This study was conducted using the following three breast cancer cell lines: estrogen receptor (ER)- and progesterone receptor (PR)-positive human breast cancer cell line MCF-7 (Korean Cell Line Bank, Korea), human epidermal growth factor receptor 2 (HER2)-positive human breast cancer cell line MDA-MB-453 (Korean Cell Line Bank), and triple-negative human breast cancer cell line HCC-70 (Korean Cell Line Bank) [10]. All cell lines were maintained in RPMI-1640 medium supplemented with 10% fetal bovine serum and 1% penicillin (Welgene, Korea). Because sevoflurane is a strong organic solvent capable of interacting with plastic materials to produce impurities [11], breast cancer cells were cultured in poly-L-lysine-coated (Sigma-Aldrich, USA) glass culture dishes.

The human NK cell line, NK92-MI (ATCC, USA), was maintained in an α-minimum essential medium containing fetal bovine serum (12.5%), horse serum (12.5%), 2 mercaptoethanol (0.1 mM), and L-glutamine (2 mM). All cell lines were cultured according to their specifications and incubated at 37°C in humidified air containing 5% CO₂.

As previously described [12–14], 100 μl sevoflurane (Sevoprane; Ilsung, Korea) was diluted in 10 ml RPMI-1640 medium and stirred for a half-hour in an airtight, amber-colored glass bottle. The concentration of sevoflurane was determined using the preliminary data obtained from our gas chromatography-mass spectrometry analysis (GCMS-QP2010 Plus; Shimadzu, Japan) as (mean [SD]): 3.92 (1.26) mM. The sevoflurane stock was serially diluted to 1200 and 600 µM (S12 and S6, respectively) immediately before the experiments. Kharasch et al. [15] simultaneously measured the end-tidal sevoflurane concentration and plasma concentration of sevoflurane during general anesthesia and revealed that the average peak plasma concentration of sevoflurane reached 772 μM at an end-tidal sevoflurane concentration of 2.7% (equivalent to 1.3 minimum alveolar concentration). Consequently, we assumed clinically relevant concentrations of sevoflurane at 600 μM during general anesthesia. Furthermore, to investigate the dose-response relationship, we administered an additional dose of sevoflurane at a concentration of 1200 μM, which was higher than the recommended dose. The corresponding concentration of distilled water in the RPMI 1640 media were used as controls (0 μM).

The MCF-7, MDA-MB-453, and HCC-70 cells were then exposed to sevoflurane for 4 h. To account for evaporation-induced concentration reduction, both the sevoflurane and control group solutions were replaced every hour (Fig. 1) [13,14]. Previous studies [12,13] have shown that despite the volatility of sevoflurane, the concentration of the sevoflurane solution dissolved in cell culture media remains stable, with a < 10% loss over a one-hour period.

mRNA expression analysis was performed 18 h after treatment completion, while the other experiments (cell viability test, flow cytometry assay for surface expression of NKG2D ligands, western blotting analysis, enzyme-linked immunosorbent assay [ELISA] for soluble NKG2D ligands, and flow cytometry assay for NK cell-mediated cytotoxicity) were performed 24 h after treatment completion (Fig. 1).

The MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide) assay is based on the principle that mitochondrial activity in living cells converts MTT into formazan crystals. These crystals are dissolved following the addition of dimethyl sulfoxide (DMSO) and detected spectrophotometrically at 540 nm, with the absorbance directly proportional to cell viability [16]. As previously described [14], the MCF-7, MDA-MB-453, and HCC-70 cells were plated in 96-well plates (1 × 10⁴ cells/well) and incubated with the control or S12 solution for 4 h. Twenty-four hours later, the cells were incubated with the MTT solution (Sigma-Aldrich) for 4 h. The supernatant was discarded and the formazan crystals were dissolved using DMSO. The absorbance at 540 nm was determined using a microplate spectrophotometer (μQuant; Bio Tek, USA).

