Phosphoinositide 3-Kinase/Akt Inhibits MST1-Mediated Pro-apoptotic Signaling through Phosphorylation of Threonine 120*

The protein kinase mammalian sterile 20-like kinase 1 (MST1) is a mammalian homologue of the Drosophila hippo and plays a critical role in regulation of programmed cell death. MST1 exerts pro-apoptotic function through cleavage, autophosphorylation-Thr183 and subsequent translocation to the nucleus where it phosphorylates a number of molecules, including LATS1/2, FOXO, JNK, and histone H2B. Here, we show that the cleavage of MST1 is inhibited by the phosphatidylinositol 3-kinase/Akt pathway. Akt interacts with MST1 and phosphorylates a highly conserved residue threonine 120 of MST1, which leads to inhibition of its kinase activity and nuclear translocation as well as the autophosphorylation of Thr183. Phospho-MST1-Thr120 failed to activate downstream targets FOXO3a and JNK. Further, inverse correlation between pMST1-Thr120 and pMST1-Thr183 was observed in human ovarian tumors. These findings indicate that the phosphorylation of MST1-Thr120 by Akt could be a major mechanism of regulation of the Hippo/MST1 pathway by cell survival signaling.

The protein kinase mammalian sterile 20-like kinase 1 (MST1) is a mammalian homologue of the Drosophila hippo and plays a critical role in regulation of programmed cell death. MST1 exerts pro-apoptotic function through cleavage, autophosphorylation-Thr 183 and subsequent translocation to the nucleus where it phosphorylates a number of molecules, including LATS1/2, FOXO, JNK, and histone H2B. Here, we show that the cleavage of MST1 is inhibited by the phosphatidylinositol 3-kinase/Akt pathway. Akt interacts with MST1 and phosphorylates a highly conserved residue threonine 120 of MST1, which leads to inhibition of its kinase activity and nuclear translocation as well as the autophosphorylation of Thr 183 . Phospho-MST1-Thr 120 failed to activate downstream targets FOXO3a and JNK. Further, inverse correlation between pMST1-Thr 120 and pMST1-Thr 183 was observed in human ovarian tumors. These findings indicate that the phosphorylation of MST1-Thr 120 by Akt could be a major mechanism of regulation of the Hippo/MST1 pathway by cell survival signaling.
Mammalian sterile 20-like kinase 1 (MST1) 4 contains an Ste20-related kinase catalytic domain in the N-terminal region followed by a non-catalytic tail, which contains a successively autoinhibitory domain, a dimerization domain, two nuclear export sequences, and a nuclear localization signal (1). It has been shown that the non-catalytic tail of MST1 is cleaved by caspase upon various apoptotic stimuli, including death receptor triggering by CD95/FasL, staurosporine, etoposide, and ceramide as well as heat shock and arsenite. The cleaved MST1 translocates into the nucleus, where it promotes chromatin condensation and herein apoptosis (2)(3)(4). Previous studies suggest that MST1 induces apoptosis by phosphorylation of Histone H2B-Ser 14 in mammalian cells and Ser 10 in Saccharomyces cerevisiae (5,6). MST1 has also been implicated in the control of cell death via phosphorylation of FOXO3a-Ser 207 and the corresponding site of FOXO1-Ser 212 (7,8). In addition, Thr 183 within the N-terminal MST1 activation loop has been defined as a primary autophosphorylation site. The autophosphorylation of Thr 183 is essential for MST1 activation (9).
Recent studies showed that Hippo in Drosophila, a homolog of mammalian MST1/2, inhibits cell growth and survival by interaction with Salvador (Sav) and Warts (Wts) resulting in inhibition of transcription and/or degradation of cyclin E and DIAPs (Drosophila inhibitor of apoptosis) (10, 11) through phosphorylation of Yorkie, the Drosophila ortholog of the mammalian transcription co-activator yes-associated protein (12). Yorkie and yes-associated protein have recently been shown to be negatively regulated by the Hippo/MST1/2 pathway and play an important role in mediating cell contact inhibition, organ size, and tumorigenesis (13,14). In mammals, MST1 has also been shown to activate JNK and p38 kinase pathways through MKK4/MKK7 and MKK3/MKK6, respectively (2).
