Purpose:

Aberrant overexpression of SFN (stratifin) plays an oncogenic role in lung adenocarcinoma. We have shown previously that SKP1, an adapter component of E3 ubiquitin ligase forming an SCF complex, is a unique SFN-binding protein in lung adenocarcinoma cells.

Experimental Design:

In silico simulation and in vitro mutagenesis analysis were performed to identify the SFN-binding domain on SKP1. We examined expression, localization, and stability of SKP1 after knockdown of SFN using lung adenocarcinoma cells including A549. In silico library screening and experimental validation were used for drug screening. Daily oral administration of each candidate drugs to A549-injected tumor-bearing mice was performed to evaluate their in vivo antitumor efficacy.

Results:

Suppression of SFN upregulated the stability of SKP1 and accelerated its cytoplasm-to-nucleus translocation. Consistently, IHC analysis revealed that cytoplasmic expression of SKP1 was significantly associated with SFN positivity, tumor malignancy, and poorer patient outcome. After SFN suppression, ubiquitination of oncoproteins, including p-cyclin E1, p-c-Myc, p-c-Jun, and cleaved Notch 1, which are target proteins of SCFFBW7, was strongly induced. These results indicate that SFN–SKP1 binding results in SCFFBW7 dysfunction and allows several oncoproteins to evade ubiquitination and subsequent degradation. Because inhibition of SFN-SKP1 binding was expected to have antitumor efficacy, we next searched for candidate SFN inhibitors. Aprepitant and ticagrelor were finally selected as potential SFN inhibitors that dose dependently reduced SFN-SKP1 binding and tumor progression in vivo.

Conclusions:

As overexpression of SFN is detectable in most adenocarcinoma, we believe that SFN inhibitors would be novel and promising antitumor drugs for lung adenocarcinoma.

This article is featured in Highlights of This Issue, p. 2677

Translational Relevance

This study demonstrates SFN (stratifin, 14-3-3 sigma) as a druggable potent oncogene that blocks the formation of E3 ubiquitin ligase SCFFBW7 and induces abnormal stabilization of several oncoproteins, including cyclin E1 and c-Myc. Using in silico drug screening, we identified 2 potentially promising candidate SFN inhibitors, aprepitant and ticagrelor, which are already approved as antiemesis or antiplatelet agents, respectively. Drug repositioning is an efficient approach to develop new treatments using preexisting drugs without the possibility to fail during preclinical trials. Although the large numbers of early-stage adenocarcinoma, referred to as ground glass nodules (GGN), are being detected by low-dose CT screening and GGNs can be surgically resected, this can be dangerous or sometimes impossible in patients with multiple GGNs, or in those who are elderly or complicated by chronic interstitial lung disease. SFN inhibitors would provide a novel chemotherapeutic strategy for early-stage lung adenocarcinoma.

Lung adenocarcinoma is the most common subtype of non–small cell lung cancer (NSCLC), accounting for about 50% of all such cases. Although driver mutations and translocations in EGFR, EML4–ALK, and BRAF and several other specific oncogenes have been identified in advanced adenocarcinoma, the mechanism that triggers its early progression has remained unclear. Noguchi and colleagues have demonstrated that adenocarcinoma in situ (AIS), which has an extremely favorable prognosis, progresses in a stepwise manner to early but invasive adenocarcinoma (eIA), with a 5-year survival rate of 75% (1). Previously, we identified stratifin (SFN, 14-3-3 sigma) as a differentially expressed gene that shows aberrant overexpression in most cases of invasive adenocarcinoma, whereas normal lung epithelium and AIS show negative or very low expression (2). We subsequently clarified that the abnormality of SFN expression was triggered by DNA demethylation in the promoter region of SFN (3).

14-3-3 is a highly conserved, ubiquitously expressed protein family associated with many cellular processes. Among the 7 human 14-3-3 isoforms (beta, epsilon, eta, gamma, tau, zeta, and sigma), 14-3-3 sigma (SFN) has a distinct function and has been linked to cancer most directly. SFN was originally identified as a negative regulator of the cell cycle, particularly in response to p53-sensitive DNA damage (4, 5). Recently, it was also reported that SFN negatively regulated the tumor-promoting metabolic program by enhancing c-Myc ubiquitination and subsequent degradation (6). On the other hand, many previous reports have indicated that SFN is a positive mediator of cell proliferation. In breast cancer, SFN induces G1/S progression by increasing the expression of cyclin D1 (7). Moreover, its overexpression is significantly associated with poorer prognosis in patients with breast cancer (8), hepatocellular carcinoma (9), colorectal carcinoma (10), and endometrial carcinoma (11). Our in vitro and in vivo functional analysis has also indicated that SFN regulates cell-cycle progression in a positive manner and promotes the progression of lung adenocarcinoma (12). Importantly, 30% of SFN-transgenic mice with lung-specific expression of human SFN spontaneously developed lung tumors, suggesting that SFN is a potent oncogene. Moreover, because SFN is known to be an anchor protein that binds to various intercellular molecules, we previously screened SFN-specific binding partner(s) in lung adenocarcinoma cells using pulldown assay and LC-MS/MS analysis in order to clarify the molecular mechanism by which SFN facilitates tumor progression (13). Among the listed proteins, in the present study we focused on S-phase kinase-associated protein 1 (SKP1), which is an adaptor part of the SCF E3 ubiquitin ligase complex, as our previous results had shown that suppression of SFN reduced the S-phase subpopulation, implying an association of SFN with S-phase of the cell cycle (12).

SCF ubiquitin ligases constitute a subset of the cullin ring ligase family and are the largest family of E3 ubiquitin ligases in mammals. Each individual SCF E3 ligase consists of an adaptor protein SKP1, a scaffold protein cullin-1, an F-box protein from a 69-member family that serves as a substrate-determining component, and Rbx1, which contains the RING (Really Interesting New Gene) finger domain for recruitment of the E2 enzyme (14). Combinations of these components make up a large number of SCF E3s, including SCFFBW7, SCFSKP2, and SCFβ-TrCP1, which play crucial roles in numerous cellular processes by promoting ubiquitination and subsequent proteasome-mediated degradation of diverse substrates involved in cell-cycle regulation, signal transduction, and transcription (15). Particularly, SCFFBW7 regulates the turnover—and thereby the activity—of important cellular regulatory proteins including cyclin E1 (16), c-Myc (17), c-Jun (18), and Notch1 (19). FBW7 recognizes substrate proteins only when they are phosphorylated (20). Additionally, the FBW7 gene is frequently mutated in a broad spectrum of human malignancies (21). This is easily explained by the fact that many targets of SCFFBW7 are oncoproteins that are expected to accumulate to abnormal levels when expression of FBW7 is lost or reduced. However, in lung adenocarcinoma, mutation of FBW7 seems to be rare. On the basis of this background, we speculated that another mechanism for inactivation of SCFFBW7 function may exist in lung adenocarcinoma, probably in association with SFN.

The aim of the present study was to clarify the molecular mechanism of tumor progression in lung adenocarcinoma associated with SFN binding to SKP1. We hypothesized that SFN binds to SKP1 and specifically blocks SCFFBW7 function to enhance the ubiquitination of oncoproteins such as cyclin E1, c-Myc, c-Jun, and Notch 1. Because our previous study had shown that depletion of SFN led to a significant reduction of cell growth and tumor progression in lung adenocarcinoma, we speculated that inhibition of SFN might have potential efficacy against lung adenocarcinoma. Here, we revealed the molecular mechanism of SFN-SKP1 binding and showed that it is one of the key events occurring in lung adenocarcinogenesis. Finally, using in silico drug screening, we identified 2 small molecules that inhibit SFN-SKP1 binding and could have potential therapeutic application.

