Chemical modifier screen identifies HDAC inhibitors as suppressors of PKD models

Abstract

Polycystic kidney disease (PKD) is a common human genetic disease with severe medical consequences. Although it is appreciated that the cilium plays a central role in PKD, the underlying mechanism for PKD remains poorly understood and no effective treatment is available. In zebrafish, kidney cyst formation is closely associated with laterality defects and body curvature. To discover potential drug candidates and dissect signaling pathways that interact with ciliary signals, we performed a chemical modifier screen for the two phenotypes using zebrafish pkd2(hi4166) and ift172(hi2211) models. pkd2 is a causal gene for autosomal dominant PKD and ift172 is essential for building and maintaining the cilium. We identified trichostatin A (TSA), a pan-HDAC (histone deacetylase) inhibitor, as a compound that affected both body curvature and laterality. Further analysis verified that TSA inhibited cyst formation in pkd2 knockdown animals. Moreover, we demonstrated that inhibiting class I HDACs, either by valproic acid (VPA), a class I specific HDAC inhibitor structurally unrelated to TSA, or by knocking down hdac1, suppressed kidney cyst formation and body curvature caused by pkd2 deficiency. Finally, we show that VPA was able to reduce the progression of cyst formation and slow the decline of kidney function in a mouse ADPKD model. Together, these data suggest body curvature may be used as a surrogate marker for kidney cyst formation in large-scale high-throughput screens in zebrafish. More importantly, our results also reveal a critical role for HDACs in PKD pathogenesis and point to HDAC inhibitors as drug candidates for PKD treatment.

Fig. 1.

Suppression of pkd2 mutants/morphants by TSA. (A and B) Treatment of 500 nM TSA from the shield stage leads to slight ventral curvature (B) on 2 dpf in wild-type embryos (wt) as compared to embryos treated with DMSO (A). (C–E) TSA can suppress the body curvature of pkd2/hi4166 mutant embryos. (C) shows mutant embryos treated with DMSO from 27 hpf on 2 dpf. Red lines depict the definition of curvature angle. (D) shows embryos treated with 500 nM TSA from 27 hpf on 2 dpf. With this treatment, mutant embryos are indistinguishable from wild-type siblings. (E) Average curvature angle in embryos treated with DMSO, 200 and 500 nM TSA. n = 5, *, P < 0.05, **, P < 0.0005. (F–H) Inhibition of kidney cyst formation in pkd2 morphants (pkd2mo) on 3 dpf by treatment of 100 nm TSA from the bud stage. (F) shows a pkd2 morphants treated with DMSO. The arrow points to the cyst. (F) shows a morphant treated with TSA. Insets show kidney cyst in F or the lack of cyst in G. (H) Average of percentage of embryos with kidney cysts from 3 independent experiments. Twenty-five to fifty embryos were examined in each experiment. ***, P < 0.005.

Fig. 2.

Inhibition of Class I HDACs can suppress the phenotypes of pkd2 mutants/morphants. (A and B) Treatment of 20 μm VPA from the shield stage leads to slight ventral curvature (B) on 2 dpf in wild-type embryos (wt) as compared to embryos treated with DMSO (A). (C–E) VPA can suppress the body curvature of pkd2/hi4166 mutant embryos. (C) shows mutant embryos treated with DMSO on 2 dpf. (D) shows mutant embryos treated with 20 μm VPA from 27 hpf on 2 dpf. (E) Average curvature angle in embryos treated with 20 μm VPM and 20 μm VPA. n = 5, *, P < 0.005, (F–H) Inhibition of kidney cyst formation in pkd2 morphants (pkd2mo) on 3 dpf by treatment of 20 μm VPA from the 20-somite stage. (F) shows a pkd2 morphants treated with 20 μm VPM. The arrow points to the cyst. (G) shows a morphant treated with VPA. White box: area shown in Insets. Insets show kidney cyst in F or the lack of cyst in G. (H) Average of percentage of embryos with kidney cysts from three independent experiments. Twenty-five to one hundred embryos were examined in each experiment. *, P < 0.005. 4166: embryos from pkd2/hi4166 heterozygous cross.

Fig. 3.

VPA reduces cyst growth and improves kidney function in mouse models of polycystic kidney disease based on Pkd1. (A) Gross (Top) and cut surface (Bottom) appearance of vehicle and VPA (200 mg/kg) treated Pkd1^flox/flox;Pkhd1-Cre mouse kidneys at P25 showing reduced kidney size in the VPA treated mice. (B) Representative scans using the image splicing feature of MetaMorph software applied to sagittal sections from vehicle and VPA treated kidneys. Images of one section from the other seven mice studied are provided in Fig. S5. (C) Aggregate data from vehicle treated (n = 5 mice, 10 kidneys) and VPA treated (n = 4 mice, 8 kidneys) mice showing significant differences in kidney weight-to-body weight ratio (P < 0.01), cystic index (P < 0.001) and blood urea nitrogen (P < 0.001). Absolute kidney and body weights are presented in Fig. S5. (Scale bar, 2 mm.)

Fig. 4.

Depletion of hdac1 can suppress the phenotypes of pkd2 mutants. (A and B) Phenotypes of hdac1/hi1618 mutants on 2 dpf in side views. (A) shows a wild-type (WT) sibling, while (B) shows a mutant embryo (hi1618). (C–E) Injection of hdac1 morpholino can suppress body curvature of pkd2/hi4166 mutants on 2 pdf. (C) shows mutant embryos injected with the five base-pair mismatch control oligo. (D) shows mutant embryos injected with the hdac1 morpholino. (E) shows average curvature angle in embryos injected with control or hdac1 morpholino. n = 5, *, P < 0.05. (F–H) Injection of hdac1 morpholino can suppress kidney cyst formation in pkd2 morphants (pkd2mo) on 3 pdf. 4166: embryos from pkd2/hi4166 heterozygous cross; ctrolMO: embryos injected with control morpholino. HDAC1 MO: embryos injected with hdac1 morpholino.