Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2016 Sep 13.
Published in final edited form as: Cell Rep. 2012 Dec 27;3(1):103–115. doi: 10.1016/j.celrep.2012.11.020

S6K1 alternative splicing modulates its oncogenic activity and regulates mTORC1

Vered Ben-Hur 1, Polina Denichenko 1, Zahava Siegfried 1, Avi Maimon 1, Adrian Krainer 2, Ben Davidson 3, Rotem Karni 1
PMCID: PMC5021319  NIHMSID: NIHMS814986  PMID: 23273915

Abstract

Ribosomal S6 Kinase 1 (S6K1) is a major mTOR downstream signaling molecule which regulates cell size and translation efficiency. Here we report that short isoforms of S6K1 are over-produced in breast cancer cell lines and tumors. Overexpression of S6K1 short isoforms induces transformation of human breast epithelial cells. The long S6K1 variant (Iso-1) induced opposite effects: It inhibits Ras-induced transformation and tumor formation, while its knockdown or knockout induced transformation, suggesting that Iso-1 has a tumor suppressor activity. We further found that S6K1 short isoforms bind and activate mTORC1, elevating 4E-BP1 phosphorylation, cap-dependent translation and Mcl-1 protein levels. Both a phosphorylation-defective 4E-BP1 mutant and the mTORC1 inhibitor rapamycin partially blocked the oncogenic effects of S6K1 short isoforms, suggesting that these are mediated by mTORC1 and 4E-BP1. Thus, alternative splicing of S6K1 acts as a molecular switch in breast cancer cells elevating oncogenic isoforms that activate mTORC1.

Keywords: Alternative splicing, S6K1, 4E-BP1, Tumor suppressor, mTORC1, Breast Cancer

Introduction

The PI3K/Akt/mTOR pathway is one of the major signaling pathways hyper activated in many cancers, and leads to uncontrolled proliferation, increased survival, motility and invasiveness of cancer cells (Mamane et al., 2006; Manning and Cantley, 2007; Yuan and Cantley, 2008). mTOR resides in two distinct complexes: mTOR complex-1 (mTORC1) and complex-2 (mTORC2) (Zoncu et al., 2011). mTORC1 core contains mTOR, Raptor, G-β-L and is considered to be sensitive to rapamycin. mTORC2 contains Rictor, as the mTOR partner instead of Raptor, and depending on the cell type, is less sensitive to rapamycin (Sarbassov et al., 2004; Sarbassov et al., 2006). The best-characterized substrates of mTORC1 are S6 Kinase 1(S6K1) and eukaryotic initiation factor 4E (eIF4E)-binding protein 1 (4E-BP1), while Akt is a substrate of mTORC2 (Sarbassov et al., 2006; Shiota et al., 2006). Several components of the mTOR signaling cascade have been identified as oncogenes or tumor suppressors that activate or repress this pathway respectively (Zoncu et al., 2011) (Shaw and Cantley, 2006). Among the two well characterized mTORC1 substrates, S6K1 and 4E-BP1, the latter has been shown to be important for efficient protein translation, proliferation, and for oncogenic transformation (Hsieh et al., 2010; Ohanna et al., 2005; She et al., 2010). S6K1 has been implicated in the regulation of cell size (Ohanna et al., 2005) and its role in transformation is controversial; in some reports it seems dispensable for transformation (Hsieh et al., 2010) while in others it appears to be required (Alliouachene et al., 2008). A link between S6K1 function and cancer was suggested by the finding that RPS6KB1, the gene encoding for S6K1, resides in the chromosomal region 17q22-17q23, which is often amplified in breast and lung cancers (Bepler and Koehler, 1995; Monni et al., 2001). However, direct evidence that S6K1 expression or activity is sufficient to lead to cellular transformation is lacking. We have previously reported that the splicing factor oncoprotein SRSF1 is up regulated and in some cases amplified in several types of tumors, including breast and lung (Karni et al., 2007). SRSF1 modulates the splicing of RPS6KB1 which encodes S6K1 in both mouse and human (Karni et al., 2007). Furthermore, SRSF1 activates the mTORC1 pathway which phosphorylates S6K1 downstream of Akt (Karni et al., 2008). The spliced variants of S6K1 yield a long active kinase p85/p70 S6K1 (will be referred to as Iso-1) and shorter splicing variants (will be referred to as Iso-2 in mouse and h6A and h6C in human). We have shown previously that SRSF1 increases the expression of the shorter S6K1 isoform, and that this isoform possesses oncogenic activity and can transform immortal mouse fibroblasts (Karni et al., 2007).

In this study we examine the oncogenic and signaling activities of S6K1 splicing isoforms and their expression in cancer. Our findings suggest that while Iso-1 is tumor suppressive in vitro and in vivo and can block Ras-induced transformation, the short kinase inactive S6K1 splicing isoforms possess oncogenic properties. We show that the short isoforms of S6K1 bind mTOR and activate mTORC1, leading to increased 4E-BP1 phosphorylation, cap-dependent translation and upregulation of the antiapoptotic protein, Mcl-1. Furthermore, mTORC1 activation is critical for the oncogenic activity of S6K1 short isoforms as the mTORC1 inhibitor rapamycin or expression of a phosphorylation-defective mutant of 4E-BP1 (Hsieh et al., 2010; She et al., 2010) partially inhibit the oncogenic properties of these isoforms. Taken together our results suggest that S6K1 alternative splicing acts as a switch between a tumor suppressor protein and an oncoprotein, which is deregulated in breast cancer and modulates mTORC1 activity.

Results

S6K1 short isoforms are up-regulated in breast cancer cell lines and tumors

The gene RPS6KB1 encoding for p85/p70 S6K1 can be alternatively spliced to form a number of truncated isoforms. In mouse cells the splicing factor SRSF1 induces the inclusion of three additional exons (a-b-c) located between exon 6 and 7 (Fig. 1A). By PCR, cloning and sequencing we have discovered that in human there are two alternative exons in this region: a and c, which can be included together or individually generating two protein isoforms which we have termed h6A and h6C (Figs. 1A, S1D–E and Table S1). All of these isoforms in mouse or human which include combinations of exons 6 (a–c), are termed S6K1 short isoforms. Inclusion of the alternative exons mentioned above results in exposure of alternative poly adenylation sites and alterations in the reading frame that in turn generate a stop codon in exon 6c in mouse and exons 6a or 6c in humans. The presence of these stop codons creates transcripts containing approximately half of the original S6K1 coding sequence (Iso-1), and lacking more than half of the conserved kinase domain (Fig. 2A).

Figure 1. Increased expression of human S6K1 short variants 6A and 6C in breast cancer cell lines and tumors.

Figure 1

Open in a new tab

A. Schematic representation of RPS6KB1 pre-mRNA and its splicing isoforms. Isoform-1(encoding for p70/p85 S6K1 protein) is composed of 15 exons (blue boxes). S6K1 mouse spliced variant-2 (Iso-2) contains three alternative exons between exon 6 and 7; a, b, c (red boxes) a different poly adenylation site and 3′ UTR region (yellow area) and a stop codon in exon 6C. Human short S6K1 isoforms, h6C and h6A, lack the alternative 6b exon and contain a combination of two alternative exons: exon a followed by 3′ UTR or followed by exon c. In both cases, a stop codon in exon a terminates translation at the same amino acid and both transcripts give rise to one protein, h6A. The other isoform includes only exon 6C followed by a poly A tail and encodes a protein with a different C-terminus. All S6K1 short variants are identical up to the 6th exon, and differ from Iso-1, and from each other in their C-terminus. B. RNA from the indicated immortalized breast cell lines (HMLE, MCF-10A) or breast cancer cell lines (MCF-7, BT474, T47D, ZR-75-1, MDA-MB-231, MDA-MB-468, SUM149, SUM159) was extracted and the levels of S6K1 h6A and h6C short isoforms and Iso-1 were detected by RT-PCR with primers specific for each isoform. The splice variants are indicated by boxes at the right side of each transcription variant. C–D. qRT-PCR quantitation of S6K1 isoforms: Iso-1, h6A and h6C. Quantitation was done on total RNA extracted from breast cell lines (C) or tumors (D). All samples were normalized to β-actin mRNA levels and to the average expression of the immortal breast cell lines (HMLE and MCF-10A)(Error bars represents standard deviation (SD) of 3 repeats).

Fig. 2. S6K1 short isoforms enhance transformation of breast epithelial cells.