The mRNA expression analysis method used has been described previously [14]. Briefly, after a 4-h treatment and an additional 18-h incubation period, cancer cells were harvested (Fig. 1). The total RNA was extracted from the cells using a RNeasy^®Mini kit (Qiagen GmbH, Germany). Reverse transcription polymerase chain reaction (RTPCR) and multiplex PCR were performed. For denaturation, 3 μg extracted total RNA and 100 pmol random primers (Takara Shuzo, Japan) were incubated at 65°C for 5 min and chilled at 4°C for 4 min. Next, 6 μl of the 5x reaction buffer, 4 μl deoxynucleotide triphosphate (10 mM; Promega Co., USA), and 1.2 μl M-MLV RT (Promega Co.) were added and incubated at 37°C for 60 minutes. Multiplex PCR was performed using a QIAGEN® Multiplex PCR kit (Qiagen GmbH). The primer sets used to evaluate NKG2D gene expression were as follows: 1) MICA: ribosomal protein L19 (RPL19), MICA, and β-actin genes and 2) MICB and ULBP 1–3: RPL19, MICB, ULBP1–3, and β-actin genes. The primer sequences are listed in Table 1. All the experiments were performed according to the manufacturer’s instructions. PCR products were quantified using a microchip electrophoresis system MCE^®-202MultiNA (Shimadzu, Japan). For normalization, the mRNA band intensity of each NKG2D ligand was divided by that of the β-actin. To assess relative gene expression ratios, the normalized mRNA band intensity of the treated samples was divided by that of the controls.

The flow cytometry assay for the surface expression of NKG2D ligands was performed using the method described in our previous study [14]. Briefly, after a 4-h treatment and an additional 24-h incubation period (Fig. 1), the cells were harvested and incubated with 10 μg/ml mouse anti-MICA/B and ULBP1–3 or the corresponding isotype controls (anti-IgG2a or anti-IgG2b; R&D Systems, USA). Samples were then incubated with goat anti-mouse phycoerythrin (PE)-conjugated antibodies (BD Biosciences, USA). The mean fluorescence intensity (MFI) was measured using a FACSCanto II flow cytometer (BD Biosciences) and quantified using FlowJo (version 10.6.1; TreeStar, USA). The MFIs of the treated samples were divided by those of the controls to assess the relative expression ratios.

After a 4-h treatment and an additional 24-h incubation period (Fig. 1), western blot analysis was performed to evaluate MMP-1 and -2 expression. The cells were washed three times with cold phosphate-buffered saline and lysed in a PRO-PREP protein extraction solution (Intron, Korea). Equal amounts of cell extracts were resolved by 4%–20% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and analyzed by western blotting. The separated proteins were transferred onto polyvinylidene difluoride membranes (Millipore, USA). The membranes were then blocked with 3% BSA in Tris-buffered saline containing 0.1% Tween 20 at room temperature. The proteins of interest were detected using primary antibodies (MMP-1 and -2; Cell Signalling, USA) and horseradish peroxidase-conjugated secondary antibodies (Enzo Life Sciences, USA) according to the manufacturer’s instructions, and a chemiluminescence imaging system (AE-9150 Ez-capture II; Atto, Japan) was used to analyze the results. Each blot was probed with anti-β-actin antibodies (Sigma-Aldrich). Band intensity was quantified using ImageJ software (version 1.53; National Institutes of Health, USA). Protein expression in the treated cells was divided by that in the control cells to calculate relative protein expression ratios.

Breast cancer cells (4 × 106 cells) were plated on 60-mm glass culture dishes. After a 4-h treatment and an additional 24-h incubation period (Fig. 1), cell culture supernatants were centrifuged at 5000 rpm for 5 min at 4°C, and aliquots were stored at −80°C until further use. The levels of soluble NKG2D ligands (MICA for MCF-7 and MDA-MB-453; MICB for HCC-70) in the cell culture supernatant were measured using ELISA kits (MICA Human ELISA kit, Invitrogen, USA; Human MICB ELISA kit, MyBioSource, USA) following the manufacturer’s protocol. The absorbance at 450 nm was measured using a microplate spectrophotometer (Synergy H1; BioTek, USA). The samples were loaded in duplicate, and the mean soluble NKG2D ligand values were used for analysis.

After a 4-h treatment and an additional 24-h incubation period (Fig. 1), target cancer cells (MCF-7, MDA-MB-453, and HCC-70) were stained with carboxyfluorescein diacetate succinimidyl ester (CFSE; CellTrace^TM; Invitrogen) and co-cultured with NK92-MI cells, effector cells, for 4 h. With cellular cytotoxicity assays, optimizing the effector-to-target cell (E:T) ratio is crucial for maximizing and distinguishing differences in cytotoxicity among treatment groups. Typically, higher E:T ratios result in enhanced NK cell cytotoxicity because an increased number of effector cells relative to target cells leads to greater cytotoxicity [17]. The recommended E:T ratio for flow cytometry cytotoxicity assays is ≤ 10:1 [18]. As previously reported [14], we determined the E:T ratio to be between 1:1 (e.g., E:T = 1 × 10⁵ : 1 × 10⁵) and 10:1 (e.g., E:T = 1 × 10⁶ : 1 × 10⁵), with 10:1 as the optimal ratio. Co-cultured cells were labelled with 1 μg/ml propidium iodide (PI; Invitrogen). A FACSCanto^TM II flow cytometer and BD FACSDiva^TM Software (BD Biosciences) were used. The percentage of NK cell-mediated lysis (%) was calculated using the following equation:

where Q2 represents CFSE-positive and PI-positive cells and Q3 represents CFSE-positive and PI-negative cells.

