Deficiency of Yes-Associated Protein Induces Cataract in Mice

Cataract is a major cause of blindness worldwide, its complicated and unclear etiopathogenesis limit effective therapy. Here, we found that Yap, a downstream effector of the Hippo pathway, is specifically expressed in lens epithelial cells and Yap conditional knockout (cKO) in the lens leads to cataract. Histologically, Yap deficient lens show fewer epithelial cells, retention of nuclei and accumulation of morgagnian globules in the transitional zone and the posterior area. Mechanistically, GFAP-mediated Yap cKO leads to the reduced proliferation of epithelial cells, delayed fiber cell denucleation and increased cellular senescence in lens. Further RNA profiling analysis reveals Yap cKO results in a significant alteration in gene transcription that is involved in eye development, lens structure, inflammation, cellular proliferation and polarity. Collectively, our data reveal a novel function of Yap in the lens and links Yap deficiency with the development of cataract, making Yap a promising target for cataract therapy.

reported. However, the function of Yap-mediated signaling in lens epithelial cells and the development of cataract remains elusive.
Here, we first characterized the expression of Yap and GFAP-Cre recombinase in the developing and adult mouse lens and found that both Yap and GFAP are mainly expressed in the lens epithelium. Next, we demonstrated that conditional knockout (cKO) of Yap using GFAP-Cre leads to the development of cataract, with 69.4% incidence at 1.5-month and 98.2% at 3-month. Furthermore, we observed that Yap cKO inhibited proliferation as well as promoted abnormal differentiation and senescence of the lens epithelial cells, which ultimately led to abnormal lens structure and cataract formation. Thus, our findings revealed that Yap plays a crucial role in the development of cataract, which might be a potential target for cataract treatment.

Animals
Mice were maintained under a 12-h light/dark cycle at 23 °C and were provided with food and water ad libitum in the Animal Care Facility at the Institute of Biophysics (Beijing, China). All experiments involving animals were approved by and conformed to the guidelines of the Institutional Animal Care and Use Committee at the Institute of Biophysics of the Chinese Academy of Sciences (Beijing, China).

Immunohistochemical assays
Animals were euthanized by overdose of CO 2 and their whole eyes removed. Tissues were fixed overnight (O/N) in 4% paraformaldehyde at room temperature, processed, and frozen or embedded in paraffin. Serial sections were cut at 5 μm (paraffin) or 15 μm (frozen) and either used for hematoxylin and eosin (H&E) staining or immunohistochemical analysis. Visualization and imaging were performed with a Nikon Tie-A1 confocal microscope (Nikon Instruments Inc., Melville, NY, USA) and NanoZoomer Digital Pathology software (Hamamatsu, Iwata City, Shizuoka Pref., Japan).

ATP detection
ATP levels were determined using a bioluminescent ATP assay kit (Promega, Madison, WI, USA) according to the manufacturer's instructions.

RNA-Seq
Lenses from WT and Yap cKO mice were dissected. RNA was extracted for RNA-Seq according to standard protocols. EdgeR [18,19] was used to evaluate the statistical significance of differentially transcribed genes. The Gene Ontology (GO) enrichment analysis was performed using the Enrichr (http://amp.pharmmssm. edu/Enrichr/) [20,21].

Statistical analysis
The data were presented as the mean ± SEM. Statistical analyses were performed using the GraphPad Prism 7 software (San Diego, CA, USA). Student's t-test or Twoway RM ANOVA was used to determine significance (*P < 0.05, **P < 0.01).

Expression of Yap and GFAP Cre recombinase in postnatal mouse eyes
To investigate the role of Yap in mouse eyes, we first defined the expression of Yap in eyes by immunofluorescent staining and found that Yap expression was restricted to the lens epithelium (Fig. 1C), inner nuclear layer (INL), and ganglion cell layer (GCL) of the retina (Fig. 1B). It has been reported that Cre recombinase under the control of the GFAP gene promoter (GFAP-Cre) is mainly expressed in the developing lens epithelium starting at E18 [22], when the formation of the lens vesicle is almost complete [23]. To analyze the expression pattern of GFAP-Cre, we utilized R26R Tomato reporter mice for lineage-tracing of GFAP positive cells and found that GFAP Cre recombinase was mainly expressed in lens epithelium, INL and GCL of the retina, which was similar to Yap's expression pattern ( Fig.  1D and Fig. S1). Taken together, the Yap conditional knockout mice (Yap cKO, Yap f/f ; GFAP-Cre) provide us a tool to investigate the role of Yap in mouse eyes, especially in the lens epithelium.