Fig. 2

Open in a new tab

A. The predicted protein structure of mouse and human RPS6KB1 splicing isoforms. All isoforms contain Raptor binding motif mTOR-signaling (TOS) at the N terminus (white boxes), and a lysine residue (K123) at the ATP binding site in the catalytic domain that is essential for its protein kinase activity. S6K1 isoform-1 kinase dead version (Iso-1 K123>A) contains a lysine to alanine (K123A) substitution. S6K1 short isoforms lack 6 out of the 12 conserved kinase helical domains as well as the C-terminal autoinhibitory domain which harbors the mTOR activatory phosphorylation site at threonine 389. B. MCF-10A cells were transduced with retroviruses encoding for the indicated S6K1 isoforms and protein extracts subjected to Western blotting. Membranes were probed with a monoclonal antibody against the N′ terminus of S6K1 to detect the endogenous and exogenous isoforms. β-actin was analyzed as a loading control. C. Pools of MCF-10A cells transduced with the indicated retroviruses as in (B) were seeded into soft agar in duplicates and colonies were allowed to grow for 14 days. Data represent the average number ± SD of colonies per well. n=2. * p=0.0049 ** p≤0.005 relative to empty vector. The results shown are a representative experiment out of three individual experiments. D. MCF-10A pools of cells transduced with the indicated retroviruses as in (B) were stimulated to migrate by physical wounding of cells seeded in monolayer. n=3. E. Representative images of the wound area of cells described in (D).

In all of these alternative splicing events, the presence of a poly adenylation sequence, and in the case of h6A also a premature stop codon (PTC) located less then 55bp from the next exon junction complex, prevents degradation of the generated transcripts by the Nonsence Mediated Decay (NMD) mechanism (Figs 1A. S1E) (Schoenberg and Maquat, 2012). We found that while in immortal breast cells (MCF-10A, HMLE) the expression of S6K1 short isoforms is relatively low, in breast cancer cell lines inclusion of exons 6a and 6c is significantly increased, especially in metastatic breast carcinoma cell lines (Figs. 1B, S1A). Indeed, while in both primary and immortal breast cells S6K1 short protein isoforms were hardly detected at the protein level, in breast cancer cell lines elevated protein levels of S6K1 short isoforms were detected (Fig. S1C), (Karni et al., 2007; Rosner and Hengstschlager, 2011). In human breast tumor samples we found elevated expression of S6K1 h6A and h6C isoforms compared to the immortal breast cell lines (Figs. 1D, S1F–G). Interestingly, whereas most analyzed breast cancer cell lines and tumor samples presented high expression of S6K1 short isoforms, we did not find elevated expression of the full length isoform, Iso-1 in most tumors (Fig. 1B–D). Two of the cell lines that showed elevated Iso-1 expression (MCF-7, BT474) possess amplification of the RPS6KB1 gene and except for MCF-7, all tumors and cell lines showed an increase in short isoforms/Iso-1 ratio (Fig. S1F–G) indicating that an alternative splicing switch in S6K1 occurs in breast cancer.

All S6K1 protein isoforms are identical in their N-terminus but share only partial homology in their kinase domain and differ from each other in their C-terminus. Iso-1, Iso-2, h6A and h6C contain distinct sequences in their C-terminus consisting of 330,121, 12 or 24 amino acids respectively (Fig. 2A).

S6K1 short isoforms enhance motility and anchorage independent growth

We sought to examine the oncogenic activity of S6K1 isoforms in human breast epithelial cells. We found that Iso-2, h6A and h6C were able to transform human immortal breast MCF-10A cells and mouse NIH3T3 cells enabling them to form colonies in soft agar despite the relatively low expression of the short isoforms in comparison to Iso-1 (Figs. 2B–C and S2A–C). The kinase dead version of Iso-1 (Iso-1 K123>A), enhanced transformation and increased the ability of MCF-10A and NIH 3T3 cells to form colonies in soft agar as well (Figs. 2C and S2C). An in vitro kinase assay using S6 as a substrate shows that the short isoform Iso-2 and the kinase dead K123>A version of Iso-1 have no kinase activity (Fig. S2G). The oncogenic effects of S6K1 short isoforms did not require a functional kinase activity since all short S6K1 isoforms share truncated kinase domains and cannot phosphorylate ribosomal protein S6, or mTOR, both known S6K1 substrates (Figs. S2E–G and 5A) (Chiang and Abraham, 2005; Holz and Blenis, 2005). In contrast to the expression of S6K1 short isoforms, overexpression of Iso-1, the long active kinase, did not enhance colony formation in soft agar and even reduced the basal level of colony formation (Fig. 2C).

Fig. 5. S6K1 kinase short isoforms interact with mTOR, enhance cap-dependent translation and increase Mcl-1 expression.

Fig. 5

Open in a new tab

A–B. HEK293 cells were co-transfected with myc- tagged mTOR and the indicated T7-tagged S6K1 isoforms. Whole cell lysates were examined for construct expression (A) and for immunoprecipitation of T7-tagged S6K1 isoforms (B). Myc-tagged bound mTOR and T7-tagged S6K1 isoforms were detected by immunoblotting using anti-myc or monoclonal antibody against the N′ terminus of S6K1, recpectively. First two left lanes represent anti T7 anibody alone, and pull down from untransfected HEK293 cells, respectively. C. Schematic representation of pLPL Cap-Renilla-IRES-Luciferase bicistronic dual reporter vector (Gerlitz et al., 2002). D. MCF-10A cells were co-transfected with dual reporter vector (C) and with the indicated S6K1 isoforms and starved for serum and growth factors for 24h post transfection. Cap-dependent translation (Renilla luciferase activity) and IRES-mediated translation (Firefly luciferase activity) were measured (n=4 experiments,*p 0.019, **p=0.008, ***p=0.033). E. MCF-10A cells were transduced with retroviruses encoding for the indicated S6K1 isoforms. 3×105 transductants were seeded in 6 well plates and starved for 24h for serum and growth factors. After Western blotting the membranes were probed with the indicated antibodies. β-catenin was analyzed as a loading control. F. MCF-10A cells described in (Fig. 3F) were seeded (3×105 cells/well) in 6-well plate. 24h later cells were lysed, total protein was extracted and separated by SDS-PAGE. The membranes were probed with the indicated antibodies. β-actin was analyzed as a loading control.

One characteristic of cellular transformation is enhancement of cell motility (Hanahan and Weinberg, 2011). MCF-10A cells expressing the shorter isoforms or the kinase dead version of Iso-1 (Iso-1 K123>A) showed accelerated migration rate compared to cells transduced with S6K1 Iso-1 or the empty vector in a tissue culture wound healing assay (Fig. 2D–E). In order to exclude the possibility that these effects are the result of changes in proliferation we measured the proliferation rate of the cells. Cells overexpressing the shorter variants, empty vector or Iso-1 K123>A did not show enhanced proliferation rate in comparison to Iso-1 (Fig. S2D). Altogether, these results suggest that S6K1 short kinase inactive isoforms promote transformation of MCF-10A cells.

Expression of S6K1 short isoforms enables growth factor-independent three dimensional acini formation and elevates 4E-BP1 phosphorylation

The immortal breast epithelial cells MCF-10A possess the ability to grow into spheroid structures (acini) when grown on matrigel. To study the effects of S6K1 isoforms on growth of three dimensional (3D) reconstituted basement membrane cultures, we seeded MCF-10A cells transduced with S6K1 isoforms in the presence or absence of growth factors necessary for acini formation in MCF-10A cells (Arias-Romero et al., 2010). Cells overexpressing S6K1 truncated isoforms, as well as the kinase-dead version of Iso-1, formed large, hyper proliferative three-dimensional structures (Figs. 3A–C) with slight morphological disruption (Fig. S3A), even when grown in the presence of only one of the growth factors. This suggests that S6K1 isoforms can replace the strong proliferative signal provided by either EGF or Insulin/IGF1. This observation is further underscored by the fact that the kinase active isoform of S6K1 (Iso-1) failed to support growth factor independent acinus formation and may even inhibit basal proliferation in matrigel (Figs. 3A–B and S3A).

Fig. 3. S6K1 inactive isoforms enable growth factor-independent three dimensional growth of MCF-10A cells and enhance 4E-BP1 phosphorylation.

