Polycystin-1 is required for insulin-like growth factor 1-induced cardiomyocyte hypertrophy

Cardiac hypertrophy is the result of responses to various physiological or pathological stimuli. Recently, we showed that polycystin-1 participates in cardiomyocyte hypertrophy elicited by pressure overload and mechanical stress. Interestingly, polycystin-1 knockdown does not affect phenylephrine-induced cardiomyocyte hypertrophy, suggesting that the effects of polycystin-1 are stimulus-dependent. In this study, we aimed to identify the role of polycystin-1 in insulin-like growth factor-1 (IGF-1) signaling in cardiomyocytes. Polycystin-1 knockdown completely blunted IGF-1-induced cardiomyocyte hypertrophy. We then investigated the molecular mechanism underlying this result. We found that polycystin-1 silencing impaired the activation of the IGF-1 receptor, Akt, and ERK1/2 elicited by IGF-1. Remarkably, IGF-1-induced IGF-1 receptor, Akt, and ERK1/2 phosphorylations were restored when protein tyrosine phosphatase 1B was inhibited, suggesting that polycystin-1 knockdown deregulates this phosphatase in cardiomyocytes. Moreover, protein tyrosine phosphatase 1B inhibition also restored IGF-1-dependent cardiomyocyte hypertrophy in polycystin-1-deficient cells. Our findings provide the first evidence that polycystin-1 regulates IGF-1-induced cardiomyocyte hypertrophy through a mechanism involving protein tyrosine phosphatase 1B.

We recently determined that polycystin-1 (PC1), encoded by the Pkd1 gene, is a critical regulator of cardiomyocyte hypertrophy elicited by mechanical stretch [16]. PC1 is a plasma membrane protein expressed in various tissues that has a large N-terminal extracellular domain and a short cytoplasmic C-terminal domain and modulates several pathways, such as calcineurin/nuclear factor of activated T-cells (NFAT) and Signal transducer and activator of transcription 6 (STAT6) [17]. PC1 also interacts with various membrane proteins, including cell-cell communication proteins [18,19], cell-matrix communication proteins [20], the inositol trisphosphate receptor (IP3R) [21], serine/threonine phosphatase 1 alpha protein [22] and tyrosine phosphatases [23]. Recently, we identified PC1 as a mechanosensor in cardiomyocytes that governs L-type Ca 2+ channel protein stability [16]. PC1 is crucial for myocardial function, as well as mechanisms involving mechanical stretch (in vitro) and pressure overload (in vivo) that may induce cardiomyocyte hypertrophy [16]. However, our previous data show that PC1 is not necessary for phenylephrine-dependent cardiomyocyte hypertrophy, suggesting that PC1 activity is stimulus-dependent. Whether PC1 is a general mediator of cardiomyocyte hypertrophy or a mediator of IGF-1-induced cardiomyocyte hypertrophy has not been previously studied. Here, we describe the role of PC1 in IGF1-induced cardiomyocyte hypertrophy. Our results show that PC1 knockdown negatively regulates the IGF-1R, Akt, and ERK signaling pathways, preventing IGF-1-induced cardiomyocyte hypertrophy. Finally, we provide evidence of PTP1B dysregulation upon PC1 knockdown.

Animals
All animals were handled according to the guidelines stated in the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health (NIH Publication, 8th Edition, 2011). The protocol was approved by the Institutional Ethics Review Committee of the Faculty of Chemical and Pharmaceutical Sciences, Universidad de Chile. We used Sprague-Dawley rat pups on postnatal days 1-3 to isolate neonatal rat ventricular myocytes (NRVM).

Cardiomyocyte transfection
Small interfering RNAs (siRNA) for PC1 and scrambled (Scr; negative control) siRNA were used according to the manufacturer's instructions. NRVM were cultured and transfected with siRNAs (120 nM) with Oligofectamine™ or Lipofectamine RNAiMax. After 12 h of incubation, the medium was removed and replaced with DMEM/M199. Experiments were performed 24 h after siRNA transfection. Knockdown efficiency was determined by Western blot after 48 h.

