed and cells were resuspended in 1 ml PBS containing 1% BSA with Fc blocking solution for 30 minutes on ice. A total of 2×10⁶ cells were stained for 30 minutes at room temperature with primary antibody (Supplemental Table 1). Cells were washed, fixed in 2% PFA for 30 minutes, and resuspended in FACS staining buffer.

To determine intracellular Ki67 staining, cell suspensions were incubated with Foxp3 Fixation/Permeabilization Solution (eBioscience) for 30 minutes at room temperature after addition of primary antibodies. Cells were incubated with anti-mouse Ki67, washed, and resuspended in 1× PBS. After immunostaining, cells were analyzed on a BD LSR II Flow Cytometer. Data analysis was performed using FlowJo version 10 software.

RNA Isolation and qRT-PCR

RNA was isolated using TRIzol and qRT-PCR was performed using TaqMan probes (Supplemental Table 2). To detect membrane-bound and secreted CSF-1, we used SYBR Green Real-Time PCR Master Mixes with the following primers: common forward primer, 5′-TCTCCTTGAAAAGGACTGGAAC-3′; secreted reverse primer, 5′-TGGTGAGGGGAGGCAGAG-3′; and cell surface reverse primer, 5′-GCAAGACTAGGATGATGGCCCG-3′. The products from membrane-bound CSF-1 and secreted CSF-1 were verified by DNA sequencing.

TaqMan Low-Density Array and RNA Sequencing

RNA was isolated from FACS-sorted cells using the RNeasy Plus Mini Kit (catalog number 74134; Qiagen) according to standard procedures. Samples were loaded onto the TaqMan low-density array (TLDA) card (PN 4342253; Applied Biosystems) and run using a 7900HT Real-Time PCR System. A full list of the TaqMan probes and associated genes is listed in Supplementary Table 3.

For RNA sequencing (RNA-seq), the NEBNext mRNA Library Prep Kit for Illumina with ribosome reduction was used to generate the library which was subsequently sequenced using Illumina NextSeq500 with paired-end, 75-bp sequencing and aligned to the University of California, Santa Cruz GRCm38/mm10 reference genome using the STAR software package.²⁶ At least 25 million reads were obtained for each sample. After alignment, HTSeq-Count (version 0.9.1) was used to count the number of reads mapping to each gene.²⁷ Fragments per kilobase of transcript per million mapped reads were calculated using the Cufflinks suite (https://github.com/cole-trapnell-lab/cufflinks) and differential expression was then applied to the count files using DESeq2.²⁸ RNA sequencing data have been deposited in the Gene Expression Omnibus under accession number GSE131896.

Statistical Analyses

Data are presented as the mean±SEM. ANOVA, analysis of covariance, two-way ANOVA, and t tests were used for statistical analysis, and differences were considered significant for P values <0.05. For studies in which a significant difference was observed between male and female mice within a group, a two-way ANOVA was performed with sex as a covariate. Briefly, analyses began by summarizing measures of central tendency (sample mean, sample median) and measures of dispersion (sample variance, sample SD). Factorial ANOVA models, including factors for day of experiment (0, 3, 7, 14, 21, 28, and 35), group (control, cilia mutant), and sex of mouse (male, female) were constructed. Specifically, main effects for each factor, each two-way interaction between factors, and the three-way interaction among factors were examined. Beginning with the highest-order interaction, nonsignificant interaction terms were removed until a parsimonious model was achieved. Normal probability plots, histograms, and residual plots were used to examine distributional assumptions for the models. Once the parsimonious model was identified, linear contrasts were used to test for marginal differences between mouse strains per day of the experiment. All analyses were conducted using SAS 9.4.

