Reducing YAP expression in Pkd1 mutant mice does not improve the cystic phenotype

Abstract The Hippo pathway is a highly conserved signalling route involved in organ size regulation. The final effectors of this pathway are two transcriptional coactivators, yes‐associated protein (YAP) and transcriptional coactivator with PDZ‐binding motif (WWTR1 or TAZ). Previously, we showed aberrant activation of the Hippo pathway in autosomal‐dominant polycystic kidney disease (ADPKD), suggesting that YAP/TAZ might play a role in disease progression. Using antisense oligonucleotides (ASOs) in a mouse model for ADPKD, we efficiently down‐regulated Yap levels in the kidneys. However, we did not see any effect on cyst formation or growth. Moreover, the expression of YAP/TAZ downstream targets was not changed, while WNT and TGF‐β pathways' downstream targets Myc, Acta2 and Vim were more expressed after Yap knockdown. Overall, our data indicate that reducing YAP levels is not a viable strategy to modulate PKD progression.

Medical Center and the Commission Biotechnology in Animals of the Dutch Ministry of Agriculture.
Inducible kidney-specific Pkd1 deletion mice (iKspPkd1 del ) and tamoxifen administration have been described before. 6 Pkd1 gene has been knocked out at post-natal day 18 (PN18). 32 male mice have been divided into two experimental groups of 16 animals each: one received scrambled antisense oligonucleotide (ASO), and the other received Yap-specific ASO. Both groups received an injection of 100 mg/kg of ASO via i.p. injection, starting two weeks after Pkd1 inactivation (PN18 + 2 weeks), once a week, until week 7 after gene inactivation. Mice were sacrificed at 8 weeks after gene inactivation (PN18 + 8 weeks). ASOs were provided by Ionis Pharmaceuticals.
Both ASOs were 16mer S-constrained ethyl gapmers with a 3-10-3 chimeric design and a phosphorothioate backbone. Yap ASO sequence was as follows: 5′-AACCAACTATTACTTC-3′; scrambled ASO sequence was as follows: 5′-GGCCAATACGCCGTCA-3′. The Yap ASO was selected from leads identified following in vitro screens which were then evaluated in vivo for renal activity and tolerability.
Scrambled ASO did not bind to any known target and was included as a control for non-specific effects.
At sacrifice, both kidneys were collected and used for immunohistochemistry (IHC) or snap-frozen for RNA and protein extraction.
Blood urea nitrogen level (BUN) was measured using the Reflotron Plus (Roche Basel). Three age-matched wild-type (Wt) mice were also included for IHC purposes.

| Gene expression analysis
Total RNA was isolated from snap-frozen kidneys using TRI Reagent (#T9424; Sigma-Aldrich) according to manufacturer's protocol, and gene expression analysis was performed by quantitative PCR (qPCR) as described previously. 8

F I G U R E 1
In vivo down-regulation of Yap with ASOs. A, Schematic representation of the in vivo experimental pipeline. Pkd1 gene inactivation was achieved with three consecutive administrations of tamoxifen at post-natal day 18 (PN18). Two weeks after gene inactivation, mice were injected weekly intraperitoneally with Yap-specific ASO or scrambled ASO as control. The last ASO injection was performed at 7 weeks after gene inactivation, and the mice were sacrificed one week later (+8 wk). B, Gene expression (fold change) of Yap and Taz at the sacrifice in mice treated with Yap ASO and scrambled ASO. Each symbol represents a mouse. Mean with ± SD. **** P < .0001, t-test. C, Representative IHC of renal tissue from mice treated with scrambled ASO and Yap ASO, showing YAP and TAZ. Scale bar 1 mm. D, Total kidney protein lysates of mice treated with scrambled ASO and Yap ASO blotted for endogenous YAP, TAZ and GAPDH. E, Quantification of YAP and TAZ protein level in total kidney normalized on GAPDH. Each symbol represents a mouse. Mean with ± SD. * P < .05, n.s. not significant, t-test. F, Representative YAP and TAZ IHC on sequential slides of Wt mice kidneys at post-natal day 100. Arrowheads show the same tubules stained for the two different proteins. Scale bar 200 µm. G, Quantification of kidney size using two kidney weight/bodyweight ratio. n.s. not significant. H, Blood urea nitrogen (BUN) level at the sacrifice. n.s. not significant

| Statistical analysis
Data were analysed using GraphPad Prism 7.00 for Windows (GraphPad Software, www.graph pad.com).

