Background
Ovarian cancer among the leading causes of cancer-related mortality in females in the world, with an estimated annual diagnosis of more than 200,000 cases and more than 150,000 deaths [
1]. The delayed onset symptoms, lack of diagnostic methods in the early stage, and tumor recurrence after chemotherapies all contribute to the high mortality rate and poor clinical management of patients with ovarian cancer [
2,
3]. The type 1 neoplasms include clear, mucinous, transitional cell, and low-grade serous ovarian carcinoma, which are typically developed and progressed in a step-wise manner [
4]. Whereas, high-grade serous ovarian carcinoma (HGSOC), as the type 2 neoplasm, is rapidly developed and widely disseminated, which is associated with the poorest overall prognosis among patients with all types of ovarian cancer [
4,
5]. In fact, the aggressive HGSOC is the dominant subtype of clinically diagnosed ovarian cancer, and nearly 80% of the deaths caused by all types of ovarian cancer are attributed to HGSOC [
5,
6]. Moreover, HGSOC is rarely diagnosed at its early stage, and compared with all the other subtypes of ovarian cancer, it possesses extensively higher cell proliferation rate, lower rate of programmed cell death, and extremely rapid cell invasiveness and metastasis [
7,
8]. Understanding the molecular pathogenesis of HGSOC and developing efficient diagnostic/prognostic biomarkers are crucial steps for improving the long-term survival outcomes in patients with HGSOC by facilitating early diagnosis and proper and effective clinical management of the patients [
6,
9].
Suppressors of cytokine signaling (SOCS) is a class of proteins that involve in the regulation of signal transduction [
10,
11]. SOCS proteins can facilitate the proteasome-mediated degradation of activated cytokine-receptor complex, and therefore, further inhibit specific cytokine signaling and reduce the cell sensitivity towards other types of cytokines or even hormones [
12,
13]. The SOCS family is consisted of eight protein members [
14], among which, SOCS7 has been implicated in the regulation of downstream pathways of cytokines and the receptors of growth factors [
15]. Meanwhile, SOCS7 may also possess an intriguing role in regulating cell cycle and tumor suppression [
16]. It was reported that SOCS7 facilitates the nuclear translocation and accumulation of non-catalytic tyrosine kinase adaptor protein, which further induces DNA damage and cell cycle arrest [
17]. At the same time, SOCS7 is also able to directly interact with growth factor receptor-bound protein, subsequently regulating cellular proliferation and differentiation [
18,
19]. Previously, it was detected that SOCS7 is differentially expressed in benign and malignant breast tissues and has critical regulatory functions in insulin-like growth factor I- and hepatocyte growth factor-induced breast cancer [
20,
21], indicating its involvement in the pathogenesis of human cancers. However, the function of SOCS7 in the pathogenesis of ovarian cancer, especially in HGSOC, is still unknown.
Therefore, we designated to investigate the relationship between SOCS7 and HGSOC, the roles of SOCS7 in inhibiting HGSOC, as well as the underlying molecular mechanisms. Through correlation analyses based on bioinformatics and patient cohorts, we demonstrated that SOCS7 expression was correlated with the prognosis of HGSOC patients. Further, by establishing SOCS7-knockdown and SOCS7-overexpression cellular models of HGSOC and analyzing the interactions between its downstream proteins, we revealed the anti-HGSOC effects of SOCS7 and illustrated the underlying molecular mechanism. This study establishes the foundation for the development of SOCS7 as potential biomarker for HGSOC.
Materials and methods
The gene expression data were obtained from The Cancer Genome Atlas (TCGA) dataset for ovarian serous cystadenocarcinoma. Pathways that were substantially enriched in both SOCS7 high- and low- expression groups were identified by Gene set enrichment analysis (GSEA). The survival data were downloaded from the Kaplan Meier plotter website (
http://www.kmplot.com/), using the best available cut-off value, including the GSE9891 (probe ID: 214015_at) and GSE26193 (probe ID: 226572_at) datasets for HGSOC [Grades 2 and 3; International Federation of Gynecology and Obstetrics (FIGO) stages III and IV].