Accumulating evidence indicates that PI3K/Akt signaling is a major cell survival pathway. It suppresses apoptosis through regulation of a number of molecules (15). The first anti-apoptotic Akt target identified was BAD, a pro-apoptotic protein in the Bcl-2 family (16). Akt phosphorylates BAD on Ser 136 , which promotes the BAD releasing from Bcl-xL/BcL2 complex and binding 14-3-3 proteins in the cytosol, thus inactivating its pro-apoptotic function. It has been shown that Akt phosphorylates and inactivates FOXO family transcription factors (17), including FOXO1 (also called FKHR), FOXO3a (also named FKHRL1), and FOXO4 (also called AFX). The phosphorylation by Akt negatively regulates FOXO activity by relocalizing FOXO from the nucleus to the cytoplasm, where it is sequestered away from target genes through interacting with 14-3-3 (18). In addition, several pro-apoptotic and anti-apoptotic proteins are also phosphorylated by Akt, including ASK1 (19,20), XAIP (21), , BAX (23,24), and HtrA2 (25), which leads to direct activation of cell survival pathway. Moreover, Akt has been shown to activate NFB pro-survival signaling by phosphorylation of IKK␣ (26,27). Although these targets could play important roles in Akt survival signaling, FOXO/ DAF-16 is the only major target that has genetically been proven across different species (15).
A previous study showed that epidermal growth factor stimulation caused a transient drop of MST1 kinase activity (28). However, the one or more upstream signaling regulators of MST/Hippo are still largely unknown. In this study, we demonstrate that MST1 is regulated by Akt. Akt phosphorylates MST1 at Thr 120 in vitro and in vivo, which leads to inhibition of MST1 cleavage and kinase activity as well as nuclear translocation. We also found that Akt phosphorylation of MST1 reduces caspase 3 activity, JNK activation, and MST1-induced phospho-FOXO3-Ser 207 . Consistent with these results, Akt-null cells or cancer cells with knockdown of Akt show elevated levels of MST1 kinase activity and decreased levels of phospho-MST1-Thr 120 , when compared with the cells expressing Akt. Further, Akt activation is inversely correlated with autophosphorylation of MST1-Thr 183 but paralleled with MST1-Thr 120 phosphorylation in cancer cell lines and tumor samples examined. Collectively, our findings suggest that MST1 is a bona fide substrate of Akt and that Akt could play a critical role in regulation of the Hippo/MST1/2-Sav/WW45-Wats/LATS1/2 pathway.
Immunoprecipitation and Immunoblotting-Immunoprecipitation and immunoblotting analysis was performed as described previously (27). Briefly, protein lysates were incubated with appropriate antibodies indicated in the figure legends in the presence of 25 l of protein A-protein G (2:1)agarose beads for 3 h at 4°C. After washing three times, the immunoprecipitates were subjected to in vitro kinase assay (see below). Protein expression was determined by probing Western blots of total cell lysate or immunoprecipitates with the appropriate antibodies as noted in the figure legends. Detection of antigen-bound antibody was carried out with the ECL Western blotting Analysis System (Amersham Biosciences).
In Vitro Kinase Assay-Protein kinase assay was performed as previously described (20). Briefly, reactions were carried out in the presence of 10 Ci of [␥-32 P]ATP (Amersham Biosciences) and 3 M cold ATP in 30 l of buffer containing 20 mM HEPES (pH 7.4), 10 mM MgCl 2 , 10 mM MnCl 2, 1 mM dithiothreitol), and 2 g of myelin basic protein (MBP) as substrate. After incubation at room temperature for 30 min, the reaction was stopped by adding protein loading buffer and then separated on a SDS-PAGE gel. Each experiment was repeated three times, and the relative amounts of incorporated radioactivity were determined by autoradiography and quantified with a PhosphorImager (Amersham Biosciences).