Cell lines and culture conditions

The A549 cell line was purchased from RIKEN Cell Bank, HCC827 was obtained from the ATCC, and PL16T and PL16B were established in our laboratory from human lung AIS and human bronchial epithelium, respectively (22). A549 cells were maintained in D-MEM/F12 (Thermo Fisher Scientific) supplemented with 10% fetal bovine serum (FBS; Sigma-Aldrich). HCC827 cells were cultured with RPMI-1640 (Thermo Fisher Scientific) with 10% FBS. PL16B and PL16T cells were maintained in MCDB153HAA (Wako) supplemented with 2% FBS (Sigma-Aldrich), 0.5 ng/mL human-EGF (Toyobo), 5 μg/mL human insulin (Wako), 72 ng/mL hydrocortisone (Wako), 40 μg/mL human transferrin (Sigma-Aldrich), and 20 ng/mL sodium selenite (Sigma-Aldrich). All cells were cultured in a 5% CO2 incubator at 37°C and passaged every 3 days.

Patients and sample collection

We used all 191 specimens of human lung adenocarcinomas that had been surgically resected at the University of Tsukuba Hospital between 1999 and 2014. Follow-up information for all of the corresponding patients was obtainable from the medical records. All of the patients provided informed consent for use of their materials. The study was approved by the institutional ethics review committee, and the lung adenocarcinoma cases were classified according to the UICC TNM classification of malignant tumors (seventh edition) and the World Health Organization classification of malignant tumors (fourth edition; refs. 23, 24).

Computational modeling of the SFN–SKP1 complex

The proposed model for the SFN–SKP1 complex was constructed using protein–protein docking. The X-ray structures of SFN (Protein Data Bank ID code: 3IQU) and SKP1 (Protein Data Bank ID code: 1FS1) were used as starting models. Both PDB entries were refined for protein–protein docking simulations using the Protein Preparation Wizard (25). ClusPro version 2.0 (26) was used to generate docked conformations with the lowest docking energies and clustering properties. During the protein–protein docking procedure, phosphorylation sites were utilized as constraints for residues (Lys49, Arg56, Arg129, Tyr130 of SFN, and Thr131 of SKP1) involved in complex formation at the protein–protein interaction interface. Putative druggable sites at the interaction interface of the SFN–SKP1 complex resulting from protein–protein docking were also detected and represented by small dummy atoms with the PLB index (27) incorporated in the SiteFinder program of MOE (Chemical Computing Group ULC).

In silico library screening

We carried out in silico library screening based on the putative druggable sites of SFN as well as the interaction interface with SKP1 using molecular docking against 7,133 bioactive compounds from the DrugBank database ver. 5.0 (28) and 9,101 bioactive compounds from the Namiki bioactive compound library set (Namiki Shoji Co., Ltd).

For all compounds with molecular weights ranging from 200 Da to 800 Da, ionization and energy minimization were performed by the OPLS3 force field in the LigPrep Script of Maestro (Schrödinger, LLC). These minimized structures were used as input structures for docking simulations. Docking simulations were performed using the Glide (29, 30) SP docking program (Schrödinger, LLC) with a grid box defined by small dummy atoms at putative druggable sites on SFN from computational modeling of the SFN–SKP1 complex step.

In vivo xenograft experiments

Animal experiments were carried out humanely in accordance with the Regulations for Animal Experiments of the University of Tsukuba and the Fundamental Guidelines for Proper Conduct of Animal Experiments and Related Activities in Academic Research Institutions under the jurisdiction of the Ministry of Education, Culture, Sports, Science, and Technology of Japan and with approval from the Institutional Animal Experiment Committee of our university (#17-112).

First, 6-week-old female immune-deficient BALB/cAJcl-nu/nu mice were purchased from CLEA Japan, Inc., and 2.5 × 106 A549 cells in 0.1 mL PBS were subcutaneously injected into the right flank. From the following day (day 1), mice were treated daily with the compounds (aprepitant, ticagrelor, ezetimibe, and chlorhexidine) dissolved in 5% DMSO and 95% corn oil at 80 mg/kg/day. The control group received 200 μL of 5% DMSO + 95% corn oil. Tumor growth was controlled and measured daily, and mice were checked for their vital status and weight. Tumor volume was calculated using the formula: mm3 = 1/2 × L × W2. All mice were sacrificed at the end of the experiment.

Secondly, 2.5 × 106 A549 cells in 0.1 mL PBS were subcutaneously injected into the right flank. Mice were randomized into groups once the tumors had attained a volume of 100 mm3. The treatment groups were treated daily with 3 doses of aprepitant or ticagrelor solution dissolved in 5% DMSO and 95% corn oil. All of the mice were sacrificed at the end of the experiment. Resected tumors were partially fixed with formalin for histologic analysis and frozen in liquid nitrogen for molecular analysis.

SFN binds to SKP1 at Thr131

Our previous pulldown assay and LC-MS/MS analysis had identified SKP1 as an SFN-binding partner in A549 lung adenocarcinoma cells (13). Here, we also confirmed binding of their endogenous proteins using immunoprecipitation (IP) and Western blotting (Supplementary Fig. S1A). Moreover, among 14-3-3 members, only SFN (14-3-3 sigma) showed binding to SKP1 (Supplementary Fig. S1B). To exclude the possibility that a lack of association between other 14-3-3 members and SKP1 was due to low expression of endogenous 14-3-3 proteins, we overexpressed 14-3-3 epsilon, zeta, or sigma and subsequently performed IP with SKP1. As a result, even overexpressed 14-3-3 epsilon or zeta showed no binding with SKP1 (Supplementary Fig. S1C). Because binding of SFN with SKP1 has not yet been reported, we carried out in silico docking simulation to identify the binding surface of each protein. The results indicated that Lys49, Arg56, Arg129, and Tyr130 in SFN and Thr131 in SKP1 were important residues for binding (Supplementary Fig. S2A). We created a mutant SKP1 (T131V) with a Halo-tag and performed a pulldown assay using Halo-tag and subsequent Western blotting with SFN antibodies. Although WT SKP1 bound with SFN, T131V impaired the interaction with SFN (Supplementary Fig. S2B), indicating that Thr131 is an SFN-binding site critical for SFN–SKP1 binding. Moreover, we detected phosphorylated SKP1 in A549 cells using antiphosphorylated threonine (pThr) antibodies (Supplementary Fig. S2C), and T131V decreased expression of phosphor-SKP1 (Supplementary Fig. S2D). Phosphorylated threonine at codon 131 is supposed to be associated with SFN–SKP1 binding. We then generated a quadruple mutant of SFN including K49A, R56A, R129A, and Y130A using a Flag-tag and subsequently performed IP with anti-Flag antibodies. The quadruple mutant of SFN was found to have lost its ability to bind with SKP1 (Supplementary Fig. S2E). Moreover, each single mutant of SFN also showed impaired binding with SKP1 (Supplementary Fig. S2F), suggesting that the residues Lys49, Arg56, Arg129, and Tyr130 form the surface for binding of SKP1 with SFN.

SFN alters the protein stability and subcellular localization of SKP1

In order to investigate the influence of SFN on SKP1 protein, a specific siRNA against SFN (siSFN) and overexpression of SFN were used. Although both suppression and overexpression of SFN do not show conspicuous change of SKP1 mRNA (Supplementary Fig. S3A), suppression of SFN led to an increase of SKP1 protein expression, whereas overexpression of SFN conversely reduced it (Supplementary Fig. S3B). The cycloheximide chase assay indicated that the half-life of SKP1 protein was markedly longer in the cells transfected with siSFN than in the controls (Fig. 1A). Additionally, ubiquitination of SKP1 was upregulated by overexpression of SFN (Fig. 1B), suggesting that SFN might regulate the half-life or stability of SKP1. Next, various cell lines such as PL16B (immortalized normal bronchial epithelial cells), PL16T (immortalized AIS cells), HCC827, and A549 (advanced lung adenocarcinoma cells) were subjected to immunofluorescence (IF) staining. Although PL16B showed no SFN expression and only nuclear SKP1 expression, 3 adenocarcinoma cell lines displayed colocalization of SFN and SKP1 in the cytoplasm (Supplementary Fig. S3C). Furthermore, we found that depletion of SFN in A549 or HCC827 cells reduced the expression of cytoplasmic SKP1 and increased that of nuclear SKP1 (Fig. 1C, top; Supplementary Fig. S3D). Conversely, overexpression of SFN in PL16B cells increased the cytoplasmic expression of SKP1 (Fig. 1C, bottom). We also performed fractionation of subcellular components and subjected the fractions to Western blotting. Consistently, transfection with siSFN increased the expression of SKP1 in the nucleus and reduced it in the cytoplasm (Fig. 1D). These data suggest that SFN alters the protein stability and subcellular localization of SKP1.