Yap deficiency in GFAP positive cells leads to cataract
In GFAP-Cre mediated Yap cKO, the expression of Yap was greatly decreased in the lens but not the retina (Fig.  S2A-C). The mRNA level of Yap in the lens was significantly reduced at P7 (Fig. S2D). Interestingly, Yap deficient mice developed lens opacities at 1.5-month of age ( Fig. 2A), the lenses were smaller than that of wildtype (WT, Yap f/f ) and manifested a typical nuclear cataract phenotypecentral opacity surrounded by a relatively transparent cortical region ( Fig. 2A). This cataract became progressively severe as the mouse aged (Fig. 2B).
Histological changes in the eyes were visualized by H&E staining from WT or Yap cKO mice at different stages. We found that the morphology of Yap deficient eyes appeared normal before 21-days of age (Fig. S3). However, Yap deficiency led to a dramatic age-dependent cataract development (Fig. 3A-H), specifically, 1) a decreased number of lens epithelial cells from 21-day to 3-month ( Fig. 3A1-H1); 2) an accumulation of morgagnian globules from 21-day to 3-month ( Fig. 3A2-H2); 3) a disordered denucleation featured with increased numbers of hematoxylin positive cell in the transitional zone and the posterior area of the lens from 21-day to 3month ( Fig. 3A2-H2); 4) a cloudy organelle-free-zone and decreased size of the lens from 1-month to 3-month ( Fig.  3C-H). There was no significant abnormality in the retina of Yap deficient mice (Fig. S4). Next, we used aquaporin-0 (AQP0) immunofluorescence staining to examine the lens structure. We found that AQP0 was exclusively and evenly localized in fiber cell membranes in WT lens (Fig.  3I). In contrast, within Yap cKO lens (Fig. 3J), AQP0 was not detected in the organelle-free-zone (Fig. 3J1) and was markedly reduced in the transitional zone and the posterior lens (Fig. 3J2-J3). Similar to H&E staining, we observed an abnormal denucleation in the transitional zone and the posterior lens (Fig. 3J2-J3). Taken together, morphological and histological analysis of the lens from Yap cKO mice clearly demonstrated that Yap deficiency leads to cataract features, including reduced number of lens epithelial cells, abnormal fiber cell denucleation and morgagnian globules accumulation.

Yap is essential for the proliferation of lens epithelial cells
Lens epithelial cells maintain the ability to proliferate throughout a mammal's lifespan, which is essential for the lens function. Previous studies showed that the Hippo signaling pathway plays an important role in cell proliferation [24][25][26][27]. Here, we found that there is a significant decrease in the number of lens epithelial cells in Yap cKO mice (Fig. 3A1-H1). We performed Ki67 immunohistochemical staining to further determine the role of Yap in lens epithelial cells proliferation. We found that the number of Ki67 positive lens epithelial cells was significantly decreased in Yap cKO lenses at 0, 7, 14postnatal day (Fig. 4A-G). Moreover, we used αTN4 cells to examine the role of Yap on cell proliferation in vitro. Yap knockdown in αTN4 cells strongly suppressed cell proliferation (Fig. 4H-K). Taken together, our results reveal that Yap deficiency inhibits the proliferation of lens epithelial cells in vivo and in vitro.

