Research ArticleCancer

The p85 isoform of the kinase S6K1 functions as a secreted oncoprotein to facilitate cell migration and tumor growth

See allHide authors and affiliations

Science Signaling  27 Mar 2018:
Vol. 11, Issue 523, eaao1052
DOI: 10.1126/scisignal.aao1052

Messaging by oncogenic kinase

Cancer cells generate an environment amenable for their growth by signaling through various pathways to other cancer cells and nontransformed cells. Zhang et al. noticed that p85, the longest isoform of the growth-promoting kinase S6K1, had a six-arginine residue motif in its N terminus that was not present in shorter isoforms. This motif resembles a motif in the HIV TAT protein that enables it to be released from cells and enter surrounding cells. In cultured breast cancer or nontransformed cells, adding the p85 isoform of S6K1 to the medium enhanced the phosphorylation of S6K1 targets and behaviors characteristic of cancer cells, such as increased cell size and migration. Furthermore, injection of this S6K1 isoform into mice increased the growth and metastasis of xenografts formed by breast cancer cells. Because the S6K1 gene is amplified in some breast cancers, an antibody that targets this longer isoform may attenuate tumor growth and metastasis.


Cancer cells can remodel surrounding microenvironments to facilitate cell growth, invasion, and migration by secreting proteins that educate surrounding stromal cells. We report that p85S6K1, the longest isoform of S6K (ribosomal protein S6 kinase), but not the shorter isoform p70S6K1 or p56S6K2, was secreted from cancer cells through its HIV TAT-like, N-terminal six-arginine motif. The exogenously produced p85S6K1 protein entered cultured transformed and nontransformed cells to promote or confer malignant behaviors, leading to increased cell growth and migration. When injected into mice, the p85S6K1 protein enhanced the growth of xenografted breast cancer cells and lung metastasis. Hence, our findings reveal a role for p85S6K1 as a secreted oncogenic kinase and provide a mechanism by which cancer cells remodel their microenvironment by transforming the surrounding cells to drive tumorigenesis.


Microenvironmental remodulation, including the induction of cancer-surrounding cells such as tumor-associated macrophages and cancer-associated fibroblasts (14) or evasion from immunosurveillance (5, 6), provides the niche to facilitate tumor initiation and progression. Notably, secreted macromolecules play important roles in mediating the cross-talk among distinct types of cells within tumors (7, 8). To this end, tumor-secreted proteins, especially the proteins released in exosomes, play crucial roles in enabling tumor cells to communicate with surrounding cells (9). As a reflection of their important role in tumorigenesis, these exosome-secreted proteins have been used as potential biomarkers for cancer diagnosis and prognosis (10, 11). However, the physiological role of cancer cell–secreted, non-exosomal proteins has been elucidated to a lesser extent.

Similar to exosomes, the HIV envelope encoding the TAT protein can penetrate mammalian cell membranes (12). Six-tandem arginine repeats confer most, if not all, of the biological ability of the TAT protein to penetrate the mammalian cell membrane and have been adopted in the design of optimized delivery system for macromolecular drugs to enter cells (1216). However, whether endogenous proteins in mammalian cells use a TAT-like protein motif to facilitate protein secretion remains unknown.

Here, we sought to identify the secreted oncoproteins with properties similar to TAT proteins. We found that the longest form of S6K, p85S6K1, was secreted in a manner that depended on a six-tandem arginine repeat in the N terminus of the protein. Furthermore, the secreted p85S6K1 could enter surrounding cells to facilitate cell proliferation and transformation, further leading to increased cancer cell growth and migration in vitro, xenograft cancer growth, and lung metastasis in vivo.


To identify the potential secreted proteins with properties similar to TAT proteins, we screened the proteome with the bioinformatics tool from Cell Signaling Technology (, searching for proteins that harbor an arginine-enriched region of at least six arginines (6R) (fig. S1). Only 10 proteins had a 6R motif, including the long-form PTEN (phosphatase and tensin homolog), which is secreted and enters cells to influence tumorigenesis (17), and p85S6K1, the longest form of S6K kinase, which harbors a 6R motif in its extreme N terminus (Fig. 1A). Because the 6R motif is required to be exposed for its biological functions, it is typically fused to the N terminus of a peptide for drug delivery purposes (15), and we therefore chose to focus on understanding the physiological role of the N-terminal 6R motif to the biochemical and biological properties of p85S6K1.

