Screening for Inhibitors of YAP Nuclear Localization Identifies Aurora Kinase A as a Modulator of Lung Fibrosis

Abstract

Idiopathic pulmonary fibrosis is a progressive lung disease with limited therapeutic options that is characterized by pathological fibroblast activation and aberrant lung remodeling with scar formation. YAP (Yes-associated protein) is a transcriptional coactivator that mediates mechanical and biochemical signals controlling fibroblast activation. We previously identified HMG-CoA (3-hydroxy-3-methylglutaryl coenzyme A) reductase inhibitors (statins) as YAP inhibitors based on a high-throughput small-molecule screen in primary human lung fibroblasts. Here we report that several Aurora kinase inhibitors were also identified from the top hits of this screen. MK-5108, a highly selective inhibitor for AURKA (Aurora kinase A), induced YAP phosphorylation and cytoplasmic retention and significantly reduced profibrotic gene expression in human lung fibroblasts. The inhibitory effect on YAP nuclear translocation and profibrotic gene expression is specific to inhibition of AURKA, but not Aurora kinase B or C, and is independent of the Hippo pathway kinases LATS1 and LATS2 (Large Tumor Suppressor 1 and 2). Further characterization of the effects of MK-5108 demonstrate that it inhibits YAP nuclear localization indirectly via effects on actin polymerization and TGFβ (Transforming Growth Factor β) signaling. In addition, MK-5108 treatment reduced lung collagen deposition in the bleomycin mouse model of pulmonary fibrosis. Our results reveal a novel role for AURKA in YAP-mediated profibrotic activity in fibroblasts and highlight the potential of small-molecule screens for YAP inhibitors for identification of novel agents with antifibrotic activity.

Keywords: Aurora kinase A inhibitor; YAP; drug screening; human lung fibroblast; idiopathic pulmonary fibrosis.

Figure 1.

High-throughput small-molecule screen for inhibitors of YAP (Yes-associated protein) nuclear localization identifies Aurora A inhibitors as novel hits. (A) Scheme of small-molecule screen. (B) Dose–response curve of 14 top hits of the screen generated on plastic or 25 kPa Matrigel. (C) Western blots and (D) quantifications showing effects of pan-Aurora inhibitors danusertib and PHA-680632 in YAP/TAZ (Transcription coactivator with PDZ-binding motif) phosphorylation and degradation. Phosphorylated YAP/TAZ and total YAP/TAZ were both normalized to the corresponding GAPDH, and the ratio of phosphorylated YAP/TAZ to total YAP/TAZ was normalized to DMSO control. Data represent mean ± SD, n = 3. *P < 0.05, ***P < 0.001, and ****P < 0.0001, unpaired t test. (E) Quantitative PCR (qPCR) results showing that pan-Aurora inhibitors danusertib and PHA-680632 significantly reduced expression of profibrotic genes CTGF (connective tissue growth factor), CYR61 (cysteine rich angiogenic inducer 61), COL1A1 (collagen type I α 1 chain), and ACTA2 (α-smooth muscle actin). Data represent mean ± SD, n = 2. ****P < 0.0001, two-way ANOVA with Sidak’s multiple comparisons test. QC = quality control.

Figure 1.

High-throughput small-molecule screen for inhibitors of YAP (Yes-associated protein) nuclear localization identifies Aurora A inhibitors as novel hits. (A) Scheme of small-molecule screen. (B) Dose–response curve of 14 top hits of the screen generated on plastic or 25 kPa Matrigel. (C) Western blots and (D) quantifications showing effects of pan-Aurora inhibitors danusertib and PHA-680632 in YAP/TAZ (Transcription coactivator with PDZ-binding motif) phosphorylation and degradation. Phosphorylated YAP/TAZ and total YAP/TAZ were both normalized to the corresponding GAPDH, and the ratio of phosphorylated YAP/TAZ to total YAP/TAZ was normalized to DMSO control. Data represent mean ± SD, n = 3. *P < 0.05, ***P < 0.001, and ****P < 0.0001, unpaired t test. (E) Quantitative PCR (qPCR) results showing that pan-Aurora inhibitors danusertib and PHA-680632 significantly reduced expression of profibrotic genes CTGF (connective tissue growth factor), CYR61 (cysteine rich angiogenic inducer 61), COL1A1 (collagen type I α 1 chain), and ACTA2 (α-smooth muscle actin). Data represent mean ± SD, n = 2. ****P < 0.0001, two-way ANOVA with Sidak’s multiple comparisons test. QC = quality control.

Figure 2.

Inhibiting Aurora A, but not Aurora B, significantly decreased YAP nuclear localization and TGFβ-induced fibroblast activation. (A) Representative images and (B) quantification of YAP staining in human lung fibroblasts (HLFs) treated with MK-5108 or barasertib, in the presence or absence of TGFβ. Data represent mean ± SD, n = 3. ***P < 0.001, unpaired t test. Scale bar, 10 μm. (C) qPCR results showing the expression of profibrotic genes in HLFs after MK-5108, barasertib, or both, in the presence or absence of TGFβ. Data represent mean ± SD, n = 2. *P < 0.05, **P < 0.01, and ****P < 0.0001, two-way ANOVA with Sidak’s multiple comparisons test. (D) qPCR results showing that only siRNAs against Aurora A kinase, but not Aurora B or Aurora C, significantly reduced expression of profibrotic genes in HLFs. Data represent mean ± SD, n = 3. *P < 0.05, **P < 0.01, ***P < 0.001, and ****P < 0.0001, two-way ANOVA with Sidak’s multiple comparisons test.