Fig. 3

Open in a new tab

A–C. Phase images of MCF-10A cells transduced with the indicated retroviruses seeded in matrigel as described in Materials and methods, in the absence of EGF (A), Insulin (B) or in the presence of both (C). Cells were allowed to grow for two weeks to form acini structures. These results were obtained in at least 3 individual experiments. (D–G) Total protein was isolated from cells and subjected to Western blotting using the indicated antibodies. D–E. MCF-10A transduced cells described above were seeded in 6 well plates (3×105cells/well). Cells were starved for 24h and then induced with EGF (D) or IGF-1 (E) for 4h. Fold increase of 4E-BP1 was normalized (Phosphorylated/total protein levels) to that of untreated (starved) pB(−) (empty vector) which was arbitrarily set at 1 (Quantitation values are shown under each panel). F. MCF-10A cells were transduced with the indicated retroviruses expressing empty vector (mlp(−)) or the indicated S6K1 Iso-1-specific shRNAs (Karni et al., 2007). G. MCF-10A cells transduced with the indicated retroviruses as described in (F) were grown for 24h in the presence of 0.5% or 5% serum. Arrows show the phophorylation states of 4E-BP1, where gamma is the fully phosphorylated form. β-actin was analyzed as a loading control.

Insulin/IGF-1 and EGF activate their corresponding receptors leading to activation of several mitogenic signaling cascades, among them the Ras-Raf-MAPK-ERK and the PI3K-Akt-mTOR pathways (Shaw and Cantley, 2006) (Yuan and Cantley, 2008) (Mendoza et al., 2011). We hypothesized that S6K1 short isoforms affect acinus proliferation by activating the Insulin/IGF-1 and EGF signaling downstream of the receptors, bypassing the need for growth factor activation (Fig. 3A–B). Thus, we sought to investigate the activities of these signaling pathways. We measured the phosphorylation state of known downstream effectors of these pathways; Akt, ERK and 4E-BP1, (Shaw and Cantley, 2006) (Yuan and Cantley, 2008) (Mendoza et al., 2011) in MCF-10A cells overexpressing S6K1 isoforms with and without growth factor activation. We found that overexpression of S6K1 kinase inactive isoforms stimulated the phosphorylation of 4E-BP1, even under serum starvation conditions (Figs. 3D–E, S3D). Under these conditions, we observed only a slight and inconsistant activation of Akt or ERK (S3B–D). Notably, while overexpression of S6K1 Iso-1 failed to increase 4E-BP1 phosphorylation under serum starvation conditions compared to cells transduced with empty vector (Figs. 3D–E and S3D), knockdown of this isoform in MCF-10A cells increased 4E-BP1 phosphorylation (Fig. 3F–G), supporting the notion that Iso-1 plays an opposite role than S6K1 kinase inactive isoforms. Thus, the increased level of 4E-BP1 phosphorylation we observed, even in serum- and growth factor-deprived cells, may contribute to the transforming phenotype of cells harboring S6K1 short isoforms (Figs. 2C–E, 3A–B and S2C). Moreover, 4E-BP1 phosphorylation is higher in most breast cancer cells compared to immortal non transformed cells (Avdulov et al., 2004) in correlation with elevated levels of S6K1 short isoforms (Figs. 1B–C, S1A and S3E).

S6K1 Iso-1 inhibits Ras-induced transformation in vitro and in vivo

Our results, unexpectedly, indicated that the full length S6K1 Iso-1 did not support growth factor-independent acinus formation, as opposed to the short isoforms (Fig. 3A–C). To further study the possibility that S6K1 Iso-1 acts as a tumor suppressor, we silenced its expression in the immortal MCF-10A cell line using Iso-1 specific shRNAs (Karni et al., 2007) (Fig. 3F). We measured reduced phosphorylation of mTOR S2448, a known substrate of p70 S6K1 (Holz and Blenis, 2005) (Chiang and Abraham, 2005), in these cells (Fig. 3F). Knockdown of S6K1 Iso-1 in MCF-10A cells increased both colony formation in soft agar and acinus number and size in matrigel (Figs. 4A, S4C). Similar results were obtained using NCI-H460 lung carcinoma cells (Fig. S4A–B). Even though overexpression of Iso-1 did not significantly decrease cell motility (Fig. 2D–E) its knockdown in MCF-10A cells was sufficient to increase motility (Fig. S4D).

Fig. 4. S6K1 Iso-1 knockdown increases transformation and its over expression blocks RAS-induced transformation in vitro and in vivo.

Fig. 4

Open in a new tab

A. MCF-10A pools of cells were transduced as in (Fig. 3F), seeded into soft agar in duplicates and colonies were allowed to grow for 14 days. Data represents the average number ± SD of colonies per well. n=2. B. MCF-10A cells were transduced with retroviruses encoding pB(−) empty vector or S6K1 Iso-1 followed by transduction with an active Ras mutant (H-Rasv12). Cells were seeded into soft agar in duplicates and colonies were allowed to grow for 14 days. C. MCF-10A pools of cells transduced with the indicated retroviruses as in (B) were photographed 24h after seeding (X100 magnification). D. RAS-transformed MCF-10A cells expressing empty vector (pB(−)) or S6K1 Iso-1 were injected into NOD-SCID mice (2×106 cells/injection). Tumor volume was measured weekly and tumor growth curve was calculated as described in Materials and methods; error bars indicate SD of 8 tumors. n/n= number of tumors per number of injections. E. Ras transformed MCF-10A cells expressing the indicated retroviruses as described in (B). Total protein from stable pools was extracted and subjected to Western blotting. The membranes were probed with the indicated antibodies. β-actin was analyzed as a loading control.

To examine if Iso-1 can suppress the oncogenic potential of transformed cells, we co-expressed S6K1 Iso-1 with an active RAS mutant (H-RasV12) in MCF-10A cells (Fig. S4E). Cells co-expressing the kinase active S6K1 isoform (Iso-1) and oncogenic RAS formed fewer and smaller colonies in soft agar compared to cells co-expressing Ras and empty vector, suggesting that S6K Iso-1 possesses tumor suppressive activity (Fig. 4B). Moreover, cells co-expressing Iso-1/Ras did not gain the spindle/fibroblastic shape that is characteristic of cells transformed with an oncogene such as Ras (Janda et al., 2002) (Fig. 4C). The same cell pools were injected subcutaneously into NOD-SCID mice. Supporting the results in vitro, we found that RAS-transformed cells co-expressing Iso-1 did not form tumors in vivo, (0 tumor formed of 8 injections) as opposed to the empty vector (8/8 tumors formed) (Figs. 4D and S4F). 4E-BP1 phosphorylation levels were lower in cells co-expressing Iso-1 and H-RasV12 than in cells expressing H-RasV12 alone (compare the γ bands in Fig. 4E). Moreover, cells co-expressing Iso-1 and RasV12 showed about 20% decrease in Cap-dependent translation in correlation with the decreased phosphorylation of 4E-BP1 (Fig. S4G). Taken together, this data suggests that S6K1 Iso-1 is a putative tumor suppressor.

S6K1 short isoforms bind mTORC1 and enhance cap-dependent translation and Mcl-1 expression

We hypothesized that S6K1 short isoforms might directly bind and activate mTORC1. To test this possibility we co-transfected myc-tagged mTOR and T7 tagged Iso-1 or Iso-2 (as a representative of the short isoforms). Immunoprecipitation of transfected S6K1 with anti-T7 antibody revealed that only S6K1 Iso-2 interacts with mTOR (Fig. 5A–B). The fact that we did not see an interaction between S6K1 Iso-1 and mTOR might be explained by T389 phosphorylation of S6K1 (Fig. 5B). This phosphorylation activates S6K1 Iso-1 (Kim et al., 2002) leading to its release from mTORC1 (Holz et al., 2005).

Cells overexpressing S6K1 short inactive isoforms showed elevated levels of cap-dependent translation under serum starvation conditions as measured by a reporter gene measuring Cap- versus IRES-mediated translation from the same transcript (Fig. 5C–D). Indeed, Iso-1 overexpressing MCF-10A cells exhibited only a slight decrease in 4E-BP1 phosphorylation that was consistent with a slight decrease of cap-dependent translation, compared to cells expressing empty vector (Figs. 3D–E and 5D). Moreover, on the background of H-RasV12 transformation, Iso-1 decreased both 4E-BP1 phosphorylation and cap-dependent translation (Figs. 4E and S4G), suggesting that Iso-1 tumor suppressive effect can be clearly detected on the background of a strong oncogene such as mutant Ras but not in a non-transformed cell system such as MCF-10A cells.

We next examined if S6K1 short isoforms alter the expression of proteins known to be controlled by mTORC1-4E-BP1 translational regulation and can contribute to cell transformation. One such protein is Mcl-1 (Hsieh et al., 2010), a key antiapoptotic protein that was shown to be translationaly controlled by mTORC1 (Mills et al., 2008). We found that Mcl-1 protein levels are elevated in MCF-10A cells overexpressing S6K1 kinase inactive isoforms, but not Iso-1 or empty vector (Fig. 5E). Interestingly, Mcl-1 was also elevated by Iso-1 knockdown (Fig. 5F). S6K family consists of two kinases, S6K1 and S6K2 (RPS6KB1 and RPS6KB2 respectivly) (Lee-Fruman et al., 1999), that are known to have redundant activities (Nardella et al., 2011; Pende et al., 2004). We did not detect any changes in S6K2 protein levels in either Iso-1 knockdown or S6K1 kinase inactive isoforms overexpression (Fig. S5A–B), indicating that the biological effects we observe are due to up or downregulation of S6K1 Iso-1 or the short S6K1 isoforms.

