Galectin-1-mediated high NCAPG expression correlates with poor prognosis in gastric cancer

Patient information and tissue samples

A total of 145 gastric adenocarcinoma patients, with detailed pathological and follow-up data were enrolled in this study. All patients were treated with radical gastrectomy and D2 lymphadenectomy at the Gastrointestinal Surgery Department, Taizhou People’s Hospital of Jiangsu province between January 2015 and May 2017. No patient received radiotherapy or chemotherapy prior to surgery, had no serious diseases or other synchronous malignancies, or distant metastases prior to surgery. Overall survival (OS) and disease free survival (DFS) times were the primary endpoints. OS was calculated from the date of surgery to the date of death or the final follow-up. DFS was calculated from the date of surgery to disease recurrence. The clinicopathological features of the patients are described in Table 1. Primary GC tissues (GCTs) and matched normal gastric mucosa tissues (NGCTs) were formalin-fixed and paraffin-embedded for hematoxylin and eosin (HE) staining and immunohistochemistry (IHC), while fresh GCTs and NGCTs were collected from eight patients for molecular analysis in April 2022.

IHC staining

Formalin-fixed, paraffin-embedded specimens were used for IHC analysis. The speciments were sliced at 4-μm thickness. Xylene and gradient ethanol were used to deparaffinize and rehydrate the tissues. Endogenous peroxidases were blocked in methanol for 10 min using 3% hydrogen peroxide. Phosphate-buffered saline (PBS) was used to wash the slides three times, and citrate buffer (pH 6.0) was used for antigen retrieval for 20 min at 95°C. Rabbit monoclonal antibodies, anti-Gal1 (dilution, 1:250, Abcam, Cambridge, UK) and NCAPG (dilution, 1:200, Abcam), were incubated with the slides overnight at 4°C. The slides were washed thrice with PBS and incubated with biotin-conjugated secondary antibodies, followed by horseradish peroxidase-conjugated streptavidin. The sections were then stained with diaminobenzidine, counterstained with hematoxylin, dehydrated, cleared, and cover-slipped.

IHC analysis

Two pathologists, who were blinded to the clinical data, scored the IHC staining for Gal1 and NCAPG in GCTs and NGCTs using semi-quantitative immunoreactivity scores (IRS). Immunostaining intensity was documented as category A with scores of 0–3 (0, negative; 1, weak; 2, moderate; and 3, strong). The percentage of immunoreactive cells was documented as category B with scores of 1–4 (1, 0–25%; 2, 26–50%; 3, 51–75%; and 4, 76–100%). IRS was calculated by multiplying the scores for categories A and B, ranging from 0 to 12. Receiver operating characteristic (ROC) analysis was used to obtain the optimum cutoff values for IRS and to distinguish high and low expression of Gal1 and NCAPG.

Cell lines and culture

Human GC cells, SGC-7901 (Cat No. C6795), and undifferentiated GC cells, HGC-27 (Cat No. C6365), were provided by Shanghai Biyuntian Biological Co., Ltd. (Shanghai, China). RPMI-1640 (Gibco-BRL, Gaithersburgh, MD, USA) with 10% (v/v) fetal bovine serum (Gibco-BRL), 100 mg/mL of streptomycin, and 100 U/mL of penicillin (Gibco-BRL) were used to culture GC cells. All cells were cultured in a humidified atmosphere containing 5% (v/v) CO₂ at 37°C, and passaged by trypsinization when 80% confluence was reached.

Lentiviral transduction

A lentiviral vector for LGALS1 overexpression and a corresponding non-targeting negative control lentiviral vector were constructed by Genechem Co., Ltd. (Shanghai, China). The GV358 (Ubi-MCS-3FLAG-SV40-EGFP-IRES-puromycin) lentiviral vector was constructed to upregulate LGALS1 expression. SGC-7901 and HGC-27 cells were seeded at a concentration of 5 × 10⁴ cells per well in six-well plates before lentiviral transduction. The cells were transduced with lentiviral vector and 10 μg/mL of polybrene (Sigma-Aldrich, St. Louis, MO, USA) based on an infection multiplicity of 10. The corresponding non-targeting negative control lentiviral vector was transduced through the same approach. The medium was replaced 12 h after transduction and puromycin (Sigma-Aldrich) was added to select stable transduced cell lines at the concentration of 2 μg/mL after another 48 h. The stable transduced cells were then cultured in the presence of 0.5 μg/mL of puromycin. Fluorescent microscopy (OLYMPUS-U-HGLGPS-IX73), qualitative reverse-transcriptase polymerase chain reaction (qRT-PCR), and Western blotting were used to evaluate transduction efficiency after 72 h.