Two cohorts of patients with HGSOC (Grades 2 and 3; FIGO stages III and IV) who underwent surgery in Obstetrics and Gynecology Hospital of Fudan University from 2012 to 2021 were included for the study. None of these patients had received chemotherapy or radiotherapy before surgery. Cohort 1 includes fresh 30 tumor and paired normal tissues and cohort 2 includes paraffin-embedded 68 tumor and 15 normal tissues. Approval of this study was obtained from the Ethics Committee of Fudan University.
Immunohistochemistry (IHC) evaluation
Paraffin-embedded HGSOC tissue sections in cohort 2 were used for IHC studies. The IHC staining analysis was performed using anti-SOCS7 (bs-20151R), anti-HuR (bs-2651R), or anti-FOXM1 (bs-2687R; all from Bioss Antibodies Inc. Woburn, MA, USA) primary antibodies, followed by secondary antibody incubation (D-3004; Shanghai Long Island Biotechnology). Immunoreactivity was scored using the H-score system based on percentage of positively stained cells (0, < 5%; 1, 5%-25%; 2, 25%-50%; 3, 50%-75%; 4, > 75%) and staining intensity (0, negative; 1, weak; 2, moderate; 3, strong), which ranged from 0–12. HGSOC patients were divided into low expression (H-score < 4) or high-expression (H-score ≥ 4) group.
Immunofluorescence microscopy
HGSOC cell lines were fixed by using 4% formaldehyde and then permeabilized by using 0.5% Triton X-100 in PBS. Subsequently, they were blocked for 30 min by using 1% bovine serum albumin in PBS, and the cells were further incubated with anti-SOCS7 (ab224589; Abcam), anti-HuR (ab200342; Abcam), Alexa Fluor 555-labeled Donkey Anti-Mouse IgG (H + L) (A0460, Beyotime Biotechnology, Nanjing, China), or Alexa Fluor 488-labeled Goat Anti-Rabbit IgG (H + L) (A0423, Beyotime Biotechnology) antibody before nuclear staining with DAPI (C1002, Beyotime Biotechnology). The stained cells were visualized and examined by using laser scanning confocal microscopy (Leica Microsystems Inc., Buffalo Grove, IL, USA).
Cell culture
Human normal ovarian epithelial cell line IOSE80 and human HGSOC cell lines SNU119, CAOV3, OVCAR3, COV318, and OVCAR4 were obtained from the American Type Culture Collection (ATCC, Rockville, MD, USA) and cultured in 10% fetal bovine serum-supplemented RPMI-1640 medium in the presence of Penicillin–Streptomycin Solution (100X; SolarbioScience & Technology Co., Ltd., Shanghai, China) in 5% CO2 at 37 °C.
Cell transfection
SOCS7 and HuR shRNA and scrambled shRNA lentivirus were purchased from Applied Biological Materials Company (Vancouver, Canada). Lentivirus production and purification were conducted as we previously described [
22]. Full-length coding sequences of human HuR or FOXM1 genes were cloned into the expression vector pLVX-Puro (Addgen, Cambridge, MA, USA) for the construction of HuR or FOXM1 overexpression plasmids. Blank pLVX-Puro vector served as a negative control. Transfection was carried out based on Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) following the protocols from the manufacturer (Life Technologies Corporation, Carlsbad, CA, USA). Following the transfection, the HGSOC cells were incubated for 48 h and subsequently transduced with the recombined expression vectors.
Cell viability assay
Approximately 3 × 103 CAOV3, OVCAR3, or COV318 cells/well were seeded in separate 96-well plates and then transduced with indicated lentiviral plasmids. Cell viability was measured at different time points (0, 24, 48, and 72 h) post-transduction, by using CCK-8 (cell counting kit-8; SAB Biotech, College Park, MD, USA). The absorbance at 450 nm was measured using a microplate reader.
Analysis of cell cycle
Approximately 3 × 105 CAOV3, OVCAR3, or COV318 cells/well were seeded in separate 6-well plates and then transduced with indicated lentiviral plasmids. At 48 h post-transduction, the cells were collected and incubated with 1 mg/ml RNase A (Sigma) and 50 μg/ml propidium iodide (Sigma) for 30 min at 25 °C for cell cycle analysis. Subsequently, the fluorescence intensities were determined by Accuri C6 flow cytometry (BD Biosciences, San Diego, CA, USA).