Cell Death Analysis-TUNEL assay was used to detect apoptosis using an in situ cell death detection kit (Roche Applied Science). The experiments were performed three times in triplicate. Total cell death was analyzed with trypan blue exclusion assay as previously described (20).
Immunohistochemistry-Immunostaining was performed as previously described (27). Briefly, formalin-fixed, paraffin-embedded sections were subjected to antigen retrieval by boiling in 0.01 M sodium citrate buffer (pH 6.0) in a microwave oven after dewaxing and rehydration. The Vectastain ABC Kit for sheep IgG (Vector Laboratories) was used to immunostain the tissue sections with anti-pAkt-Ser 473 , -pMST1-Thr 120 , and -pMST1-Thr 183 antibodies. Endogenous peroxidase and biotin were blocked, and sections were incubated for 1 h at room temperature with individual antibody. The remainder of the staining procedure was performed according to the manufacturer's instructions using diaminobenzidine tetrahydrochloride as the chromogen and hematoxylin for counterstaining.

PI3K/Akt Pathway Inhibits MST1-mediated Apoptosis as
Well as MST1 Cleavage and Activation-Previous studies have demonstrated that the MST1 is cleaved and activated upon apoptotic stimuli and that the activation of MST1 is required for DNA damageand FasL-induced apoptosis (2,3). However, the regulation of MST1 cleavage and activation still remains elusive. We first examined whether extracellular cell survival signal inhibits DNA damage-induced MST1 cleavage. As shown in Fig. 1A and supplemental Fig. S1, MST1 was cleaved after exposure to STS or CD95/FasL. However, the cleavage was considerably reduced by IGF-1 treatment. Because IGF-1 exerts its cell survival action primary through the PI3K/Akt pathway (29), we further demonstrated that IGF-1-inhibited MST1 cleavage was abrogated with PI3K inhibitor, LY294002. Consistent with MST1 cleavage status, IGF-1 rescues STS-and CD95/ FasL-induced apoptosis, which was inhibited by LY294002 (Fig. 1A). These data suggest that the PI3K/ Akt pathway mediates the IGF-1-inhibited MST1 cleavage.
It has been shown that overexpression of MST1 is able to induce its cleavage and cell death (6). To further determine the involvement of PI3K/Akt in regulation of MST1 cleavage and MST1-induced apoptosis, FLAG-MST1 was expressed in COS7 cells, and the ϳ36-kDa cleavage bands were readily detected (Fig. 1B). However, co-expression of constitutively active p110␣, a catalytic subunit of PI3K, or constitutively active Akt significantly reduces MST1 cleavage to the similar extent of Z-VAD treatment, a pan-caspase inhibitor. Further, expression of dominant negative-Akt or treatment with LY294002 or Akt inhibitor API-2 (30) increased the MST1 cleavage and MST1induced apoptosis (Fig. 1B). Therefore, we conclude that MST1 cleavage is regulated by IGF-1 through a PI3K/Akt-dependent pathway.
The autophosphorylation of MST1 at Thr 183 has been shown to be a hallmark of MST1 activation (9). To determine the inhibitory effect of Akt on MST1 activity, we treated Akt1-knock-out MEF and Akt1-reconstituted cells with STS and probed with anti-pMST1-Thr 183 antibody (Fig. 1C). The results show that autophosphorylation of MST1 is much higher in Akt1 Ϫ/Ϫ cells. To further confirm the inhibition of MST1 by Akt, we knocked down AKT2 in OVCAR8 cells, which overexpress endogenous AKT2 (31). Fig. 1D shows that phospho-MST1-Thr 183 was increased when AKT2 was depleted. Together, these results indicate that the PI3K/Akt pathway inhibits the MST1 cleavage and activation as well as MST1-induced cell death.