Figure 1.

SFN alters the protein stability and localization of SKP1. A, Cycloheximide chase assay to examine the half-life of SKP1 protein using A549 cells. Suppression of SFN prolonged the half-life of SKP1. We performed 3 independent experiments and quantified each band to show SKP1 expression as a line graph after normalization of each by ACTB (β-actin). B, Ubiquitination assay for SKP1. A549 cells were transfected with siRNA-resistant pSFN for 24 hours followed by Ub-HA for 24 hours. The cells were then treated with 10 μmol/L MG132, a proteasome inhibitor, for 6 hours at 37°C in a 5% CO2 incubator. The total protein was extracted using IP Lysis Buffer containing 10 mmol/L N-ethylmaleimide (NEM), a deubiquitinating enzyme inhibitor. Ubiquitinated proteins in the cell lysate were collected by co-IP and then subjected to Western blotting. Overexpression of SFN induced ubiquitination of SKP1. C, A549 and PL16B cells transfected with siSFN or pSFN were subjected to immunofluorescence (IF) using SKP1 or SFN antibodies. Suppression of SFN in A549 cells led to translocalization of SKP1 from the cytoplasm to the nucleus. On the other hand, overexpression of SFN in PL16B cells induced translocalization of SKP1 from the nucleus to the cytoplasm. Scale bar, 50 μm. D, Western blotting using subcellular fractionated protein extracted from A549 cells. Consistent with the IF results, depletion of SFN increased the nuclear expression of SKP1 and reduced the cytoplasmic expression. Lamin B was used as a nuclear protein marker, and α-tubulin as a cytoplasmic protein marker. NU; nuclear fraction, CY; cytoplasmic fraction.

Figure 1.

SFN alters the protein stability and localization of SKP1. A, Cycloheximide chase assay to examine the half-life of SKP1 protein using A549 cells. Suppression of SFN prolonged the half-life of SKP1. We performed 3 independent experiments and quantified each band to show SKP1 expression as a line graph after normalization of each by ACTB (β-actin). B, Ubiquitination assay for SKP1. A549 cells were transfected with siRNA-resistant pSFN for 24 hours followed by Ub-HA for 24 hours. The cells were then treated with 10 μmol/L MG132, a proteasome inhibitor, for 6 hours at 37°C in a 5% CO2 incubator. The total protein was extracted using IP Lysis Buffer containing 10 mmol/L N-ethylmaleimide (NEM), a deubiquitinating enzyme inhibitor. Ubiquitinated proteins in the cell lysate were collected by co-IP and then subjected to Western blotting. Overexpression of SFN induced ubiquitination of SKP1. C, A549 and PL16B cells transfected with siSFN or pSFN were subjected to immunofluorescence (IF) using SKP1 or SFN antibodies. Suppression of SFN in A549 cells led to translocalization of SKP1 from the cytoplasm to the nucleus. On the other hand, overexpression of SFN in PL16B cells induced translocalization of SKP1 from the nucleus to the cytoplasm. Scale bar, 50 μm. D, Western blotting using subcellular fractionated protein extracted from A549 cells. Consistent with the IF results, depletion of SFN increased the nuclear expression of SKP1 and reduced the cytoplasmic expression. Lamin B was used as a nuclear protein marker, and α-tubulin as a cytoplasmic protein marker. NU; nuclear fraction, CY; cytoplasmic fraction.

Close modal

Overexpression of SKP1 in the cytoplasm of human lung adenocarcinoma

To clarify the association of SFN with SKP1 in human lung adenocarcinoma tissue, SKP1 IHC was performed using 191 cases of lung adenocarcinoma. We scored nuclear and cytoplasmic SKP1 positivity separately. Consequently, 92 cases showed high SKP1 expression in the cytoplasm. Although SFN-negative cases tended to show SKP1 positivity only in the nucleus, most SFN-positive cases showed positivity not only in the nucleus but also in the cytoplasm (Fig. 2A). H-score for cytoplasmic SKP1 was significantly higher in SFN-positive tumors than in those with SFN negativity (Fig. 2B). Moreover, we found that cytoplasmic expression of SKP1 was significantly associated with gender, the Noguchi classification for small adenocarcinomas of the lung (2 cm or less in diameter), pathologic subtype, pathologic stage, lymphatic permeation, and vascular invasion (Supplementary Table S1). Furthermore, cytoplasmic SKP1 positivity was significantly associated with a poorer outcome relative to cytoplasmic SKP1-negative cases (P < 0.001; Fig. 2C), whereas nuclear SKP1 positivity was not (P = 0.132, data not shown). We also performed multivariate survival analysis using Cox proportional hazards model using 6 factors selected on the basis of the results of univariate analysis: sex (female vs. male), vascular invasion (negative vs. positive), lymphatic permeation (negative vs. positive), pathologic stage (p stage I, II vs. III, IV), smoking history (never-smoker vs. smoker), and cytoplasmic SKP1 H-score (low vs. high). The results showed that vascular invasion, lymphatic permeation, and pathologic stage but not cytoplasmic SKP1 expression were independent prognostic factors (Supplementary Table S2). Taken together, although cytoplasmic SKP1 expression was not an independent prognostic factor, our findings suggest that cytoplasmic SKP1 expression is significantly associated with SFN expression and malignancy of lung adenocarcinoma.

Figure 2.

SKP1 shows overexpression in the cytoplasm of human lung adenocarcinoma. A, Image of SFN or SKP1 IHC for representative cases showing SFN positivity or negativity in IHC. Whereas the SFN-negative case showed SKP1 expression only in the nucleus and not in the cytoplasm, the SFN-positive case had both nuclear and cytoplasmic SKP1 expression in the tumor cells. B, Scatter diagram for H-score of cytoplasmic SKP1 expression. SFN-positive tumor showed significantly higher expression of cytoplasmic SKP1 than negative tumors (P < 0.001). C, Disease-free survival depicted as Kaplan–Meier curves showing the correlation between outcome and cytoplasmic SKP1 expression. Positive expression of cytoplasmic SKP1 was associated with a poorer outcome than the case for negative expression (P < 0.001).

Figure 2.

SKP1 shows overexpression in the cytoplasm of human lung adenocarcinoma. A, Image of SFN or SKP1 IHC for representative cases showing SFN positivity or negativity in IHC. Whereas the SFN-negative case showed SKP1 expression only in the nucleus and not in the cytoplasm, the SFN-positive case had both nuclear and cytoplasmic SKP1 expression in the tumor cells. B, Scatter diagram for H-score of cytoplasmic SKP1 expression. SFN-positive tumor showed significantly higher expression of cytoplasmic SKP1 than negative tumors (P < 0.001). C, Disease-free survival depicted as Kaplan–Meier curves showing the correlation between outcome and cytoplasmic SKP1 expression. Positive expression of cytoplasmic SKP1 was associated with a poorer outcome than the case for negative expression (P < 0.001).

Close modal

Interaction of SFN with SKP1 is competitive with FBW7

Although 69 types of F-box protein have been reported, FBW7, SKP2, and β-TrCP1 are the best-characterized members and therefore were studied further. We expected that the SFN-binding domain might overlap with the FBW7-binding domain on the SKP1 protein. IP with anti-SKP1 antibodies and subsequent Western blotting indicated that suppression of SFN increased SKP1–FBW7 binding but not SKP1–SKP2 or –β-TrCP1 binding (Supplementary Fig. S4A), suggesting that SFN may compete with FBW7 for binding with SKP1. In silico docking simulation analysis also supported this finding (Fig. 3A). In PL16B cells, which do not express SFN, SKP1 and FBW7 were colocalized only in the nucleus (Supplementary Fig. S4B, left). However, in HCC827 cells, which have high SFN expression, SKP1 but not FBW7 was partially localized in the cytoplasm (Supplementary Fig. S4B, right), implying that SFN might sequester SKP1 in the cytoplasm so that SCFFBW7 does not form in the nucleus. SCFFBW7 specifically degrades several phosphorylated oncoproteins that function in cellular growth, including cyclin E1, c-Myc, c-Jun, and Notch1. Therefore, we analyzed the expression of target proteins of SCFFBW7, SCFSKP2, and SCFβ-TrCP1 after transfection with siSFN. The results showed that target proteins of SCFFBW7 such as p-cyclin E1, p-c-Myc, p-c-Jun, and NICD (Notch intracellular domain) were downregulated after suppression of SFN, whereas p-p27 (target of SCFSKP2) and p-IκBα (target of SCFβ-TrCP1) were unaffected (Fig. 3B). Additionally, we found that suppression of SFN clearly increased the ubiquitination of p-cyclin E1, p-c-Myc, p-c-Jun, and NICD (Fig. 3C). These results indicate that SFN might obstruct formation of the SCFFBW7 complex and inhibit ubiquitination of target oncoproteins.