Quantitative real-time polymerase chain reaction (qRT-PCR)
Total RNA was isolated from HGSOC tissues and cell lines by TRIzol reagent (Life Technologies, Grand Island, NY, USA), and reversely transcribed into cDNA by following the First Strand cDNA Synthesis Protocols (New England Biolabs, Ipswich, MA, USA). SYBR Green PCR Master Mix (Agilent Technologies, Santa Clara, CA, USA) was used for carrying out qRT-PCR, based on an ABI7300 System (Applied Biosystems, Carlsbad, CA, USA). The following primers were used for qRT-PCR: GAPDH-F:5ʹ-AATCCCATCACCATCTTC-3ʹ, GAPDH-R:5ʹ-AGGCTGTTGTCATACTTC-3ʹ; SOCS7-F:5ʹ-GGTTTGTGGCTTCCTGATG-3ʹ, SOCS7-R:5ʹ-GTGGGCTGTGTTTATGGTG-3ʹ; HuR-F:5ʹ-TCTGCGCTTGGCCTTAGTC-3ʹ, HuR-R:5ʹ-AACCGTCTTCGGGTGCTTC-3ʹ; FOXM1-F:5ʹ-AATGGCAAGGTCTCCTTC-3ʹ, FOXM1-R:5ʹ-AGCAGTGGCTTCATCTTC-3ʹ. The 2−ΔΔCT method was applied for determining the fold changes of target genes at the transcript level.
Western blot assay
Total protein was isolated and collected from HGSOC cell lines with radioimmunoprecipitation assay lysis buffer (Santa Cruz Biotech, Santa Cruz, CA, USA). The isolated proteins with equal quantity were further separated by sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS-PAGE; 10% or 12%), followed by transferring onto separate polyvinylidene fluoride membranes (Beijing Solarbio Science& Technology, Beijing, China). The antibodies used for immunoblotting were the following: GAPDH (Cell Signaling Technology, Beverly, MA, USA), SOCS7 (ab224589; Abcam, Cambridge, MA, USA), HuR (ab238528; Abcam), Cyclin D1 (ab16663; Abcam), Survivin (ab76424; Abcam), CDC25B (ab124819; Abcam), and FOXM1 (ab207298; Abcam). After incubating with horseradish peroxidase-conjugated secondary antibody (Beyotime Institute of Biotechnology), the signals were further examined based on an enhanced chemiluminescence system (Bio-Rad, Richmond, CA, USA).
Immunoprecipitation and liquid chromatography/mass spectrometry (LC/MS) analyses
Full-length coding sequence of human SOCS7 was cloned into a pCMV-FLAG vector. The 293 T cells expressing FLAG-tagged SOCS7 were constructed by transfection with Lipofectamine 2000 (Invitrogen) according to the instructions from the manufacturer and lysed with lysis buffer (1% Triton X-100, 150 mM NaCl, 20 mM Tris pH7.5, and 1 mM EDTA) supplemented with protease inhibitor cocktail (Sigma-Aldrich, MO, USA). The cell lysates were then incubated with anti-FLAG beads (Sigma-Aldrich) overnight incubated at 4 °C. After washing with lysis buffer five times, the immunoprecipitated protein complex was eluted by using FLAG peptide (Sigma-Aldrich) for 1 h at 4 °C. Protein samples were then resolved by SDS-PAGE for Coomassie Blue staining. Protein bands were excised from the gel and subjected to in-gel digestion with trypsin, and the resulting peptides were dissolved in a solution comprising 0.1% trifluoroacetic acid and 2% acetonitrile for analysis with Orbitrap Fusion mass spectrometer (Thermo Fisher Scientific, San Jose, CA) coupled with Easy nLC (Thermo Fisher Scientific, Waltham, MA, USA) as previously described. The raw mass files generated by the Orbitrap Fusion instrument were processed using Proteome Discoverer software (Thermo Scientific, version 1.4) integrated with the MASCOT (Matrix Science, London, UK; version 2.3.2) search engine for protein identification. Data were searched against the Human UniProtKB/Swiss-Prot database. Proteins and peptides with a false discovery rate less than 1% were selected.