Fig. 1 p85S6K1 and p70S6K1 share similar biochemical and biological properties.

(A) Alignment of domains between S6K1 and S6K2. As indicated, p85S6K1 contains an extra 6R residues and nuclear localization sequence (NLS) in its extreme N terminus. KD, kinase domain. (B) In vitro kinase assays were performed with S6K1 purified from mammalian cells and with the bacterially purified His-S6 as the substrate. n = 2 biological replicates. (C) Immunoblot (IB) analysis of whole-cell lysates (WCLs) derived from human embryonic kidney (HEK) 293 cells transfected with the indicated constructs encoding various forms of S6K1. n = 2 biological replicates. P, phosphorylated. (D to K) IB analysis of WCL derived from S6k1/2−/− MEFs (D) or S6K1-depleted MDA-MB-231 cells (H) stably expressing the indicated constructs. The cells were subjected to cell size detection (E and I), colony formation (F and J, left), and invasion (G and K, left) assays. Relative colony numbers or numbers of invasive cells were calculated in the right panels of (F), (G), (J), and (K). EV, empty vector. Data are means ± SD; n = 3 biological replicates for (D) to (K). *P < 0.05, **P < 0.01 (t test).

As the major downstream target of mammalian target of rapamycin complex 1 (mTORC1), S6K phosphorylates multiple downstream substrates including S6, eIF4B, and GSK3 (glycogen synthase kinase 3) to regulate various key cellular processes including protein synthesis, cell size, cell proliferation (1821), and cell migration (22). However, the biochemical and biological functions of different S6K family proteins including S6K1 (which consists of the p70S6K1 and p85S6K1 isoforms) and S6K2 (also known as p56S6K2) are relatively undefined (23). In keeping with a possible oncogenic role for S6K kinases, bioinformatics analysis indicated that both S6K1 and S6K2 were amplified in multiple human cancers, especially in breast cancers (fig. S2, A and B). The amplification of S6K1 and S6K2 appeared to be mutually exclusive in breast cancers (fig. S2C), suggesting possibly redundant functions of S6K1 and S6K2. Structurally, p85S6K1 shared a high degree of homology in amino acid sequence with p56S6K2, except the extreme N-terminal 6R motif that is unique to p85S6K1 (Fig. 1A). Moreover, the acquisition of an extra exon coding 23 amino acids starting with the “MRRRRRR” motif in the p85S6K1 N terminus appears to be a relatively late event during evolution because it is present only in mammalian isoforms (fig. S2D). In addition to the MRRRRRR motif, the extra 23–amino acid N terminus of p85S6K1 also contains a nuclear localization signal (NLS) (Fig. 1A), which does not affect S6K1 protein stability (fig. S2E).

To distinguish between the physiological roles of p85S6K1 and p70S6K1, plasmids encoding p70S6K1 (which lacks the N-terminal 23–amino acid sequence of p85S6K1) and a form of p85S6K1 lacking the N-terminal 6R motif (termed Δ6R-p85) were constructed (Fig. 1A). Consistent with previous findings, we observed that the absence of either the 23–amino acid sequence or the 6R motif minimally affected p85-mediated phosphorylation of its downstream target S6 both in vitro (Fig. 1B) and in cells (Fig. 1C) (23). In addition, ectopic expression of p70, p85, and Δ6R-p85 in S6k1/2−/− mouse embryonic fibroblasts (MEFs) had comparable effects on the phosphorylation status of S6K downstream targets, including pS6 and pEIF4B (Fig. 1D). As a result, compared to empty vector–expressing S6k1/2−/− MEFs, we observed similar increases in cell size, cell proliferation, colony formation, and migration upon expression of p70, p85, and Δ6R-p85 in S6k1/2−/− MEFs (Fig. 1, E to G). Similar results were obtained in S6K1-depleted breast cancer MDA-MB-231 cells (Fig. 1, H to K, and fig. S3A). Together, these results suggest that neither the 6R motif nor the extra 23–amino acid N terminus of p85 was indispensable for the biochemical function of S6K1, at least in terms of phosphorylating its downstream targets to govern the cell oncogenic behaviors we examined.

