The Michigan Cancer Foundation-7 (MCF7) and Metastatic Derivative of Anaplastic BC-231 (MDA-MB-231) cell lines were obtained from The Cell Bank of Type Culture Collection of The Chinese Academy of Sciences and maintained in Dulbecco's modified Eagle's medium (HyClone; Cytiva) supplemented with 10% fetal bovine serum (FBS; HyClone; Cytiva) and 1% penicillin/streptomycin at 37°C in a humidified incubator with 5% CO₂. Interference with PTPN18 expression levels was achieved by transfecting (37°C, 6 h) cells with PTPN18-3HA (1 µg/µl) and PTPN18 small interfering RNA (20 µM) (siRNA; Suzhou GenePharma Co., Ltd.) using Lipo6000 (Beyotime Institute of Biotechnology), according to the manufacturer's instructions. To achieve an improved knockdown efficiency, the knockdown siRNA was transfected again after 24 h of initial transfection. The siRNA sequences used are listed in Table SI.

First, total protein was extracted from cells using precooled radioimmunoprecipitation assay (RIPA) buffer (Beyotime Biotechnology) containing 1% protease inhibitor and phosphatase inhibitor (Selleck Chemicals) and then quantified by bicinchoninic acid assay. Equal amounts of protein (10 µg) were separated by 10% SDS-PAGE at a constant voltage (80 V) until the blue band of the loading buffer approached the lower edge of the gel. Proteins were then transferred to PVDF membranes using the sandwich method at a constant voltage of 100 V for 100 min at 4°C. Membranes were blocked with 5% bovine serum albumin (Beyotime Institute of Biotechnology) in Tris-buffered saline with 0.1% Tween 20 (TBST) for 1 h at room temperature, then incubated overnight at 4°C with blocking buffer plus primary antibodies. Subsequent washing with TBST was followed by incubation with secondary antibodies for 1 h at room temperature. Goat anti-rabbit IgG (heavy and light chain) antibody or horse anti-mouse IgG antibody conjugated with horseradish peroxidase secondary antibody was selected for chemiluminescence detection according to the primary antibody species. Finally, the protein bands were visualized and semi-quantified using the ChemiDoc XRS + Gel Imaging Analysis System (Bio-Rad Laboratories, Inc.) and Image Lab 5.2.1 software (6). The antibody information is listed in Table SII.

Treated cells were lysed in pre-chilled RIPA lysis buffer containing protease inhibitor and phosphatase inhibitor cocktails for 10-30 min. Experiments involving tyrosine phosphorylation or PTPN18 C229S were stimulated with EGF 100 ng/ml for 30 min before cells were collected. The samples were centrifuged at 12,000 × g for 15 min at 4°C to remove cell debris, then the supernatant was incubated with the appropriate primary antibody for 2-4 h or overnight at 4°C. The conjugated product was incubated with 30-50 µl Protein A/G agarose beads (Beyotime Biotechnology) for 2-4 h at 4°C the following day. The agarose beads were washed five times with cold lysis buffer, the supernatant was discarded and the beads were boiled with 1X SDS loading buffer for 10 min prior to western blotting.

The apoptosis of BC cells was observed using flow cytometry. Cells were washed twice with pre-chilled phosphate-buffered saline (PBS) and centrifuged at 500 × g for 5 min at 4°C to collect 1-2.5×10⁶ cells. The cells were resuspended in 100 µl of 1X Binding Buffer (BD Biosciences) and divided into two aliquots, one without Annexin V-FITC/PI and the other with 5 µl Annexin V-FITC and 5 µl PI Staining Solution (BD Biosciences), then mixed gently. The individual groups were stained with PI or FITC only for subsequent adjustment of compensation. The samples were protected from the light and the reaction was allowed to proceed at room temperature for 15 min. Next, 400 µl of 1X Binding Buffer was added and mixed well with the samples, which were then filtered with gauze and detected by BD Accuri^™ C6 Plus flow cytometry (BD Biosciences) within 1 h. For each sample, 10,000 events were analyzed and the experiment was independently repeated three times (24). Data were analyzed and processed using the FlowJo 10.8.1 (FlowJo LLC; BD Biosciences) software.