Co-immunoprecipitation (co-IP) and ubiquitination assay
Cell lysates were prepared with lysis buffer (1% Triton X-100, 150 mM NaCl, 20 mM Tris pH7.5, and 1 mM EDTA) supplemented with protease inhibitor cocktail (Sigma-Aldrich), incubated with normal IgG (ab172730; Abcam), anti-HuR (ab200342; Abcam), or anti-SOCS7 (ab224589; Abcam) antibody, followed by further incubating with Protein A/G PLUS-Agarose beads (sc-2003; Santa Cruz Biotechnology) at 4 °C for 2 h. Then, the immunocomplex was three-time washed by the lysis buffer for Western blot analysis with antibodies against SOCS7 (ab224589; Abcam), HuR (ab200342; Abcam), and ubiquitin (ab7780; Abcam).
RNA immunoprecipitation (RIP) assay
Magna RNA immunoprecipitation (RIP) RNA-Binding Protein Immunoprecipitation kit (Millipore) was used for the RIP assay following the manufacturer’s instructions. Cells transduced with indicated lentiviral plasmids were prepared using RIP lysis buffer, the lysed cells were centrifuged at 12,000 × g for 10 min at 4˚C, and the RNA–protein complexes were incubated with anti-HuR (ab200342; Abcam) or anti-IgG antibody (ab172730; Abcam) at 4 °C for 1 h. Once incubation was complete, agarose beads and 50 µl of protein A/G were added and cells were incubated for a further 60 min at 4˚C. Subsequently, the precipitated beads were washed with RIP-wash buffer for 10 min at 4 °C and then RIP-lysis buffer for 5 min at 4 °C. The RNA in the immunoprecipitated complex and the RNA in the previously saved input fraction were released by incubating cells at 65˚C for 2 h with 200 mM NaCl and 20 µg proteinase K, which reversed the cross-linking. The amount of FOXM1 mRNA bound by HuR was determined by qRT-PCR.
Plasmid construction and dual-luciferase reporter assay
The wide-type and mutant FOXM1 3' UTR reporter plasmid was constructed by cloning PCR-amplified sequences from the 3' UTR of FOXM1 cDNA into a pGL3-basic firefly luciferase reporter plasmid (Promega, Madison, WI, USA). CAOV3 cells transduced with HuR-expressing vector were seeded into 24-well plates and transfected with 400 ng of FOXM1 3' UTR reporter plasmid (WT or Mutant). To normalize the transfection efficiency, cells were cotransfected with 50 ng of pRL-TK containing renilla luciferase. After 48 h, cells were washed with PBS and lysed using passive lysis buffer. Luciferase activity was assessed using a Dual-Luciferase Reporter Assay system (Promega) following the manufacturer’s instructions.
Measurement of mRNA stability
Cells transduced with the indicated lentiviral plasmids were incubated with 5 μg/ml transcription inhibitor actinomycin D (Sigma, St. Louis, MO, USA). Total RNA was isolated at time intervals of 0, 2, 4 and 6 h following actinomycin D addition. FOXM1 mRNA was determined using qRT-PCR, and the relative amount of FOXM1 mRNA at 0 h following actinomycin D treatment was set to 100%.
Tumor xenografts in animals
CAOV3 cells stably expressing pLVX-Puro-SOCS7 or blank pLVX-Puro were subcutaneously injected into male nude mice (6-week-old; n = 6 per group). At the 33rd day post-inoculation, tumors were collected, weighed, photographed, and analyzed by qRT-PCR, Western blot and immunofluorescence staining by using anti-Ki67 antibody (ab15580; Abcam). All the animal studies were approved by the Ethics Committee of Fudan University.
Statistical analyses
Data were presented in the form of ‘mean ± standard deviation’, and validated by three independent experiments or multiple independent mice trials, as indicated. Experimental results were analyzed with GraphPad Prism Software Version 8.4.2 (GraphPad Software Inc., La Jolla, CA, USA). Two-tailed Student's t-test and analysis of variance were applied for two groups comparisons and multiple groups comparisons, respectively. Wilcoxon test and nonparametric Kruskal–Wallis test were both utilized for evaluating the differences of tumor tissues and controls. Pearson's or Spearman's correlation was used for measuring association between expression levels of SOCS7/FOXM1, FOXM1/HuR or SOCS7/HuR. A statistical significance was considered when the P value was less than 0.05.