Results

Renal Resident Macrophages Undergo a Subtype Transition during Normal Postnatal Renal Maturation

To assess whether a specific macrophage subtype is associated with the differential rate of cyst formation observed in juvenile versus adult-induced cilia mutant mice, we first analyzed macrophage populations isolated from wild type kidneys at different periods during postnatal development using flow cytometry. The gating strategy (Figure 1A) used to distinguish bone marrow–derived infiltrating (hereafter referred to as region 1 [R1] macrophages) and embryonic-derived resident macrophages (hereafter referred to as region 2 [R2] macrophages) was based on differential expression of CD11b and F4/80 as established by Schulz et al.¹⁶ R2 macrophages were further subgated based on expression of CD11c (R2a CD11c^hi; R2b CD11c^lo) (Figure 1A). Analysis of R2 cells from kidneys of wild-type mice at P4, P19, and P28 reveals a transition from R2b (CD11c^lo) to R2a (CD11c^hi) during this period (Figure 1B). This change in R2 macrophage subtype overlaps with the period in which there is a switch from a rapid to slow rate of cyst formation in response to cilia dysfunction.

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.

To broadly characterize the R2 macrophage subtypes that are present during these developmental time points, we analyzed gene expression profiles using bulk RNA-seq on flow-sorted macrophage subtypes followed by verification of macrophage-specific genes using TLDAs. There is strong correlation between the RNA-seq and TLDA data (0.71), confirming the reproducibility of these studies. Heat-map analysis generated from RNA-seq of R2a and R2b populations isolated from 8-week-old C57Bl/6J mice indicates a unique gene expression profile between these populations (Figure 1C). Further analysis comparing TLDA and RNA-seq data shows that R2a cells have increased expression of genes encoding CD11c and CCR2, whereas R2b cells have increased expression of genes encoding IL-10, matrix metallopeptidase 9, and CD206—cytokines associated with renal development, injury repair, and cyst formation (Supplemental Table 4).^29–31

Loss of Primary Cilia in Adult Mice Leads to the Reappearance of Juvenile-like, CD11c^lo R2b Macrophages after Renal Injury

To evaluate the effect of cilia loss on resident macrophage subtype and number, we compared macrophage profiles in wild-type and conditional Ift88 mice in which cilia loss was induced through tamoxifen injection at P7, followed by harvesting and analysis at P28. The data reveal an increased proportion and number of the R2b macrophages as a percentage of total cells in cilia mutant mice compared with control mice (Figure 2A). In contrast, there were no differences in the number of R2a macrophages between control and cilia mutant mice.

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.

To determine if the accumulation of R2b macrophages in cilia mutant mice was due to intrinsic defects in resident macrophages lacking Ift88, we flow sorted R1 and R2 macrophages as well as ciliated LTA+ proximal tubule epithelium and performed qRT-PCR for Ift88. Our data indicate that R1 macrophages lack Ift88 gene expression, whereas Ift88 was detected in R2 macrophages and the LTA+ epithelium (Supplemental Figure 2). The low level of Ift88 expression detected in the R2 population is likely due to minor contamination of the R2 cells with CD45-ciliated epithelium (Supplemental Figure 2; note CD45 cells with acetylated tubulin staining showing a cilium). Therefore, the accumulation of the R2b population in the cilia mutant background is unlikely to be due to an intrinsic defect in macrophages.

Because induction of cilia loss at P7 leads to rapid cystogenesis with pathology already present at the time of analysis (P28), we performed flow cytometry on adult-induced (8- to 10-week-old) conditional Ift88 mice at different time points after undergoing 30 minutes of unilateral IRI. This model allows us to analyze changes in macrophage subtypes and number before cyst initiation and during cyst progression. In this model, the first renal cysts become evident around 14 days after injury, followed by rapid progression (Supplemental Figure 3). Analysis of flow cytometry data indicates that the number of R2b macrophages was significantly increased in cilia mutant mice beginning 7 days post renal injury and remains significantly increased throughout the time course of the analysis (Figure 2B). In contrast, the number of R2a macrophages was only increased in injured cilia mutant mice at 7 and 14 days post injury compared with injured wild-type mice (Figure 2B). In these studies, we found that the number of R2b macrophages was significantly increased in injured cilia mutant males compared with injured cilia mutant females at 7 and 14 days post injury. Additional analysis of flow cytometry data reveals a significant increase in R1 infiltrating macrophages in cilia mutant mice at 7 days post injury compared with control injured mice (Supplemental Figure 4). These data indicate that both infiltrating and resident macrophages are increased in cilia mutant mice before the onset of renal cystogenesis.