In vitro apoptosis was examined by DAPI staining. MCF-7 cells that had previously been transfected for 48 h were seeded into 24-well plates for overnight incubation and attachment. The cells were then washed with PBS and fixed with 4% paraformaldehyde for 10 min at room temperature. The fixative was removed, then the cells were washed with PBS and stained with 2 µg/ml DAPI (Beyotime Institute of Biotechnology) for 15 min at room temperature in the dark. Changes in nuclear morphology were examined by fluorescence microscopy and images were collected.

Cell cycle distribution was determined using an Accuri C6 flow cytometer (BD Biosciences). The collected cells were fixed with pre-chilled 70-80% ethanol overnight at 4°C, centrifuged to remove the fixative and washed once with 1 ml PBS. The samples were then divided into two aliquots after the supernatant had been removed: One aliquot was incubated at room temperature without PI/RNase for 15 min; the other was incubated at room temperature with 500 µl PI/RNase (BD Biosciences) in the dark for 15 min. The cells were then filtered with a 200-mesh filter and detected by flow cytometry. FlowJo 10.8.1 (FlowJo LLC; BD Biosciences) or ModFit LT 4.0 software (Verity Software House) was used to process the data (25).

Cell viability was monitored using the MTS assay, a modified version of the classical MTT assay that has the same core principle (26). The treated cells (2,000 cells/100 µl medium) were digested and transferred to 96-well plates, with four duplicate wells set up for each treatment group, and the absorbance was detected after 0, 12, 24, 36 and 48 h. In each well, an average of 20 µl of CellTiter 96^® AQueous One Solution Reagent (Promega Corporation) was added to 100 µl of medium. After gentle mixing and 2 h of reaction in a 37°C and 5% CO₂ incubator, the absorbance at 490 nm was measured.

Treated cells were seeded into 6-well plates at a density of 2,000 cells/well. After the cells adhered overnight, the medium was replaced every 3 days for 14 days. The cells were then fixed with 4% paraformaldehyde (room temperature, 30 min) and stained with 0.1% crystal violet (Beyotime Institute of Biotechnology) for 20 min at room temperature. Finally, the plates were washed with water and air dried at room temperature. The colonies were recorded using ImageJ 1.53e software (National Institutes of Health).

Conditioned medium (500 µl) containing 20% FBS was added to the lower chamber of a 24-well plate. Next, 100 µl of diluted BD Matrigel was added to each upper chamber. After the Matrigel had solidified at 37°C, a digested and diluted serum-free cell suspension was added to the upper chamber, and the plates were incubated at 37°C for 24-48 h. The inserts (8 µm) were then washed with PBS and fixed with a 5% paraformaldehyde solution for 30 min. The cells were stained with crystal violet (0.1%) at room temperature for 30 min, then observed and counted under an inverted fluorescence microscope (Leica Microsystems GmbH) operating in white light mode.

After creating a scratch with a 200-µl sterilized tip, the cells were washed three times with sterilized PBS, the impurities and cell debris were washed off, and fresh low serum (1%) culture medium was added to continue the cell culture. Images were collected at 0, 24 and 48 h and the change in scratch size was recorded. The experiment was repeated three times. ImageJ software was used to assess the scratch area and for subsequent experimental data processing.

Total cellular RNA was extracted using TRIzol and cDNA synthesis was performed using the All-in-One cDNA Synthesis SuperMix kit (Selleck Chemicals) following the manufacturer's instructions. In a total reaction volume of 20 µl per well, the 2X SYBR Green qPCR Master Mix kit (Selleck Chemicals) was used for qPCR. The PCR amplification protocol consisted of pre-denaturation at 95°C for 5 min; followed by 40 cycles of denaturation at 95°C for 15 sec, annealing at 55°C for 30 sec, and extension at 72°C for 30 sec. A total of 3-5 replicate wells were set up per group. The mean quantification cycle (Cq) values were collected for each reaction and relative expression was quantified using the 2^−ΔΔCq method (27). The sequences of the qPCR primers used in the present study are listed in Table SIII.