siRNA transduction

The siRNA against NCAPG and matched negative control siRNA were purchased from Biomics Biotechnologies Co. Ltd. (Nantong, China). OE-LGALS1 SGC-7901 and OE-LGALS1 HGC-27 cells were seeded in six-well plates at a concentration of 5 × 10⁴, and were transfected with NCAPG siRNA or control siRNA using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA). The manufacturer’s instructions were followed for all the steps. After 24 h, the cells were harvested for further experiments. The three siRNA sequences were as follows: hs-NCAPG-si-1 sense (5′–3′): CACGAUGGAUGAUAAGACA, hs-NCAPG-si-1 antisense (3′–5′): UGUCUUAUCAUCCAUCGUG; hs-NCAPG-si-2 sense (5′–3′): GGAGUUCAUUCAUUACCUU, hs-NCAPG-si-2 antisense (3′-5′): AAGGUAAUGAAUGAACUCC; hs-NCAPG-si-3 sense (5′–3′): GCUGAAACAUUGCAGAAAU, and hs-NCAPG-si-3 antisense (3′–5′): AUUUCUGCAAUGUUUCAGC.

Wound healing assay

All cells were seeded in six-well plates at concentrations of 1 × 10⁵. When the cell monolayer reached a confluence of 80–90%, a 10-μL pipette tip was scored across it to create a wound. PBS (Corning, Manassas, VA, USA) was used to wash the plates and remove cellular debris. The cells were photographed using a Leica DMIRB microscope (100× magnification; Leica, Wetzlar, Germany). Then, the cells were incubated with a serum-free medium containing 10 μg/mL of mitomycin C to block proliferation, and the wounds were photographed after 24 h. The number of migrated wild type cells was defined as 100% to calculate the percentage of cell migration.

In vitro invasion assay

The invasion bility of SGC-7901 and HGC-27 cells was measured using 24-well transwell units with polycarbonate filters (pore size, 8.0 μm; Corning, NY, USA). Matrigel^® basement membrane (BD Biosciences, San Diego, CA, USA) was mixed with the serum-free RPMI at a ratio of 1:8, and 100 μL of the mixture was used to coat the upper transwell inserts. Then, the cells were seeded at 1 × 10⁵ in the upper insert with 100 μL serum-free RPMI medium, and placed in the lower chambers with 600 μL of complete media. The cells were allowed to culture for 24 h at 37°C, and non-invasive cells were removed using a cotton swab. The filters were fixed using 4% (v/v) paraformaldehyde, the cells were stained with 0.05% (v/v) crystal violet solution, and counted under a microscope.

RNA extraction and real-time PCR

The RNeasy Mini Kit (Invitrogen) was used to isolate real-time PCR RNAs, and the Revert Aid RT reverse transcription kit (Thermo Fisher Scientific, Waltham, MA, USA) was used to reverse-translate the purified RNAs into cDNA. A SYBR Green dye kit (RocheDiagnostics, Mannheim, Germany) was used to perform qRT-PCR, and the products were analyzed using an iQ5 Multicolor real-time PCR Detection System (Bio-Rad, Hercules, CA, USA). GAPDH was used as the reference control gene and analyzed using the 2^−ΔΔCt method. The following primers were used: NCAPG (forward): ACCCAAGCATCAAAGTCTACTCAGC and (reverse) TGACACCTCCTGTTCGTCCTAGC; LGALS1 (forward): GCCAGATGGATACGAATTCAAG and GCCACACATTTGATCTTGAAGT; and GAPDH (forward) CCAGCAAGAGCACAAGAGGAAGAG and (reverse) GGTCTACATGGCAACTGTGAGGAG.