Loss of S6K1/2 enhances cap-dependent translation, Mcl-1 expression and transformation

In order to establish definitive proof that S6K1 Iso-1 is a tumor suppressor and to rule out any possible compensation or interference by S6K2, we analysed S6K1 and S6K2 double-knockout (DKO) mouse embryonic fibroblasts (MEFs) (Fig. 6A) (Dowling et al., 2010; Pende et al., 2004). Consistent with our previous results, serum starved S6K DKO MEFs showed elevated phosphorylation levels of 4E-BP1 (Fig. 6A). In addition, S6K DKO MEFs formed significantly higher number of colonies in soft agar and also presented 5-fold increase in cap-dependent translation, indicating that these cells are transformed (Fig. 6B–C). Furthermore, Western blot analysis of S6K DKO MEFs revealed high levels of expression of Mcl-1 (Fig. 6D) in agreement with what was observed in Iso-1 knockdown in MCF-10A cells (Fig. 5F). The fact that the DKO cells show minor/negligible mTOR phosphorylation (Fig. 6A) and no S6 phosphorylation (Fig. S2F), but were transformed, supports the notion that mTOR S2448 phosphorylation is not essential for its ability to phosphorylate 4E-BP1.

Fig. 6. Loss of S6K1/2 enhances cap-dependent translation, Mcl-1 expression and transformation.

Fig. 6

Open in a new tab

A. Wild type (WT) or S6K1 and S6K2 double-knockout (DKO) mouse embryonic fibroblasts (MEFs) were seeded at 80% confluency and serum starved for 5 hours. Total protein was extracted and subjected to Western blotting. The membranes were probed with the indicated antibodies. B. WT and DKO MEFs were seeded into soft agar in triplicates and colonies were allowed to grow for 14 days. Data represents the average number ± SD of colonies per well. n=3. The results shown are a representative experiment out of three individual experiments. C. WT and DKO MEFs were transfected with dual reporter vector (Cap-Renilla-IRES-Luciferase described in Fig 5C). 24h post transfection cells were serum starved for another 24h. Cap-dependent translation (Renilla luciferase activity) and IRES-mediated translation (Firefly luciferase activity) were measured (Data represents the average number ± SD of n=6 experiments). D. Total protein from WT and DKO MEFs described in (A) was separated by SDS-PAGE. After Western blotting the membranes were probed with the indicated antibodies. E. S6K1 and S6K2 DKO MEFs cells transduced with retroviruses encoding for empty vector (pB(−)) or S6K1 Isoform-1(Iso-1) were transfected with dual reporter vector (Cap-Renilla-IRES-Luciferase) 24h post transfection cells were serum starved for another 24h. Cap-dependent translation (Renilla luciferase activity) and IRES-mediated translation (Firefly luciferase activity) were measured (n=3 experiments,*p<0.004). F. MEF cells described in (E) were seeded (200 cells/well) in 6 well plates and grown for 14 days. Colonies were fixed and stained with methylene blue. G. MEF cells described in (E) were seeded into soft agar in triplicates and colonies were allowed to grow for 14 days. Data represent the average number ± SD of colonies per well. n=3. The results shown are a representative experiment out of three individual experiments.

Expression of Iso-1 in S6K DKO MEFs (Fig. S2E) partially reduced cap-dependent translation, colony survival and growth of colonies in soft agar showing that Iso-1 harbors tumor suppressive ability (Fig. 6E–G). The transformed phenotype of S6K DKO MEFs was not completely reversed by the ectopic expression of S6K1 Iso-1, probably due to the absence of S6K2 in this cell system that might contribute to the tumor suppressive phenotype of S6K1.

The oncogenic activities of S6K1 short isoforms are mediated by mTORC1 inactivation of 4E-BP1

S6K1 short isoforms activated mTORC1 as measured by 4E-BP1 phosphorylation (Figs. 3D–E, S3D). We next examined, using two complementary strategies, if this activation is required for their oncogenic activities: 1) Inhibition of mTORC1 by the pharmacological inhibitor rapamycin and 2) expression of a phosphorylation-defective 4E-BP1 which is mutated in its phosphorylation sites and cannot dissociate from eIF4E upon mTORC1 activation (Avdulov et al., 2004; She et al., 2010). We transduced MCF-10A cells expressing the mouse S6K1 short isoform (Iso-2) with phosphorylation-defective 4E-BP1 (4E-BP15A or 4E-BP14A), where each of the five (or four) insulin and rapamycin-responsive 4E-BP1 phosphorylation sites have been mutated to alanine (Hsieh et al., 2010).

Co-expression of both Iso-2 and mutant 4E-BP1 decreased colony formation in soft agar two fold relative to cells expressing Iso-2 alone (Fig. 7A–B and S6A–B). Similarly, rapamycin reduced the number of colonies formed in soft agar in cells overexpressing Iso-2 to the background level of cell expressing empty vector (Fig. 7B). These results suggest that mTORC1 activation and 4E-BP1 phosphorylation plays a major role in the oncogenic capabilities of S6K1 short isoforms. Expression of mutant 4E-BP1 together with Iso-2 also partially decreased cell motility compared to cells expressing Iso-2 alone (Fig. 7C–D).

Fig. 7. 4E-BP1 inactivation and mTORC1 activity is required for the oncogenic activities of S6K1 short isoforms.

Fig. 7

Open in a new tab

A. MCF-10A cells were co-transduced with retroviruses encoding for empty vector pWZl-Hygro (pW) or pW-4E-BP1 phosphorylation defective mutant in which all five phosphorylation sites were mutated to alanine (4E-BP15A) (Hsieh et al., 2010) and empty vector pBABE (pB(−)) or pB- Iso-2. Total protein from stable pools was extracted and separated by SDS-PAGE. Membranes were probed with monoclonal antibodies against 4E-BP1 or S6K1 to detect the endogenous and exogenous isoforms. β-actin was analyzed as a loading control. B. MCF-10A pools of cells transduced with the indicated retroviruses as in (A) were seeded into soft agar in duplicates, with or without 100 nM Rapamycin and colonies were allowed to grow for 14 days. Colonies from ten fields of each well were counted and representative fields of colonies were photographed in phase image (X100 magnification). Data represents the average number ± SD of colonies per well. n=2. C. MCF-10A pools of cells transduced with the indicated retroviruses as in (A) were stimulated to migrate by physical wounding of cells seeded in monolayer. Data represents the average number of quantified wound area and SD from three individual experiments. D. Representative images of the wound area of cells described in C. E. Phase images of MCF-10A cells transduced with the indicated retroviruses seeded in matrigel. Cells were allowed to grow for two weeks to form acini structures. F. A proposed model of mTORC1 regulation by S6K1 isoforms. In non-transformed cells upon mitogen stimulation, S6K1 Iso-1 is activated by mTORC1 and generates a feedback signal loop resulting in phosphorylation of mTOR at S2448 in the repressor domain. This might attenuate mTOR’s ability to phosphorylate and repress 4E-BP1 leading to decreased cap-dependent translation. Other cellular substrates might contribute to the tumor suppressive activity of Iso-1 independently of cap-dependent translation (left panel). In transformed cells S6K1 short isoforms are up-regulated, bind mTORC1 and increase its activity. mTORC1 activation leads to enhanced 4E-BP1 phosphorylation, cap-dependent translation elevation of the anti-apoptotic protein Mcl-1 and other proliferating or anti apoptotic proteins and increased cell survival and transformation (right panel).

Next, we investigated whether 4E-BP1 phosphorylation is important for Iso-2 mediated growth factor independent 3D proliferation in matrigel. Expression of 4E-BP15A strongly inhibited Iso-2 ability to induce growth factor independent acinus formation in matrigel (Fig. 7E). Taken together, our results suggest that the mechanism of action of S6K1 short isoforms is mediated mostly by 4E-BP1 phosphorylation through mTORC1 activation.