To determine the spatial relationship between R2b macrophages and renal cysts, we conducted immunofluorescence microscopy analysis on kidney sections isolated from mice 35 days postinjury using the macrophage markers F4/80 and CD11c. To identify R2b macrophages, we used a fluorescently labeled antibody against CD206, a gene (Mrc1) whose expression is enriched in R2b macrophages relative to R2a cells (Supplemental Table 4). To further confirm that this marker is specific for R2b macrophages at this time point, we analyzed flow cytometry data collected at 35 days postinjury. Our data indicate that R2b resident macrophages are the predominant cell type that express CD206 at this time point (Supplemental Figure 5). Analysis of confocal images indicates that most macrophages in regions adjacent to cysts coexpress F4/80 and CD206 (Figure 2C). In contrast, we did not observe CD206-positive macrophages around noncystic tubules in injured cilia mutant mice. These data indicate that R2b macrophages are located adjacent to cysts and raise the possibility that they are involved in paracrine/juxtacrine signaling with the cilia mutant epithelium.

Injured Cilia Mutant Mice Express Increased Proinflammatory Transcripts Compared with Control Injured Mice

To identify differences in the renal microenvironment between control and cilia mutant mice after injury, we compared expression levels of several macrophage chemoattractants, proinflammatory cytokines, and extracellular matrix genes by qRT-PCR using RNA isolated from whole kidney tissue. In injured cilia mutant kidneys (day 14 postinjury), there was increased gene expression of the monocyte chemoattractant protein (Ccl2) compared with control injured kidneys (Figure 3A). Analysis of gene expression of proinflammatory cytokines reveals increased expression of Tumor necrosis factor-α (Tnfa) and Arginase 1 in injured cilia mutant mice compared with injured cilia wild-type mice (Figure 3B). Our data also indicate increased expression of InhibinBa, Fibronectin 1, and Col3a1 in injured cilia mutant mice compared with injured wild-type mice although the values were not significant (Figure 3C). Several of these genes, including those encoding TNFα and Inhibin β_A, are known to promote cyst formation.^32,33 There were no overt changes in genes encoding total Csf1, membrane-bound Csf1, IL1b, nitric oxide synthase 2 (Nos2), and Col1a2 (Figure 3). We did not observe a difference in expression of inflammatory genes between injured cilia mutant male and female mice at any time point analyzed.

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.

Because iNos (Nos2) gene expression is increased after AKI in mice receiving contralateral nephrectomy³⁴ and our data did not reveal an increase in Nos2 expression between uninjured (day 0) and unilaterally injured mice 1–3 days postinjury, we analyzed gene expression of Kidney Injury Molecule-1 (Kim1) to confirm injury in our model. Kim1 is a commonly used marker of tubular damage after kidney injury.³⁵ Our data indicate that Kim1 expression is significantly elevated in both injured control and cilia mutant mice 1 day postinjury compared with uninjured controls (Supplemental Figure 6). These data indicate that 30 minutes of unilateral IRI results in significant tubular damage in injured wild-type and cilia mutant mice.

Accumulation of Resident Macrophages in Cilia Mutant Kidney Is Independent of Peripheral Blood Input after Injury

After injury there is an accumulation of R2b macrophages in cilia mutant mice compared with control injured mice. This accumulation could be driven either through recruitment of circulating monocytes which differentiate into resident-like macrophages or through local paracrine-driven self-proliferation. To address this question, we performed parabiosis experiments by joining the circulation of a CD45.2 control or cilia mutant mouse with a congenic CD45.1 wild-type C57BL/6J mouse. The mice were allowed to equilibrate for 4 weeks before performing unilateral IRI or sham surgery on the CD45.2 mouse (Figure 4A). Our results indicate that R1 infiltrating macrophages from both control and cilia mutant kidneys have approximately equal levels of chimerism compared with blood monocytes, indicating a complete exchange of these cells with circulating precursor cells (Figure 4B). The level of chimerism in infiltrating R1 macrophages in the kidney is similar to previously published reports.^36–38 In contrast to R1 macrophages, both kidney R2a and R2b macrophages from control and cilia mutant mice after injury had 3%–4% chimerism, similar to that observed in resident microglia, a population that is independent of peripheral monocyte recruitment (Figure 4B).³⁹ Thus, our data indicate that the observed accumulation of R2b macrophages in injured cilia mutant kidneys is largely independent of peripheral blood monocyte recruitment. These data are in agreement with previously published data indicating that resident macrophage accumulation after injury is mainly driven through self-proliferation.^36,40,41

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.