Akt Phosphorylates MST1 at Thr 120 and Interacts with MST1-The majority of Akt substrates contain Akt phosphorylation consensus motif RXRXX(S/T) (R ϭ arginine, X ϭ any amino acid, and S/T ϭ serine/threonine). Sequence analysis revealed that threonine 120 ( 115 RLRNKT 120 ) of human MTS1 matches the consensus motif, which is conserved among mouse, rat, Xenopus, Drosophila (Hippo), and yeast (Mess1, Fig. 2A). As constitutively active Akt inhibits, whereas dominant negative-Akt enhances MTS1 cleavage, we reasoned that MST1 could be a substrate of Akt. In vitro Akt kinase assay showed that wild-type MST1 but not MST1-T120A mutant, converting Thr 120 into alanine, was highly phosphorylated by Akt (Fig. 2B). Further, immunoblotting analysis with anti-Akt substrate antibody revealed that Akt1 phosphorylates wild-type MST1 but not MST1-T120A (Fig.  2C). We next generated an antibody designed to specifically recognize Akt phosphorylation site of MST1-Thr 120 . Upon immunoblotting of lysates from HEK293T cells transfected with WT-MST1 or MST1-T120A together with Akt1, the pMST1-Thr 120 antibody recognized the phosphorylated WT-MST1, but not the MST1-T120A (Fig.  2D). Elevated levels of endogenous phospho-MST1-Thr 120 were also detected in WT-Akt1-and Myr-Akt1-but not kinase-dead Akt1 (KD-Akt1)-transfected HEK293T cells (Fig. 2E). Further, the phospho-MST1-Thr 120 was induced by IGF-1 (Fig. 2F). Moreover, re-expression of Akt1 in Akt1-knock-out MEFs increased the MST1-Thr 120 phosphorylation (Fig. 2G). In addition, we also demonstrated that Akt interacts with the full length and N terminus of MST1 (Fig. 2, H and I). These data indicated that Akt interacts with and phosphorylates MST1 at threonine 120 in vitro and in vivo. Akt Inhibits MST1 Nuclear Translocation via a Phosphorylation-dependent Mechanism-It has been well documented that cleavage and nuclear translocation are critical steps for MST1-induced programmed cell death (32). Because Akt phosphorylates MST1, we reasoned that Akt phosphorylation of MST1 could regulate its nuclear translocation. To this end, we created GFP-tagged wild-type and mutant forms of MST1, including Akt phosphomimic (MST1-T120D) and non-phosphorylatable (MST1-T120A) full-length and C-terminal truncated MST1 (Fig. 3A). After transfection of the various forms of GFP-MST1 together with and without constitutively active Akt, subcellular localization of GFP-MST1s was examined under a fluorescence microscope. Constitutively active Akt reduced WT-MST1 nuclear translocation. Phosphomimic MST1-T120D located at the cytoplasm, whereas near 40% of cells expressing MST1-T120A exhibited nuclear localization. Expression of constitutively active Akt had no effect on MST1-T120A (left panels of Fig. 3B). Furthermore, C-terminal truncated mutants MST1-WT⌬C and MST1-T120A⌬C were highly restricted to the nucleus, whereas MST1-T120D⌬C was retained in the cytoplasm. Notably, expression of constitutively active Akt effectively blocked nuclear localization of MST1-WT⌬C but not MST1-T120A⌬C (Fig. 3C). In addition, STS treatment induced full-length wild-type MST1 and MST1-T120A but not MST1-T120D nuclear translocation (supplemental Fig. S2). Cell fractionation assays also revealed that endogenous nuclear (e.g. cleaved) MST1 induced by STS was largely reduced by expression of constitutively active Akt (Fig. 3D). Thus, we conclude that Akt inhibition of MST1 nuclear translocation is regulated by Akt through phosphorylation of Thr 120 .

Phosphorylation of Thr 120 Is Required for Akt Inhibition of MST1 Cleavage and Kinase Activity-
Having demonstrated Akt inhibition of MST1 nuclear translocation through phosphorylation of Thr 120 , we next examined if Akt inhibition of MST1 cleavage and kinase activity depends on phosphorylation of Thr 120 . Following transfection with FLAG-tagged wild-type and mutant MST1 together with and without constitutively active Akt, COS7 cells were treated with STS or vehicle. Immunoblotting analysis showed that STS treatment enhanced the cleavage of wild-type MST1 and non-phosphorylatable MST1-T120A, but had no significant effect on the cleavage of phospho- mimic MST1-T120D. Moreover, ectopic expression of constitutively active Akt reduced STS-stimulated wild-type MST1 but not MST1-T120A cleavage (Fig. 4A).