Consistent with the critical role of the 6R motif to enable proteins to penetrate into or to be secreted from cells [including long-form PTEN (17)], we observed that the addition of the 6R motif to the short-form PTEN enabled its secretion from cells (fig. S3, B to D) and entry into cells to inhibit downstream Akt signaling (fig. S3E). Similar to long-form PTEN, p85, but not Δ6R-p85 or p70, was enriched by concentrating columns or immunoprecipitations from cellular medium derived from cells ectopically expressing different forms of S6K1 (Fig. 2, A and B). Secretory proteins are synthesized in the endoplasmic reticulum (ER) (24, 25), and we found that treating p85-expressing cells with the ER transport inhibitor brefeldin A efficiently blocked the secretion of p85 from 293T cells (Fig. 2C). This result indicates that p85 might be secreted through ER modification, a process that likely depends on the N-terminal 6R motif of p85.

Fig. 2 p85S6K1 is secreted from cells and can enter cells in a manner dependent on the N-terminal 6R motif.

(A) Schematic workflow for the identification of secreted proteins. (B) The supernatant medium harvested from HEK293T cells transfected with indicated constructs was concentrated by column or hemagglutinin (HA) immunoprecipitation (IP) and then subjected to IB analysis. n = 2 biological replicates. (C) HEK293T cells were transfected with p85S6K1 and treated with or without brefeldin A (1 μg/ml) for 20 hours. The supernatant medium was harvested and subjected to HA IP and analyzed by IB. n = 2 biological replicates. (D) Different S6K1 proteins were eluted and purified from HEK293T cells transfected with the indicated constructs by HA IP, and the purified proteins were applied to S6K1-depleted MDA-MB-231 cells for 20 hours before harvesting for HA IP and IB analysis. n = 2 biological replicates. (E) S6K1 proteins were collected by HA IP from HEK293T cells expressing p85-GFP-HA constructs, eluted, and applied to S6K1-depleted HeLa cells, which were immunostained to show cell penetration of p85. n = 2 biological replicates. Scale bar, 50 μm. (F) Workflow for the analysis of secreted p85 protein entering the surrounding fibroblasts in vivo. (G) Tumors dissected from mice harboring HA-p85– or p70-expressing MDA-MB-231 cells treated with or without brefeldin A were stained with HA antibody. Scale bars, 25 μm. (H) Bacterially produced, purified recombinant S6K1 protein was applied to S6K1-depleted MDA-MB-231 cells for 20 hours before harvesting for HA IP and IB analysis. n = 2 biological replicates. IHC, immunohistochemistry; GFP, green fluorescent protein.

Next, we characterized the potential functions of secreted p85 on surrounding cells. To this end, p85, Δ6R-p85, and p70 were purified from cells ectopically expressing different forms of S6K1, among which p85, but not other purified proteins including Δ6R-p85 and p70, entered cells as detected by immunoprecipitations (Fig. 2D and fig. S4, A to C) or immunofluorescence staining (Fig. 2E). Consequently, only the administration of p85, but not Δ6R-p85 or p70, in S6k1/2−/− MEFs or S6K1-depleted MDA-MB-231 cells resulted in increased phosphorylation of S6K1 downstream targets including S6 and EIF4B (fig. S4, D to F), which was associated with enhanced cell proliferation and migration in S6K-deficient cells (fig. S5, A to F). p85 staining was detected in fibroblasts within tumors containing p85-expressing MDA-MB-231 cells, an effect that was abrogated by brefeldin A treatment (Fig. 2, F and G). In contrast, p70 staining was not detected in fibroblasts within tumors containing p70-expressing cells. These results validate the secretion and penetration capability of p85 in vivo.