Changes in the expression levels of total PTPN18 protein were analyzed through the 'Proteomics' catalog of the UALCAN database (https://ualcan.path.uab.edu/cgi-bin/CPTAC-Result.pl?genenam=PTPN18&ctype=Breast) (28). 'Expression DIY' in the Gene Expression Profiling Interactive Analysis (GEPIA) database (http://gepia2.cancer-pku.cn/#index) was used to obtain gene expression related information in the box plot and stage plot formats. Correlation analysis (GEPIA) was used to test the association of PTPN18 or CDKN1A/B with CCNE1/2 gene. Survival analysis (GEPIA) was used to obtain gene and OS or disease-free survival data (29). Additionally, gene expression levels and OS or relapse-free survival data were obtained from the BC category of the Kaplan-Meier Plotter database (https://kmplot.com/analysis/index.php?p=service&cancer=breast) (30). PTPN18 correlation analysis was performed using breast invasive carcinoma and HiSeq RNA data from the LinkedOmics platform (https://www.linkedomics.org/admin.php) (31). Protein interaction networks for cyclin E1 and cyclin E2 were obtained from the STRING database (https://cn.string-db.org/) (32).

Data are presented as the mean ± SD of at least three independent experiments. Data analysis was performed using GraphPad Prism software 8.0.2 (Dotmatics). Data were analyzed using unpaired Student's t-test or one-way analysis of variance with Tukey's multiple comparison test, as appropriate. P<0.05 was considered to indicate a statistically significant difference.

Proteomic expression profiling using the UALCAN database showed that PTPN18 protein expression levels are decreased in BC (Fig. 1A). In addition, a close link between PTPN18 gene expression and the stage of BC was identified (Fig. 1B). Statistical analyses using the Kaplan-Meier Plotter and GEPIA databases showed that high levels of PTPN18 improved the survival and prognosis of patients with BC (Fig. 1C-F), indicating that PTPN18 may play a tumor suppressor role in this disease. Therefore, the function and mechanism of PTPN18 in BC was further investigated.

PTPN18 is widely expressed in BC cell lines (15,16). In the present study, the western blot results showed that PTPN18 was expressed at higher levels in MCF7 and MDA-MB-231 cells (Fig. S1A), which represent estrogen receptor and progesterone receptor double-positive and triple-negative BC cell lines, respectively. Overexpression and knockdown of PTPN18 in MCF7 cells at different time points after transfection is shown in Fig. S1B and C. Next, the effect of PTPN18 on cancer cell invasion was investigated using a Transwell chamber assay. The results showed that overexpression of PTPN18 effectively inhibited the invasion of cancer cells, while knockdown of endogenous PTPN18 expression enhanced invasion (Fig. 2A and B). Consistent with these results, in vitro cell wound healing assays demonstrated that PTPN18 expression inhibited tumor cell migration (Fig. 2C and D). Analysis of the mRNA levels of genes involved in cell metastasis revealed that E-cadherin gene expression increased and Snail family transcriptional repressor 1 expression decreased in the PTPN18 overexpression group compared with the vector group, while the TWIST basic helix-loop-helix transcription factor 1 and Notch receptor 3 levels remained essentially unchanged (Fig. 2E). The trend was roughly the opposite between the knockdown and overexpression groups (Fig. 2E). Subsequent verification of the translation levels of the E-cadherin EMT marker protein revealed changes consistent with the transcription levels (Fig. 2F).

The key to cancer treatment is inhibiting the metastasis and proliferation of cancer cells and inducing apoptosis. The MTS assay results demonstrated that PTPN18 overexpression inhibited cell viability, an inhibitory effect that exhibited an increasing cumulative trend over time (Fig. 3A). Although the MCF7 and MDA-MB-231 cells in the PTPN18 knockdown groups showed a high cell viability at different time points, the overall trend was the opposite to that of the overexpression group (Fig. 3B). This conclusion was further verified by a colony formation assay (Fig. 3C and D).

The flow cytometric results identified that PTPN18 overexpression significantly increased the proportion of BC cells undergoing apoptosis, while the three knockdowns of PTPN18 revealed different degrees of apoptosis' inhibition (Fig. 3E and F). In addition, the MCF7 BC cell line was stained with DAPI and subsequent fluorescence microscopy showed significantly increased signs of apoptotic cell death, such as chromatin condensation and fragmentation, in the PTPN18 overexpression group compared with the control cells (Fig. 3G). Furthermore, RT-qPCR (Fig. 3H) and western blot (Fig. 3I) results showed that PTPN18 inhibited the expression of mechanistic target of rapamycin kinase (mTOR) mainly at the protein level. Additionally, PTPN18 promoted Bax expression and suppressed Bcl-2 expression (Fig. 3H and I). Collectively, these results confirmed that PTPN18 inhibited proliferation and promoted the apoptosis of BC cells.