To determine the effect of Akt phosphorylation of MST1-Thr 120 on MST1 kinase activity, we carried out MST1 in vitro and in-gel kinase assays. Fig. 4B shows that Akt is capable of inhibiting kinase activity of WT-MST1 but not MST1-T120A. The phosphomimic MST1-T120D exhibited very low level of kinase activity with no cleavage band detected, which is similar to the cells co-transfected with Akt and MST1 (bottom panel of Fig. 4B). Further, cells expressing MST1-T120D exhibited considerably lower levels of auto-pMST1-Thr 183 as compared with those of the cells transfected with MST1-T120A and WT-MST1 (Fig. 4C). In addition, MST1-WT⌬C and MST1-T120A⌬C had higher levels of kinase activity as compared with MST1-T120D⌬C (supplemental Fig. S3). Taking into consideration the findings that knockdown of Akt increases and expression of Akt decreases autophospho-MST1-Thr 183 (Fig. 1, C and D), we conclude that Akt phosphorylation of Thr 120 represents a key step for the regulation of MTS1.
Akt Inhibits Signaling Downstream of MST1-We also assessed whether Akt reduces downstream signaling of MST1. Previous studies showed that JNK is a major target of MST1 and that MST1 activates JNK to induce apoptosis (32). To determine if Akt phosphorylation of MST1 affects JNK activation, COS7 cells were transfected with different forms of MST1. Following 48-h incubation, immunoblotting analysis revealed that JNK activation was induced by expression of wildtype MST1, MST1⌬C, non-phosphorylatable MST1-T120A and MST1⌬C-T120A, but not Akt phosphomimic MST1-T120D and MST1⌬C-T120D (Fig. 5A).
Recently, it has been shown that MST1 promotes cell death through phosphorylation of FOXO3a at residue Ser 207 and enhancing its nuclear translocation by disrupting MST1/14-3-3 complex (7). We therefore assessed the phosphorylation of FOXO3a-Ser 207 and its capability of binding to 14-3-3 upon the Akt phosphorylation of MST1. As shown in Fig. 5B, ectopic expression of MST1/FOXO3a considerably increased phospho-FOXO3a-Ser 207 and reduced FOXO3a/14-3-3 complex when compared with expression of FOXO3a alone (lane 2 versus lane 3). However, constitutively active Akt decreased MST1-induced FOXO3a-Ser 207 phosphorylation and, as a result, increased the FOXO3a association with 14-3-3 (lane 4). Further, phosphomimic MST1-T120D significantly reduced its ability to phosphorylate FOXO3a-Ser 207 leading to increase of FOXO3a/14-3-3 complex formation, FIGURE 4. Akt inhibits MST1 cleavage and kinase activity through phosphorylation of Thr 120 . A, MST1 cleavage inhibited by Akt depends on phospho-MST1-Thr 120 . COS7 cells were transfected/treated with the indicated plasmids and STS (300 M for 1 h) and then immunoblotted with indicated antibodies. The cleaved MST1 bands were quantified. B, Akt inhibits MST1 kinase activity through a pMST-Thr 120 -dependent manner. COS7 cells were transfected/treated with the indicated expression constructs and STS and lysed. A portion of cell lysates was immunoprecipitated with anti-FLAG (MST1) antibody, and the immunoprecipitates were subjected to in vitro kinase assay using histone H2B as substrate (top). The rest of the lysates were used for in-gel kinase assay using MBP as substrate (bottom). C, Akt phosphorylation of MST1-Thr 120 inhibited autophospho-MST1-Thr 183 . COS7 cells were transfected with indicated plasmids and then immunoblotted with anti-pMST1-183 (top) and -FLAG (bottom) antibodies.
which resembled the finding of expression of FOXO3a alone (lane 2 versus lane 5). Collectively, these results suggest that MST1 downstream targets are negatively regulated by Akt through phosphorylation of MST1.