To pinpoint the physiological functions of secreted p85, we purified recombinant p85, Δ6R-p85, and p70 from bacteria and observed that, similar to p85 purified from 293T cells, bacterially purified recombinant p85, but not Δ6R-p85 or p70, entered target cells (Fig. 2H and fig. S6, A and B). Moreover, application of bacterially purified p85, but not Δ6R-p85 or p70, significantly increased the phosphorylation of S6K downstream targets, including S6 and eIF4B, in S6K1-depleted MDA-MB-231 and HeLa cells (Fig. 3A and fig. S6, C to E), subsequently resulting in increases in cell size, colony formation, and migration (Fig. 3, B to D, and fig. S6, F to H). Moreover, in support of a role for secreted p85 in remodeling tumor microenvironment, we found that bacterially purified recombinant p85, but not Δ6R-p85 or p70, entered nontransformed, S6k1/2−/− MEFs to promote the phosphorylation of downstream targets (Fig. 3E) and enhance cell size, colony formation, and migration (Fig. 3, F to H). Thus, these findings suggest that recombinant p85 could induce the malignant phenotypes of both breast cancer cells and fibroblasts.

Fig. 3 Purified recombinant p85S6K1 enters cells and activates downstream signaling to promote behaviors typical of malignant cells.

(A to D) IB analysis of WCL derived from S6K1-depleted MDA-MB-231 cells treated with different doses of bacterially purified S6K1 proteins (A). The treated cells were assessed for cell size (B), colony formation (C, left), and invasion (D, left), and colony numbers and the number of invasive cells were quantified in the right panels of (C) and (D). Data are means ± SD; n = 3 biological replicates from (A) to (D). *P < 0.05, **P < 0.01, ***P < 0.001 (t test). (E to H) IB analysis of WCL derived from S6k1/2−/− MEFs treated with different doses of bacterially purified S6K1 proteins (E). The treated MEFs were assessed for cell size (F), colony formation (G, left), and invasion (H, left), and colony numbers and the number of invasive cells were quantified in the right panels of (G) and (H). Data are means ± SD; n = 3 biological replicates. **P < 0.01, ***P < 0.001 (t test).

In further support of a role for secreted p85 in microenvironment remodeling, we found that media conditioned by control cells, but not S6K1-depleted MDA-MB-231 cells, induced cellular transformation phenotypes, including colony formation and migration in the immortalized, nontransformed, breast epithelial cell line MCF10A (Fig. 4, A to E) and in MEFs (fig. S7, A to E). Moreover, reconstitution of S6K1-depleted MDA-MB-231 cells with p85, but not Δ6R-p85 or p70, restored the ability of media conditioned by these cells to promote MCF10A cell colony formation and migration (Fig. 4, A to E). In further support of this notion, we observed that media conditioned by S6K1-depleted MDA-MB-231 cells reconstituted with p85, but not Δ6R-p85 or p70, enhanced colony formation and migration in nontransformed MEFs (fig. S7, A to E) or WI-38 lung fibroblasts (fig. S7, F to J). Furthermore, to investigate the crucial roles of secreted p85S6K1 from conditioned media of MDA-MB-231 in promoting cell growth and proliferation, we used antibodies to immunodeplete secreted p85 proteins from conditioned media (Fig. 4F). In MCF10A cells exposed to conditioned media from hemagglutinin (HA)-p85–expressing MDA-MB-231 cells, HA antibody, but not the control immunoglobulin G (IgG) antibody, significantly decreased colony formation and migration (Fig. 4, G and H). Similar results were observed in MEFs exposed to conditioned media (fig. S8, A to C). These results suggest that secreted p85 in conditioned medium could induce cells surrounding breast tumors, including normal breast epithelial cells or fibroblasts, to adopt tumor-associated malignant phenotypes.

Fig. 4 Media conditioned by p85S6K1-expressing MDA-MB-231 cells promote malignant behaviors in MCF10A cells.

(A to E) Application of media conditioned by cancer cells onto normal epithelial cells or fibroblasts (A). Where indicated, the conditioned media from S6K1-depleted MDA-MB-231 cells transfected with p85S6K1, p70S6K1, or Δ6Rp85S6K1 were applied to nontransformed MCF10A cells. The treated cells were subjected to colony formation assays (B) and wound-healing assays (C), and colony numbers and relative open area were quantified in (C) and (E). Data are means ± SD; n = 3 biological replicates for (B) and (D). *P < 0.05, **P < 0.01 (t test). (F to H) Immunodepletion of p85S6K1 from media conditioned by MDA-MB-231 cells before application of conditioned media to MCF10A cells (F). The treated cells were subjected to colony formation (G) and wound-healing (H) assays. The relative colony numbers and open area were quantified in the right panels of (G) and (H). Data are means ± SD; n = 3 biological replicates for (G) and (H). **P < 0.01 (t test).

