Clinical samples

GC tissues were collected from 106 patients who underwent neoadjuvant chemotherapy with the 5-Fu-based regimen and radical gastrectomy between 2016 and 2018 at Peking Union Medical College Hospital. The clinical and demographic characteristics of patients were listed in Supplementary Table S1. Patients were divided into two groups based on the degree of pathological response, as evaluated according to the guidelines of the College of American Pathologists (CAP) [22]. CAP 0, CAP 1 and CAP 2 were defined as exhibiting pathological response (n = 63), whereas CAP 3 indicated no pathological response (n = 43). Additionally, fifteen normal gastric mucosal tissue samples were obtained via endoscopic biopsy. Tumor samples were collected with patients' written informed consent, and the study was reviewed and approved by the Institutional Review Board of Peking Union Medical College Hospital.

Immunohistochemistry staining

A total of 106 collected tumor specimens were fixed in 10% formaldehyde solution, embedded in paraffin, and then serially severed into slices (4 μm). After being deparaffinized and rehydrated by xylene and graded ethanol, antigen was retrieved using microwave heating with sodium citrate retrieval buffer (pH 6.0). The endogenous peroxidase was inactivated by treatment with 3% H2O2 for 10 min. Following treatment with blocking buffer to obstruct nonspecific bindings, the samples were incubated with primary antibodies: Anti-MELK (1:100, #11403-1-AP, Proteintech Group), Anti-CSF-1 (1:100, #ab233387, Abcam), and Anti-CD206 (1:100, #18704-1-AP, Proteintech Group) at 4 °C overnight. The next day, after PBS washing, horseradish peroxidase (HRP)-conjugated antibody anti-Rabbit IgG (1:100, #7074, Cell Signaling Technology) was added for 30 min’ incubation at room temperature. 3, 3ʹ-diaminobenzidine (DAB) regent was taken for brown staining subsequent to PBS washing, then all slices were re-dyed with hematoxylin staining solution, dehydrated, and sealed for microscopic examination. At least three slices were taken from each specimen and five independent fields were randomly chosen from each slice for microscopic examination. The immunoreactivity of MELK and CSF-1 was scored by multiplying the scores of percentage of positive cells (< 5% scores 0, 5–25% scores 1, 25–50% scores 2, 50–75% scores 3, and 75–100% scores 4) with the scores of staining intensity (no staining scores 0, weak staining scores 1, moderate staining scores 2, and strong staining scores 3). The overall scores of 0–3 were defined as low expression, while scores above 3 were considered high expression. For CD206 evaluation, the Olympus microscope was used to count the number of positive cells in five random 400-fold fields.

Cell lines and culture

Human gastric cancer cell lines MKN45 and HGC27, human mononuclear cells.

THP-1, and human gastric epithelial cell line GES-1 were obtained from the Cell Resource Center of Peking Union Medical College (Beijing, China). Parental gastric cancer cells were regarded as 5-Fu-sensitive (MKN45-S and HGC27-S) and 5-Fu-resistant cells (MKN45-R and HGC27-R) were generated as previously reported [21]. All cells were grown in RPMI-1640 medium (Gibco, Carlsbad, CA, USA) containing 10% fetal bovine serum (FBS) (Gibco) and 1% penicillin/streptomycin (Gibco) at 37 °C in a humidified incubator with a volume fraction of 5% CO2. For the generation of M0 macrophages, THP-1 cells were induced by 100 ng/ml phorbol 12-myristate 13-acetate (PMA) (Sigma-Aldrich, St. Louis, MO, USA) for 24 h.

Co-culture of cancer cells and macrophages

Transwell chambers (6-well plates, 0.4-μm pore size; Corning, NY, USA) were used for co-culture experiment. GC cells were seeded onto the membrane of upper chambers with 0.4-μm pore, while M0 macrophages were seeded onto the lower chambers of 6-well plates for 48 h co-culture. Macrophages were then harvested for subsequent experimental analysis.