In addition, we investigated the effect of Akt phosphorylation of MST1 on the programmed cell death. After transfection and treatment with/without STS, COS7 cells were analyzed by annexin V-fluorescein isothiocyanate/fluorescence-activated cell sorting. Fig. 5C shows that ectopic expression of Akt-phosphomimic MST1-T120D or MST1⌬C-T120D not only failed to induce apoptosis and caspase-3 cleavage but also considerably reduced the programmed cell death and caspase-3 activity induced by STS. In contrast, expression of non-phosphorylatable MST1-T120A or MST1⌬C-T120A induced apoptosis and enhanced STS-stimulated cell death as well as caspase-3 activation, indicating that Akt decreases MST1-induced apoptosis through phosphorylation of Thr 120 .

DISCUSSION
Aberrant activation of the PI3K/Akt pathway is one of the most common genetic alterations in human malignancy (33,34). In this study, we demonstrated that Akt phosphorylates MST1-Thr 120 in vitro and in vivo, and inhibits MST1 cleavage, autophosphorylation (e.g. pMST1-Thr 183 ), kinase activity, and nuclear translocation. As a result, MST1 fails to induce apoptosis and activate JNK1 and FOXO3a. Further, the activation of Akt largely correlates with elevated levels of pMST1-Thr 120 , and an inverse relationship between pMST1-Thr 120 and pMST1-Thr 183 was detected in human ovarian cancer. In addition, elevated levels of pMST1-Thr 120 are associated with poor prognosis in this malignancy. Our results indicate that Akt is a key upstream regulator of MST1 and that MST1 pro-apoptotic function is regulated by the PI3K/Akt pathway through direct phosphorylation of a highly conserved residue Thr 120 .
The Hippo pathway was initially identified in the fly to control organ size. Its core components are evolutionally conserved in mammals. Hippo, Sav, Wts, and Mats in the fly are homologous to mammalian MST1/2, WW45, LATS1/2, and Mob1, respectively (35). The pathway impinges on transcriptional co-activator Yorkie in fly and yes-associated protein in mammals to coordinate cell proliferation and apoptosis (13,14,36). Previous studies also showed that the Drosophila insulin receptor transduces signals that positively regulate cell and organ growth through its downstream molecule Chico/Dp110/ Dakt1 (37,38). Overexpression of the Dakt1 dramatically increases clonal size in wing imaginal disc through an enlargement of the cells (37). These studies suggest that the cross-talk between Akt and Hippo/ MST regulates cell growth and survival. Our study shows that MST1 possesses an Akt phosphorylation site ( 115 RLRNKTL 121 ) within its N-terminal kinase domain that is highly conserved among yeast, Drosophila, Xenopus, mouse, and human. Akt phosphorylates Thr 120 of MST1 in vitro and in vivo. Although we did not examine its canonical downstream target LATS1/ 2, we demonstrated that Akt phosphorylation of MST1-Thr 120 considerably reduces MST1-induced pFOXO3a-S207 and JNK activation (Fig. 5).
In addition, MST1 has been reported to inhibit Akt activity through direct interaction (39). Thus, this suggests that a feedback regulation loop exists between Akt and MST1, to keep a balance between Akt and MST1 in the cells. Furthermore, a recent study showed that Akt also phosphorylates Thr 387 at the C-terminal region of MST1 to abrogate MST1 function (40). Thr 387 , unlike Thr 120 , is not conserved in Hippo and found only in mammalian cells. Because Akt inhibits the nuclear translocation and the kinase activity of C-terminal-truncated form MST1 (MST1⌬C), which lacks Thr 387 , and because pMST1-Thr 120 was detected in primary tumor cells expressing hyperactive Akt, we concluded that Akt1 regulates MST1 primarily through phosphorylation of Thr 120 .