| YAP knockdown using ASOs does not improve cystic phenotype in vivo
We showed in the past that cyst-lining epithelia have intense nuclear YAP localization, both in Pkd1-mutant mouse models and in ADPKD patients. 3 Therefore, we hypothesized that YAP could actively contribute to cyst formation or cyst growth, through up-regulation of target genes involved in cell proliferation and apoptosis.
To check the effect of YAP on the cystic phenotype in vivo, we knocked down Yap using ASOs in young adult iKspPkd1 del mice.
The Pkd1 gene was inactivated in 18-day-old mice (PN18), and 2 weeks after gene inactivation they were injected via i.p., with However, the sporadic cysts that are formed did not seem to show impaired growth, consistently with in vivo data. Nuclear staining of TAZ was clearly observed in Yap KO cells, suggesting that lack of YAP increases TAZ shuttling in vitro ( Figure 2C).

| YAP and TAZ downstream targets expression is not changed by Yap knockdown in vivo
YAP and TAZ are transcriptional coactivators and can translocate into the nucleus where they can drive gene expression. To study the effect of Yap knockdown on the expression of its target genes, we quantified the expression of known YAP/TAZ targets, Wtip, Ajuba, Cyr61 and Amotl2. 9 Despite the consistent Yap reduction at the mRNA level, the expression of target genes is not changed in Yap ASO-treated compared to scrambled ASO-treated mice ( Figure 2D).
Additionally, we evaluated the expression of Ki-67, a marker for cell proliferation, as YAP and TAZ can regulate transcriptional programmes that control cell proliferation. 9 We did not find significant differences between Yap ASO and scrambled ASO treated mice ( Figure 2E). In conclusion, the knockdown of Yap does not affect the expression of the downstream targets we tested.