Figure 3.

Effect of Aurora A inhibition in YAP inactivation is independent of effects on mitosis. (A) Western blot and quantification showing the decrease in phosphorylated AURKA on T288 1 hour after changing to serum-free media and MK-5108 treatment. ***P < 0.001. (B) Average percentage of cells with positive nuclear YAP localization during a time course of MK-5108 treatment, after overnight prestarvation of the cells in serum-free medium. Data represent mean ± SD. (C) Average cell numbers of three individual wells after immunostaining in 96-well plates, after MK-5108 treatment in HLFs, with or without prestarvation for 16 hours. Data represent mean ± SD. (D) Average percentage of cells with YAP nuclear localization correlation coefficient greater than 0.7, in DMSO- or MK-5108–treated HLFs, with or without prestarvation for 16 hours. Data represent mean ± SD.

Figure 4.

YAP inactivation and antifibrotic effect of Aurora A inhibition is not dependent on canonical Hippo pathway kinases LATS1 and LATS2 (Large Tumor Suppressor 1 and 2). (A) Western blot images and (B) quantification showing the amount of total and phosphorylated YAP. Phosphorylated YAP/TAZ and total YAP/TAZ were both normalized to the corresponding GAPDH, and the ratio of phosphorylated-to-total YAP/TAZ was normalized to DMSO control. Data represent mean ± SD, n = 2. *P < 0.01, unpaired t test. (C) quantification of the YAP staining in HLFs after MK-5108 treatment combined with siNT2 or siLATS1 + 2, in the presence or absence of TGFβ. Data represent mean ± SD, n = 3. ***P < 0.001 and ****P < 0.0001, unpaired t test. (D) Quantification of the YAP staining after treating HLFs with a dose curve of MK-5108 combined with MST (Mammalian Ste20-like kinase) inhibitor XMU-MP-1. (E) qPCR results showing the effect of MK-5108 on CTGF, CYR61, and COL1A1 expression with control and LATS1 and LATS2 siRNA, in the presence or absence of TGFβ. Data represent mean ± SD. ns = non-significant. **P < 0.01, ***P < 0.001, and ****P < 0.0001, unpaired t test.

Figure 5.

Aurora A inhibition inactivates YAP and decreases fibrotic gene expression partly through regulating actin polymerization. (A) Rhodamine phalloidin (yellow) and YAP (green) fluorescent images showing F-actin fibers and YAP localization after treating HLFs with DMSO or MK-5108, combined with TGFβ (5 ng/ml), Rho activator II (0.5 μg/ml), lysophosphatidic acid (LPA) (10 μM), or sphingosine 1-phosphophate (S1P) (1 μM) treatment. Scale bars, 10 μm. (B) Quantification of YAP nuclear localization in A. Percentage of cells with positive nuclear YAP staining normalized to control are shown as percentage of control value. Data represent mean ± SD, n = 3. ****P < 0.0001 and ***P < 0.001, one-way ANOVA with Sidak’s multiple comparisons test. (C and D) Profibrotic genes CTGF and CYR61 expression at 4 hours or 16 hours normalized to DMSO shown as percentage of control value, after treating HLFs with 10 μM MK-5108 alone or combined with TGFβ and Rho activator II, LPA, or S1P. Data represent mean ± SD, n = 3. ****P < 0.0001, one-way ANOVA with Sidak’s multiple comparisons test. (E) Fluorescent images (scale bar, 10 µm) and (F) quantifications showing MRTFA (Myocardin-Related Transcription Factor A) localization after treating HLFs with DMSO or MK-5108, combined with TGFβ (5 ng/ml). Scale bar, 10 μm. Data represent mean ± SD, n = 6. ****P < 0.0001, unpaired t test.

Figure 6.

Aurora A–specific inhibition protects mice from bleomycin-induced lung fibrosis. (A) Scheme of administering bleomycin and MK-5108. Mice were given 1 unit/kg of bleomycin intratracheally. Starting the following day, 30 mg/kg MK-5108 was given by oral gavage two times a day for 14 days, before whole lungs were harvested and assayed for hydroxyproline amount. (B) Weight curve of mice treated with vehicle or MK-5108. (C) Masson’s trichrome staining (on 5X lung sections) showing significant reduction of hydroxyproline amount in the lungs of the mice treated with MK-5108. (D) Quantification of hydroxyproline content in the whole lungs of mice treated with vehicle or MK-5108, after injection of PBS or bleomycin. Data represent mean ± SD. **P < 0.01 and ****P < 0.0001, one-way ANOVA with Sidak’s multiple comparisons test.