Resident Macrophages from Injured Cilia Mutant Mice Have Increased Proliferation Rates Compared with Injured Wild-Type Mice

As there is limited peripheral blood monocyte contribution to the R2 population in cilia mutant kidneys after injury, we next compared the proliferation rate of kidney-resident macrophages in control and cilia mutant mice after injury using flow cytometry and antibodies against Ki67, a marker of active cell proliferation.⁴² Our data indicate that total R2, R2a, and R2b macrophages from cilia mutant mice have significantly increased proliferation rates at 3 days postinjury compared with injured cilia wild-type mice (Figure 5). Total R2 and R2a macrophages from injured cilia mutant mice also have increased proliferation rates compared with control injured mice at 14 and 35 days postinjury (Figure 5). Because our data indicate that infiltrating macrophage numbers are also increased in cilia mutant kidneys compared with control kidneys after injury, we analyzed proliferation rates of infiltrating, bone marrow–derived R1 macrophages via flow cytometry. Our data indicate that the proliferation rate of R1 macrophages was not different between injured control and cilia mutant kidneys at any time point tested (Supplemental Figure 7). There were no differences in proliferation rates between injured male and female cilia mutant mice at any time point analyzed. These data indicate that the increased accumulation of R1 cells in injured cilia mutant kidneys is likely due to enhanced Ccl2-dependent peripheral blood monocyte recruitment and that the increased accumulation of R2b macrophages in injured cilia mutant kidneys is due to increased self-proliferation.

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.

Injured Cilia Mutant Epithelium Expresses Increased Levels of Membrane-Bound Csf-1 Compared with Control Injured Epithelium

A key factor driving resident macrophage proliferation and survival is the cytokine CSF-1, which is expressed in a secreted or membrane-bound (mbCSF-1) form generated through alternative RNA splicing.²¹ Although Csf1 expression from whole kidney tissue demonstrated minimal changes between injured control and cilia mutant mice, when we analyzed expression in flow cytometry–isolated cell types we noted there was a significant increase in mbCsf-1 expression in proximal tubule cells of the injured cilia mutant kidneys compared with controls (Figure 6A). This increase in mbCsf-1 was not observed in the collecting duct (Dolichos biflorus agglutinin positive) or in R1 or R2 macrophage populations. Full-length CSF-1 did not differ significantly between any of the groups tested (Figure 6B). Both R1 and R2 macrophages express CSF-1R (Figure 6C), whereas receptor expression was not detected in either epithelial cell population. These data indicate excess accumulation and proliferation of R2 macrophages observed in the cilia mutant mice may be due to local paracrine/juxtacrine signaling between the injured proximal tubule epithelium and R2 macrophages.

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.

CSF-1R Inhibition Reduces Resident Macrophage Proliferation and Number of R2b Macrophages in Cilia Mutant Mice after IRI

To test the model that R2b resident macrophage accumulation in the injured cilia mutant kidney occurs in part through mbCSF-1–driven local proliferation, we inhibited CSF-1R activity in vivo using GW2580, a selective small-molecule inhibitor of the CSF-1R kinase.⁴³ Injured adult–induced cilia mutant mice were treated by daily gavage starting 1 day after injury and continuing throughout the 35 day course of the experiment.⁴³ At 3 days postinjury, GW2580 treatment caused a significant reduction in the proliferation rate of both R2a and R2b resident macrophages compared with vehicle-treated mice (Figure 7A). In contrast, R1 infiltrating macrophage proliferation was increased after GW2580 treatment (Supplemental Figure 8). Analysis of resident macrophage numbers at 3 days postinjury indicates that GW2580 treatment significantly reduced the accumulation of R2b macrophages in injured cilia mutant mice (Figure 7B). The number of R2a and R1 macrophages in injured cilia mutant mice was not affected by GW2580 treatment (Figure 7B, Supplemental Figure 8). Thus, GW2580 specifically reduced the accumulation of R2b macrophages in cilia mutant mice after renal injury.

