Inhibiting centrosome clustering reduces cystogenesis and improves kidney function in autosomal dominant polycystic kidney disease

Abstract

Autosomal dominant polycystic kidney disease (ADPKD) is a monogenic disorder accounting for approximately 5% of patients with renal failure, yet therapeutics for the treatment of ADPKD remain limited. ADPKD tissues display abnormalities in the biogenesis of the centrosome, a defect that can cause genome instability, aberrant ciliary signaling, and secretion of pro-inflammatory factors. Cystic cells form excess centrosomes via a process termed centrosome amplification (CA), which causes abnormal multipolar spindle configurations, mitotic catastrophe, and reduced cell viability. However, cells with CA can suppress multipolarity via "centrosome clustering," a key mechanism by which cells circumvent apoptosis. Here, we demonstrate that inhibiting centrosome clustering can counteract the proliferation of renal cystic cells with high incidences of CA. Using ADPKD human cells and mouse models, we show that preventing centrosome clustering with 2 inhibitors, CCB02 and PJ34, blocks cyst initiation and growth in vitro and in vivo. Inhibiting centrosome clustering activates a p53-mediated surveillance mechanism leading to apoptosis, reduced cyst expansion, decreased interstitial fibrosis, and improved kidney function. Transcriptional analysis of kidneys from treated mice identified pro-inflammatory signaling pathways implicated in CA-mediated cystogenesis and fibrosis. Our results demonstrate that centrosome clustering is a cyst-selective target for the improvement of renal morphology and function in ADPKD.

Keywords: Cell Biology; Cytoskeleton; Nephrology.

Figure 1. Human ADPKD cells with amplified centrosomes form pseudo-bipolar mitotic spindles.

(A) Cartoon schematic of centrosome amplification (CA) and its consequences. Mitotic spindles in cells with excess centrosomes typically form more than 2 poles, leading to multipolar spindle configurations and activation of the spindle assembly checkpoint (SAC). This is followed by activation of the p53-mediated surveillance pathway, leading to caspase-mediated apoptosis. Centrosome clustering and the formation of pseudo-bipolar spindles comprise a survival mechanism adapted by some of these cells to avoid cell death. These cells subsequently demonstrate enhanced secretion of pro-inflammatory cytokines and growth factors. Inhibition of centrosome clustering can push cells toward the apoptotic pathway. (B) Immunofluorescence staining of human ADPKD kidney section with antibodies to highlight centrioles (centrin), microtubules (α-tubulin), and DNA (DAPI). Right panel shows magnified region indicated by white box. Arrows point to cyst-lining cells with amplified centrosomes. Scale bars = 10 μm. (C) Left panel: Representative immunofluorescence images of cells in mitosis in human ADPKD cyst. Tissue was stained with antibodies to highlight centrioles (centrin), spindle microtubules (α-tubulin), and DNA (DAPI). Right panel: Representative magnified immunofluorescence images of human ADPKD cystic cells in mitosis. Arrows point to centrosomes at spindle poles. Scale bars = 10 μm. Scale bar = 5 μm for the 2 magnified center images. (D) Quantification of spindle configurations in dividing cells from wild-type and ADPKD kidneys. Wild-type: n = 366 mitotic cells from 4 patient samples; ADPKD: n = 789 mitotic cells from 8 patient samples. ***P < 0.001 (1-way ANOVA).

Figure 2. Induction of CA accelerates cystogenesis in slow-onset ADPKD mice.