Next, the effect of PTPN18 on the cell cycle of BC cells was further investigated. The results of the cell cycle flow cytometry experiments revealed that PTPN18 overexpression significantly increased the number of cells in the S phase and decreased the number of cells in the G2/M phase but did not significantly change the number of cells in the G0/G1 phase (Fig. 4A and C). However, the PTPN18 knockdown groups showed the opposite trend (Fig. 4B and D), indicating that PTPN18 could induce cell cycle arrest in the S phase and affect the cell cycle progression of BC cells. In addition, PTPN18 overexpression significantly decreased cyclin E1 expression and increased cyclin D1, cyclin B1 and CDK4 expression at both the transcriptional and translational levels (Fig. 4E-H). These results suggested that significant downregulation of cyclin E may be an important factor in PTPN18 overexpression-induced S phase cell cycle arrest in BC. The aforementioned phenotypic functional experiments demonstrated that PTPN18 could play a tumor suppressive role in BC cells by promoting apoptosis and inhibiting proliferation, cell cycle, migration and invasion.

To illustrate the biological function of PTPN18 in BC development, RNA sequencing gene enrichment analysis of PTPN18 in BC was performed using the LinkedOmics database. The results revealed that PTPN18 showed a close and significantly negative correlation with the cell cycle in BC (Fig. 5A). Additionally, the results of the cell cycle gene enrichment analysis showed that PTPN18 may affect cell cycle progression by inhibiting gene expression of cell cycle pathways (Fig. 5B). Volcano plots revealed the genes that were positively or negatively correlated with PTPN18 (Fig. 5C). Pearson correlation coefficient analysis suggested that PTPN18 was significantly negatively correlated with cyclins E1 and E2 (Fig. 5C). Furthermore, the GEPIA database gene correlation analysis results also demonstrated that PTPN18 was negatively correlated with cyclins E1 and E2 (Fig. 5D). The cyclin E1 and E2 expression levels were found to be significantly increased in BC tissues compared with normal tissues (Fig. 5E). Cyclin E1 expression was significantly different in different clinical stages of BC and may play a role in disease stage progression; however, cyclin E2 expression was weakly associated with stage, suggesting that the functions of these two proteins in disease progression may be different (Fig. 5F). The results of the survival analyses using the Kaplan-Meier Plotter and GEPIA databases identified that high expression of cyclin E was associated with a reduced OS and the poor prognosis of patients with BC (Fig. 5G and H; Fig. S2A and B). These analyses support the hypothesis that PTPN18 may affect the BC cell cycle through regulation of the cyclin E pathway. These results echo the aforementioned experimental results and reinforce the rationality and necessity of studying the mechanism of cyclin E downregulation due to PTPN18 overexpression.

In view of the notable role of PTPN18 in the BC cell cycle, the present study further explored the molecular mechanism by which PTPN18 causes cell cycle arrest in the S phase. First, treatment with cycloheximide (CHX), an inhibitor of protein synthesis, revealed that overexpression of PTPN18 promoted cyclin E1 degradation and knockdown of PTPN18 inhibited cyclin E1 degradation (Fig. 6A and B). Furthermore, MG132 proteasome inhibitor inhibited cyclin E1 degradation induced by PTPN18 overexpression and enhanced the inhibitory effect on cyclin E1 degradation in the knockdown group (Fig. 6C and D). When CHX and MG132 were combined, the effect of PTPN18 on the cyclin E1 protein levels was not significantly different from that of the control (Fig. 6E and F).