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.

To further confirm that GW2580 reduced proliferation and accumulation of macrophages at later time points, we stained kidney sections of vehicle- and GW2580-treated cilia mutant mice with LTA, F4/80, and Ki67 35 days post-IRI. In vehicle-treated kidneys, we detected dual-labeled Ki67+, F4/80+ macrophages that are frequently localized to regions surrounding epithelial-derived renal cysts (Figure 7C, white arrows). In contrast, the number of F4/80+ cells per square millimeter of tissue was markedly reduced in injured cilia mutant mice receiving GW2580 treatment (Figure 7C). These data indicate that the number of macrophages in regions adjacent to tubule epithelial cells is decreased in injured cilia mutant mice treated with GW2580 at 35 days post injury.

Inhibition of Resident Macrophage Proliferation with GW2580 Reduces Cystic Severity

Next, we assessed whether preventing the accumulation of R2b macrophages affected cyst number and severity after IRI. Kidneys were isolated from male and female cilia mutant mice 35 days post injury. Because our analyses indicate that female cilia mutant mice failed to develop severe renal cysts after injury compared with their male counterparts (Supplemental Figure 9), we could only analyze the effect of GW2580 treatment on injured male cilia mutant mice at 35 days post injury. Histologic examination of kidneys from male cilia mutant mice at 35 days post-IRI revealed a significant reduction in cystic severity in GW2580-treated mice compared with vehicle-treated cilia mutant mice (Figure 8A). Importantly, this was due to both a reduction in the number and size of cysts compared with vehicle controls (Figure 8A). In addition to the cystic phenotype, there was also a significant reduction in renal fibrosis as indicated by picrosirius red staining in mice treated with GW2580 compared with vehicle-treated mice (Figure 8B). These data indicate that CSF1R-dependent proliferation and accumulation of R2b resident macrophages is a driving factor in renal cyst formation.

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.

To confirm the relevance of CSF-1 signaling during cystic renal disease in the absence of kidney injury and to test whether inhibition of CSF1R is efficacious in slowing cyst progression in mice with prominent renal cysts, we tested the effect of GW2580 in the Cys1^cpk/cpk mouse model. Cys1^cpk/cpk mice phenocopy aggressive autosomal recessive polycystic kidney disease with rapidly progressive cyst growth.⁴⁴ In contrast to the above-described studies in the Ift88-conditional IRI model where GW2580 treatment was started at the time of renal injury and well before cyst initiation, a 10-day GW2580 gavage treatment of Cys1^cpk/cpk mice could only be started at P11 due to the small size of the mice. At this point, renal cystic disease is already moderately severe. Using MRI, we tracked changes in total kidney volume (TKV) as a surrogate for cyst expansion in longitudinal studies on the same animals. GW2580 treatment reduced the severity of renal cystic disease as reflected by the reduced rate of increase (slope) in TKV (TKV reduction by 37%, adjusted means 2697 versus 1782 mm³, P=0.021) and body length–adjusted TKV (TKV/body length, by 34%, adjusted means 511 versus 358, P=0.033 (Figure 8, C and D). Body weight–adjusted TKV was also reduced by GW2580 treatment although the value was not significant (Supplemental Figure 10). Body weight and body length were significantly reduced in the GW2580-treated mice compared with vehicle-treated mice (Supplemental Figure 10). Similar to the Ift88-conditional IRI model, GW2580 treatment decreased the number of renal F4/80-positive macrophages compared with vehicle-treated mice (Figures 8E).

Discussion

The clinical importance of cilia in cystic kidney disease is well established; however, the mechanism by which cilia dysfunction leads to cyst formation is still largely unknown. Additionally, it remains poorly understood why cysts develop rapidly in juvenile-induced cilia mutant mice (before P12–14), but slowly in adult-induced cilia mutants (after P14), and why injury in adult-induced cilia mutants reinitiates rapid cystogenesis. Our data indicate that R2b resident macrophages accumulate in both juvenile-induced and injured adult–induced cilia mutant mice before the onset of renal cystogenesis. Inhibition of CSF1R-dependent R2b resident macrophage accumulation reduces both the number and severity of renal cysts after IRI. These data suggest that proliferation and accumulation of R2b macrophages trigger renal cyst formation and that targeting this population of cells in human patients may prevent cyst formation and expansion.