Depletion of S6K1 decreased growth and lung metastasis of MDA-MB-231 xenografts in mice (Fig. 5, A to F), advocating for a critical role of S6K1 in breast cancer proliferation and metastasis. Moreover, tumor growth retardation and lung metastasis induced by S6K1 deficiency were partially reversed by treating the xenografted mice with bacterially purified recombinant p85, but not p70 or Δ6R-p85 (Fig. 5, A to F). These results indicate that 6R motif–induced membrane penetration is critical for recombinant p85 to function as an oncoprotein to enhance cancer growth and distant metastases.

Fig. 5 Purified recombinant p85S6K1 enhances in vivo tumorigenesis and lung metastasis in xenograft mouse model.

(A to C) S6K1-depleted MDA-MB-231 and control cells were subjected to mouse xenograft assays, in which the tumor-grafted mice were treated or not with bacterially purified His-S6K1 proteins. Tumor sizes were monitored (A), and tumor mass was weighed (B) and quantified (C). Data are means ± SD; n = 6 mice per condition. **P < 0.01 (t test). (D to F) The lung metastases from mice treated or not with recombinant S6K1 proteins were presented in (D) and (E), and the metastatic nodules of the lung were quantified in (F). Data are means ± SD; n = 12 mice per condition. **P < 0.01 (t test). Scale bar, 100 μm.


Tumor-secreted proteins, including (but not limited to) exosomes, are critical in educating the tumor microenvironment (10). Here, based on the ability of arginine-enriched peptides to penetrate cell membranes and to be secreted from mammalian cells, we screened the whole proteome and found multiple proteins with an arginine-enriched region, including p85S6K1. Similar to the long-form PTEN protein, long-form S6K1 (p85) was secreted and entered cells in a manner dependent on its N-terminal 6R motif. Although there are no obviously different biological functions among the S6K1 isoforms in cells, p85S6K1, but not other isoforms, was secreted and entered surrounding cells to re-educate distinct types of cells including normal epithelial cells (MCF10A), fibroblasts (MEFs), and lung-derived fibroblasts (MI-38) in a paracrine- and N-terminal 6R motif–dependent manner. In turn, this re-education process may build up a tumor-associated microenvironment that promotes cancer growth, invasion, and further migration (Fig. 6). However, whether and how the secreted p85S6K1 enters the infiltrated immune cells and contributes to tumor immune evasion warrant further investigation.

Fig. 6 A proposed model for the re-education of surrounding cells by p85S6K1 secreted by cancer cells.

A proposed model to illustrate the roles of secreted p85S6K1 in mediating the cross-talk between cancer cells and surrounding stromal cells. In brief, cancer cells harboring amplified S6K1, especially in breast tumors, may secrete p85S6K1, which may enter surrounding cells and have a paracrine effect on surrounding cells including fibroblasts, adjacent normal breast epithelial cells, and other cells (such as immune cells or endothelial cells) to induce cancer-associated phenotypes in these cells. Subsequently, these tumor-associated converted cells could re-educate the microenvironment to facilitate proliferation, invasion, and migration of cancer cells.

Although p85S6K1 has an extra NLS domain compared to p70S6K1, p70S6K1 can translocate into the nucleus to phosphorylate nuclear substrates such as histone proteins (26), which would be expected to further modulate epigenetic alterations, and contributes to early adipogenesis (26). Here, we demonstrated that recombinant p85S6K directly enhances growth and migration in cultured and xenografted breast cancer cells, highlighting a potential strategy of targeting secreted p85 with specific antibodies to combat abnormal p85S6K1-driven breast cancer growth and lung metastasis.