Figure 3.

Interaction of SFN with SKP1 is competitive with FBW7. A, Structural models of (a) the F-box/SKP1/Cul1 complex, (b) the SFN/SKP1/Cul1 complex, and (c) an overlapped view of the F-box/SKP1/Cul1 complex with SFN. The F-box/SKP1/Cul1 complex was constructed using heterodimer complexes from a Cul1–Rbx1–Skp1-F boxSkp2 SCF ubiquitin ligase complex (Protein Data Bank ID code: 1LDK) and a FBW7–SKP1–Cyclin E1 complex (Protein Data Bank ID code: 2OVR) by structural alignment with SKP1. The SFN/SKP1/Cul1 complex was constructed using a predicted model of SFN/SKP1 in this work and a Cul1–Rbx1–Skp1–F boxSkp2 SCF ubiquitin ligase complex (Protein Data Bank ID code: 1LDK). Filled yellow circle indicates the area of structural overlap between the F-box and SFN (c). B, Western blotting using A549 cells transfected with 3 kinds of siSFN (1–3). Substrate proteins of FBW7 such as p-cyclin E1, p-c-Myc, p-c-Jun, and NICD (Notch intracellular domain) were suppressed after knockdown of SFN, although p-p27 (substrate protein of SKP2) and p-IκBα (substrate protein of β-TrCP1) showed no change in expression. C, Ubiquitination assay for substrate proteins of FBW7. A549 cells were transfected with siSFN for 24 hours followed by Ub-HA for 24 hours. The cells were then treated with 10 μmol/L MG132, a proteasome inhibitor, for 6 hours at 37°C in a 5% CO2 incubator. The total protein was extracted using IP Lysis Buffer containing 10 mmol/L N-ethylmaleimide (NEM), a deubiquitinating enzyme inhibitor. Ubiquitinated proteins in the cell lysate were collected by co-IP and then subjected to Western blotting. Suppression of SFN induced marked ubiquitination of p-cyclin E1, p-c-Myc, p-c-Jun, and NICD.

Figure 3.

Interaction of SFN with SKP1 is competitive with FBW7. A, Structural models of (a) the F-box/SKP1/Cul1 complex, (b) the SFN/SKP1/Cul1 complex, and (c) an overlapped view of the F-box/SKP1/Cul1 complex with SFN. The F-box/SKP1/Cul1 complex was constructed using heterodimer complexes from a Cul1–Rbx1–Skp1-F boxSkp2 SCF ubiquitin ligase complex (Protein Data Bank ID code: 1LDK) and a FBW7–SKP1–Cyclin E1 complex (Protein Data Bank ID code: 2OVR) by structural alignment with SKP1. The SFN/SKP1/Cul1 complex was constructed using a predicted model of SFN/SKP1 in this work and a Cul1–Rbx1–Skp1–F boxSkp2 SCF ubiquitin ligase complex (Protein Data Bank ID code: 1LDK). Filled yellow circle indicates the area of structural overlap between the F-box and SFN (c). B, Western blotting using A549 cells transfected with 3 kinds of siSFN (1–3). Substrate proteins of FBW7 such as p-cyclin E1, p-c-Myc, p-c-Jun, and NICD (Notch intracellular domain) were suppressed after knockdown of SFN, although p-p27 (substrate protein of SKP2) and p-IκBα (substrate protein of β-TrCP1) showed no change in expression. C, Ubiquitination assay for substrate proteins of FBW7. A549 cells were transfected with siSFN for 24 hours followed by Ub-HA for 24 hours. The cells were then treated with 10 μmol/L MG132, a proteasome inhibitor, for 6 hours at 37°C in a 5% CO2 incubator. The total protein was extracted using IP Lysis Buffer containing 10 mmol/L N-ethylmaleimide (NEM), a deubiquitinating enzyme inhibitor. Ubiquitinated proteins in the cell lysate were collected by co-IP and then subjected to Western blotting. Suppression of SFN induced marked ubiquitination of p-cyclin E1, p-c-Myc, p-c-Jun, and NICD.

Close modal

We had previously found that depletion of SFN decreased cell growth in vitro and tumor progression or development in vivo, suggesting that SFN might be a promising therapeutic target for lung adenocarcinoma (12). Here, we revealed that SFN enhances tumor progression through binding with SKP1 and blocking of oncoprotein ubiquitination. Thus, it is also expected that inhibition of SFN–SKP1 binding might have antitumor efficacy through recovery of normal ubiquitination and degradation of oncoproteins.

Aprepitant and ticagrelor are potential SFN inhibitor candidates

To identify potential SFN inhibitor(s), we focused on preexisting low-molecular-weight drugs, because SFN is an intracellular molecule. In silico simulation revealed a druggable pocket (PLB score 2.86, exceeding the druggable threshold of 2.0) around the SKP1 binding site in SFN protein (Fig. 4A). In order to identify compounds that can specifically bind to the pocket and inhibit SFN-SKP1 binding, we performed in silico screening using DrugBank (https://www.drugbank.ca/) and the Namiki bioactive compounds library (Namiki Shoji Co., Ltd). From a total of 7,133 drugs, we extracted 5,597 whose molecular weights were 200 to 800 and subjected them to in silico screening. Among 5,597 drugs, we obtained 46 available compounds that had high ranking (Supplementary Table S3). We then carried out experimental validation of these 46 compounds using the WST-8 proliferation assay and IP/Western botting to analyze SFN–SKP1 dissociation using lung adenocarcinoma cells after treatment with each compound. Among these compounds, 4 (aprepitant, ticagrelor, ezetimibe, and chlorhexidine) reduced both cell proliferation and SFN–SKP1 binding in a dose-dependent manner and were therefore subjected to further analysis (Fig. 4B and C). Moreover, in silico analysis indicated that the patterns of aprepitant and ticagrelor binding to the SFN druggable pocket were similar, broadly blocking the SKP1-binding surface on the SFN protein (Fig. 4Da–b). However, ezetimibe lacked a polar functional group and chlorhexidine had only few interactions with the pocket (Fig. 4Dc–d). Therefore, we considered aprepitant and ticagrelor to be the final SFN inhibitor candidates. Even though the original targets of these 2 compounds are completely different, they use common pharmacophores to link with the druggable pocket on the SFN protein (Fig. 4De). Additionally, these 2 compounds were confirmed to suppress cell growth in a dose-dependent manner in another 4 lung adenocarcinoma cell lines (Fig. 4E). However, importantly, Calu-6 (large cell lung carcinoma cells) and PL16B (immortalized bronchial epithelial cells), which have no SFN expression, did not reduce cell number even after administration of aprepitant or ticagrelor (Fig. 4E), indicating that their effects are mediated by inhibition of SFN.

Figure 4.