Treatments
NRVM were treated with IGF-1 (10 nM) to study both hypertrophy (48 h) and the IGF-1R signaling pathway (5 min). When using phosphatase inhibitors, NRVM were treated with the inhibitors 1 h prior to IGF-1 stimulation and were maintained throughout the experiment. CinnGEL was dissolved in 0.05% DMSO, and the same solution without CinnGEL was used as a control condition.

Cell morphology and sarcomerization
After treatments, cells were fixed with paraformaldehyde and permeabilized with Triton X-100. Rhodamine-phalloidin was used to stain F-actin (1:500, Thermo Fisher Scientific, MA) [24]. Cardiomyocyte hypertrophy, area, perimeter, the percentage of sarcomerization and sarcomere fluorescence profile were calculated from fluorescent images obtained using an epifluorescence microscope (Carl Zeiss Axiovert 135, LSM Microsystems). At least 50 cells from randomly selected fields were analyzed using the ImageJ software (NIH, USA).

Statistical analysis
Depending on the type of experiment, results are shown as representative images or mean ±SD of at least three independent experiments. The Mann-Whitney-U or Kruskal-Wallis tests were used for data analysis. A p value <0.05 was set as the level of statistical significance.

PC1 knockdown blunts IGF-1-induced cardiomyocyte hypertrophy
To elucidate whether PC1 is required for IGF-1-dependent cardiomyocyte hypertrophy, PC1 expression was reduced in neonatal rat ventricular myocyte (NRVM) using specific siRNA constructs (siPC1) and scrambled (Scr) siRNA as a negative control. In our previous work, we tested two different siRNAs to knockdown PC1 in NRVM. Both siRNAs decreased PC1 expression in the same level [16], so in this work, we used only one of them. siPC1 decreased PC1 protein content by 50% compared to the control after 24 h of transfection ( Fig 1A) and the Pkd1 mRNA levels (Fig 1B). To determine hypertrophy in vitro, we measured the following markers: β-myosin heavy chain (MHC) protein expression, brain natriuretic peptide (BNP) mRNA levels, and [ 3 H]-leucine incorporation. Interestingly, PC1 knockdown reduced IGF-1 dependent increases of β-MHC protein content, BNP mRNA abundance, and [ 3 H]-leucine incorporation (Fig 1C-1E). Moreover, IGF-1 did not change Pkd1 mRNA levels in cardiomyocytes (Fig 1F), suggesting that IGF-1 does not alter PC1 expression.
In vitro cardiomyocyte hypertrophy can be defined phenotypically as an increase in cellular area, perimeter, and the degree of sarcomerization. NRVM have a very rudimentary sarcomeric structure, which turns into a highly organized "stair-like" structure after a pro-hypertrophic stimulus [24][25][26]. IGF-1 treatment increased cardiomyocyte area, perimeter, and percentage of cells with highly organized sarcomeres (Fig 2A and 2B), while PC1 knockdown completely prevented these IGF-1 dependent changes (Fig 2C-2E). Taken altogether, these results suggest that PC1 is required for IGF-1-induced cardiomyocyte hypertrophy.

PC1 regulates the IGF-1 receptor signaling pathway
IGF-1 activates its receptor (IGF-1R) by inducing tyrosine autophosphorylation, subsequently activating downstream signaling pathways such as Akt and ERK1/2 [27][28][29]. It is well known that IGF-1 exerts its pro-hypertrophic effects via the Akt and ERK1/2 signaling pathways [27,29]. To study the role of PC1 in the activation of IGF-1R and its downstream signaling pathways, we measured the phosphorylation status of IGF-1R, ERK1/2, and Akt upon IGF-1 stimulation. IGF-1 increased IGF-1R phosphorylation at 5 min, and PC1 knockdown partially prevents this activation (Fig 3A). Total IGF-1R content did not change upon PC1 knockdown in NRVM (Fig 3A). Next, we measured both Akt and ERK1/2 activation in NRVM after IGF-1  (Fig 3B) and ERK1/2 ( Fig 3C) phosphorylation upon IGF-1 treatment. In order to evaluate whether these changes were specific for the IGF-1 receptor, we assessed insulin receptor (IR) activation upon insulin stimulation. Insulin triggered similar IR phosphorylation levels in NRVM transfected with either Scr or siPC1 (Fig 3D). We did not detect IGF-1R or IR phosphorylation in basal condition. Taken together, these results suggest that PC1 has a specific effect on the IGF-1R signaling pathway in cardiomyocytes.