Cell culture and transfection

HEK293, HEK293T, HEK293FT, HeLa, and MDA-MB-231 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum (FBS), 100 U of penicillin, and streptomycin (100 mg/ml). S6k1/2−/− MEFs were a gift from D. Gao (Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences) and were also maintained in DMEM supplemented with 10% FBS. The human immortalized breast epithelial cell line MCF10A was cultured in keratinocyte serum-free medium (Keratinocyte-SFM, Gibco, Life Technologies) supplemented with epidermal growth factor (5 ng/ml) and bovine pituitary extract (40 pg/ml). WI-38 lung fibroblasts were obtained from the American Type Culture Collection. Cell transfection was performed using Lipofectamine and Plus reagents. Packaging of lentiviral short hairpin RNA (shRNA)– or complementary DNA (cDNA)–expressing viruses and retroviral cDNA–expressing viruses as well as subsequent infection of various cell lines were performed according to standard protocols. After viral infection, cells were maintained in the presence of hygromycin (200 μg/ml) or puromycin (1 μg/ml), depending on the viral vectors used to infect cells. The ER transport inhibitor brefeldin A (S2758) was obtained from Selleck.

Plasmid construction and shRNAs

The pcDNA3-PTEN constructs have been previously described (27). pcDNA-p85S6K1 and pcDNA-S6 were obtained from Addgene. pcDNA3-PTEN-HA, pcDNA3-6R-PTEN-HA, pcDNA3-p85S6K1-HA, pcDNA3-Δ6R-p85S6K1-HA, pcDNA3-p70S6K1-HA, pcDNA3-p85S6K1-HA-GFP (green fluorescent protein), pET-p85S6K1-HA-His, pET-Δ6R-p85S6K1-HA-His, pET-p70S6K1-HA-His, and pET-His-S6 were cloned into the appropriate vectors. Details of how these plasmids were constructed are available upon request. shRNA vectors against endogenous S6K1 have been previously described (28) and target both the p85 and p70 isoforms of S6K1.


All antibodies were used at a 1:1000 dilution in TBST (tris-buffered saline–Tween 20) buffer with 5% nonfat milk for Western blotting. Anti–phospho-Ser473-Akt antibody (4060), anti-Akt1 antibody (2938), anti–total Akt antibody (4691), anti–phospho-S6 antibody (4858), anti-S6 antibody (2217), anti–phospho-eIF4B (Ser406) antibody (8151), anti-eIF4B antibody (3592), anti–His-tag antibody (12698), anti–phospho-S6K1 (Thr389) antibody (9205), and anti-S6K1 antibody (2708) were obtained from Cell Signaling Technology. Polyclonal anti-HA antibody (sc-805) was obtained from Santa Cruz Biotechnology. Anti-tubulin antibody (T-5168), anti-HA agarose beads (A-2095), peroxidase-conjugated anti-mouse secondary antibody (A-4416), and peroxidase-conjugated anti-rabbit secondary antibody (A-4914) were obtained from Sigma. Monoclonal anti-HA antibody (MMS-101P) was obtained from Covance.

Bioinformatics analysis

The screen of 6R-enriched whole-genome encoding proteins was performed with a Cell Signaling Technology web tool ( Genetic alterations in S6K1 and S6K2 were analyzed using the cBioPortal database (29, 30).

Immunoblot and immunoprecipitation analyses

Cells were lysed in EBC buffer [50 mM tris (pH 7.5), 120 mM NaCl, and 0.5% NP-40] supplemented with protease inhibitors (cOmplete Mini, Roche) and phosphatase inhibitors (Phosphatase Inhibitor Cocktail Set I and II, Calbiochem). The protein concentrations of whole-cell lysates were measured by the Beckman Coulter DU-800 spectrophotometer using the Bio-Rad protein assay reagent. Equal amounts of whole-cell lysates were resolved by SDS–polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotted with indicated antibodies. For immunoprecipitation analysis, 1000 μg of lysates was incubated with the indicated beads at 4°C for 3 to 4 hours. The recovered immunocomplexes were washed five times with NETN buffer [20 mM tris (pH 8.0), 150 mM NaCl, 1 mM EDTA, and 0.5% NP-40] before being resolved by SDS-PAGE and immunoblotted with indicated antibodies.