Yap deficiency in GFAP positive cells leads to senescence in lens
Early cataract is accompanied by accumulated morgagnian globules and cellular senescence [28]. Furthermore, inhibition of cellular senescence could postpone cataract formation [29]. In our previous study, we found that Yap is crucial for the regulation of cellular senescence in normal diploid cells [30]. Here, we observed that morgagnian globules accumulation, a process associated with lens senescence, occurred in Yap cKO lens (Fig. 3B). Thus, we examined the senescenceassociated β-galactosidase activity in the lens and found that β-galactosidase activity was significantly increased in 1.5-month-old Yap cKO compared to WT lens (Fig. 5A). Moreover, we examined the expression levels of p21, p53 and p16 in the lens using quantitative RT-PCR (qRT-PCR) and found that mRNA levels of p21 and p53 were significantly increased in Yap cKO mice (Fig. 5B-C). Taken together, these results suggest that Yap deficiency in GFAP positive cells increased the mRNA levels of senescence-associated genes and promoted cellular senescence in lens.

RNA profiles in Yap cKO lens
To further illustrate the molecular mechanism underlying cataract formation induced by Yap deletion, we examined RNA expression profiling using RNA-seq in the lenses from WT and Yap cKO mice at 1.5-month-old. A set of differentially expressed genes have been identified in Yap deficient mice versus age-matched littermate controls (Fig. 6A). Gene Ontology (GO) enrichment analysis showed that these differentially expressed genes were frequently enriched in biological processes such as epithelial cell proliferation, differentiation and migration, inflammatory response, camera-type eye development as well as apoptotic process (Fig. 6B). These altered genes were also functionally linked with eye development, lens structure, inflammation, cell proliferation and polarity (Fig. 6C). Furthermore, the expression of 24 genes were validated using RT-qPCR (Fig. 6D-E) and we found that: 1) the expression levels of Sox2 and Pax6, which are important for lens development and cataract formation [31,32], were significantly decreased in Yap cKO lens (Fig. 7A-B); 2) Dnase2b, an enzyme of DNA degradation was significantly downregulated upon Yap knockout from 21-days of age, which correlates with abnormal denucleation of fiber cells (Fig. 7C); 3) Crystallins, the lens structure proteins, were dramatically reduced in Yap cKO lens (Fig. 7D-E); 4) Inflammation genes such as Tnfα and Il6 were significantly increased in Yap cKO lens (Fig. 7F-G). Taken together, Yap deficiency led to the reduced expression levels of Sox2, Pax6 and Dnase2b at an early stage, which may initiate cataract; and at a later stage, the downregulation of lens structural genes and increased inflammation might exacerbate the cataract process. Therefore, we argue that Yap regulates the expression of genes related to lens development and Yap deficiency leads to abnormal lens development and cataract formation.