Our data suggest that the primary cilium is an important regulator of the injury and repair process in the kidney. Further, we propose that the primary cilia-dependent repair mechanism after injury is driven through paracrine/juxtacrine signaling with nearby tissue-resident macrophages. One of the candidate pathways involved in this process is mbCSF-1. Our data show that mbCsf-1 is elevated in injured cilia mutant proximal tubule epithelium compared with control injured tubule epithelium. Based on the outcome of the CSF1R inhibitor studies (GW2580), our data indicate that CSF-1 is responsible for local proliferation of R2 macrophages and that this process is increased in cilia mutant mice after injury, resulting in enhanced and persistent R2b macrophage accumulation. This indicates that injured epithelial cells lacking a primary cilium are stuck in a state of persistent injury and that they are signaling to nearby resident macrophages in an attempt to promote tubule repair. In the case of injured cilia mutant mice, the continued attempt by resident macrophages to repair the epithelium results in enhanced and prolonged epithelial proliferation and renal cyst formation.

An interesting finding from our studies was that the number of R2b macrophages was increased in male cilia mutant mice at 7 and 14 days after injury compared with injured female cilia mutant animals. In addition, we found that injured female cilia mutant mice failed to develop severe renal cysts 35 days postinjury compared with their injured male counterparts. These data provide correlative evidence that the number of R2b macrophages that are present at early periods after injury are associated with the severity of renal cysts at later time points. Interestingly, although the number of R2b macrophages differed between male and female cilia mutant mice after injury, we did not observe a difference in the proliferation rate between macrophages that were isolated from these same groups. This suggests that the accumulation of R2b macrophages observed in male cilia mutant mice is due to a reduced levels of apoptosis. This hypothesis is further strengthened by our data showing that the number of R2 macrophages was not different between male and female mice on the day of IRI. Studies that address the effect of sex on apoptosis in resident macrophages after injury are warranted.

These studies indicate that renal resident macrophages undergo a phenotypic switch during normal postnatal renal maturation. Following injury, we observe the reappearance of juvenile-like macrophages in cilia mutant mice, indicating a potential reprogramming of these macrophages toward a juvenile-like state. This finding agrees with our previous data indicating that renal injury induces a transcriptional reprogramming in wild-type mice³⁶ and suggests that macrophages which are present during juvenile periods and after injury in cilia mutant mice likely share some phenotypic and functional overlap that promotes rapid cyst formation. Furthermore, we show that renal resident macrophages self-proliferate independently of peripheral blood input after injury to maintain the resident macrophage niche. Interestingly, both R2a and R2b macrophages proliferate after injury. Based on our data showing that GW2580 treatment inhibits the proliferation of all R2 macrophages but only prevents the accumulation of R2b macrophages, we propose that R2a macrophages proliferate to become R2b macrophages. Future fate mapping studies to confirm these findings are warranted.

Our parabiosis data indicate that the majority of resident macrophages found in injured cilia mutant kidneys are independent of blood monocyte recruitment. However, there is a small proportion of resident macrophages (3%–4% of total residents) that are derived from the peripheral blood. Previous work from our group and others has shown that an emptied resident macrophage niche can be repopulated through local self-proliferation and influx of infiltrating bone marrow–derived macrophages.^36,38,41 This suggests that renal injury induces a minor depletion of the resident macrophage niche immediately after injury and that influx of bone marrow–derived infiltrating macrophages and resident macrophage self-proliferation combine to fill that niche. Because the contribution of peripheral monocytes to the resident niche is minor and the proliferation of resident macrophages 3 days postinjury reaches 70%–80%, it is likely that resident macrophage self-proliferation is the major source of resident macrophage repopulation. Further, based on the results of our parabiosis studies showing a similar level of chimerism in R2a/R2b macrophages between injured control and cilia mutant mice, it is likely that the increase in R2b resident macrophages in injured cilia mutant mice is driven by local self-proliferation and not enhanced recruitment from peripheral blood monocytes. The fact that treatment of injured cilia mutant mice with a CSF1R inhibitor prevents the accumulation of R2b macrophages, but does not affect the number of R1 macrophages in injured cilia mutant mice, provides further support of this idea. Additional studies using a lineage tracing approach in which resident macrophages were permanently labeled with a tdTomato reporter show that a majority of resident macrophages that proliferated and accumulated after injury were tdTomato-positive and that only a minor number of cells were tdTomato-negative.³⁸ Collectively, these data suggest negligible contribution of peripheral monocytes to the resident macrophage pool after injury and that the accumulation of R2b macrophages in cilia mutant mice is driven through self-proliferation.