MedCalc^® (version 20; MedCalc Ltd., Belgium) and IBM SPSS Statistics (version 25; IBM Corp., USA) were used for statistical analyses. Variables are presented as medians with the first and third quartiles (Q1, Q3). For comparisons between groups, Mann-Whitney U tests or Kruskal-Wallis tests were performed. If the Kruskal-Wallis test was significant, post-hoc comparisons using the Conover method were conducted.

Our trial consisted of a zero-dose control group (C) and two sevoflurane treatment groups (S6 and S12) that received increasing doses of sevoflurane in the following order: C, S6, S12. As a secondary outcome, the dose-response relationship was evaluated using the Jonckheere-Terpstra trend test. Statistical significance was set at P < 0.05.

Sevoflurane was not found to affect the viability of MCF-7, MDA-MB-453, or HCC-70 cells. The relative cell viabilities (%) related to the controls measured by MTT assay were as follows (median [Q1, Q3]): MCF-7 at S12: 100.2 (92.4, 118.9), P = 1.000; MDA-MB-453 at S12: 101.4 (96.2, 109.0), P = 1.000; and HCC-70 at S12: 89.4 (87.3, 107.0), P = 0.305 (n = 6 per group).

The results of sevoflurane exposure on the mRNA expression of NKG2D ligands are summarized in Fig. 2 (n = 6 per group). In MCF-7, MDA-MB-453, and HCC-70 cell lines, mRNA expression of MICA, MICB, ULBP2, and ULBP3 was observed, but the expression of ULBP1 was not observed.

In MCF-7 cells, the relative mRNA expression ratios of MICB at S6 and S12 were lower than those in the controls (Kruskal-Wallis test: P = 0.013). In MDA-MB-453 cells, the relative mRNA expression ratios of MICA at S6 and S12 were lower than those in the controls (Kruskal-Wallis test: P = 0.013) and the relative surface expression ratio of ULBP3 was downregulated at S12 compared with the controls (Kruskal-Wallis test: P = 0.044). In HCC-70 cells, the relative mRNA expression ratios of ULBP2 and ULBP3 at S6 and S12 were lower than those in the controls (Kruskal-Wallis test: P = 0.002 and P = 0.003, respectively).

The secondary outcome results showed a dose-response relationship in MCF-7 cells, demonstrating significantly lower relative mRNA expression levels of MICA, MICB, ULBP2, and ULBP3 in response to increasing concentrations of sevoflurane (Jonckheere-Terpstra trend test: P = 0.015, P = 0.008, P = 0.036, and P = 0.043, respectively). Similarly, sevoflurane downregulated the mRNA expression of MICA and ULBP3 in MDA-MB-453 cells in a dose-dependent manner (Jonckheere-Terpstra trend test: P = 0.002 and P = 0.029, respectively). Likewise, sevoflurane downregulated the mRNA expression of MICA, MICB, ULBP2, and ULBP3 in HCC-70 cells in a dose-dependent manner (Jonckheere-Terpstra trend test: P = 0.043, P = 0.043, P = 0.006, and P = 0.012, respectively).

Fig. 3 summarizes the results of the flow cytometry analysis of NKG2D ligand surface expression (n = 6 per group). Consistent with mRNA expression, ULBP1 was rarely expressed on the surface of the MCF-7, MDA-MB-453, and HCC-70 cell lines.

In MCF-7 cells, MICA and ULBP2 were predominantly expressed; the relative surface expression ratios of MICA and ULBP2 at S6 and S12 were downregulated compared to those in the controls (Kruskal-Wallis test: P = 0.002 and P ≤ 0.001, respectively).

MDA-MB-453 cells predominantly exhibited surface expression of MICA and ULBP2. The relative surface expression ratios of MICA, ULBP2, and ULBP3 at S6 and S12 were lower than those in the controls (Kruskal-Wallis test: P < 0.001, P = 0.003, and P = 0.004, respectively).