Western blotting

The tissues and cells were lysed using the RIPA buffer (Thermo Fisher Scientific) to prepare total cells extracts. Denatured proteins were separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and the separated proteins were transferred onto nitrocellulose membranes (GE Healthcare Life Sciences, Pittsburgh, PA, USA). Antibodies against Gal1, NCAPG, and GAPDH (dilution, 1:2000) were used to probe the proteins on the blots at 4°C overnight. They were then incubated with peroxidase-conjugated secondary anti-bodies (Sigma), and a West Pico chemiluminescent substrate (Pierce, Carlsbad, CA, USA) was used to visualize the immunoreactive protein bands. A densitometric image analysis software (Image Master VDS; Pharmacia Biotech, Little Chalfont, UK) was used to quantify the proteins. GAPDH levels were determined as an internal reference. All experiments were independently performed thrice.

Statistical analyses

SPSS Statistics version 25.0 software (IBM Corp., Armonk, NY, USA) was used to analyze the data. The clinicopathological features and protein expression levels were compared using the chi-square test. Wilcoxon test (grouped) was used to analyze the IRS for Gal1 and NCAPG in GCTs and NGCTs. OS and DFS were analyzed through Kaplan-Meier survival analysis. The hazard ratios (HRs) and 95% confidence intervals (CIs) were estimated by univariate or multivariate Cox regression analysis. Continuous variables were expressed as means ± standard error of the mean. Multiple comparisons were performed through one-way analysis of variance (ANOVA) and Dunnett’s t test. P-values < 0.05 were considered significant.

Gal1 and NCAPG expression in GC and non-cancer tissues

Western blotting was used to analyze Gal1 and NCAPG expressions in eight paired specimens from GC patients, including primary GCTs and matched NGCTs. GCTs had increased Gal1 and NCAPG expressions compared to the paired NGCTs (Figure 1A, 1B). Immunohistochemistry of the GCTs and NGCTs was used to further investigate Gal1 and NCAPG expressions in 145 primary GC patients. Gal1 staining was mainly localized in the cytoplasm and extracellular matrix, whereas NCAPG was mainly expressed in the cytoplasm and nuclei (Figure 1C). Typical images of Gal1 and NCAPG expression in non-cancer tissues are shown in Figure 1D. The IRS distribution for Gal1 and NCAPG expressions in primary GCTs and NGCTs are shown in Figure 1E, 1F. Compared to NGCTs, Gal1 and NCAPG expressions in GCTs were significantly increased (both P < 0.001, Figure 1E, 1F). NCAPG expression correlated significantly with Gal1 expression in cancerous tissues (r = 0.653, P < 0.001, Figure 1G).

Figure 1. Gal1 and NCAPG expression predict prognosis of gastric cancer (GC). (A, B) Expression of Gal1 and NCAPG proteins detected by Western blotting in GC cancer tissues (GCTs) and non- gastric cancer tissues (NGCTs). (C, D) Representative immunohistochemistry images for Gal1 and NCAPG in (C) GC tissues, and (D) matched non-cancerous tissues. (E) Immunoreactivity score (IRS) for Gal1 compared between GCTs and matched NGCTs. (F) IRS for NCAPG compared between GCTs and matched NGCTs. (G) The expression of Gal1 was positively correlated with NCAPG expression in GC tissues. Abbreviations: T: tumor tissue; N: non-tumor tissue; GCTs: gastric cancer tissues; NGCTs: non- gastric cancer tissues. ^**P < 0.01.

Gal1 and NCAPG expression correlated with clinicopathological characteristics

IHC cutoff scores were used to classify Gal1 and NCAPG levels as high and low through ROC analysis. IRS ≥ 9.17 and ≥ 3.17 indicated high Gal1 (Figure 2A) and NCAPG (Figure 2B) expressions, respectively. Based on this standard, high Gal1 and NCAPG expression was 83 (57.24%) and 110 (75.86%) in GC cases, respectively. Gal1 expression in GCTs correlated with the clinicopathological features (Table 1), including tumor diameter, pathological classification, depth of invasion, lymph node metastasis, and TNM stage (all P < 0.05). NCAPG expression in GCTs was significantly associated with gender, tumor diameter, pathological classification, depth of invasion, lymph node metastasis, and TNM stage (all P < 0.05).