PC1 regulates IGF-1R activation through protein tyrosine phosphatase 1B
Tyrosine phosphatases fine-tune tyrosine kinase receptor function, modulating their phosphorylation status [30,31]. Some evidence suggests that PC1 interacts with several isoforms of receptor protein tyrosine phosphatases, forming multimeric complexes that regulate PC1-dependent signaling [23]. Therefore, we hypothesized that PC1 knockdown might impair phosphatase activity. Using the phosphatase inhibitors PhoSTOP and Na 3 VO 4 , we disrupted the inhibitory effect of PC1 knockdown on IGF-1R phosphorylation upon IGF-1 stimulation ( Fig  4A). These data suggest that tyrosine phosphatases are involved in the mechanism through which PC1 regulates IGF-1R phosphorylation in NRVM.

PC1 knockdown impairs PTP1B activation and IGF-1-induced cardiomyocyte hypertrophy
To assess the effect of PC1 knockdown on PTP1B activation in IGF-1-induced cardiomyocyte hypertrophy, NRVM were exposed to IGF-1 in the presence or absence of CinnGEL, and hypertrophic parameters were measured. CinnGEL reduced the inhibitory effect of PC1 knockdown on BNP mRNA upon IGF-1 stimulation (Fig 5A). Moreover, morphologic studies and fluorescence intensity profiles (Fig 5B and 5C) confirmed that CinnGEL also prevented the decrease in cell area (Fig 5D), perimeter (Fig 5E), and sarcomerization (Fig 5F) induced by PC1 knockdown upon IGF-1 stimulation. All these data are presented as fold with respect to Scr. These results suggest that PTP1B is involved in PC1-dependent regulation of IGF-1-induced cardiomyocyte hypertrophy.

Discussion
In this study, we showed that IGF-1-induced cardiomyocyte hypertrophy requires PC1. PC1 knockdown using siRNA blunted the IGF-1-dependent upregulation of hypertrophic β-MHC protein levels, BNP mRNA abundance, [ 3 H]-leucine incorporation, sarcomerization, cell area, and perimeter of NRVM. Furthermore, we provided evidence that PC1 knockdown affects the phosphorylation status of IGF-1R depending on PTP1B activity.
IGF-1 participates in the initiation and development of physiological cardiac hypertrophy [32,33], and its cardiomyocyte-specific overexpression leads to cardiac hypertrophy in transgenic mice [7]. IGF-1 is critical for cardiomyocyte survival due to its potent cardioprotective effect [32,33]. Our group and others have shown that IGF-1 activates multiple signal transduction pathways in cardiomyocytes [8,27,28,34]. To our knowledge, this is the first study showing the involvement of PC1 in IGF-1 signaling. Our data suggest that PC1 knockdown specifically regulates the IGF-1 signaling pathway, but without affecting other tyrosine kinase receptors, such as the IR, or participating in phenylephrine-mediated cardiomyocyte hypertrophy [27].
Tyrosine phosphorylation is regulated by phosphatases to terminate the activation of tyrosine kinase receptors. Moreover, regulation of both IGF-1 and insulin signaling involves serine phosphorylation of Insulin receptor substrate 1/2 (IRS1/2) [35]. Insulin-or IGF-1-induced serine phosphorylation of IRS1/2 dissociates these proteins from their receptors, preventing