Next, the effect of PTPN18 on cyclin E1 ubiquitination was examined. The results demonstrated that PTPN18 overexpression markedly increased the ubiquitination level of cyclin E1 and that PTPN18 knockdown reduced the cyclin E1 ubiquitination level after MG132 treatment (Fig. 6G and H). Since PTPN18 is a PTP, it was found that the ubiquitination of cyclin E1 by PTPN18 requires enzymatic activity (Fig. 6I). Using western blotting to detect the pan-tyrosine phosphorylation level of cyclin E1, it was found that PTPN18 was able to dephosphorylate the tyrosine of cyclin E1, indicating that cyclin E1 is a substrate of PTPN18 (Fig. 6J and K). Next, the interaction between PTPN18 and cyclin E1 was verified by forward and reverse Co-IP experiments as well as by examination of the endogenous interaction in MCF7 cells (Fig. 6L-N). In summary, the results revealed that PTPN18 can bind to cyclin E1 and promotes its degradation through the ubiquitin-proteasome pathway.

Next, the present study explored the effect of PTPN18 on cyclin E2, another isoform of cyclin E. It was found that overexpression of PTPN18 decreased the cyclin E2 protein levels after treatment with CHX (Fig. 7A). However, the cyclin E2 protein levels were restored when MG132, a proteasome inhibitor, was added (Fig. 6C). When PTPN18 expression was knocked down with siRNA, the opposite effects on the cyclin E2 protein levels were observed (Fig. 7B and D). The combined use of CHX and MG132 nearly abolished the effect of PTPN18 on the cyclin E2 protein levels compared with the control (Fig. 7E and F). Examination of the cyclin E2 ubiquitination levels revealed that MG132 enhanced the PTPN18-mediated protein ubiquitination of cyclin E2, whereas knockdown of PTPN18 reduced the cyclin E2 ubiquitination levels (Fig. 7G and H). Similar to the cyclin E1 results, inactivation treatment with PTPN18 C229S reduced the ubiquitination and increased the phosphorylation of cyclin E2, indicating that cyclin E2 is also a substrate for PTPN18 (Fig. 7I-K). Additionally, forward and reverse Co-IP experiments verified the interaction of PTPN18 with cyclin E2 (Fig. 7L and M). These results indicated that PTPN18 can also promote the degradation of cyclin E2 through the ubiquitin-proteasome pathway. Given that both cyclins E1 and E2 can bind PTPN18, it was further explored whether there is a competitive relationship between their interactions. The Co-IP results showed that the addition of cyclin E2 did not reduce the binding of PTPN18 to cyclin E1, suggesting cyclin E2 did not interfere with the binding of cyclin E1 to PTPN18 (Fig. 7N).

To further explore the signaling pathways through which PTPN18 regulates the cell cycle, key genes associated with the cell cycle according to the KEGG database were selected to examine the effects of PTPN18 on their transcript levels. The results demonstrated that PTPN18 overexpression produced significant changes in the phosphoinositide-dependent protein kinase 1 (PDPK1), phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit γ (PI3CG), forkhead box O3 (FOXO3), mitogen-activated protein kinase 8 (MAPK8), CDKN1A, growth arrest and DNA damage-inducible protein α (GADD45A), adenomatous polyposis coli protein (APC) and E1A-binding protein p300 (EP300) transcript levels, indicating that PTPN18 may have a wide range of multi-target and multi-pathway roles in gene transcription regulation (Fig. S3A-H). Analysis using STRING, a functional protein association network database, showed that CDKN1A and CDKN1B closely interacted with cyclins E1 (CCNE1) and E2 (CCNE2) (Fig. 8A and B). The Kendall statistical analysis results from the GEPIA database also showed that CDKN1A and CDKN1B had significant negative correlations with cyclin E1 and CDKN1A had a negative correlation with cyclin E2 (Fig. 8C and D). Further protein level experiments showed that PTPN18 overexpression significantly decreased the tyrosine phosphorylation levels of PI3K p85 and AKT and the protein expression levels of PI3K P110β compared with the controls, whereas the protein levels of FOXO1, CDKN1A and CDKN1B were significantly increased (Fig. 8E-H). Decreased levels of AKT tyrosine phosphorylation resulted in reduced inhibition of FOXO and increased FOXO expression, followed by increased CDKN1A and CDKN1B protein levels, resulting in cyclin inhibition and cell cycle arrest (Figs. 8E-H and 9). The results of the endogenous PTPN18 knockdown experiments were generally consistent with those of the overexpression experiments (Fig. 8E-H). Taken together, the protein level results confirmed that the antitumor effect of PTPN18 may be closely related to the PI3K/AKT signaling pathway.