The dark side of hippo signaling: A cancer promoter role

ABSTRACT

The Hippo signaling pathway regulates organ size and tissue homeostasis. Given this role it is unsurprising that dysregulation of this pathway has implications for cancer progression. A convincing body of literature shows that the Hippo pathway serves a tumor suppressive function with its inactivation leading to massive overgrowth. However, additional studies have also shown that activation of Hippo signaling can promote tumor progression. It remains unknown how a single pathway can produce such diametrically opposed effects. This lack of knowledge is in part due to our inability to make meaningful comparisons from studies which have taken place in a variety of cell types, tissues, and organisms. Recently however, we have published 2 studies using the Drosophila wing disk to study the Hippo pathway and have found that Hippo pathway activation can promote cell migration and invasion while Hippo pathway inactivation leads to overgrowth. Thus we propose here that Drosophila can provide a research platform with which to begin addressing how the Hippo pathway can both enhance and suppress tumor progression due to published pro- and anti-tumor functionalities of the Hippo pathway in the same tissue.

Hippo pathway – tumor promoter or suppressor?

Central to multicellular life is the ability of cells within a tissue to coordinate their growth and behaviors to collectively generate a functional tissue. The Hippo pathway has been found to be central to this collective regulation. This pathway has been extensively studied for its function in regulating proper organ size and shape through its control on proliferation, apoptosis, and stem cell identity.^1,2 The founding member of the Hippo pathway, Warts (wts), was discovered in 1995 in 2 independent papers using Drosophila as a model system.^3,4 In both studies, loss of wts led to dramatic tissue overgrowth and dysregulated morphogenesis. Since then, Drosophila has played a key role in elucidation of other Hippo pathway components and characterization of Hippo signaling in various healthy and diseased contexts with remarkable molecular conservation with mammals.^2,5-12

Due to its roles in regulating organ size, the Hippo pathway has logically been studied in relation to cancer progression. For example, dysregulation of YAP, the mammalian homolog of Yki, has been found to lead to cell transformation and tumor development in the mouse liver.¹⁰ Many additional studies have been performed examining a putative role of Hippo signaling in cancer with most of these studies finding that YAP and Yki act as oncogenes to promote cancer progression while upstream components such as Hippo act as tumor suppressors.² YAP activation acts as an oncogene by inducing transcriptional programs which promote proliferation and survival through allowing cells to escape contact inhibition and grow in an anchorage-independent manner.^9,13 In addition, YAP has also been shown to play a role in affecting cell identity through increasing expression of ZEB1/1 which increases an epithelial-to-mesenchymal transition and may play a role in tumor progression.^14,15 Along with these experimental findings, the Hippo pathway has also been found to be mutated or dysregulated in various human cancers,^16,17 which highlights the need to boost our understanding of this pathway.

Despite the overwhelming evidence that Hippo pathway dysregulation promotes tumor growth and progression, several studies have also emerged highlighting a different side of the story. For example, YAP is frequently deleted in haematological cancers and restoration of YAP expression leads cell death both in vitro and in vivo.¹⁸ Even in breast epithelial tissue, YAP has been found to display tumor suppressor abilities as knockdown in breast cell lines resulted in anchorage independent survival, increased migration, and invasiveness.¹⁹ Recently, the upstream regulator of Hippo Kibra has also been shown to promote cancer growth and motility in prostate cancer.²⁰ In addition to the above described cell intrinsic mechanisms, the Hippo pathway has recently been found to have a role in suppressing anti-tumor immune responses as loss of LATS1/2 kinases increased the innate and adaptive immune responses to cancer.²¹

Taken altogether, it is clear that the Hippo pathway can both enhance and impede cancer growth and invasion, depending on the models and contexts. However, we do not yet understand how a single pathway can display these dual and seemingly contradictory roles. This question will no doubt be an area of great research interest in the future due to its implications for human health. Currently however, there is no clear research path to address this seeming paradox due to the diversity of models used in previous studies and the inherent complexity and dangers in trying to make conclusions and generalizations across organisms, tissues, or cell identities. Recently however, we have published 2 studies using the “ptc” stripe model in the Drosophila wing imaginal disk to dissect in vivo function of the Hippo pathway and have found that Hippo pathway activation can promote cell migration and invasion²² while Hippo pathway inactivation leads to overgrowth.²³ Thus Drosophila represents an ideal model with which we can begin to discover how the Hippo pathway can produce such divergent results on cancer progression within a single cellular and tissue context.