Discussion

We have previously shown that mouse S6K1 is alternatively spliced to form a short isoform (Iso-2) which is essential for the oncogenic activity of the splicing factor oncoprotein SRSF1(Karni et al., 2007). Here we report that in human cells S6K1 has two alternatively spliced short isoforms that are over-produced in breast cancer cell lines and tumors. Furthermore, all of S6K1 short splicing variants lack an autoinhibitory C-terminus domain (Ali and Sabatini, 2005), half of the kinase domain and do not exibit kinase activity, at least on the known S6K1 substrate, rpS6 (Figs. 2A, S2E–G). Overexpression of mouse or human S6K1 short isoforms enhanced transformation, anchorage-independent growth, cell motility and growth factor-independent three-dimensional acinus formation of human breast epithelial cells (Figs. 2, 3A–C, S2C and S3A). Surprisingly, the long, kinase active S6K1 isoform (Iso-1), inhibited 3D acinus formation, reduced 4E-BP1 phosphorylation, cap-dependent translation and transformation in vitro and in vivo, demonstrating many properties that characterize a tumor suppressor. Our results suggest that only S6K1 short isoforms, but not Iso-1, interact with and activate mTORC1 leading to elevated 4E-BP1 phosphorylation, enhanced cap-dependent translation and upregulation of the anti-apoptotic protein Mcl-1. Inhibition of mTORC1 or 4E-BP1 phosphorylation can partially reverse the oncogenic activity of S6K1 short isoforms suggesting that their oncogenic properties are at least in part mediated by this pathway.

A switch in RPS6KB1 alternative splicing up regulates oncogenic isoforms in breast cancer

We found that many breast cancer cell lines and tumors switch the splicing of RPS6KB1 to elevate the human short isoforms of S6K1 h6A and h6C (Figs. 1B–D, S1F–G). There are several documented examples of an alternative splicing switch which induces a pro-oncogenic isoform at the expense of a tumor suppressive isoform. In many cancers tumor suppressive isoforms are down regulated, while pro-oncogenic ones are up regulated (David and Manley, 2010). We found that while all the short S6K1 isoforms, as well as the kinase-dead form of Iso-1, induced anchorage independent growth, enhanced motility and growth factor-independent three dimensional acinus formation in matrigel, the active S6K1isoform (Iso-1) did not (Figs 2–3 and S2C). Moreover, in most of these assays Iso-1 showed an opposite effect, indicating that these splicing isoforms possess antagonistic activities. Several lines of evidence suggest that Iso-1 acts as a functional tumor suppressor: a) when co-expressed with an active RasV12 mutant, Iso-1 inhibited Ras-induced transformation in vitro and in vivo (Fig. 4) suggesting it possesses an anti tumorigenic activity. b) Iso-1 knockdown in MCF-10A cells induced transformation, colony formation in soft agar, acinus formation, increased motility and elevated Mcl-1 levels (Figs. 4A, 5F and S4C–D). c) Loss of S6K1 and S6K2 had dramatic effect on cellular transformation, as immortalized MEFs from S6K1/2 DKO mice were transformed, formed large numbers of colonies in soft agar and showed increased levels of the anti-apoptotic protein Mcl-1 (Fig. 6), which is translationally regulated by mTORC1(Hsieh et al., 2010). d) Re-introduction of S6K1 Iso-1 partially inhibited the transformed phenotypes of S6K1/2 DKO cells indicating that Iso-1 possesses tumor suppressive activities, but might require also the presence of S6K2 in these cells to restore its full capability as an anti tumorigenic protein (Fig. 6). It has been shown previously that S6K1 is required for insulinoma development in a mouse model where an active Akt mutant (Myr-Akt) was overexpressed in pancreatic beta cells under the insulin promoter (Alliouachene et al., 2008). Similarly, S6K1 deletion on the background of PTEN+/− and low epression of S6K2 in pheochromocytoma cells impaired adrenal tumorigenesis, (Nardella et al., 2011). In both of these reports the cancer was induced by activation of the same pathway (PI3K-Akt) and in a specific tissue. It is conceivable that under other experimental settings the outcome might be different, as already demonstrated for S6K1 deletion in other tissues (Nardella et al., 2011). The lack of an effect on tumor development in the latter cases is possibly due to redundancy between S6K1 and S6K2. Moreover, to the best of our knowledge, no study has directly shown that overexpressed or mutated S6K1 or S6K2 acts as an oncogene. Hence, the present report provides the first direct evidence for the pro- and anti-tumorigenic effects of stable knockdown or overexpression of S6K1 long isoform (Iso-1) respectively. It shoud be noted that the pro-oncogenic effect of Iso-1 knockdown does not involve S6K2, as the latter remained unchanged (Fig. S5). Notably, the catalytic activity of S6K1 is essential for its tumor suppressive activity, as the kinase-dead point mutant completely abrogated this activity. In addition, the fact that Iso-1 inhibited Ras transformation, but only partially inhibited cap-dependent translation raises the possibility that Iso-1 phosphorylates additional substrates other than those in the mTOR pathway, that contribute to its tumor suppressor activity independently of cap-dependent translation. Importantly, S6K1short isoforms are catalytically inactive and did not induce phosphorylation of the known S6K1 substrates, the ribosomal protein S6 or mTOR itself in an in vitro kinase assay and upon transfection into cells (Figs. S2E–G, 5E, 5A and S3C). Altogether, these results suggest that Isoform-1 of S6K1 has tumor suppressive properties, while the short isoforms are pro-oncogenic. Our results suggest that the gain of h6A and h6C observed in breast cancer cells and tumors is a mechanism to switch-off a tumor suppressive isoform and to turn-on an oncogenic one.

RPS6KB1 splicing isoforms modulate the activity of mTORC1

Surprisingly, we found that most S6K1 short isoforms activated mainly 4E-BP1 phosphorylation without a significant activation of Akt or ERK and in a growth factor-independent manner (Figs. 3 and S3). In accordance with increased 4E-BP1 phosphorylation, cells expressing S6K1 short isoforms showed elevated cap-dependent translation and upregulation of Mcl-1, while Iso-1 expressing cells showed low or basal cap-dependent translation, similar to the control cells expressing empty vector or reduced cap-dependent translation in Ras-transformed cells (Figs. S4G, 5). Moreover, loss of S6K1 and S6K2 enhanced cap-dependent translation in MEF cells which was partially inhibited by S6K1 Iso-1 re-introduction (Fig. 6) indicating that S6K1 and S6K2 activities suppress cap-dependent translation and transformation (Fig. 6). Although surprising, previous studies support this result: Cells from a knockin mouse of a mutant S6 gene (rpS6P−/−) where all five phosphorylation sites were replaced to alanines, showed a two fold increase in global translation as measured by methionine incorporation and a two fold increase in proliferation rate as compared to MEFs from WT littermates (Ruvinsky et al., 2009). Combining these two findings, both inactivation of the “kinase” (S6K1/2 DKO) or inhibition of phosphorylation of the substrate (rpS6P−/−), gave similar results of increased translation. This suggests that S6 phosphorylation might inhibit rather than increase translation. Other studies did not find reduced translation upon S6K1/2 loss eventhough they did not observe increased translation (Mieulet et al., 2007). These results suggest that the growth factor independent growth observed in matrigel (Fig. 3) could be explained by the ability of S6K1 short isoforms to activate mTORC1 that in turn inactivates 4E-BP1 even under low nutrient conditions. Enhanced cap-dependent translation contributes to cancer development by elevating the translation of several oncogenes and anti-apoptotic genes such as Mcl-1, Bcl-X, MDM2, HIF-1α, β-catenin, c-myc, cyclin D1 and others, which are translationally regulated by cap-dependent translation and the mTORC1 pathway (Hsieh et al., 2010) (Karni et al., 2002) (Karni et al., 2005) (Mamane et al., 2004). We found that, at least in the case of the anti-apoptotic protein Mcl-1, overexpression of S6K1 short isoforms, Iso-1 knockdown or S6K1/2 knockout enhanced its expression (Figs. 5E–F, 6D). To examine if S6K1 short isoforms can interact with mTOR we co-expressed mTOR with S6K1 Iso-1 or Iso-2 and found that only the short isoform could pull down mTOR, whereas Iso-1 did not, even though its expression was much higher (Fig. 5A–B).