Figure 2. Aberrant Gal1 and NCAPG expression in GC tissues indicates poor prognosis. (A, B) Receiver operating characteristic statistics were employed to estimate the cut-off points of the IRS for (A) Gal1 and (B) NCAPG in GC tissues. (C–E) Kaplan-Meier curves for Gal1, NCAPG, and combined Gal1/NCAPG expression in the training cohort for overall survival. (F–H) Kaplan-Meier curves for Gal1, NCAPG, and combined Gal1/NCAPG expression in the training cohort for disease free survival.

High Gal1 and NCAPG expression inversely correlated with survival

All patients follow-up ended at death or at least 5 years, the median follow-up time was 36.37 months. Kaplan-Meier analysis revealed that high Gal1 or NCAPG expression in GCTs was significantly correlated with poor OS (all P < 0.001, Figure 2C, 2D) and DFS (all P < 0.001, Figure 2F, 2G). Univariate Cox regression analysis indicated that tumor diameter, pathological classification, depth of invasion, lymph node metastasis, TNM stage, and Gal1 and NCAPG expression were associated with OS in GC patients (all P < 0.001, Table 2). Multivariate Cox regression analysis indicated that tumor diameter, TNM stage, Gal1 and NCAPG expression were independent prognostic factors for GC (Gal1: HR, 0.072, 95% CI, 0.035–0.145, P < 0.001; NCAPG: HR, 0.185, 95% CI, 0.081–0.420, P < 0.001, Table 3).

Synergistic effect of Gal1 and NCAPG on OS and DFS

The patients were stratified into three groups based on the IRS for Gal1 and NCAPG, i.e., both high (n = 73), either Gal1 or NCAPG high (n = 47), and both low (n = 25). Kaplan-Meier analysis revealed that the both low expression group had more favorable OS and DFS, while the both high expression group had poorer OS and DFS, compared to the Gal1 or NCAPG high group (P < 0.001, Figure 2E, 2H).

Gal1 regulates NCAPG at mRNA and protein levels

We constructed lentivirus that overexpressed LGALS1 in SGC-7901 and HGC-27 cells to obtain four cell lines, over-expressed LV-LGALS1 SGC-7901 (OE-LGALS1 SGC-7901), over-expressed LV-LGALS1 HGC-27 (OE-LGALS1 HGC-27), and corresponding transfected negative control (NC) cells. Successful overexpression of LGALS1 was confirmed by green fluorescent protein signal (Figure 3A, 3B), Western blotting (Figure 3C, 3D), and qRT-PCR (Figure 3E, 3F). Western blotting indicated that OE-LGALS1 SGC-7901 and OE-LGALS1 HGC-27 cells had higher Gal1 levels than NC cells and wild cells (P < 0.05 and < 0.01, respectively, Figure 3C, 3D). mRNA expression for LGALS1 was detected using qRT-PCR; the expression of LGALS1 mRNA was consistent with Gal1 protein expression (all P < 0.01, Figure 3E, 3F).

Figure 3. LGALS1 regulates NCAPG at the mRNA and protein levels. Transduction efficiency confirmed by green fluorescent protein (GFP) assay in (A) SGC-7901 cells and (B) HGC-27 cells. Original magnification 100×. (C, D) Western blot confirmation of stable overexpression of Gal1 and NCAPG in (C) SGC-7901 and (D) HGC-27 cells when LGALS1 was overexpressed. (E, F) Quantitative real-time PCR (qRT-PCR) analysis of LGALS1 and NCAPG expression in (E) SGC-7901 and (F) HGC-27 cells when LGALS1 was overexpressed. Abbreviations: OE-LGALS1: Overexpression of LGALS1; NC: negative control (empty vector); NS: not significant. ^*P < 0.05; ^**P < 0.01.

In all cell lines, NCAPG expression was consistent with Gal1 expression on Western blotting, and was significantly higher in OE-LGALS1 GC cells than in NC cells and wild cells (all P < 0.01, Figure 3C, 3D). We then performed qRT-PCR to evaluate NCAPG mRNA expression in all cell lines, and the results were consistent with those of Western blotting (all P < 0.01, Figure 3E, 3F). This indicated that LGALS1 could positively regulate NCAPG expression at mRNA and protein levels.