Synopsis of our recently published work

Using the wing disk as a major model, we recently demonstrated that activated Hippo signaling under the “ptc” promoter induces cell invasion and epithelial mesenchymal transition via the JNK pathway (Fig. 1B).²² We showed that Hippo signaling activation via overexpression of Hpo or Wts, or knock down of yki triggered invasive migration toward the posterior compartment of the discs, along with increased matrix metalloproteinase 1 (MMP1) expression, a protein required for basement membrane degradation. We found that JNK signal activation is responsible for these phenotypes as inhibition of JNK completely abolished the Hippo pathway activation-induced cell invasion and MMP1 activation. Mechanistically, we further identified the bantam (ban)-Rox8 module as an essential linker downstream of Yki. Loss of ban upregulates Rox8 expression, which in turn induces JNK dependent cell invasion (Fig. 1B). Additionally, we also demonstrated a conserved role of YAP in regulating cell invasion and TIA1 (Rox8 homolog) expression,²² highlighting the importance of Drosophila as an excellent organism to address cancer-related signaling questions.

Figure 1.

Hippo signal inactivation induces growth, whereas Hippo activation promotes cell invasion via different mechanisms in Drosophila wing disk. (A) Hippo signal depleted cells directly upregulates rho1 transcription, which in turn induces JNK-dependent growth. (B) Elevated Hippo activity acts though ban-Rox8 module to induce JNK-mediated cell invasion.

Intriguingly, using the same “ptc” stipe model, we recently showed that impaired Hippo signaling also activates the JNK pathway.²³ Rather than prompting invasive migration, we found that inhibition of Hippo pathway transcriptionally up-regulates Rho1 to induce JNK-dependent growth (Fig. 1A). It is worth noting that in spite of the dramatic overgrowth and JNK activation upon ectopic Yki expression, we detected no GFP positive cells in the posterior part of the wing disk, indicating that excess Yki activity is not sufficient to drive cell invasion, in line with our recent model that Yki is a negative regulator of cell invasion.²² As the interplay between the JNK and Hippo pathways is highly context dependent,²⁴ and increasing studies show that both pathways possess components that are potentially attractive drug targets for cancer treatment,^25,26 it will be vital in the future to focus on one tissue/organ to carefully dissect the outcome of Hippo activity alternation in tumorigenesis.

YAP overexpression induces JNK-mediated growth, not invasion

YAP is commonly upregulated in a wide spectrum of human cancer cell lines as well as primary tumors,¹⁰ and it can functionally substitute for Yki in regulating cell proliferation and apoptosis in Drosophila.⁸ To further explore the consequence of ectopic YAP expression on cell migration in Drosophila wing disk, here we temporarily expressed YAP^S127A, an activated form of YAP, with ptc-Gal4 tublin-Gal80^ts,,^10,27 and observed robust MMP1 expression (Fig. 2A-B), JNK phosphorylation (Fig. 2C-D) and overgrowth in a JNK-dependent manner (Fig. 3B-C), in accordance with our model that JNK is essential for Yki/YAP induced growth.²³ However, it is noteworthy that despite the intensive MMP1 expression, the boundary of “ptc” stripe remains intact, and no cell invasion behavior observed, which is phenotypically different from JNK activation induced cell invasion behavior,^22,28-32 suggesting that YAP^S127A expression is not sufficient to induce cell invasion in the Drosophila wing disk “ptc” model. Furthermore, consistent with YAP's molecular interactions with TEAD, the mammalian ortholog of Sd, we also found that Sd is required for YAP^S127A induced MMP1 expression and overgrowth (Fig. 3B and D), as well as Rho1 transcription, as shown by in situ hybridization (Fig. 3E-G). Taken together, our data demonstrate that Yki's role in regulating Rho1-JNK signaling mediated overgrowth is evolutionary conserved.

Figure 2.

Ectopic YAP^S127A expression activates JNK signaling in Drosophila. Third instar wing discs are shown. Compared with control (A, C), ectopic YAP^S127A dramatically activates MMP1 expression (B) and JNK phosphorylation (D). Genotype: hs-Flp; act>y+>Gal4 UAS-GFP/+ (A), hs-Flp; act>y+>Gal4 UAS-GFP/UAS-Yap^S127A (B), ptc-Gal4 UAS-GFP/tub-Gal80^ts (C), ptc-Gal4 UAS-GFP/+; UAS-YAP^S127A/tub-Gal80^ts (D).