Aprepitant and ticagrelor as candidate SFN inhibitors. A, Putative druggable sites at the interaction interface of the SFN–SKP1 complex from protein–protein docking and small dummy atoms (red = hydrophilic, white = hydrophobic) determined using the SiteFinder program of MOE (Chemical Computing Group ULC). The PLB score at this site was 2.86. B, IP with SFN antibodies after treatment with each compound (10 μmol/L). #1, aprepitant; #2, ticagrelor; #3, ezetimibe; #4, chlorhexidine. All 4 compounds inhibited the binding of SFN to SKP1. As a negative control, cells treated with DMSO were used. C, WST-8 cell proliferation assay after treatment with each compound. As a negative control, cells treated with DMSO were used. Semi-log graph was used to show cell viability. All 4 compounds reduced cell growth dose-dependently. D, Predicted interactions between 4 hit ligands and the SFN druggable pocket (a)–(d) and a pharmacophore model for aprepitant and ticagrelor on the SFN druggable pocket (e). Circles with dashed lines in (c) and (d) indicate the interaction gaps between the ligands (ezetimibe and chlorhexidine) and key residues of the SFN druggable pocket. Three-feature pharmacophore model generated for aprepitant and ticagrelor using MOE: hydrogen bond acceptor (cyan feature), hydrophobic region (dark green feature), and hydrophobic region or aromatic ring center (green feature; e). E, WST-8 cell proliferation assay after treatment with aprepitant or ticagrelor. PC9, H1975, HCC827, PL16T, PL16B, and Calu-6 were used. As a negative control, cells treated with DMSO were used. Semi-log graph was used to show cell viability. In accordance with the previous pharmacokinetics reports on aprepitant and ticagrelor, which indicated that their degree of plasma protein binding is 99.6% and 99.9%, respectively, we calculated the unbound drug concentration from the loading concentration of inhibitors. Both aprepitant and ticagrelor reduced the cell growth dose-dependently in PC9, H1975, HCC827, and PL16T. However, Calu-6 or PL16B cells, which have no SFN expression, did not reduce cell number.

Figure 4.

Aprepitant and ticagrelor as candidate SFN inhibitors. A, Putative druggable sites at the interaction interface of the SFN–SKP1 complex from protein–protein docking and small dummy atoms (red = hydrophilic, white = hydrophobic) determined using the SiteFinder program of MOE (Chemical Computing Group ULC). The PLB score at this site was 2.86. B, IP with SFN antibodies after treatment with each compound (10 μmol/L). #1, aprepitant; #2, ticagrelor; #3, ezetimibe; #4, chlorhexidine. All 4 compounds inhibited the binding of SFN to SKP1. As a negative control, cells treated with DMSO were used. C, WST-8 cell proliferation assay after treatment with each compound. As a negative control, cells treated with DMSO were used. Semi-log graph was used to show cell viability. All 4 compounds reduced cell growth dose-dependently. D, Predicted interactions between 4 hit ligands and the SFN druggable pocket (a)–(d) and a pharmacophore model for aprepitant and ticagrelor on the SFN druggable pocket (e). Circles with dashed lines in (c) and (d) indicate the interaction gaps between the ligands (ezetimibe and chlorhexidine) and key residues of the SFN druggable pocket. Three-feature pharmacophore model generated for aprepitant and ticagrelor using MOE: hydrogen bond acceptor (cyan feature), hydrophobic region (dark green feature), and hydrophobic region or aromatic ring center (green feature; e). E, WST-8 cell proliferation assay after treatment with aprepitant or ticagrelor. PC9, H1975, HCC827, PL16T, PL16B, and Calu-6 were used. As a negative control, cells treated with DMSO were used. Semi-log graph was used to show cell viability. In accordance with the previous pharmacokinetics reports on aprepitant and ticagrelor, which indicated that their degree of plasma protein binding is 99.6% and 99.9%, respectively, we calculated the unbound drug concentration from the loading concentration of inhibitors. Both aprepitant and ticagrelor reduced the cell growth dose-dependently in PC9, H1975, HCC827, and PL16T. However, Calu-6 or PL16B cells, which have no SFN expression, did not reduce cell number.

Close modal

Next, the in vivo antitumor effect was evaluated using 2 types of experimental schedule in a tumor-bearing mouse model. First, we subcutaneously injected A549 cells into nude mice (day 0) and started daily oral administration of each compound from the following day (day 1) to mimic the effect of adjuvant chemotherapy. Surprisingly, aprepitant completely blocked tumor formation, and ticagrelor markedly reduced tumor progression (Supplementary Fig. S5). These results are consistent with our in silico data indicating that aprepitant and ticagrelor broadly blocked SKP1 binding surface and were more appropriate candidates than the others.

For the second in vivo validation procedure, we subcutaneously injected A549 cells into nude mice. The mice were randomized into 7 groups when tumors had attained a volume of 100 mm3 (day 1) and started on daily oral administration of each compound at 3 different doses for 21 days. It was found that both aprepitant and ticagrelor significantly suppressed tumor progression in a dose-dependent manner (Fig. 5A and B). Histologically, the resected tumors were poorly differentiated adenocarcinomas with a medullary appearance. Notably, tumors that developed in the mice treated with ticagrelor had acellular spaces, and those administered with aprepitant also had spaces filled with lymphocytes but not tumor cells, apparently attributable to regression of tumor cells after treatment. Moreover, the tumor cells surrounding these spaces showed lower expression of Ki67, and even SFN, in comparison with those from mice treated with vehicle alone (Fig. 5C). This suggested that aprepitant and ticagrelor may exert an antitumor effect by reducing tumor cell proliferation.

Figure 5.

Aprepitant and ticagrelor suppressed tumor progression in vivo. A and B, Nude mice were given a subcutaneous injection of A549 cells and randomly divided into groups, each comprising 8 individuals. Three different doses of aprepitant or ticagrelor dissolved into 5% DMSO + 95% corn oil were administered orally every day when the tumors had attained a volume of 100 mm3. 5% DMSO + 95% corn oil was administered in the vehicle group. Tumor volume was measured with a caliper at the indicated time points. Both aprepitant and ticagrelor decreased tumor growth dose-dependently. C, HE and IHC for SFN and Ki-67 were conducted using the collected subcutaneous tumors. Red arrows indicate spaces filled with only lymphocytes or acellular spaces inside the tumors. Green lines indicate the tumor cells surrounded by acellular spaces. *P < 0.05, **P < 0.01.

Figure 5.

Aprepitant and ticagrelor suppressed tumor progression in vivo. A and B, Nude mice were given a subcutaneous injection of A549 cells and randomly divided into groups, each comprising 8 individuals. Three different doses of aprepitant or ticagrelor dissolved into 5% DMSO + 95% corn oil were administered orally every day when the tumors had attained a volume of 100 mm3. 5% DMSO + 95% corn oil was administered in the vehicle group. Tumor volume was measured with a caliper at the indicated time points. Both aprepitant and ticagrelor decreased tumor growth dose-dependently. C, HE and IHC for SFN and Ki-67 were conducted using the collected subcutaneous tumors. Red arrows indicate spaces filled with only lymphocytes or acellular spaces inside the tumors. Green lines indicate the tumor cells surrounded by acellular spaces. *P < 0.05, **P < 0.01.

Close modal

Using the collected subcutaneous tumors, we extracted the total proteins and subjected them to IP with SKP1 antibodies and subsequent Western blotting for SFN, and vice versa. The tumors treated with aprepitant or ticagrelor showed reduced association of SFN with SKP1 (Supplementary Fig. S6A). Additionally, Western blotting indicated that expression of p-cyclin E1, p-c-Jun, p-c-Myc, and NICD was lower in tumors from mice treated with aprepitant or ticagrelor than in those from vehicle-treated mice (Supplementary Fig. S6B), suggesting that they may exert its antitumor effect through inhibition of SFN–SKP1 binding, leading to ubiquitination and degradation of oncoproteins. Because those 2 compounds suppressed cell growth even in PL16T cells (Fig. 4E), which had been established from AIS, we considered that aprepitant and ticagrelor might show antitumor efficacy against not only advanced, but also early-stage lung adenocarcinoma.