Purification of His-tagged proteins from bacteria and HA-tagged proteins from mammalian cells

Recombinant His-conjugated S6K1 or S6 was generated by transforming BL21 (DE3) Escherichia coli with pET-28 vectors. Starter cultures (5 ml) grown overnight at 37°C were inoculated (1%) into larger volumes (500 ml). Cultures were grown at 37°C until an optical density of 0.8 was reached, following which protein expression was induced for 12 to 16 hours using 0.1 mM isopropyl-β-d-thiogalactopyranoside at 16°C with vigorous shaking. Recombinant proteins were purified from harvested pellets. Pellets were resuspended in 5 ml of EBC buffer and sonicated (five cycles of 10 s each at 50% output). Insoluble proteins and cell debris were discarded following centrifugation in a tabletop centrifuge (13,000 rpm at 4°C for 15 min). Each 1 ml of supernatant was incubated with 50 μl of 50% Nico beads (Qiagen) at 4°C for 3 hours. The Nico beads were washed three times with wash buffer [50 mM tris-HCl (pH 8.0) and 20 mM imidazole] and eluted by elution buffer (wash buffer containing 100 mM imidazole). Recovery and yield of the desired proteins (or complexes) were confirmed by analyzing 10 μl of beads by Coomassie blue staining and quantified against bovine serum albumin standards.

HA-tagged S6K1 were transfected into HEK293T cells. Forty-eight hours after transfection, cells were harvested and lysed with EBC buffer, and HA immunoprecipitation was performed at 4°C for 2 hours. The resulting beads were washed four times with NETN buffer and once with phosphate-buffered saline (PBS) and then eluted with 50 mM HA peptides (Sigma) as instructed. The eluted proteins were validated and quantified by immunoblot analysis.

S6K in vitro kinase assay

S6K1 proteins purified from transfected cells were used for in vitro kinase assays. Two micrograms of bacterially purified His-S6 was used as the substrate. The reactions were performed in the presence of 1 mM adenosine triphosphate (ATP) and kinase reaction buffer [10 μM tris-HCl (pH 7.5), 10 mM MgCl2, 0.1 mM EDTA, and 2 mM dithiothreitol] at 30°C for 30 min, and the reactions were stopped with 3× SDS loading buffer and then subjected to immunoblot analysis.

Cell size calculation and MTS assays

Cells were trypsinized and centrifuged at 1000 rpm for 10 min, and then cell pellets were resuspended in 1 ml of PBS. Fluorescence-activated cell sorting (FACS) analysis was performed, and the mean forward scatter height (FSC-H) was determined. For MTS [3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium] cell proliferation assays, 2000 cells per well were seeded into 96-well plates and maintained in 100 μl of DMEM containing 10% FBS. Twenty microliters of CellTiter 96 AQueous One Solution Reagent (Promega) was added to each well at the indicated time point. After incubation at 37°C for 1 hour, OD490 (optical density at 490 nm) was measured and viability rate was quantified. Each experiment with triple wells was repeated three times independently.

Cell invasion and wound-healing assays

For cell invasion assays, 3 × 104 to 2 × 105 cells were plated in an 8-μm 24-well plate chamber insert (Corning BioCoat GFR Matrigel Invasion Chambers, 354483) with serum-free medium on the top of the insert and medium containing 10% FBS in the well. Each sample was performed in triplicate. Cells were incubated for 24 hours and fixed with 4% paraformaldehyde for 15 min. After washing with PBS, cells on the top of the insert were scraped with a cotton swab. Cells that adhered to the well were stained with 0.5% crystal violet blue for 15 min and examined under the microscope. Cell migration was investigated by cell wound-healing assays. Briefly, a line was scratched through confluent cultures in six-well plates. The gap was inspected by microscopy over time as the cells moved in and filled the damaged area.

Colony formation assays

Cells were seeded into six-well plates (300 or 600 cells per well) and left for 8 to 12 days until visible colonies formed. Colonies were washed with PBS, fixed with 10% acetic acid/10% methanol for 20 min, and stained with 0.4% crystal violet in 20% ethanol for 20 min. After staining, the plates were washed and air-dried, and colony numbers were counted. Three independent experiments were performed to generate the SE of the difference.

Concentration from supernatant medium

Three 15-cm plates of cells were transfected with S6K1-encoding construct for 24 hours in normal medium, and 12 hours later, the cells were washed with PBS twice and the medium was changed to Opti-MEM (Gibco). After 36 hours, the supernatant media were harvested and centrifuged at 4000 rpm for 20 min to eliminate the floating cells, and then the media were subjected to HA immunoprecipitation or column concentration (10-kDa cutoff), respectively. The concentrated proteins were separated by SDS-PAGE and detected by Western blot.