DISCUSSION
Cataract is the most common disease that causes blindness and its complicated etiopathogenesis limits effective therapies. Therefore, it is of great medical significance to explore the mechanisms of cataract formation and progression so that a novel therapeutic target(s) can be developed to prevent or treat cataract．Here, we analyzed the temporal and spatial expression patterns of Yap in the murine lens. We found that Yap is mainly expressed in the lens epithelium. GFAP-Cre mediated Yap knockout in the lens leads to abnormal lens structure including morgagnian globules accumulation, reduced lens epithelial cells and abnormal denucleation in the transitional zone and posterior of the lens. Further mechanistical studies indicated that (1) Yap knockout inhibited the proliferation of lens epithelial cells in vivo and in vitro; (2) Yap deficiency led to senescence in lens; (3) Sequential denucleation of fiber cells was disordered in Yap deficient mice; (4) Increased inflammation in the lens upon Yap knockout (Fig. 8). Additional transcriptional profiling analysis found that the downregulated Sox2 and Pax6 at early stages might initiate cataract, and the increased inflammation factors and decreased levels of lens structural proteins in the later stages could exacerbate cataract formation. Collectively, our data reveals a novel function of Yap in the lens and links Yap deficiency with the development of cataract, with an implication of a diagnostic and therapeutic target for cataract.
This study is the first to demonstrate that Yap deficiency causes cataract in mice. Here, we categorized the pathological alterations of Yap deficiency-induced cataract into two stagesthe primary effects and the secondary effects. The primary effects include the decrease in proliferation of lens epithelial cells and in this stage, Yap deficiency downregulated the expressions of genes related to lens proliferation and developmentrelated genes (Sox2, Pax6 and Dnase2b), which initiated cataract. In the secondary stage, Yap deletion accelerated cellular senescence and reduced the expression levels of lens structural proteins as well as increased inflammation, which finally promoted cataract formation.
The initial disorder of cataract development is the suppressed proliferation of lens epithelial cells in Yap cKO mice, which leads to fewer lens epithelial cells from P21. Lens epithelial cell has been reported to be essential for lens development and its decreased number leads to cataract in alpha-crystalline knockout or FoxE3 knockout mice [33,34]. Further RNA-seq and subsequent RT-qPCR analysis reveal the downregulation of Sox2 and Pax6 in the early stage, which could be the molecular mechanism underlying the decreased proliferation. Moreover, as key regulators in lens development, Sox2 and Pax6 could activate a battery of genes for early lens development such as δ1-crystallin [35]. Mutations of Sox2 in mice leads to anophthalmia or microphthalmia [31] and patients with Sox2 mutation are more likely to develop posterior cortical cataract [32]. In addition, Yap could also transcriptionally induce Sox2 expression through physically interacting with transcription factor Oct4 in stem-like cells from non-small cell lung cancer [36]. We also identified Sox2 as a key molecule in Yap deficient lens by functional analysis with the MetaCore software. Therefore, in addition to decreased proliferation, Yap-deficiency downregulated Sox2/Pax6 expression that might lead to developmental dysfunction in the lens.
During lens development, optical clarity is ensured by the degradation of intracellular nuclei in fiber cells, in which Dnase2b plays an important role [37]. Many reports showed that the decrease of Dnase2b expression is closely related to cataract formation [38][39][40]. In our study, the mRNA level of Dnase2b in Yap deficient lens decreased at P21, indicating Yap could transcriptionally regulate Dnase2b during lens development and Yap deficiencyinduced downregulation of Dnase2b led to the abnormal denucleation and finally cataract.
We have reported that Yap transcriptionally regulates Cdk6 in diploid cells and knockdown of Yap induces cellular senescence [30]. However, we observed that Cdk6 level was decreased only at P7 (Fig. S5), suggesting that Cdk6 might not be a main target in lens development. We further found that β-galactosidase activity was significantly increased in Yap deficient lens, and the mRNA levels of p21 and p53 were also significantly elevated since P21, which is correlated with senescenceinduced morgagnian globules accumulation during cataract formation.
Many studies show that inflammation [9,41,42] and mutation of crystalline proteins [43][44][45] are involved in the etiology of cataract. We also observed that there were increased levels of inflammation factors and the decreased mRNA levels of lens structural genes in Yap deficient lens at P30. Thus, we argue that inflammation further accelerates cataract formation. However, whether Yap deficiency directly or indirectly induced inflammation remains to be further investigated.
Yap and Taz are twin homologues in the mammalian system and mostly have redundant functions. In our study, we found that the lenses in GFAP-mediated Taz cKO mice are normal. Similar to Yap cKO alone, GFAPmediated Taz and Yap double cKO mice showed cataract phenotype (Fig. S6). Previous studies have demonstrated that Yap and Taz also have distinct roles. For example, Yap global knockout is embryonically lethal. However, Taz conventional knockout mice only exhibit some developmental defects such as polycystic kidney and emphysema [46][47][48]. In addition, Yap and Taz exert different biological functions in myogenic differentiation and they are differently regulated through distinct interaction with Parafibromin [49,50]. Not surprisingly, we indeed found that the expression level of Taz in mouse lens was much lower than Yap. Our results reveal that Yap, rather than Taz, plays a critical role in lens development and cataract formation.
As GFAP-Cre recombinase is also expressed in the INL and GCL of retina, it is possible that the cataract that occurred in the Yap cKO model may secondary to Yap cKO-induced retinal defects. However, the general retinal morphology was not altered in Yap cKO mice. Additionally, we observed that the gross brain structure in Yap cKO mice appears normal. Therefore, we argue that Yap cKO lens phenotypes are autonomously developed (Fig. S4).
In summary, our results show that Yap plays an essential role in the formation of cataract in the mouse. The pathological examination and transcriptional profiling analysis provide insight into the development of cataract as a consequence of Yap deficiency.