In the present study, we revealed that SFN bound with SKP1 and inhibited formation of the SCFFBW7 complex, allowing several oncoproteins to evade ubiquitination. Our data showed that SKP1 bound only with SFN and not with other 14-3-3 family members (Supplementary Fig. S1B). This might be because SFN has unique amino acids (Met202, Asp204, and His206) that may be responsible for binding to particular ligands that are not recognized by other 14-3-3 members (31). Because Lys49, Arg56, Arg129, and Tyr130 in the SFN protein are reportedly residues important for binding with phosphorylated binding partners, and we confirmed that SKP1 also used the same interface for binding with SFN (Supplementary Fig. S2E and S2F), we had speculated that SKP1 might require phosphorylation for binding to SFN. Based on the previous phosphoproteomics study that had revealed phosphorylated threonine at residue 131 of SKP1 (32), we clarified that the Thr131 residue in SKP1 protein was indeed phosphorylated and acted as a binding site for SFN (Supplementary Fig. S2B–S2D), suggesting that the binding of SKP1 to SFN might be phosphorylation dependent.

14-3-3 proteins are known to be anchor proteins that alter the stability, activity, or subcellular localization of binding partners (33). We also found that SFN reduced the stability of SKP1 protein and accelerated its ubiquitination (Fig. 1A and B). Based on these data, we expected that interaction with SFN might induce a conformational change in the SKP1 protein structure, leading to rapid ubiquitination and degradation. Our data also indicated that SFN binding to SKP1 altered the latter's subcellular localization (Fig. 1C and D; Supplementary Fig. S3D). SFN has been reported to bind with constitutive photomorphogenic 1 (COP1), a p53-specific E3 ubiquitin ligase involved in regulating the level of p53 protein, resulting in redistribution of COP1 to the cytoplasm in colon cancer cells (34). It is assumed that SFN similarly binds with newly-synthesized SKP1 proteins in the cytoplasm and blocks the translocation of SKP1 to the nucleus, thus preventing formation of the SCF complex. We expect that SFN might have particular function to regulate the activity or subcellular localization of various E3 ubiquitin ligases.

Our present IHC analysis showed that cytoplasmic SKP1 expression was significantly associated with tumor malignancy or poor prognosis (Supplementary Table S1; Fig. 2). Although the physiologic location of SKP1 is the nucleus, cytoplasmic SKP1 was shown to be more associated with tumor progression. From our in vitro results, we speculate that SKP1 expression in the cytoplasm might result from sequestration of SKP1 by SFN. These IHC results might allow our in vitro findings to be generalized to human tumor tissue. As we have previously reported that expression of SFN is related to poor prognosis (13), SKP1 expression in the cytoplasm might also serve as a prognostic marker for lung adenocarcinoma.

SFN appears to compete with only FBW7, and not SKP2 or β-TrCP1, for binding with SKP1 protein (Fig. 3; Supplementary Fig. S4). There is increasing evidence to suggest that cyclin E1, one of the FBW7 target proteins, is physiologically activated in late G1 phase to facilitate G1/S transition, and degraded in S-phase by SCFFBW7. On the other hand, p27, the target substrate of SKP2, is known to be degraded in late G1 phase by SCFSKP2. These findings indicate that each E3 ubiquitin ligase may function at a distinct time point in the cell cycle. Here, we showed that SFN selectively inhibited SCFFBW7 by competing with FBW7, suggesting that SFN might bind with SKP1 in S-phase but not in G1 phase. In support of this, a previous report has indicated that SFN enhances G1/S progression through upregulation of cyclin D1 in breast cancer (7). On the other hand, the specificity of SFN for FBW7 might also be related to the difference in localization of each F-box protein. Although we showed that FBW7 was located in the nucleus of lung adenocarcinoma cells (Supplementary Fig. S4B), SKP2 and β-TrCP1 function mainly in the cytoplasm. As mentioned above, because SFN is thought to bind with newly-synthesized SKP1 protein in the cytoplasm and inhibit its nuclear transport, formation of the SCFFBW7 complex in the nucleus might be blocked more efficiently than that of the SCFSKP2 or SCFβ-TrCP1 complex in the cytoplasm.

Recently, Reiterer and colleagues reported that psuedophosphatase STYX suppressed SCFFBW7 function by competing with SKP1 for binding to the F-box motif of FBW7, resulting in upregulation of FBW7 substrates including c-Myc, and cyclin E1 in HeLa cells (35). Additionally, various upstream factors such as USP29, ERK and PLK1, which induce deubiquitination or autoubiquitination of FBW7, have been reported to modulate the stability and activity of FBW7 (36). However, because indirect regulation of FBW7 through blocking of SCFFBW7 complex formation by binding with SKP1 has not yet been confirmed, our findings provide insight into a new mechanism for regulation of FBW7.

In our recent report, we identified an SFN-binding partner other than SKP1, ubiquitin-specific protease 8 (USP8) (13). USP8 is a deubiquitination enzyme specific for receptor tyrosine kinases (RTK) including EGFR and MET, contributing to the proliferative activity of many human cancers including NSCLC. We found that SFN stabilizes RTKs through activation of USP8 in lung adenocarcinoma. Because 14-3-3 proteins are anchor proteins that can bind with various molecules, it is natural that SFN would have multiple binding partners. Based on these facts, inhibition of SFN is thought to lead to ubiquitination of not only oncoproteins but also RTKs. Because SFN is related to multiple pathways, we believe that inhibition of SFN would simultaneously block several pathways that contribute to lung adenocarcinogenesis, thus exerting high antitumor efficacy.

Our results indicate that inhibitors of SFN–SKP1 binding could be potential targeted drugs for treatment of lung adenocarcinoma. In order to screen potent inhibitors, we adopted an in silico approach utilizing structural data for protein–protein interaction (PPI) targets and a database of known bioactive compounds, a strategy known as “in silico drug repositioning.” This drug-repositioning strategy makes it possible to rationalize experimental protocols, prioritize the design of therapeutic compounds for PPIs, and speed up the development of promising drugs, while reducing the risk of failure in clinical trials. Although the structure of the SFN–SKP1 complex has not been clarified experimentally, a reliable computational model of the complex, which was consistent with the results of in vitro mutational experiments, and in silico detection of the druggable pockets overlapping at the SFN–SKP1 interaction interface led to the successful identification of 2 independent drugs for this particular PPI: aprepitant and ticagrelor.

Aprepitant is an NK-1R (neurokinin-1 receptor) antagonist already approved as an antiemetic drug for cancer patients who need to take antitumor drugs that cause nausea or emesis. Aprepitant has already been reported to suppress cell growth of hepatoblastoma through inhibition of NK-1R (37). However, we consider that the antitumor effect of aprepitant on lung adenocarcinoma cells may not be due to inhibition of NK-1R binding but rather that of SFN–SKP1 binding, because lung adenocarcinoma cells do not express NK-1R (data not shown). It is thought that aprepitant is a potential SFN inhibitor that could be used for treatment for lung adenocarcinoma. On the other hand, ticagrelor is a reversible P2Y12 inhibitor exerting an antiplatelet effect and is used clinically with aspirin for treatment of cardiovascular and cerebrovascular diseases. A phase III clinical study of ticagrelor showed that 9.8% of patients who took daily ticagrelor and aspirin together suffered a bleeding event within 12 months. Although we confirmed that mice given a single dose of ticagrelor showed a lower antiplatelet effect than was the case for coadministration of ticagrelor and aspirin in terms of bleeding time and platelet aggregation (Supplementary Fig. S7), we will need to carefully verify the safety of ticagrelor as an antitumor drug.

Previous studies have shown that when aprepitant (MW 534.43) was administered to humans as a 125-mg capsule, the Cmax was 1,539 ng/mL (38), and the protein binding rate was 99.6% (39), indicating that the plasma unbound drug concentration was approximately 11.52 nmol/L. Using the reported plasma Cmax of aprepitant (3,535 ng/mL) when administered at 30 mg/kg to mice (39), we estimate that the plasma unbound drug concentration of aprepitant after administration at 20 mg/kg to mice (Fig. 5A) could be approximately 17.64 nmol/L, which is similar to that in humans (11.52 nmol/L). On the other hand, the Cmax of ticagrelor (MW 522.57) when administered to humans at a single oral dose of 90 mg was reported to be 581 ng/mL, with a protein binding rate of 99.9% (40), suggesting that the plasma unbound drug concentration would be approximately 1.11 nmol/L. Because administration of ticagrelor at 180 mg/kg to mice was reported to result in a plasma drug concentration of 0.99 μmol/L (41), the dose of ticagrelor we used (100 mg/kg, Fig. 5B) would be expected to achieve a plasma unbound drug concentration of 0.56 nmol/L. This is lower than the aforementioned unbound drug concentration in humans, suggesting that the dose would be tolerated in humans.