Previous studies have shown that MST1 cleavage is a critical step for the activation of MST1 (6,41). The two caspase cleavage sites of MST1, Asp 326 and Asp 349 , have been reported. The mutation of these cleavage sites attenuates MST1 kinase activity, nuclear translocation, and ability to induce apoptosis (3,41). Furthermore, it has been reported that MST1 promotes cell death through regulation of multiple targets, including phos-FIGURE 6. pMST1-Thr 120 inversely correlates with pMST1-Thr 183 and associates with adverse prognosis in human ovarian cancer. Immunoblotting (A) and immunohistochemical staining (B) of representative primary ovarian tumors show partial inverse correlation between pMST1-Thr 120 /pAkt1-Ser 473 and pMST1-Thr 183 . C, chi-square test analysis of pMST1-Thr 120 and pMST1-Thr 183 in 70 ovarian cancer specimens examined. The inverse correlation is significant (p Ͻ 0.01). D, overall survival in patients with high pMST1-Thr 120 /low pMST1-Thr 183 (n ϭ 20) versus low pMST1-Thr 120 /high pMST1-Thr 183 patients (n ϭ 16) was plotted by the Kaplan-Meier method. Statistical comparison of survival between groups was carried out with the log-rank statistic analysis.
phorylation of LATS1/2, histone H2B, FOXO3a, as well as induction of JNK and caspase-3 activation (3,32). Interestingly, caspase-3 also acts as the upstream activator of MST1 through cleavage of the C-terminal regulatory domain primarily at Asp 326 . As a result, an active catalytic N-terminal region (e.g. MST1⌬C) is generated and accumulated in the nucleus due to lack of nuclear exporting sequence motifs (Fig. 3A). However, full activation of MST1 requires both caspase-mediated cleavage and autophosphorylation (42). Taken collectively, we proposed the model that the phosphorylation of MST1-Thr 120 by Akt inhibits MST1 conformational change and autophosphorylation (e.g. p-Thr 183 ). Therefore, caspase will not be activated by MST1 to feedback cleave MST1 (Fig. 7). With regard to the cleavage form of MST1⌬C, which lacks nuclear export sequence, constitutively activated Akt blocks its nuclear translocation through phosphorylation of Thr 120 and inactivation of its kinase activity (supplemental Figs. S2 and S3).
Raf-1 has been shown to bind to and suppress MST2 by preventing MST2 dimerization as well as recruiting a phosphatase in an Raf-1 kinase-independent manner (43). Raf-1 also binds ASK1 (apoptosis signal-regulating kinase-1), another pro-apoptotic kinase, and inhibits its activity independently of its catalytic activity (44). We and others have previously shown that ASK1 activation is inhibited by Akt through direct phosphorylation of ASK1-Ser 83 (27). In addition, Thr 120 is conserved in MST1 and MST2, and Akt also phosphorylates MST2 (data not shown). Collectively, these studies suggest that Raf-1 functions as a scaffolding protein for MST and ASK1, whereas Akt regulates these proteins through a phosphorylation-dependent mechanism.
Several genetic studies have implicated the involvement of the Hippo/MST pathway in oncogenesis (10,45). Recent reports have shown down-regulation of MTS1 and MST2 in human soft tissue sarcoma and colorectal carcinoma (46,47). However, the total protein expression level of MST1 or MST2 does not correlate with disease progression. Although MST1 kinase activity can be up-regulated by nuclear translocation, the loss of cytoplasmic MST1 is not associated with late stage and shortened survival in colorectal cancer (46). The importance of loss of cytoplasmic MST1 needs to be further examined in a large number of human tumor samples (46). Our study indicates that the correlation between pMST1-Thr 120 and pMST1-Thr 183 could be a valuable prognostic marker in human ovarian cancer (Fig. 6D). Physiological importance of Akt phosphorylation of MST1-Thr 120 needs to be further investigated in a knock-in mouse model. Figure S1. CD95/FasL-induced MST1 cleavage and cell death are inhibited by the IGF-I signaling pathway. HeLa cells were treated with or without IGF-1 (50 μM) and/or LY294002 (20 μM) prior to addition of CD95/FasL. Following 3 h incubation, the cells were subjected to immunoblotting analysis with indicated antibodies (top) and annexin V/FACS assay (bottom).   Sup. Figure 3