Cell viability assays

GC Cell (MKN45-S LV-NC, MKN45-S LV-MELK, HGC27-S LV-NC, HGC27-S LV-MELK, MKN45-R sh-NC, MKN45-R sh-MELK, HGC27-R sh-NC, HGC27-R sh-MELK) viability was measured using Cell Counting Kit-8 (CCK-8; Dojindo, Kumamoto, Japan). Briefly, GC cells were planted in 96-well plates at a density of 5 × 103 cells per well and cultured at 37 °C in a humidified incubator with 5% CO2 for 24 h. Then cells were incubated with culture medium containing varying concentrations of 5-Fu (0, 5, 10, 20, 40, 80, 120, 160, 200 μg/ml) for another 24 h. 10 μl of CCK-8 reagent was added to each well for 2 h incubation at 37 °C. The absorbance value at 450 nm was measured by microplate reader (Termo Scientific, Rockford, IL, USA).

Colony formation assay

GC cells (MKN45-S LV-NC, MKN45-S LV-MELK, HGC27-S LV-NC, HGC27-S LV-MELK, MKN45-R sh-NC, MKN45-R sh-MELK, HGC27-R sh-NC, HGC27-R sh-MELK) were seeded into 6-well plates at a density of 500 cells per well. After incubation for 24 h, the cells were treated with 5-Fu at a final concentration of 15 μg/ml, and the culture medium was replenished every 3–4 days. Following a two-week incubation period, visible colonies were rinsed with PBS twice and fixed with 4% paraformaldehyde for 15 min. 0.1% crystal violet solution was applied to staining the fixed colonies for 10 min. Colonies containing a minimum of 50 cells were later photographed and counted. The reason for using 15 μg/ml of 5-FU and the method for counting cell colonies were stated and uploaded as Supplementary Text S1.

Cell transfection

Lentiviral transfection was used to establish stable cell lines with overexpressed or silenced MELK. Lentiviral vectors were packaged with full-length MELK gene or shRNA targeting MELK by Genechem (Shanghai). MKN-45-S and HGC-27-S cells were transfected with overexpression lentivirus (LV-MELK) or its negative control (LV-NC), while MKN-45-R and HGC-27-R cells were transfected with shRNA targeting MELK lentivirus (sh-MELK) or its negative control (sh-NC). The full-length MELK gene sequence and three sequences of shRNA targeting MELK were listed in Supplementary Table S2 and S3. Multiplicity of infection (MOI) value was set at 10 based on our preliminary experiments. Cells were chosen for 2 weeks in a medium containing 2 μg/mL puromycin (Sangon Biotech, Shanghai, China), and the alive cells were defined as stable expression cells.

RNA extraction and quantitative real-time polymerase chain reaction (qRT-PCR)

Total RNA was isolated from the tissues and cells using TRIzol reagent (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s protocol. The RNA samples were reversed to cDNA using 1 μg of total RNA and 5 × PrimeScript RT reagent Kit (Takara, Dalian, China). Then qPCR was carried out by using TB Green Premix Ex Taq II (Takara, Dalian, China). The primer sequences of targeted genes are listed in Supplementary Table S4. The relative expression was calculated with GAPDH using 2−ΔΔCt method and experiment was conducted in triplicate.