Many targeted drugs for treatment of lung adenocarcinoma, such as EGFR–TKI, ALK inhibitors, and BRAF inhibitors, are available. Unfortunately, however, these drugs do not significantly improve the patient survival rate. In the present study, we focused on epigenetic changes in early-stage lung adenocarcinoma to devise a new treatment strategy and were able to confirm that SFN–SKP1 binding is one of the key events involved in the malignant progression of such cancers. As a result, 2 potentially promising candidate drugs, aprepitant and ticagrelor, were identified. Recently, it has been confirmed that low-dose CT screening is able to reduce the death rate of patients with early lung adenocarcinomas. Because this strategy is now being promoted in many countries, large numbers of small nodules, referred to as ground glass nodules (GGN), are being detected. Although GGNs can be surgically resected, this can be dangerous or sometimes impossible in patients with multiple GGNs, or in those who are elderly or complicated by chronic interstitial lung disease. Inhibitors of SFN–SKP1 binding such as aprepitant and ticagrelor are thought to be desirable candidates for the treatment of early-stage lung adenocarcinoma. Because SFN is widely associated with tumor progression from an early stage, these drugs can be applied without companion diagnosis, and used together with other antitumor drugs such as immune-checkpoint inhibitors, even for neoadjuvant chemotherapy. Not only aprepitant and ticagrelor, but other small molecules might be effective for inhibition of SFN–SKP1 binding, and efforts to identify new inhibitors with better potency would be justified.

In conclusion, we have revealed that SFN binds with SKP1 and inhibits the formation of SCFFBW7, resulting in stabilization of oncoproteins (Fig. 6). Aprepitant and ticagrelor were identified as candidate SFN inhibitors for suppression of abnormal SFN–SKP1 binding and tumor progression. Our present findings suggest that these inhibitors may provide a novel chemotherapeutic strategy for early-stage lung adenocarcinoma.

Figure 6.

Molecular mechanism at a glance. SFN binds with SKP1 to sequester it in the cytoplasm, resulting in the obstruction of SCFFBW7 complex formation. Oncoproteins including cyclin E1 and c-Jun subsequently evade from ubiquitination and proteasomal degradation and are stabilized due to lack of their specific E3 ubiquitin ligase.

Figure 6.

Molecular mechanism at a glance. SFN binds with SKP1 to sequester it in the cytoplasm, resulting in the obstruction of SCFFBW7 complex formation. Oncoproteins including cyclin E1 and c-Jun subsequently evade from ubiquitination and proteasomal degradation and are stabilized due to lack of their specific E3 ubiquitin ligase.

Close modal

No potential conflicts of interest were disclosed.

Conception and design: A. Shiba-Ishii, M. Noguchi

Development of methodology: A. Shiba-Ishii, J. Hong, T. Hirokawa

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): A. Shiba-Ishii, J. Hong, S. Sakashita, Y. Kozuma, Y. Sato

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A. Shiba-Ishii, J. Hong, T. Hirokawa

Writing, review, and/or revision of the manuscript: A. Shiba-Ishii, J. Hong, T. Hirokawa, M. Noguchi

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Y. Kim, T. Nakagawa, S. Sakashita, N. Sakamoto