PLOS ONE
Polycystin-1 and IGF-1-induced cardiac hypertrophy tyrosine phosphorylation and inhibiting binding with downstream effectors. Therefore, this process serves as negative feedback mechanism [35]. Our results showed that PC1 knockdown not only reduces IGF-1-induced Akt and ERK phosphorylation but also decreases IGF-1R phosphorylation. This effect seems to be specific to IGF-1R signaling, as PC1 knockdown did not affect insulin-induced IR phosphorylation in cardiomyocytes. On the other hand, PC1 knockdown reduces only around 50% of IGF-1-dependent IGF-1R phosphorylation, as well as Akt and ERK1/2 phosphorylations. This result could be explained, at least partly, because PC1 siRNA only reduces 50% of PC1 expression in NRVM. However, a significant amount of PC1 remains in the cell in these cells, and the knockdown of PC1 could affect its subcellular distribution or traffic to the plasma membrane, thus explaining the effects on the IGF-1R. We acknowledge as a study limitation that the subcellular location of the remaining PC1 was not studied here. Interestingly, PC1 has been located both in the plasma membrane and other subcellular compartments [36]. In aggregate, our data strongly suggest that PC1 regulates positively IGF-1R activity induced by IGF-1 in NRVM.
Receptor protein tyrosine phosphatases interact with polycystins both in the primary cilia and at adhesion complexes [20,23]. The presence of cilia in cardiomyocytes is still controversial [37][38][39], and methodological issues to identify cardiac myocytes or cilia could explain these discrepancies. However, independent of the presence of cilia, cardiomyocyte-specific knockdown of PC1 promotes cardiomyocyte alterations [16,40] that are distinct from the ones elicited by Polycystin-2 knockdown [37]. Moreover, we previously showed that NRVM knockdown to PC1 did not change the PC2 expression [41] as well as cardiomyocytes-restricted silencing of PC1 mice model [16], suggesting that our results are not influenced by PC2 alterations. Altogether, this evidence suggests that these proteins play different roles in cardiomyocytes, so we did not explore the role of Polycystin-2 in IGF-1-induced cardiomyocyte hypertrophy.
Interestingly, we found that PTP1B inhibition normalizes IGF-1R phosphorylation and the hypertrophic response upon IGF-1 stimulation in cells lacking PC1. These results suggest that PC1 acts as a negative regulator of PTP1B activity, allowing normal activation of the IGF-1R. IGF-1R can also be dephosphorylated by the tyrosine phosphatase SHP-2 [42]. Therefore, although our data support that PTP1B has a relevant role in IGF-1R dephosphorylation, we cannot rule out the participation of other tyrosine phosphatases in the PC-1-dependent effects. Previous reports have indicated that PC1 can interact and form multiprotein complexes with focal adhesion proteins [18,20], cell-cell adherens junction proteins [18,20], the IP3R [21], intermediate filaments [19], serine/threonine protein phosphatase-1α [43] and tyrosine phosphatases [31]. Tyrosine phosphatases also interact with and bind to receptor tyrosine kinases (RTKs) [31]. We hypothesize that PC1 controls at least the activity of PTP1B to regulate IGF-1R activity. Here, we have shown that the use of both general phosphatase inhibitors and specific tyrosine phosphatase inhibitors completely revert the PC1 knockdown-dependent decreases in IGF-1R, Akt and ERK phosphorylation. Therefore, these data suggest that PC1 may regulate the IGF-1 signaling pathway by controlling the function of tyrosine phosphatases. A possible mechanism through PC1 regulates PTP1B activity could involved AKT as a negative modulator of this phosphatase. PC1 stimulates AKT activity [41,44], while PTP1B phosphorylation on Ser(50) by AKT negatively regulates the activity of this phosphatase [45]. Future experiments should address this hypothesis andinvestigate a potential interaction between PTP1B and PC1 with the IGF-1R and whether PC1 regulates PTP1B availability to clarify the mechanism involved in this signaling pathway.
Furthermore, interactions between the IGF-1R and mechanosensors, such as integrins, have also been described [22,46]. This interaction is fundamental for proper IGF-1R signaling in response to IGF-1 in osteoblasts [22]. PC1 is a plasma membrane protein expressed in various tissues. As previously reported, we have identified PC1 as a mechanosensor in cardiomyocytes [16]. Whether PC1 modulates IGF-1R signaling by only regulating phosphatase levels or exerting its control as a mechanoreceptor remains unknown. Moreover, future studies are necessary to determine whether PC1 and PTB1B interact directly or indirectly and will broaden our understanding of this pathway.

Conclusion
Our results show for the first time that PC1 is required for IGF-1-dependent induction of cardiomyocyte hypertrophy. PC1 may regulate IGF-1R phosphorylation by controlling the availability of tyrosine phosphatases. A graphical model is shown in Fig 6. Supporting information S1 Fig. This is the S1 raw images.