Figure 3.

JNK is required for YAP^S127A induced growth. Compared with control (A), YAP^S127A-induced MMP1 expression and overgrowth (B) were both significantly suppressed by co-expression of Bsk^DN (C) or inhibition of Sd (D). (E-G) In situ hybridization to Rho1 mRNA of third instar wing discs. Knocking-down sd suppresses YAP^S127A-induced Rho1 transcription. Genotype: ptc-Gal4 UAS-GFP/tub-Gal80^ts (A, E), ptc-Gal4 UAS-GFP/+; UAS-YAP^S127A /tub-Gal80^ts (B, F), ptc-Gal4 UAS-GFP/+; UAS-YAP^S127A, UAS-Bsk^DN/tub-Gal80^ts (C), ptc-Gal4 UAS-GFP/UAS-sd.RNAi; UAS-YAP^S127A/tub-Gal80^ts (D, G).

Concluding remarks

In summary, using the “ptc” stripe model in the Drosophila wing disk as our predominate model we have found that Hippo pathway activation can induce migration while Hippo inactivation can induce overgrowth, both in a JNK-dependent manner. Although these seemingly contradictory roles for Hippo signaling have been published previously by many laboratories, there has been no easy experimental route to pick apart why and how these divergent outputs of Hippo signaling can occur. This difficulty arises in the fact that most of the previous studies highlighting the pro-tumor functions of Hippo signaling have used different models or cellular contexts. This diversity of models used, while it is beneficial for certain aspects of research, has made it difficult to draw broadly applicable conclusions in regards to the mechanisms governing the Hippo pathway in cancer. Since we have published both pro and anti-tumor functionalities of the Hippo pathway in the same Drosophila wing disk system, we propose that through studies of the Drosophila wing disk we can begin addressing how the Hippo pathway can both enhance and suppress tumor progression. Using the power and speed of Drosophila genetics it will be possible to discover how the Hippo pathway is regulated in either its upstream activation or downstream output to induce differing phenotypes.

Studies on how the Hippo pathway can produce differing effects on cancer progression will have implications for human health. Currently, it has been proposed that treatments aimed at the Hippo pathway will have beneficial effects on patient outcomes.^16,33 We would argue however that a more detailed understanding of the nuances that allow the Hippo pathway to both promote and inhibit various aspects of cancer progression will allow for more precise therapies and that Drosophila can play a key role in leading us to this understanding.

Experimental procedures

Drosophila genetics and stocks

For experiments involving Yap^S127A expression, flies were raised at 18ºC to restrict Gal4 activity for 6 d, shift to 29ºC for 2 d to inactivate Gal80^ts, then dissection was performed. The following stocks were used for this study: ptc-Gal4, UAS-GFP, tub-Gal80^ts were obtained from Bloomington Stock Center; UAS-sd RNAi (V101497) was obtained from Vienna Drosophila RNAi Center; UAS-Bsk^DN, UAS-Yap^S127A (gift from Duojia Pan) were described previously.^10,31

Clonal analysis

Flip-out ectopic expression clones were generated by crossing UAS-Yap^S127A with y w hs-FLP;act>y⁺>Gal4 UAS-GFP. Clones were induced at the second instar: heat shock for 10 min at 37°C 48–72 h after egg laying (AEL), dissection were performed 36h after clone induction.

Immunostaining and in situ hybridization

Wing discs of third instar larvae were fixed in 4% paraformaldehyde and stained as described previously,³⁴ using mouse anti-MMP1 (1:100, DSHB), rabbit anti-phospho-JNK (1:200, Calbiochem). Secondary antibodies were anti-rabbit-Alexa (1:500, CST) and anti-mouse-Cy3 (1:500, Jackson Immuno Research). In situ hybridization was performed on wing discs as described previously.²³

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Acknowledgments

We thank Dr. Duojia Pan, the Vienna Drosophila RNAi Center and Bloomington Stock Center for reagents. X.M. specifically thank Dr. Lei Xue (Tongji University) for his mentoring and guidance.

Funding

This work was supported by grants from the National Natural Science Foundation of China (31601024) to X. M.