Study supervision: M. Noguchi

We sincerely express our appreciation to Dr. Hidehito Horinouchi, Dr. Shun-ichi Watanabe, and Dr. Masahiko Kusumoto (National Cancer Center Hospital) for their suggestion and support from the clinical point of view. Moreover, we are grateful to Professor Kenji Irie, Dr. Hiroyuki Suzuki, and Dr. Yuji Funakoshi (Faculty of Medicine, University of Tsukuba) for research support and kind advice. This work was supported by JSPS KAKENHI grant number 17K15634 (T. Hirokawa), the Platform Project for Supporting Drug Discovery and Life Science Research [Basis for Supporting Innovative Drug Discovery and Life Science Research (BINDS)] from AMED under grant number JP18am0101114 (T. Hirokawa), and MEXT as “Priority Issue on Post-K computer” (Building Innovation Drug Discovery Infrastructure Through Functional Control of Biomolecular Systems; hp160213; T. Hirokawa).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Noguchi
M
,
Morikawa
A
,
Kawasaki
M
,
Matsuno
Y
,
Yamada
T
,
Hirohashi
S
, et al
Small adenocarcinoma of the lung. Histologic characteristics and prognosis
.
Cancer
1995
;
75
:
2844
52
.
2.
Shiba-Ishii
A
,
Kano
J
,
Morishita
Y
,
Sato
Y
,
Minami
Y
,
Noguchi
M
. 
High expression of stratifin is a universal abnormality during the course of malignant progression of early-stage lung adenocarcinoma
.
Int J Cancer
2011
;
129
:
2445
53
.
3.
Shiba-Ishii
A
,
Noguchi
M
. 
Aberrant stratifin overexpression is regulated by tumor-associated CpG demethylation in lung adenocarcinoma
.
Am J Pathol
2012
;
180
:
1653
62
.
4.
Hermeking
H
,
Lengauer
C
,
Polyak
K
,
He
TC
,
Zhang
L
,
Thiagalingam
S
, et al
14-3-3 sigma is a p53-regulated inhibitor of G2–M progression
.
Mol Cell
1997
;
1
:
3
11
.
5.
Chan
TA
,
Hermeking
H
,
Lengauer
C
,
Kinzler
KW
,
Vogelstein
B
. 
14-3-3Sigma is required to prevent mitotic catastrophe after DNA damage
.
Nature
1999
;
401
:
616
20
.
6.
Phan
L
,
Chou
PC
,
Velazquez-Torres
G
,
Samudio
I
,
Parreno
K
,
Huang
Y
, et al
The cell cycle regulator 14-3-3sigma opposes and reverses cancer metabolic reprogramming
.
Nat Commun
2015
;
6
:
7530
.
7.
Zhang
Y
,
Karas
M
,
Zhao
H
,
Yakar
S
,
LeRoith
D
. 
14-3-3sigma mediation of cell cycle progression is p53-independent in response to insulin-like growth factor-I receptor activation
.
J Biol Chem
2004
;
279
:
34353
60
.
8.
Ko
S
,
Kim
JY
,
Jeong
J
,
Lee
JE
,
Yang
WI
,
Jung
WH
. 
The role and regulatory mechanism of 14-3-3 sigma in human breast cancer
.
J Breast Cancer
2014
;
17
:
207
18
.
9.
Zhang
Y
,
Li
Y
,
Lin
C
,
Ding
J
,
Liao
G
,
Tang
B
. 
Aberrant upregulation of 14-3-3sigma and EZH2 expression serves as an inferior prognostic biomarker for hepatocellular carcinoma
.
PLoS One
2014
;
9
:
e107251
.
10.
Perathoner
A
,
Pirkebner
D
,
Brandacher
G
,
Spizzo
G
,
Stadlmann
S
,
Obrist
P
, et al
14-3-3sigma expression is an independent prognostic parameter for poor survival in colorectal carcinoma patients
.
Clin Cancer Res
2005
;
11
:
3274
9
.
11.
Nakayama
KI
,
Nakayama
K
. 
Ubiquitin ligases: cell-cycle control and cancer
.
Nat Rev Cancer
2006
;
6
:
369
81
.
12.
Shiba-Ishii
A
,
Kim
Y
,
Shiozawa
T
,
Iyama
S
,
Satomi
K
,
Kano
J
, et al
Stratifin accelerates progression of lung adenocarcinoma at an early stage
.
Mol Cancer
2015
;
14
:
142
.
13.
Kim
Y
,
Shiba-Ishii
A
,
Nakagawa
T
,
Iemura
SI
,
Natsume
T
,
Nakano
N
, et al
Stratifin regulates stabilization of receptor tyrosine kinases via interaction with ubiquitin-specific protease 8 in lung adenocarcinoma
.
Oncogene
2018
.
doi 10.1038/s41388-018-0342-9
.
14.
Lyapina
SA
,
Correll
CC
,
Kipreos
ET
,
Deshaies
RJ
. 
Human CUL1 forms an evolutionarily conserved ubiquitin ligase complex (SCF) with SKP1 and an F-box protein
.
Proc Natl Acad Sci U S A
1998
;
95
:
7451
6
.
15.
Hussain
M
,
Lu
Y
,
Liu
YQ
,
Su
K
,
Zhang
J
,
Liu
J
, et al
Skp1: Implications in cancer and SCF-oriented anti-cancer drug discovery
.
Pharmacol Res
2016
;
111
:
34
42
.
16.
Clurman
BE
,
Sheaff
RJ
,
Thress
K
,
Groudine
M
,
Roberts
JM
. 
Turnover of cyclin E by the ubiquitin-proteasome pathway is regulated by cdk2 binding and cyclin phosphorylation
.
Genes Dev
1996
;
10
:
1979
90
.
17.
Bonetti
P
,
Davoli
T
,
Sironi
C
,
Amati
B
,
Pelicci
PG
,
Colombo
E
. 
Nucleophosmin and its AML-associated mutant regulate c-Myc turnover through Fbw7 gamma
.
J Cell Biol
2008
;
182
:
19
26
.
18.
Nateri
AS
,
Riera-Sans
L
,
Da Costa
C
,
Behrens
A
. 
The ubiquitin ligase SCFFbw7 antagonizes apoptotic JNK signaling
.
Science
2004
;
303
:
1374
8
.
19.
Fryer
CJ
,
White
JB
,
Jones
KA
. 
Mastermind recruits CycC:CDK8 to phosphorylate the Notch ICD and coordinate activation with turnover
.
Mol Cell
2004
;
16
:
509
20
.
20.
Koepp
DM
,
Schaefer
LK
,
Ye
X
,
Keyomarsi
K
,
Chu
C
,
Harper
JW
, et al
Phosphorylation-dependent ubiquitination of cyclin E by the SCFFbw7 ubiquitin ligase
.
Science
2001
;
294
:
173
7
.
21.
Cancer Genome Atlas N
. 
Comprehensive molecular characterization of human colon and rectal cancer
.
Nature
2012
;
487
:
330
7
.
22.
Shimada
A
,
Kano
J
,
Ishiyama
T
,
Okubo
C
,
Iijima
T
,
Morishita
Y
, et al
Establishment of an immortalized cell line from a precancerous lesion of lung adenocarcinoma, and genes highly expressed in the early stages of lung adenocarcinoma development
.
Cancer Sci
2005
;
96
:
668
75
.
23.
Travis
WD
,
Brambilla
E
,
Nicholson
AG
,
Yatabe
Y
,
Austin
JHM
,
Beasley
MB
, et al
The 2015 World Health Organization classification of lung tumors: impact of genetic, clinical and radiologic advances since the 2004 classification
.
J Thorac Oncol
2015
;
10
:
1243
60
.
24.
Travis
WD
,
Brambilla
E
,
Noguchi
M
,
Nicholson
AG
,
Geisinger
KR
,
Yatabe
Y
, et al
International association for the study of Lung Cancer/American Thoracic Society/European Respiratory Society international multidisciplinary classification of lung adenocarcinoma
.
J Thorac Oncol
2011
;
6
:
244
85
.
25.
Sastry
GM
,
Adzhigirey
M
,
Day
T
,
Annabhimoju
R
,
Sherman
W
. 
Protein and ligand preparation: parameters, protocols, and influence on virtual screening enrichments
.
J Comput Aided Mol Des
2013
;
27
:
221
34
.
26.
Comeau
SR
,
Gatchell
DW
,
Vajda
S
,
Camacho
CJ
. 
ClusPro: an automated docking and discrimination method for the prediction of protein complexes
.
Bioinformatics
2004
;
20
:
45
50
.
27.
Soga
S
,
Shirai
H
,
Kobori
M
,
Hirayama
N
. 
Identification of the druggable concavity in homology models using the PLB index
.
J Chem Inf Model
2007
;
47
:
2287
92
.
28.
Wishart
DS
,
Feunang
YD
,
Guo
AC
,
Lo
EJ
,
Marcu
A
,
Grant
JR
, et al
DrugBank 5.0: a major update to the DrugBank database for 2018
.
Nucleic Acids Res
2018
;
46
:
D1074
D82
.
29.
Friesner
RA
,
Banks
JL
,
Murphy
RB
,
Halgren
TA
,
Klicic
JJ
,
Mainz
DT
, et al
Glide: a new approach for rapid, accurate docking and scoring. 1. Method and assessment of docking accuracy
.
J Med Chem
2004
;
47
:
1739
49
.
30.
Halgren
TA
,
Murphy
RB
,
Friesner
RA
,
Beard
HS
,
Frye
LL
,
Pollard
WT
, et al
Glide: a new approach for rapid, accurate docking and scoring. 2. Enrichment factors in database screening
.
J Med Chem
2004
;
47
:
1750
9
.
31.
Benzinger
A
,
Popowicz
GM
,
Joy
JK
,
Majumdar
S
,
Holak
TA
,
Hermeking
H
. 
The crystal structure of the non-liganded 14-3-3sigma protein: insights into determinants of isoform specific ligand binding and dimerization
.
Cell Res
2005
;
15
:
219
27
.
32.
Olsen
JV
,
Vermeulen
M
,
Santamaria
A
,
Kumar
C
,
Miller
ML
,
Jensen
LJ
, et al
Quantitative phosphoproteomics reveals widespread full phosphorylation site occupancy during mitosis
.
Sci Signal
2010
;
3
:
ra3
.
33.
Hermeking
H
. 
The 14-3-3 cancer connection
.
Nat Rev Cancer
2003
;
3
:
931
43
.
34.
Su
CH
,
Zhao
R
,
Velazquez-Torres
G
,
Chen
J
,
Gully
C
,
Yeung
SC
, et al
Nuclear export regulation of COP1 by 14-3-3sigma in response to DNA damage
.
Mol Cancer
2010
;
9
:
243
.
35.
Reiterer
V
,
Figueras-Puig
C
,
Le Guerroue
F
,
Confalonieri
S
,
Vecchi
M
,
Jalapothu
D
, et al
The pseudophosphatase STYX targets the F-box of FBXW7 and inhibits SCFFBXW7 function
.
EMBO J
2017
;
36
:
260
73
.
36.
Shimizu
K
,
Nihira
NT
,
Inuzuka
H
,
Wei
W
. 
Physiological functions of FBW7 in cancer and metabolism
.
Cell Signal
2018
;
46
:
15
22
.
37.
Berger
M
,
Neth
O
,
Ilmer
M
,
Garnier
A
,
Salinas-Martin
MV
,
de Agustin Asencio
JC
, et al
Hepatoblastoma cells express truncated neurokinin-1 receptor and can be growth inhibited by aprepitant in vitro and in vivo
.
J Hepatol
2014
;
60
:
985
94
.
38.
Majumdar
AK
,
Howard
L
,
Goldberg
MR
,
Hickey
L
,
Constanzer
M
,
Rothenberg
PL
, et al
Pharmacokinetics of aprepitant after single and multiple oral doses in healthy volunteers
.
J Clin Pharmacol
2006
;
46
:
291
300
.
39.
Endo
T
,
Saijo
T
,
Haneda
E
,
Maeda
J
,
Tokunaga
M
,
Zhang
MR
, et al
Quantification of central substance P receptor occupancy by aprepitant using small animal positron emission tomography
.
Int J Neuropsychopharmacol
2014
;
18
.
40.
Butler
K
,
Teng
R
. 
Pharmacokinetics, pharmacodynamics, and safety of ticagrelor in volunteers with mild hepatic impairment
.
J Clin Pharmacol
2011
;
51
:
978
87
.
41.
Preusch
MR
,
Rusnak
J
,
Staudacher
K
,
Mogler
C
,
Uhlmann
L
,
Sievers
P
, et al
Ticagrelor promotes atherosclerotic plaque stability in a mouse model of advanced atherosclerosis
.
Drug Des Devel Ther
2016
;
10
:
2691
9
.