Cell lines and treatment

Two human ovarian cancer cell lines (SKOV3 and A2780) were originally obtained from ATCC and kept in our lab. The cell lines were grown at 37oC in a humid atmosphere containing 5% CO2 with RPMI 1640 medium, supplemented with 10% fetal calf serum and 100U/mL penicillin plus 100 g/mL streptomycin (all from gibco). Cytokines were purchased from PeproTech. To test the regulation role of IFN-γ to HLA-E expression, the tumor cells were seeded into 6-well plate (2 × 105/well) for 24 h, then treated with 25ng/mL recombinant human IFN-γ for another 24 h. To explore the specific mechanisms of IFN-γ regulation, IFN-γ-treated tumor cells were further cultured with or without Caerulomycin A (MCE, HY-114,495) for 48 h, or MG-132 (MCE, HY-13,259) for 6 h. To completely block JAK signaling pathway, cells were pretreated with 50µM Ruxolitinib (Selleck, S1378) for 1 h, followed by IFN-γ stimulating for 24 h in the presence of the inhibitor.

Detection of HLA-E expression by real-time PCR

Cell RNA was extracted using the RNA-Quick purification kit (ES Science). Up to 1 µg total RNA was reverse transcribed to cDNA by PrimeScript RT reagent kit (Takara). The primers were as follows: HLA-E forward: 5’-CCG TCA CCC TGA GAT GGA AG-3’, HLA-E reverse: 5’-ACA GCT CCA GAG ACC ACA GA-3’; β-actin forward: 5’-ACA CTG TGC CCA TCT ACG AGG-3’, β-actin reverse: 5’-AGG GGC CGG ACT CGT CAT ACT-3’. Quantitative PCR assay was performed using Hieff qPCR SYBR Green Master Mix (YEASEN) and Applied Biosystems QuantStudio Dx system. All PCR reactions were performed in triplicate. The relative expression value for each sample was shown as 2−ΔΔCt.

Western blot analysis

Homogenization of cells was carried out in cell lysis buffer containing PMSF and protease inhibitor cocktail (all from Beyotime), then proteins were separated in SDS-PAGE gel and electroblotted onto nitrocellulose membranes. The primary antibodies were used as follows: HLA-E antibody (Abcam, MEM-E/02, 1:1000), Vinculin antibody (CST, #4650, 1:500), GAPDH antibody (Proteintech, 60004-1-Ig, 1:50000), IFN-γ signaling pathway antibody sampler kit containing antibodies of JAK1, p-JAK1, JAK2, p-JAK2, STAT1, p-STAT1 (CST, #44,902, 1:1000), which were incubated at 4 oC overnight. After washing, the secondary antibody conjugated to horseradish peroxidase (HRP) was incubated at room temperature for 1 h. The protein bands were detected with ELC reagent (YEASEN) on Amersham Imager 600 (GE Healthcare Life Sciences).

Flow cytometry

3D12-APC mAb or mouse IgG1 kappa isotype control APC (eBioscience) was incubated at 4 oC for 30 min. After washing twice, flow cytometry was used to detect surface HLA-E expression on ovarian cancer cells. The data were acquired on Cytoflex S (Beckman Coulter), and analyzed using FlowJo software.

NK lysis assay

NKL cell line was purchased from ATCC. After ovarian cancer cells were treated, NKL (5 × 105/well) cells were added and cultured for another 4 h. To observe the survived cells after killing, the ovarian cancer cells were transfected with EGFP gene. The real-time cell history recorder (Chinchilla life sciences) was used to survey and record changes at the same location. Meanwhile, LDH cytotoxicity assay kit (Beyotime) was used to evaluate the quantity of cells which were killed by NKL cells.

Proteasome activity assay

The ovarian cancer cells treated with IFN-γ were homogenized in 0.5% NP-40 (Sangon Biotech). Proteasome activity assay kit (Abnova) was used according to the manufacturer’s instructions. In brief, each sample was added into two wells (one with proteasome inhibitor MG-132), then incubated in reaction buffer and AMC-tagged peptide substrate. One hour later, the highly fluorescent AMC was measured in a microplate reader (SpectraMax iD3, Molecular Devices). The RFU generated by proteasome is ΔRFU = RFU total - RFU inhibitor. Apply the ΔRFU to AMC standard curve to get proteasome activity concentration.

Patients and tissue samples

A total of 20 ovarian cancer tissues and 20 normal ovary tissues were obtained from patients who underwent surgical resection at Fudan University Shanghai Cancer Center before receiving any other treatments. The normal ovary tissues were from patients with cervical cancer who were willing to remove ovary to avoid recurrence. Each specimen was evaluated in hematoxylin and eosin (HE)-stained sections by pathologists, and the images were available in additional file 1. All specimens were fixed in formalin and embedded in paraffin, and the tumor or normal tissues were selected to make a tissue microarray. The study’s protocol was reviewed and approved by the Ethics Committee of Fudan University Shanghai Cancer Center (certification No. 050432-4-1212B). Informed consents were obtained from each patient.

Multiplexed immunohistochemistry

The tissue microarray was stained with four fluorescent dyes by tyramide signal amplification (TSA) kit (WiSee Biotechnology). Primary antibodies to HLA-E (Abcam, MEM-E/02), IFN-γ (R&D, #25,718), CD3 (Yuanxibio, #YX22005) and CD56 (GeneTech, #GT200502) were sequentially applied, followed by HRP-conjugated secondary antibody (Yuanxibio, #A10011-60) incubation. The slides were microwave heat-treated after each TSA operation, and then stained with DAPI after all the antibodies above being bound. The stained slides were scanned to obtain whole image with Aperio Versa 8 tissue imaging system. The total cells in each sample were identified and analyzed by Indica Halo software.

Statistical analysis

Statistical analysis of the data was performed using SPSS version 22.0 software. The data were compared between two groups using an independent Student’s t-test. Pearson analysis was applied to evaluate the correlation between IFN-γ and HLA-E/CD3/CD56 or PFS. Differences were considered significant when P < 0.05.