(A) Timeline of conditional induction of CA in Pkd1^RC/RC mChPlk4 mice, including serum and kidney sample collection. (B) Immunofluorescence staining of kidney sections from 5-month-old Pkd1^RC/RC and Pkd1^RC/RC mChPlk4 mice using antibodies against mCherry, centrosomes (γ-tubulin), epithelial cells (E-cadherin), and DNA (DAPI). Arrows point to mCherry-positive cells with amplified centrosomes. Scale bars = 10 μm. Scale bar = 2 μm for all insets. (C) Quantification of the percentage of mChPlk4-positive cells and (D) the percentage of cyst-lining cells with excess (> 2) centrosomes. (E and F) Images of whole kidneys and H&E-stained sections of Pkd1^RC/RC and Pkd1^RC/RC mChPlk4 mice at 5 months of age. Scale bar = 2 mm (whole kidneys) and scale bar = 1 mm (H&E-stained sections). (G) Evaluation of kidney weight expressed as percentage of body weight at 5 months. (H) Quantification of cyst number and (I) fractional cyst area per kidney section. (J) Analysis of relative blood urea nitrogen (BUN) and (K) serum creatinine levels at 5 months. (L) Immunofluorescence staining (top) with α–smooth muscle actin (SMA; to mark myofibroblasts) and DNA (DAPI) and trichrome staining (bottom) of kidney sections from 5-month-old Pkd1^RC/RC and Pkd1^RC/RC mChPlk4 mice. Scale bar = 100 μm (SMA) and scale bar = 500 μm (trichrome). (M) Quantification of the fraction of α-SMA–positive area. n = 5 mice per group for all experiments. *P < 0.05, **P < 0.01, ***P < 0.001 (1-way ANOVA).

Figure 3. Inhibition of centrosome clustering in ADPKD cells promotes multipolar spindle formation and activates the SAC.

(A) Immunofluorescence staining of interphase wild-type (HK-2) and PKD1-null (WT9-12) cells for centrioles (Cep152), centrosomes (γ-tubulin), microtubules (α-tubulin), and DNA (DAPI). Scale bar = 2 μm. Scale bar = 2 μm for all insets. (B) Percentage of cells with amplified centrosomes. n = 293 (HK-2) and 374 (WT9-12) cells. (C) Distribution of the number of excess centrosomes per cell. n = 301 (HK-2) and 322 (WT9-12) cells. (D) Immunofluorescence staining of mitotic wild-type (HK-2) and PKD1-null (WT9-12) cells for centrioles (Cep152), centrosomes (γ-tubulin), microtubules (α-tubulin), and DNA (DAPI). Scale bar = 2 μm. Scale bar = 2 μm for all insets. (E) Top graph shows the percentage of wild-type cells that form bipolar (cells containing normal centrosome number) or multipolar (cells with > 2 centrosomes) spindles. Bottom graph shows the percentage of PKD1-null cells with CA that formed pseudo-bipolar (clustered centrosomes) versus multipolar (declustered centrosomes). For HK-2 cells: n = 207 (control), 674 (CCB02), 174 (PJ34), 101 (AZ82); for WT9-12 cells: n = 393 (control), 449 (CCB02), 205 (PJ34), 189 (AZ82). (F) Immunofluorescence staining of mitotic wild-type and PKD1-null cells for centrosomes (γ-tubulin), microtubules (α-tubulin), Bub1, and DNA (DAPI). Grayscale images provide improved contrast of the Bub1 staining. Scale bar = 2 μm. (G) Percentage of cells showing Bub1 accumulation in mitosis. For HK-2 cells: n = 167 (DMSO), 133 (CCB02); for WT9-12 cells: n = 180 (DMSO), 278 (CCB02). Results are from 2 experiments. (H) Immunoblot of wild-type and PKD1-null cells treated with vehicle or CCB02. (I) Percentage of wild-type and PKD1-null cells containing normal centrosome number following treatment with CCB02. n = 371 (HK-2) and 264 (WT9-12) cells. For all experiments, cells were incubated with each inhibitor for 24 hours. Results are from 3 experiments. *P < 0.05, **P < 0.01, ***P < 0.001 (2-way ANOVA).

Figure 4. Inhibition of centrosome clustering attenuates disease progression during rapid stages of cytogenesis in vivo.