| WNT and TGF-β pathways seem to be more active in Yap ASO mice
It is well known that both YAP and TAZ can interact with the final effectors of the WNT and TGF-β pathways. 10 For this reason, we F I G U R E 2 Effect of YAP modulation in vivo and in 3D cultures (A) Representative IHC for aquaporin-2 (AQP2-collecting duct), Tamm-Horsfall (THP-distal tubules) and megalin (MEG-proximal tubules) on sequential slides of ASO-treated mice kidneys. The staining shows the presence of dilated tubules and small cysts that are positive for each marker. Scale bar 100 µm. B, Representative periodic acid-Schiff (PAS) staining of renal tissue from mice treated with scrambled ASO and Yap ASO. Scale bar 200 µm. C, H&E staining of formalin-fixed, paraffin-embedded cysts grown in Matrigel after forskolin stimulation (top row). Representative IHC of formalin-fixed, paraffin-embedded cysts of forskolin-treated Wt and mutant mIMCD3 cells stained for YAP (middle row) and TAZ (bottom row). Wild-type epithelial cells and Pkd1 KO cells grown in Matrigel spontaneously develop cystic structures with a visible lumen and can swell after forskolin stimulation. Yap KO cells and Yap/Pkd1 double KO cells grown in Matrigel showed impaired cyst formation, with the majority of the cells developing tumour-like mass structures, which did not respond to forskolin stimulation. Only the sporadic cysts that developed a normal lumen before forskolin treatment increased in size after stimulation. For mIMCD3 Yap/Pkd1 KO, cysts from multiple fields of view are shown separated by a white line. White arrowheads indicate cytoplasmic localization of the proteins; black arrowheads indicate nuclear localization of the proteins. Scale bar 50 µm. D, Gene expression (fold-change) of YAP/TAZ targets at the sacrifice in mice treated with Yap ASO and scrambled ASO. Each symbol represents a mouse. Mean with ± SD. n.s. (not significant) refers to all the genes in the graph, t-test. E, Quantification of Ki-67-positive area. Each symbol represents a mouse. Mean with ± SD. n.s. not significant, t-test. F, Gene expression (fold change) of WNT pathway targets Axin2 and Myc at the sacrifice in mice treated with Yap ASO and scrambled ASO. Each symbol represents a mouse. Mean with ± SD. ** P < .01, n.s. not significant, t-test. G, Gene expression (fold-change) of TGF-β pathway targets, Acta2, Col1a1, Vim, Fn1, Pai1 and Mmp2, at the sacrifice in mice treated with Yap ASO and scrambled ASO. Each symbol represents a mouse. Mean with ± SD. * P < .05, ** P < .01, t-test. If no significance is indicated, the comparison is not significant checked the expression levels of several target genes regulated by β-catenin (Axin2 and Myc) and Smads (Acta2, Col1a1, Vim, Fn1, Pai1 and Mmp2). We observed increased Myc expression in Yap ASO-treated mice compared to scrambled ASO, but only a trend for Axin2 expression ( Figure 2F). Moreover, we saw significantly increased expression of alpha smooth muscle actin (Acta2) and vimentin (Vim) and a consistent trend for collagen 1 alpha 1 (Col1a1), fibronectin (Fn1), plasminogen activator inhibitor-1 (Pai1) and matrix metallopeptidase 2 (Mmp2) ( Figure 2G). Thus, although not conclusive, these results suggest that WNT and TGF-β pathways are more active upon Yap knockdown.

| D ISCUSS I ON
To investigate YAP as a potential target for therapeutic intervention in PKD, we used ASOs to selectively knockdown the expression of Yap in young adult iKspPkd1 del mice. We reached about 70% reduction in gene expression, indicating that ASOs can be a viable strategy to effectively and selectively down-regulate a target in kidneys in models for PKD. Our data clearly indicate that Yap knockdown using ASOs in a mouse model for ADPKD does not improve the cystic phenotype.  Hence, targeting TAZ together with YAP might be a viable strategy to inhibit cyst progression. However, TAZ physically interacts with PC1 and PC2, participating in common signalling routes involved in cyst formation. Moreover, knockout of Taz leads to cysts formation in mice and zebrafish, even in the absence of a Pkd1 mutation. [11][12][13] Thus, further studies are needed to clarify the role of TAZ in PKD.

Scrambled ASO
A recent paper showed that Yap KO was able to reduce PKD progression mildly in Pkd1-deficient mice. 14  interact with SMADs and β-catenin, preventing their nuclear translocation and transcriptional activity. 10 In our study, we observed increased expression of some of the downstream targets of WNT and TGF-β pathways in Yap ASO treated mice. This might suggest that reduced YAP levels cause an imbalance in the regulation of these interacting signalling pathways, either because YAP cannot physically interact with SMADs and β-catenin any more, or due to increased activation of TAZ overcompensating for YAP loss. As increased activation of WNT or TGF-β pathways is well known for promoting cyst formation in ADPKD, 15,16 it might explain why we did not see any amelioration of the phenotype with Yap knockdown in vivo. Further studies are necessary to unveil the exact molecular mechanisms.
In conclusion, although we cannot exclude that the Hippo pathway is involved in cyst growth, we believe that the strong nuclear YAP localization observed in cyst-lining epithelia is more a consequence of cell stretching rather than a driving force for cell proliferation. Indeed, down-regulation of YAP using ASOs did not affect cell proliferation nor the cystic phenotype in our Pkd1 KO mouse model.
Moreover, due to its profound interconnection with other signalling pathways, such as WNT and TGF-β, a therapeutic intervention for PKD based on the modulation of YAP levels might not be feasible, at least with the current knowledge.