Application of exogenous S6K1 protein to cultured cells

Cells were grown to 60% confluence in medium with 10% serum, washed with PBS, and incubated for another 2 hours in serum-free medium. Cells were then refed with serum-free media containing different S6K1 proteins purified from bacteria or HEK293T cells. The cells were harvested after a 2-hour treatment and subjected to immunoprecipitation and immunoblot analysis.

Conditioned medium treatment

Serum-free conditioned medium was generated from MDA-MB-231 cells transfected with the indicated constructs.

Mouse xenograft assays and treatment of mice

Nude mice were obtained from the National Rodent Laboratory Animal Resources, Shanghai Branch (China), and maintained in a pathogen-free animal facility. For animal vivo imaging technology analysis, S6K1-deleted MDA-MB-231 and control cells were injected into the tail veins of nude mice. Xenografts were then treated with or without bacterially purified His-S6K1 proteins, administered daily through intraperitoneal injection. After 72 hours, lung colonization by the cancer cells was analyzed on the basis of normalized photon emissions. Lung colonization was monitored by imaging with Xenogen Spectrum small-animal imaging system 20 min after intraperitoneal injection of 100 μl of luciferase substrate into mice. Xenograft experiments were performed by injection of 2 × 106 cells in Matrigel (BD) and were monitored every other day. Day 0 was the first day of treatment. Tumor progression was monitored by caliper measurements, as indicated. No detrimental effects on behavior and grooming were observed in the mice after protein treatments, and gross abnormalities were not observed at the time of autopsy. Mice were euthanized by asphyxiation with CO2, and the desired tissues were snap-frozen.

Hematoxylin and eosin staining

Nude mice lung tissues were fixed with formalin, embedded with paraffin, spliced into 5-μm sections, deparaffinized with xylene, and submerged into EDTA antigenic retrieval buffer for antigenic retrieval. Lung sections were stained with hematoxylin and eosin. Digital images of organs were acquired with NanoZoomer (Hamamatsu Photonics).


Differences between control and test conditions were evaluated by Student’s t test or one-way analysis of variance (ANOVA) test using the SPSS 11.5 statistical software. Values of P < 0.05 were considered statistically significant.


Fig. S1. Whole-proteome screen of arginine-enriched proteins.

Fig. S2. S6K1 and S6K2 are amplified in breast cancers.

Fig. S3. The N-terminal 6R motif facilitates PTEN secretion and cell penetration.

Fig. S4. Purified p85, but not p70 or Δ6R-p85, enters HeLa cells.

Fig. S5. Purified p85 enters cells to promote cell proliferation and migration.

Fig. S6. Bacterially purified recombinant p85 enters cells to facilitate proliferation and migration of HeLa cells.

Fig. S7. Media conditioned by p85-expressing MDA-MB-231 cells promote malignant behaviors in MEFs and WI-38 cells.

Fig. S8. Immunodepletion of p85 suppresses malignant behaviors in MEFs induced by conditioned media.


Acknowledgments: We thank the members of the Wei laboratory for critical reading of the manuscript, as well as members of the Chen and Pandolfi laboratories for helpful discussions. Funding: W.W. is a Leukemia & Lymphoma Society research scholar. This work was supported by the National Program on Key Research Project of China (grant 2016YFC0902700), the National Natural Science Foundation of China (grants 81472572 and 81772933), the Shanghai Science and Technology Commission (grants 15QA1402800 and 15DZ2292300), and the NIH (grant CA177910 to W.W. and A.T.). Author contributions: J.Z., J.G., W.C., and W.W. conceived the study and designed experiments. J.Z., J.G., and X.Q. developed the methods. J.Z., J.G., X.Q., B.W., L.Z., Y.W., and W.G. acquired the data from cells and animals. J.Z. and J.G. analyzed and interpreted the data. J.Z., J.G., X.Q., P.P.P., W.C., and W.W. wrote, reviewed, and revised the manuscript. J.Z., J.G., X.Q., B.W., L.Z., Y.W., and W.G. reported and organized data with the supervision of P.P.P., W.C., and W.W. Competing interests: The authors declare that they have no competing interests. Data and materials availability: All data needed to evaluate the conclusions in the paper are present in the paper or the Supplementary Materials.

Stay Connected to Science Signaling

Navigate This Article