The oncogenic activities of S6K1 short isoforms are partly mediated through 4E-BP1 and mTORC1 activation

We examined the contribution of mTORC1-4E-BP1 axis to S6K1 short isoform mediated oncogenesis and found that inhibition of eIF4E/4E-BP1 dissociation greatly inhibited motility, anchorage-independent growth and growth factor-independent acinus formation in matrigel (Fig. 7 and S6). These results suggest that most of the oncogenic effects of S6K1 short isoforms are mediated by mTORC1 activation. However, we can not rule out the possibility that S6K1 short isoforms affect other oncogenic signaling pathways as suggested by the fact that expression of a dominant-negative 4E-BP1 did not fully suppress colony formation in soft agar and motility induced by S6K1 short isoforms. These results also suggest that S6K1 is not only a substrate of mTOR, but that it may also modulate the activity of mTOR increasing 4E-BP1 phosphorylation when S6K1 short isoforms are elevated (Fig. 3 and S3D). In addition, S6K1 Iso-1 can phosphorylate mTOR at Serine 2448, (Figs 3F, 5A and 6A), although the consequences of this phosphorylation on mTOR activity are not fully understood (Chiang and Abraham, 2005; Holz and Blenis, 2005). S6K1 short isoforms cannot be phosphorylated by mTOR at Threonine 389 and by phosphoinositide-dependent kinase 1 (PDK1) at Threonine 229, as they lack both phosphorylation sites. However, both Iso-1 and the short isoforms should be able to bind mTORC1, since all of them contain the Raptor binding motif mTOR-signaling (TOS) motif (Schalm and Blenis, 2002). Thus, while Iso-1 activity responds to mitogenic and nutritional stimuli, as well as energy status, the short isoforms are refractory to any of these signals, yet they can still affect mTORC1. Indeed, our data suggests that these short isoforms transmit through mTORC1 activation, a constitutive mitogenic/metabolic signal, even in the absence of growth factors (Figs. 3, S3). In addition, S6K1 short isoforms cannot induce mTOR S2448 phosphorylation, as they lack a functional kinase domain and are catalytically inactive (Figs. S2E–G, and 5A). Thus, our current hypothesis is that Iso-1 phosphorylates mTOR and inhibits cap-dependent translation and transformation. Iso-1 also phosphorylates other cellular substrates that might lead to suppression of transformation. In contrast, the short isoforms bind to the same or other cellular targets and thereby induce opposite effects. S6K1 short isoforms can bind directly to mTORC1 and enhance its activity possibly by competing with Iso-1 for mTOR binding (Fig. 7F). Nevertheless, the mechanism by which S6K1 short isoforms activate mTOR is yet to be determined.

We have discovered new alternatively spliced variants of the gene encoding S6K1. Our findings provide additional insight into the oncogenic switch that can be caused by alternative splicing. Furthermore, a new mode of mTORC1 activation is identified in this study, which may open a new direction in manipulating mTORC1 activity for cancer therapy. In addition, we show that, the balance between S6K1 short isoforms and Iso-1 is tipped towards those short isoforms in most tumors examined, suggesting that this splicing event can be a useful marker for breast cancer development. Moreover, since S6K1 short isoforms activate mTORC1 and this activation is important for tumorigenesis, we expect that the newly approved active site mTOR inhibitors will act more efficiently on tumors where the S6K1 alternative splicing switch has occurred.

Materials and methods

Cells

NIH 3T3 cells were grown in DMEM supplemented with 10% (v/v) calf serum (BS), penicillin and streptomycin. Human breast cells: MCF-10A, were grown in DMEM/F12 supplemented with 5% (v/v) horse serum, 50 ng/ml epidermal growth factor (EGF), 10 μg/ml insulin, 0.5 μg/ml hydrocortisone, 100 ng/ml cholera toxin, penicillin and streptomycin (Arias-Romero et al., 2010). MCF-7, MDA-MB-231 and HEK293 cells were grown in DMEM, HMLE cells were grown in MEBM/DMEM/F12 and MDA-MB-468 cells were grown in Leibovitz-F12, supplemented with 10% (v/v) FBS, penicillin and streptomycin. SUM159 cells were grown in Ham’s F12 with 5% (v/v) calf serum, 5 μg/ml insulin and 1 μg/ml hydrocortisone. S6K1/2 double knockout Mouse Embryonic Fibroblasts (MEFs) were generated as described in (Pende et al., 2004) and were immortalized by the 3T3 protocol as described (Dowling et al., 2010).

Stable cell lines

NIH 3T3 and MCF-10A cells were infected with pBABE-puro retroviral vectors encoding T7-tagged S6K1 isoforms human and mouse cDNAs as described (McCurrach and Lowe, 2001). Infected cells were selected with puromycin (2 μg/ml) or hygromycin (200 μg/ml) for 72–96 h. In the case of double infection with pWZL-hygro-Ras (McCurrach and Lowe, 2001), cells were treated with hygromycin for 72h after selection with puromycin. In the case of infection with MLP-puro-shRNAs vectors, MCF-10A cell transductants were selected with puromycin (2 μg/ml) for 96 h.

Immunoblotting

Cells were lysed in Laemmli buffer and analyzed for total protein concentration as described (Golan-Gerstl et al., 2011).. Primary antibodies. Sigma; β-catenin (1:2,000) phospho-ERK (1:10000 T202/Y204), Santa Cruz; β-actin (1:200), Total Akt (1:200 sc1619), GAPDH (1:1000 sc25778), Myc (1:1000 sc40). Novagen; T7 tag (1:5,000);. BD Trunsduction Laboratories; Total S6K1-anti-p70. Cell Signaling Technologies; Total ERK1/2 (1:1,000); phospho-4E-BP1 Thr70 (1:1000 #2855), 4E-BP1 (1:1000 #9452), phospho-S6K1 Thr389 (1:1000 #9205), p-S6 ser240/244 (1:1000 #2215), Total S6 (1:1000 #2217), phospho-Akt ser473 (1:1000 #9271), RAS (1:1000 #3339), p-S2448-mTOR (1:1000 #2972), Secondary antibodies: HRP-conjugated goat anti-mouse, anti-rabbit or anti-goat IgG (H+L) (1:10,000 Jackson Laboratories).

Immunoprecipitation

HEK293 cells were transfected with myc-mTOR and pCDNA3-T7-Iso-2/h6A/h6C/pCDNA3 or pCDNA-T7-Iso-1 S6K1 with empty pCDNA. 48 hours after transfection cells were lysed in CHAPS buffer as described (Sarbassov et al., 2004). After protein quantitation, 1 μg of anti-T7 antibody bound to 40 μl of 50% protein G-sepharose was incubated with 800 μg of total protein lysate over night. After washing 4 times with CHAPS buffer, beads were incubated with 50 μl of 2X Laemmli buffer and separated by SDS-PAGE.

Colony survival assay

S6K1/2 DKO MEF cell populations were transduced with retroviruses encoding for empty pB(−) vector or T7-Iso-1 S6K1. After selection, cells were seeded sparsely (200 cells/well of 6-well plates) and were grown for 14 days. Colonies were fixed and stained with methylene blue as described (Karni et al., 2007).

Dual luciferase reporter assay

MCF-10A cells were co-transfected using Fugene 6 (Roche) with the dual reporter vector pLPL (Cap-Renilla-IRES-Luciferase) (Gerlitz et al., 2002) and the indicated S6K1 isoforms. Cap-dependent translation (Renilla luciferase activity) and IRES-mediated translation (Firefly luciferase activity) were measured with the Promega Stop and Glow assay kit according to the manufacturer’s instructions.

EGF and IGF-1 activation

3×105 MCF-10A cells were seeded in 6 well plates. 24h later, cells where washed with PBS and medium was replaced to a serum and growth factor free media for 24 hours. Cells were stimulated with 50 ng/ml IGF-1 or 50 ng/ml EGF for 4 hours. After 4 hours cells were lysed in Laemmli buffer for Western Blot analysis.

Anchorage-independent growth

Colony formation in soft agar was assayed as described (Karni et al., 2007; McCurrach and Lowe, 2001). After 14–21 days, colonies from ten different fields in each of two wells were counted for each transductant pool, and the average number of colonies per well was calculated. The colonies were stained as described (Karni et al., 2007; McCurrach and Lowe, 2001).

Growth curves

MCF-10A cells were infected with the indicated retroviruses. Following selection, 2000 or 3500 cells per well were seeded in 96-well plates and grown in DMEM/F12 media with 5% or 0.2% Horse Serum. Cells were fixed and stained with methylene blue as described (Karni et al., 1999), and the absorbance at λ=650nm of the acid-extracted stain was measured on a plate reader (BioRad).

Three dimensional morphogenesis of mammary epithelial cells assay

MCF-10A cells were infected with the indicated retroviruses. Following selection, 1500 cells per well were seeded in 96-well in Growth factor- reduced Matrigel (BD Bioscience, Cat# 25300), as described in (Zhan et al., 2008). Method also described in (Debnath et al., 2003). In some cases 2.5 or 5 ng/ml of EGF or 10 μg/ml of insulin were added to assay medium of seeded cells.