Discussion
SOCS family proteins inhibit related signaling pathways and regulate multiple cytokine responses and growth factor signals [
28]. The aberrant expression or activation of SOCS was correlated with the development and progression of various types of human tumors [
29]. In the current study, we observed that one of the SOCS family members, SOCS7, was downregulated in the tumor tissues in HGSOC patients. This finding was supported by previous report suggesting that the transcript level of SOCS7 was inversely correlated with the tumor, node, and metastasis (TNM) stage of cancer [
16]. Moreover, a recent study also illustrated that SOCS7 expression was connected with poor prognosis in ovarian cancer patients [
30], based on which we revealed the noticeable correlation between the expression of SOCS7 protein and survival rate or clinicopathological features of HGSOC patients. Further, the results from univariate and multivariate analyses indicate the recognition of SOCS7 as a promising independent prognostic biomarker for HGSOC patients.
The upregulation of SOCS7 was found to restrict the aggressive cellular activities of prostate cancer [
31]. Similarly, SOCS7 was also demonstrated to negatively control the growth factor signaling in breast cancer cells and limit their growth and migration [
32]. Consistent with these observations, here we observed the inhibitory effect of SOCS7 on HGSOC cell viability, cell cycle progression, and tumor development. In addition, SOCS proteins can regulate STAT signaling and facilitate p53 activation, both of which contribute to the cell cycle arrest at the G1 phase [
33], which might explain the finding in the present study that SOCS7 overexpression caused more HGSCO cells to accumulate at the G0-G1 phase.
The RNA-binding protein Hu antigen R (HuR) belongs to embryonic lethal abnormal vision-like protein family, which binds to adenine or uracil-enriched motifs to enhance mRNA stabilization and post-translational regulation or target genes [
34,
35]. For the first time, we identified the association and interaction between SOCS7 and HuR proteins in the study. As a matter of fact, the aberrant or dysregulated post-translational modifications of HuR, especially the ubiquitin-proteasomal degradation, have been implicated in carcinogenesis processes [
36,
37]. Since all SOCS proteins share a common conserved motif (SOCS box) that is recognized as the binding site for inducing ubiquitin-mediated proteasomal degradation of target substrates [
38,
39], we proposed that SOCS7 might be able to mediate HuR degradation. We further confirmed that SOCS7 could facilitate the ubiquitination of HuR and down-regulate its cellular protein level. HuR has been reported to promote oncogenesis of multiple human cancers through its RNA-regulatory effect and its significance in cell cycle modulation, cell proliferation, and cell differentiation [
36,
40,
41]. In this study, we have also revealed that HuR could facilitate HGSOC cell viability and regulate cell cycle. Meanwhile, HuR overexpression counteracted the tumor-suppressive effect of SOCS7, demonstrating the interplay between these two proteins in HGSOC pathogenesis. Collectively, SOCS7 constrains the tumorigenicity of HGSOC, potentially through mediating HuR ubiquitination and minimizing its cellular level in ovarian cancer.
FOXM1 is one of the transcriptional factors in the FOX family, and its transcriptional activation contributes to homologous recombination repair and G1-S/G2-M transitions in cell cycle [
42,
43]. Previous study has suggested the proto-oncogenic roles of FOXM1 and its downstream target genes Cyclin D1, Survivin and CDC25B in tumors [
44], which include facilitating cell cycle progression and inducing tumor cell proliferation, as well as its relation with the poor prognosis of cancer patients [
45,
46]. Similarly, we also found that FOXM1 overexpression could stimulate HGSOC cell viability and regulate tumor cell cycle. The expression level of FOXM1 can be employed to predict the survival probability of HGSOC patients. Furthermore, the interaction between HuR and FOXM1 was also revealed by the current study, for which we demonstrated that HuR could stabilize FOXM1 mRNA by binding to FOXM1 3′UTR. This finding is in line with the well-known HuR activities of specifically binding to the targeted genes for enhancing their transcription [
36,
47‐
49], and it also explains the oncogenic roles of HuR in HGSOC. Therefore, the expression levels of SOCS7, HuR, and FOXM1 are closely correlated with prognosis in HGSOC patients, and these three proteins participate in the pathogenicity and progression of HGSOC. By elucidating their interactions and regulatory functions, the overall suppressive role of SOCS7 in HGSOC was well explained.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.