Western blot analysis

Total protein was isolated out of cells and tissues using RIPA buffer (Termo Scientifc, Rockford, IL, USA) containing Halt Protease and Phosphatase inhibitor Cocktail (Termo Scientifc, Rockford, IL, USA). The lysates were sonicated followed by centrifugation at 12,000g for 15 min at 4 °C, and then the concentration of extracted protein was measured by BCA Protein Assay Kit (Beyotime, Shanghai, China). Approximately 30 µg of protein was transferred onto polyvinylidene fluoride (PVDF) membranes (Millipore, Billerica, MA, USA) using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). Membranes were blocked with TBST solution incorporating 5% non-fat milk for 1 h at room temperature and then immunoblotted overnight at 4 °C with the primary antibodies against MELK (1:1000, #11403-1-AP, Proteintech Group), Jak2 (1:1000, #3230, Cell Signaling Technology), p-Jak2 (1:1000, #3771, Cell Signaling Technology), STAT3 (1:1000, #9139, Cell Signaling Technology), p-STAT3 (1:1000, #9145, Cell Signaling Technology), GAPDH (1:1000, #5174, Cell Signaling Technology). On the following day, the membranes were incubated with corresponding HRP-labeled secondary antibody (1:5000, #7074, Cell Signaling Technology) at room temperature for 1 h. The antibody-antigen complex was visualized using SuperSignal West Pico PLUS Chemiluminescent Substrate (Termo Scientific, Rockford, IL, USA) in a Kodak Image station (Tanon, China).

Enzyme-linked immunosorbent assay (ELISA)

The concentrations of CSF-1 in the cultured media of GC cells were measured by CSF-1 ELISA kits (Cell Signaling Technology, MA, USA) in accordance with the manufacturer’s instructions. The immunoreactive was evaluated by a microplate reader (Termo Scientific, Rockford, IL, USA), and the concentrations were estimated from the standard curve.

Apoptosis assay

Cell apoptosis was measured using Annexin-V-FITC Apoptosis Detection Kit (Dojindo, Kumamoto, Japan) in accordance with the manufacturer’s instructions. Cells were harvested and washed twice with ice-cold PBS and incubated in 100 μl binding buffer in the presence of 5 μl Annexin V-FITC and 5 μl propidium iodide for 15 min in dark conditions. Then apoptosis was quantified using BD Accuri C6 Plus flow cytometer (BD Biosciences, San Jose, CA, USA). The experiment was conducted in triplicate and data were analyzed using FlowJo software (Tree Star, Oregon, OR, USA).

Flow cytometry analysis of macrophages

Macrophages were washed twice using PBS and filtered through a 100 μm cell strainer for flow cytometry analysis. Then 1 × 106 cells in 100 μl staining buffer were stained with FITC-CD11b (BioLegend, San Diego, CA, USA) and PE-CD206 antibodies (BioLegend, San Diego, CA, USA) in darkness for 15 min at 4 °C. Finally, the stained cells were analyzed by BD Accuri C6 Plus flow cytometer (BD Biosciences, San Jose, CA, USA). The process was conducted in triplicate and data were transferred and analyzed using FlowJo software (Tree Star, Oregon, OR, USA).

Mouse tumor xenograft and drug resistance assay in vivo

Male BALB/c nude mice (Sinogenetic Biotechnology, Beijing), aged 4–5 weeks, were randomly divided into four groups (five mice each group). Approximately 1 × 107 cells were resuspended in 200 µL of PBS and then inoculated subcutaneously into the armpit of each mouse. 10 days after the inoculation, the tumor-bearing mice were administrated intraperitoneally with 5-Fu (30 mg/kg) every 3 days. Tumor lengths and widths were measured using calipers and tumor volumes were calculated according to the formula: length × width2/2. Mice were sacrificed at the end of 3 weeks after the administration. Tumors were carefully separated and weighed. The procedure abided by the National Institutes of Health guide for the care and use of Laboratory animals and approved by the Ethics Committee of Animal Experiments of Peking Union Medical College Hospital.

Statistical analysis

Data were shown as mean ± standard deviation (SD) of more than three separate experiments to ensure accuracy. Statistical comparisons between groups were performed by Student’s t-test or one-way analysis of variance (ANOVA). Correlations were assessed using Spearman rank-order correlation. Actuarial rates of survival were analyzed and compared using Kaplan–Meier methods and log-rank tests. All statistical analyses were performed using SPSS 22.0 (SPSS Inc., Chicago, IL, USA) in conjunction with GraphPad Prism 8 (GraphPad Prism Software, Inc., San Diego, CA, USA) and p-value < 0.05 indicated a statistically significant difference (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001).