Wound healing assay

MCF-10A cells were seeded in a 6 well plate until formation of a confluent monolayer, and a “wound” was created by scratching the monolayer with a p200 pipette tip as described in (Liang et al., 2007). Cells were washed with PBS and growth medium was replaced to DMEM/F12 media with 5% horse serum without EGF and insulin. The wound was photographed (×100 magnification) after matching the reference point in a phase contrast microscope and wound area was measured using Photoshop record measurement analysis tools.

Xenograft tumor formation in mice

Pools of Ras-MCF-10A cells expressing the indicated S6K1 isoforms were injected sub-cutaneusly into the rear flanks of NOD-SCID mice (2×106 cells per site in 100 μl of serum free media containing 0.25 v/v matrigel (BD Bioscience) using a 26-gauge needle. Tumor growth was monitored as described (Karni et al., 2007). Veterinary care was provided to all animals by the Hebrew University animal care facility staff in accord with AAALAC standard procedures and as approved by the Hebrew University Ethics committee (No’ MD-08-11517-4).

Statistical Analysis

All data presented as histograms refer to a mean value ± SEM of the total number of independent experiments. An unpaired, two-tailed t test was used to determine p values for Figure 2C and 5B.

Supplementary Material

Supp Figs
Supplemental

Acknowledgments

The authors wish to thank Nahum Sonenberg and Jerry Pelletier (McGill University) for 4E-BP1 mutant constructs; Orna Elroy-Stein for the bi-cistronic Cap-Renilla-IRES-firefly luciferase reporter gene construct; Mario Pende (INSERM, Paris), for the S6K1/2 DKO MEF’s; Oded Meyuhas and Ittai Ben Porath for comments on the manuscript; Members of the Ben Porath lab for technical help and members of the Karni lab for discussions. This study was supported by the US-Israel Bi-national Science Foundation (BSF) (grant no’ 2009026 to R.K. and A.K) and Israel Science Foundation (ISF) (grant no’ 780/08 to R.K.).

Footnotes

Author Contribution

V.B.H. and R.K. designed the experiments, V.B.H., P.D., Z.S and A.M. performed experiments, B.D. provided breast tumor samples, A.K. provided reagents and comments on the manuscript, and V.B.H. and R.K wrote the manuscript.

Conflict of Interest

The authors declare they do not have any Conflict of Interest in publishing this paper