LGALS1 and NCAPG mutually reinforced GC cell line regulation

In order to further explore the regulatory relationship between LGALS1 and NCAPG, lentiviral was used to generate over-expressed LV-LGALS1 SGC-7901 and HGC-27 stable cell lines. We constructed an siRNA to simultaneous silence NCAPG expression in these stable cell lines, and obtained OE-LGALS1+ NCAPG-knockdown GC cells (OE-LGALS1+siNCAPG), including both SGC-7901 and HGC-27 cell lines. mRNA expressions for LGALS1 and NCAPG were detected using qRT-PCR. NCAPG knockdown in OE-LGALS1 GC cell lines significantly inhibited NCAPG mRNA expression for SGC-7901 (P < 0.01, Figure 4A) and HGC-27 (P < 0.01, Figure 4B). However, LGALS1 mRNA expression was also decreased when NCAPG was knocked down in OE-LGALS1 GC cell lines, including SGC-7901 (P < 0.01, Figure 4C) and HGC-27 (P < 0.01, Figure 4D). Thus, LGALS1 and NCAPG mutually reinforced GC cell line regulation. In order to further verify these results, Western blotting was used to detect Gal1 and NCAPG protein expressions in each GC cell line, and the results were consistent with those of qRT-PCR (all P < 0.01, Figure 4E, 4F).

Figure 4. LGALS1 and NCAPG may mutually reinforce regulation in GC cell lines. qRT-PCR analysis of (A, B) NCAPG and (C, D) LGALS1 expression in SGC-7901 and HGC-27 cells when LGALS1 was overexpressed, with or without simultaneous silencing of NCAPG. Western blot confirmation of stable overexpression of Gal1 and NCAPG in (E) SGC-7901 and (F) HGC-27 cells when LGALS1 was overexpressed, with or without simultaneous silencing of NCAPG. Abbreviations: OE-LGALS1: Overexpression of LGALS1; siNCAPG: silencing of NCAPG; OE-LGALS1+NC: overexpression of LGALS1 + NCAPG negative control (empty vector); NS: not significant. ^*P < 0.05; ^**P < 0.01.

LGALS1 promoted migration and invasion in GC cells through NCAPG regulation

Wound healing and transwell assays were used to assess migration and invasion abilities of SGC-7901 and HGC-27 cells, and compare them to the NC (OE-LGALS1 transfected NCs), OE-LGALS1, and OE-LGALS1+siNCAPG groups. The wound healing experiment demonstrated that LGALS1 over-expression enhanced SGC-7901 migration. However, this was partly rescued by reinfected NCAPG-shRNA in OE-LGALS1 GC cells (Figure 5A). Figure 5B shows the fold changes in migration. These experiments were repeated in HGC-27 cells to confirm LGALS1-mediated promotion of migration in vitro, which was reduced by simultaneous NCAPG-knockdown (Figure 5C, 5D).

Figure 5. LGALS1 promotes the migration and invasion of GC cell lines in vitro. Overexpression of LGALS1 (OE-LGALS1) significantly enhanced the migration capacity of (A, B) SGC-7901 and (C, D) HGC-27 cells compared with wild control (WC) and negative controls (NC). The migration capacity was abolished when NCAPG was simultaneously silenced. Magnification: ×100. (E) Transwell assay showing that overexpressed LGALS1 significantly enhanced the invasion ability of SGC-7901 and HGC-27 cells, and simultaneous silencing of NCAPG abolished the invasion capacity (n = 3). Magnification: ×200. Abbreviations: NC: negative control (empty vector); OE-LGALS1: Overexpression LGALS1; OE-LGALS1+siNCAPG: Overexpression of LGALS1 +silencing of NCAPG. NS: not significant. ^**P < 0.01.

Transwell assay showed significantly enhanced invasion ability in OE-LGALS1, SGC-7901, and HGC-27 cells. The cell invasion capability could be partly rescued by re-infected NCAPG-shRNA in OE-LGALS1 GC cells (Figure 5E; P < 0.01).