Our data show that the number of R2b macrophages is increased in injured cilia mutant mice compared with injured control mice and that inhibition of resident macrophage proliferation and accumulation reduced renal cysts in two independent models of cyst formation. These data suggest that targeting resident macrophages in human patients with cystic disease may be a possibility for therapeutic intervention.

Disclosures

Dr. Mrug reports grants, personal fees, and nonfinancial support from Otsuka; grants, personal fees, and nonfinancial support from Sanofi; personal fees from ClearView Healthcare Partners; personal fees from Decision Resources Group Consulting; personal fees from CLARION Healthcare; personal fees from Chinook Therapeutics, outside of the submitted work; and is the Scientific Advisory Committee Chair of the PKD Foundation. Dr. Yoder reports grants from National Institutes of Health (NIH) and grants from PKD Foundation during the conduct of the study.

Funding

These studies were supported in part by the following research grants: PKD Foundation grant 214g16a (Dr. Yoder); UAB School of Medicine AMC21 grant (Dr. Yoder, Dr. Mrug, Dr. George); NIH T32 training grant in Basic Immunology and Immunologic Disease 2T32AI007051-38 (Zimmerman); NIH R01 DK115752 (Dr. Yoder), R01 DK097423 (Dr. Mrug), and R01 NS57563 (Dr. Benveniste); and Office of Research and Development, Medical Research Service, US Department of Veterans Affairs grant 1-I01-BX002298 (Dr. Mrug). The following NIH-funded cores provided services for this project: UAB Hepatorenal Fibrocystic Disease Core Center P30-DK074038, UAB-University of California San Diego O’Brien Center for AKI Research P30-DK079337, and the UAB Comprehensive Flow Cytometry Core P30-AR048311 and P30-AI27667. Additional services were provided by the UAB Comparative Pathology Laboratory and UAB Heflin Genomic Core.

Supplemental Material

This article contains the following supplemental material online at http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ASN.2018080810/-/DCSupplemental.

Supplemental Table 1. List of relevant information for flow cytometry antibodies.

Supplemental Table 2. List of Taqman probes used for qRT-PCR analysis.

Supplemental Table 3. List of Taqman probes used for microarray.

Supplemental Table 4. Comparison of the relative expression of inflammatory genes between RNA sequencing and Taqman array data.

Supplemental Figure 1. Treatment of CaggcreERT2 Ift88^f/f mice with tamoxifen leads to reduction of Ift88 mRNA, IFT88 protein, and loss of primary cilia.

Supplemental Figure 2. Resident macrophages lack primary cilia.

Supplemental Figure 3. Cilia mutant mice begin to form cysts in cortical regions 14 days post injury.

Supplemental Figure 4. R1 infiltrating macrophages are increased in injured cilia mutant mice compared with injured wild type mice.

Supplemental Figure 5. Cilia mutant R2b macrophages are the predominant cell type that express CD206 35 days post injury.

Supplemental Figure 6. Injured wild type and cilia mutant mice have increased Kim1 gene expression one day following renal injury.

Supplemental Figure 7. The percentage of proliferating R1 macrophages does not differ between injured wild type and cilia mutant mice.

Supplemental Figure 8. GW2580 treatment does not reduce number of R1 macrophages in cilia mutant mice 3 days post injury.

Supplemental Figure 9. Female cilia mutant mice fail to develop severe renal cysts 35 days post injury.

Supplemental Figure 10. Cys1^cpk/cpk mice receiving GW2580 have reduced body length and body weight compared with vehicle treated mice.