(A) Schematic representation of CCB02 and PJ34 treatment scheme and washout in Pkd1^RC/RC mice. (B) Quantification of relative weight change during the treatment timeline. (C) Immunofluorescence staining of kidney sections from Pkd1^RC/RC mice at 11 months with antibodies to highlight centrosomes (γ-tubulin), spindle microtubules (α-tubulin), and DNA (DAPI). Scale bar = 10 μm. (D) Quantification of the percentage of spindle configurations at metaphase. n = 201 cells (vehicle group), 134 (CCB02 group), and 117 (PJ34 group). (E and F) Images of whole kidneys and H&E-stained sections of Pkd1^RC/RC mice at 11 months of age after treatment with CCB02 or PJ34 and postwashout (14 months). Scale bar = 2 mm (whole kidneys) and scale bar = 1 mm (H&E-stained sections). (G) Evaluation of kidney weight expressed as percentage of body weight. (H) Quantification of cyst number and (I) fractional cyst area per kidney section in treated and untreated mice. (J) Analysis of relative blood urea nitrogen (BUN) and (K) serum creatinine levels at 11 months and postwashout (14 months). (L) Immunofluorescence staining (top) with α–smooth muscle actin (SMA; myofibroblasts) and DNA (DAPI) and trichrome staining (bottom) of kidney sections following the indicated treatment regimen. Scale bar = 100 μm (SMA) and scale bar = 500 μm (trichrome). (M) Quantification of the fraction of α-SMA–positive area. n = 10 mice each for vehicle, CCB02, and PJ34; n = 4 animals for CCB02 washout group. *P < 0.05, **P < 0.01, ***P < 0.001 (1-way ANOVA).

Figure 5. Inhibition of centrosome clustering at earlier stages of cystogenesis shows stronger effects on disease progression.

(A) Schematic representation of CCB02 and tolvaptan treatment scheme in Pkd1^RC/RC mice. (B) Quantification of relative weight change during the treatment timeline. (C and D) Images of whole kidneys and H&E-stained sections of Pkd1^RC/RC mice at 11 months of age after treatment with CCB02 or tolvaptan. Scale bar = 2 mm (whole kidneys) and scale bar = 1 mm (H&E-stained sections). (E) Evaluation of kidney weight expressed as percentage of body weight. (F and G) Quantification of cyst number and fractional cyst area per kidney section in treated and untreated mice. (H) Analysis of relative changes in blood urea nitrogen (BUN) and (I) serum creatinine levels. (J) Immunofluorescence staining (top) with α-SMA (myofibroblasts) and DNA (DAPI) and trichrome staining (bottom) of kidney sections following the indicated treatment regimen. Scale bar = 100 μm (SMA) and scale bar = 500 μm (trichrome). (K) Quantification of the fraction of α-SMA–positive area. n = 5 mice per group for all experiments. *P < 0.05, **P < 0.01, ***P < 0.001 (1-way ANOVA).

Figure 6. Centrosome clustering inhibitors promote p53-mediated apoptosis and attenuate pro-inflammatory signaling pathways.

(A and B) Quantification of p53-positive nuclear staining in kidneys of Pkd1^RC/RC mice treated with CCB02 or PJ34 for 2 months or with CCB02 or tolvaptan for 5 months. (C and D) Quantification of TUNEL staining in Pkd1^RC/RC kidney sections treated as indicated. (E and F) Quantification of the percentage of cyst-lining cells with excess (> 2) centrosomes. (G and H) Quantification of γ-H2AX–positive cells after treatment for 2 or 5 months. n = 10 mice per group (2-month treatment); n = 5 mice per group (5-month treatment).

Figure 7. Centrosome clustering inhibitor attenuates pro-inflammatory signaling pathways.

(A) Venn diagram showing the comparison of secreted factors identified in patients with ADPKD and mouse models, secretome of cells with amplified centrosome, and cilia loss–induced pro-inflammatory and pro-fibrotic factors implicated in cystogenesis. (B) Quantification of the relative change in gene expression levels of selected factors in kidneys of wild-type or Pkd1^RC/RC mice treated with CCB02. n = 3 mice per group. *P < 0.05, **P < 0.01, ***P < 0.001 (2-tailed). The network of significantly enriched biological themes defined by a CompBio pathway analysis tool, comparing the differentially expressed genes between control and CCB02-treated groups. The size of a sphere is proportional to the CompBio enrichment score of its theme.