References

  1. Ali SM, Sabatini DM. Structure of S6 kinase 1 determines whether raptor-mTOR or rictor-mTOR phosphorylates its hydrophobic motif site. J Biol Chem. 2005;280:19445–19448. doi: 10.1074/jbc.C500125200. [DOI] [PubMed] [Google Scholar]
  2. Alliouachene S, Tuttle RL, Boumard S, Lapointe T, Berissi S, Germain S, Jaubert F, Tosh D, Birnbaum MJ, Pende M. Constitutively active Akt1 expression in mouse pancreas requires S6 kinase 1 for insulinoma formation. J Clin Invest. 2008;118:3629–3638. doi: 10.1172/JCI35237. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Arias-Romero LE, Villamar-Cruz O, Pacheco A, Kosoff R, Huang M, Muthuswamy SK, Chernoff J. A Rac-Pak signaling pathway is essential for ErbB2-mediated transformation of human breast epithelial cancer cells. Oncogene. 2010;29:5839–5849. doi: 10.1038/onc.2010.318. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Avdulov S, Li S, Michalek V, Burrichter D, Peterson M, Perlman DM, Manivel JC, Sonenberg N, Yee D, Bitterman PB, et al. Activation of translation complex eIF4F is essential for the genesis and maintenance of the malignant phenotype in human mammary epithelial cells. Cancer Cell. 2004;5:553–563. doi: 10.1016/j.ccr.2004.05.024. [DOI] [PubMed] [Google Scholar]
  5. Bepler G, Koehler A. Multiple chromosomal aberrations and 11p allelotyping in lung cancer cell lines. Cancer Genet Cytogenet. 1995;84:39–45. doi: 10.1016/0165-4608(95)00063-1. [DOI] [PubMed] [Google Scholar]
  6. Chiang GG, Abraham RT. Phosphorylation of mammalian target of rapamycin (mTOR) at Ser-2448 is mediated by p70S6 kinase. J Biol Chem. 2005;280:25485–25490. doi: 10.1074/jbc.M501707200. [DOI] [PubMed] [Google Scholar]
  7. David CJ, Manley JL. Alternative pre-mRNA splicing regulation in cancer: pathways and programs unhinged. Genes Dev. 2010;24:2343–2364. doi: 10.1101/gad.1973010. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Debnath J, Muthuswamy SK, Brugge JS. Morphogenesis and oncogenesis of MCF-10A mammary epithelial acini grown in three-dimensional basement membrane cultures. Methods. 2003;30:256–268. doi: 10.1016/s1046-2023(03)00032-x. [DOI] [PubMed] [Google Scholar]
  9. Dowling RJ, Topisirovic I, Alain T, Bidinosti M, Fonseca BD, Petroulakis E, Wang X, Larsson O, Selvaraj A, Liu Y, et al. mTORC1-mediated cell proliferation, but not cell growth, controlled by the 4E-BPs. Science. 2010;328:1172–1176. doi: 10.1126/science.1187532. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Gerlitz G, Jagus R, Elroy-Stein O. Phosphorylation of initiation factor-2 alpha is required for activation of internal translation initiation during cell differentiation. Eur J Biochem. 2002;269:2810–2819. doi: 10.1046/j.1432-1033.2002.02974.x. [DOI] [PubMed] [Google Scholar]
  11. Golan-Gerstl R, Cohen M, Shilo A, Suh SS, Bakacs A, Coppola L, Karni R. Splicing factor hnRNP A2/B1 regulates tumor suppressor gene splicing and is an oncogenic driver in glioblastoma. Cancer Res. 2011;71:4464–4472. doi: 10.1158/0008-5472.CAN-10-4410. [DOI] [PubMed] [Google Scholar]
  12. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell. 2011;144:646–674. doi: 10.1016/j.cell.2011.02.013. [DOI] [PubMed] [Google Scholar]
  13. Holz MK, Ballif BA, Gygi SP, Blenis J. mTOR and S6K1 mediate assembly of the translation preinitiation complex through dynamic protein interchange and ordered phosphorylation events. Cell. 2005;123:569–580. doi: 10.1016/j.cell.2005.10.024. [DOI] [PubMed] [Google Scholar]
  14. Holz MK, Blenis J. Identification of S6 kinase 1 as a novel mammalian target of rapamycin (mTOR)-phosphorylating kinase. J Biol Chem. 2005;280:26089–26093. doi: 10.1074/jbc.M504045200. [DOI] [PubMed] [Google Scholar]
  15. Hsieh AC, Costa M, Zollo O, Davis C, Feldman ME, Testa JR, Meyuhas O, Shokat KM, Ruggero D. Genetic dissection of the oncogenic mTOR pathway reveals druggable addiction to translational control via 4EBP-eIF4E. Cancer Cell. 2010;17:249–261. doi: 10.1016/j.ccr.2010.01.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Janda E, Lehmann K, Killisch I, Jechlinger M, Herzig M, Downward J, Beug H, Grunert S. Ras and TGF[beta] cooperatively regulate epithelial cell plasticity and metastasis: dissection of Ras signaling pathways. J Cell Biol. 2002;156:299–313. doi: 10.1083/jcb.200109037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Karni R, de Stanchina E, Lowe SW, Sinha R, Mu D, Krainer AR. The gene encoding the splicing factor SF2/ASF is a proto-oncogene. Nat Struct Mol Biol. 2007;14:185–193. doi: 10.1038/nsmb1209. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Karni R, Dor Y, Keshet E, Meyuhas O, Levitzki A. Activated pp60c-Src leads to elevated hypoxia-inducible factor (HIF)-1alpha expression under normoxia. J Biol Chem. 2002;277:42919–42925. doi: 10.1074/jbc.M206141200. [DOI] [PubMed] [Google Scholar]
  19. Karni R, Gus Y, Dor Y, Meyuhas O, Levitzki A. Active Src elevates the expression of beta-catenin by enhancement of cap-dependent translation. Mol Cell Biol. 2005;25:5031–5039. doi: 10.1128/MCB.25.12.5031-5039.2005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Karni R, Hippo Y, Lowe SW, Krainer AR. The splicing-factor oncoprotein SF2/ASF activates mTORC1. Proc Natl Acad Sci U S A. 2008;105:15323–15327. doi: 10.1073/pnas.0801376105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Karni R, Jove R, Levitzki A. Inhibition of pp60c-Src reduces Bcl-XL expression and reverses the transformed phenotype of cells overexpressing EGF and HER-2 receptors. Oncogene. 1999;18:4654–4662. doi: 10.1038/sj.onc.1202835. [DOI] [PubMed] [Google Scholar]
  22. Kim DH, Sarbassov DD, Ali SM, King JE, Latek RR, Erdjument-Bromage H, Tempst P, Sabatini DM. mTOR interacts with raptor to form a nutrient-sensitive complex that signals to the cell growth machinery. Cell. 2002;110:163–175. doi: 10.1016/s0092-8674(02)00808-5. [DOI] [PubMed] [Google Scholar]
  23. Lee-Fruman KK, Kuo CJ, Lippincott J, Terada N, Blenis J. Characterization of S6K2, a novel kinase homologous to S6K1. Oncogene. 1999;18:5108–5114. doi: 10.1038/sj.onc.1202894. [DOI] [PubMed] [Google Scholar]
  24. Liang CC, Park AY, Guan JL. In vitro scratch assay: a convenient and inexpensive method for analysis of cell migration in vitro. Nat Protoc. 2007;2:329–333. doi: 10.1038/nprot.2007.30. [DOI] [PubMed] [Google Scholar]
  25. Mamane Y, Petroulakis E, LeBacquer O, Sonenberg N. mTOR, translation initiation and cancer. Oncogene. 2006;25:6416–6422. doi: 10.1038/sj.onc.1209888. [DOI] [PubMed] [Google Scholar]
  26. Mamane Y, Petroulakis E, Rong L, Yoshida K, Ler LW, Sonenberg N. eIF4E--from translation to transformation. Oncogene. 2004;23:3172–3179. doi: 10.1038/sj.onc.1207549. [DOI] [PubMed] [Google Scholar]
  27. Manning BD, Cantley LC. AKT/PKB signaling: navigating downstream. Cell. 2007;129:1261–1274. doi: 10.1016/j.cell.2007.06.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. McCurrach ME, Lowe SW. Methods for studying pro- and antiapoptotic genes in nonimmortal cells. Methods Cell Biol. 2001;66:197–227. doi: 10.1016/s0091-679x(01)66010-2. [DOI] [PubMed] [Google Scholar]
  29. Mendoza MC, Er EE, Blenis J. The Ras-ERK and PI3K-mTOR pathways: cross-talk and compensation. Trends Biochem Sci. 2011;36:320–328. doi: 10.1016/j.tibs.2011.03.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Mieulet V, Roceri M, Espeillac C, Sotiropoulos A, Ohanna M, Oorschot V, Klumperman J, Sandri M, Pende M. S6 kinase inactivation impairs growth and translational target phosphorylation in muscle cells maintaining proper regulation of protein turnover. Am J Physiol Cell Physiol. 2007;293:C712–722. doi: 10.1152/ajpcell.00499.2006. [DOI] [PubMed] [Google Scholar]
  31. Mills JR, Hippo Y, Robert F, Chen SM, Malina A, Lin CJ, Trojahn U, Wendel HG, Charest A, Bronson RT, et al. mTORC1 promotes survival through translational control of Mcl-1. Proc Natl Acad Sci U S A. 2008;105:10853–10858. doi: 10.1073/pnas.0804821105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Monni O, Barlund M, Mousses S, Kononen J, Sauter G, Heiskanen M, Paavola P, Avela K, Chen Y, Bittner ML, et al. Comprehensive copy number and gene expression profiling of the 17q23 amplicon in human breast cancer. Proc Natl Acad Sci U S A. 2001;98:5711–5716. doi: 10.1073/pnas.091582298. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Nardella C, Lunardi A, Fedele G, Clohessy JG, Alimonti A, Kozma SC, Thomas G, Loda M, Pandolfi PP. Differential expression of S6K2 dictates tissue-specific requirement for S6K1 in mediating aberrant mTORC1 signaling and tumorigenesis. Cancer Res. 2011;71:3669–3675. doi: 10.1158/0008-5472.CAN-10-3962. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Ohanna M, Sobering AK, Lapointe T, Lorenzo L, Praud C, Petroulakis E, Sonenberg N, Kelly PA, Sotiropoulos A, Pende M. Atrophy of S6K1(−/−) skeletal muscle cells reveals distinct mTOR effectors for cell cycle and size control. Nat Cell Biol. 2005;7:286–294. doi: 10.1038/ncb1231. [DOI] [PubMed] [Google Scholar]
  35. Pende M, Um SH, Mieulet V, Sticker M, Goss VL, Mestan J, Mueller M, Fumagalli S, Kozma SC, Thomas G. S6K1(−/−)/S6K2(−/−) mice exhibit perinatal lethality and rapamycin-sensitive 5′-terminal oligopyrimidine mRNA translation and reveal a mitogen-activated protein kinase-dependent S6 kinase pathway. Mol Cell Biol. 2004;24:3112–3124. doi: 10.1128/MCB.24.8.3112-3124.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Rosner M, Hengstschlager M. Nucleocytoplasmic localization of p70 S6K1, but not of its isoforms p85 and p31, is regulated by TSC2/mTOR. Oncogene. 2011;30:4509–4522. doi: 10.1038/onc.2011.165. [DOI] [PubMed] [Google Scholar]
  37. Ruvinsky I, Katz M, Dreazen A, Gielchinsky Y, Saada A, Freedman N, Mishani E, Zimmerman G, Kasir J, Meyuhas O. Mice deficient in ribosomal protein S6 phosphorylation suffer from muscle weakness that reflects a growth defect and energy deficit. PLoS One. 2009;4:e5618. doi: 10.1371/journal.pone.0005618. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Sarbassov DD, Ali SM, Kim DH, Guertin DA, Latek RR, Erdjument-Bromage H, Tempst P, Sabatini DM. Rictor, a novel binding partner of mTOR, defines a rapamycin-insensitive and raptor-independent pathway that regulates the cytoskeleton. Curr Biol. 2004;14:1296–1302. doi: 10.1016/j.cub.2004.06.054. [DOI] [PubMed] [Google Scholar]
  39. Sarbassov DD, Ali SM, Sengupta S, Sheen JH, Hsu PP, Bagley AF, Markhard AL, Sabatini DM. Prolonged rapamycin treatment inhibits mTORC2 assembly and Akt/PKB. Mol Cell. 2006;22:159–168. doi: 10.1016/j.molcel.2006.03.029. [DOI] [PubMed] [Google Scholar]
  40. Schalm SS, Blenis J. Identification of a conserved motif required for mTOR signaling. Curr Biol. 2002;12:632–639. doi: 10.1016/s0960-9822(02)00762-5. [DOI] [PubMed] [Google Scholar]
  41. Schoenberg DR, Maquat LE. Regulation of cytoplasmic mRNA decay. Nat Rev Genet. 2012;13:246–259. doi: 10.1038/nrg3160. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Shaw RJ, Cantley LC. Ras, PI(3)K and mTOR signalling controls tumour cell growth. Nature. 2006;441:424–430. doi: 10.1038/nature04869. [DOI] [PubMed] [Google Scholar]
  43. She QB, Halilovic E, Ye Q, Zhen W, Shirasawa S, Sasazuki T, Solit DB, Rosen N. 4E-BP1 is a key effector of the oncogenic activation of the AKT and ERK signaling pathways that integrates their function in tumors. Cancer Cell. 2010;18:39–51. doi: 10.1016/j.ccr.2010.05.023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Shiota C, Woo JT, Lindner J, Shelton KD, Magnuson MA. Multiallelic disruption of the rictor gene in mice reveals that mTOR complex 2 is essential for fetal growth and viability. Dev Cell. 2006;11:583–589. doi: 10.1016/j.devcel.2006.08.013. [DOI] [PubMed] [Google Scholar]
  45. Yuan TL, Cantley LC. PI3K pathway alterations in cancer: variations on a theme. Oncogene. 2008;27:5497–5510. doi: 10.1038/onc.2008.245. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Zhan L, Rosenberg A, Bergami KC, Yu M, Xuan Z, Jaffe AB, Allred C, Muthuswamy SK. Deregulation of scribble promotes mammary tumorigenesis and reveals a role for cell polarity in carcinoma. Cell. 2008;135:865–878. doi: 10.1016/j.cell.2008.09.045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Zoncu R, Efeyan A, Sabatini DM. mTOR: from growth signal integration to cancer, diabetes and ageing. Nat Rev Mol Cell Biol. 2011;12:21–35. doi: 10.1038/nrm3025. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supp Figs
NIHMS814986-supplement-Supp_Figs.pdf (914.1KB, pdf)
Supplemental
NIHMS814986-supplement-Supplemental.doc (35.5KB, doc)

RESOURCES