In HCC-70 cells, MICB, ULBP2, and ULBP3 were predominantly expressed. The relative surface expression ratios of MICA, MICB, ULBP2, and ULBP3 at S6 and S12 were downregulated compared with those in the controls (Kruskal-Wallis test: P = 0.002, P = 0.017, P = 0.002, and P = 0.013, respectively).

In the dose-response analysis, we observed significantly lower relative surface expression ratios of MICA, MICB, and ULBP2 in MCF-7 cells with increasing concentrations of sevoflurane (Jonckheere-Terpstra trend test: P < 0.001, P = 0.024, and P < 0.001, respectively). Similarly, sevoflurane decreased the relative surface expression ratios of MICA, ULBP2, and ULBP3 in MDA-MB-453 cells in a dose-dependent manner (Jonckheere-Terpstra trend test: P < 0.001, P = 0.002, and P = 0.001, respectively). Likewise, sevoflurane downregulated the relative surface expression ratios of MICA, MICB, ULBP2, and ULBP3 in HCC-70 cells in a dose-dependent manner (Jonckheere-Terpstra trend test: P = 0.001, P = 0.010, P < 0.001, and P = 0.029, respectively).

Western blot analysis revealed no changes in the protein expression of MMP-1 and -2 between the control and sevoflurane treatment groups in MCF-7, MDA-MB-453, and HCC-70 cells (Fig. 4A; n = 6 per group).

In MCF-7 cells, the median (Q1, Q3) of the relative protein expression ratio of MMP-1 was as follows: controls, 1.7 (0.9, 2.4); S6, 1.6 (0.8, 2.5); and S12, 1.5 (0.7, 2.3). P values for the Kruskal-Wallis and Jonckheere-Terpstra trend tests were 0.519 and 0.258, respectively. The median (Q1, Q3) of the relative protein expression ratio of MMP-2 was as follows: controls, 1.3 (0.8, 2.4); S6, 1.3 (1.1, 2.0); and S12, 1.3 (0.8, 2.1). P values for the Kruskal-Wallis and Jonckheere-Terpstra trend tests were 0.854 and 0.686, respectively.

In MDA-MB-453 cells, the median (Q1, Q3) of the relative protein expression ratio of MMP-1 was as follows: controls, 1.3 (0.9, 1.5); S6, 2.0 (0.9, 5.3); and S12, 1.5 (0.7, 3.9). P-values for the Kruskal-Wallis and Jonckheere-Terpstra trend tests were 0.653 and 0.628, respectively. The median (Q1, Q3) of the relative protein expression ratio of MMP-2 was as follows: controls, 0.9 (0.5, 1.1); S6, 0.8 (0.7, 0.9); and S12, 0.8 (0.8, 1.3). P values for the Kruskal-Wallis and Jonckheere-Terpstra trend tests were 0.778 and 0.746, respectively.

In HCC-70 cells, the median (Q1, Q3) of the relative protein expression ratio of MMP-1 was as follows: controls, 1.7 (1.0, 2.3); S6, 1.2 (0.8, 1.7); and S12, 1.4 (0.8, 2.1). P-values for the Kruskal-Wallis and Jonckheere-Terpstra trend tests were 0.423 and 0.293, respectively. The median (Q1, Q3) of the relative protein expression ratio of MMP-2 was as follows: controls, 0.8 (0.4, 1.0); S6, 0.7 (0.4, 1.0); and S12, 0.9 (0.4, 1.0). P values for the Kruskal-Wallis and Jonckheere-Terpstra trend tests were 0.911 and 0.808, respectively.

According to the ELISA results, no significant changes were detected in the levels of soluble NKG2D ligands between the control and sevoflurane-treated groups in MCF-7, MDA-MB-453, or HCC-70 cells (Fig. 4B, n = 6 per group).

According to the Jonckheere-Terpstra test results, no dose-response relationship was identified. P values for the MCF-7, MDA-MB-453, and HCC-70 cells were 0.467, 0.125, and 0.686, respectively.

The flow cytometry results are shown in Fig. 5 (n = 4 per group). In all cell lines, NK cell-mediated cytotoxicity at S6 and S12 was lower than that in the controls. Compared with the controls, the P values were as follows: MCF-7 at S6: 0.005 and MCF-7 at S12: 0.005; MDA-MB-453 at S6: 0.005 and MDA-MB-453 at S12: 0.005; and HCC-70 at S6: 0.005 and HCC-70 at S12: 0.005 (Fig. 5; effect cells:target cells = 10:1, n = 4 per group).

Analysis of the secondary outcomes using the Jonckheere-Terpstra trend test revealed that sevoflurane reduced NK cell-mediated cytotoxicity in MCF-7, MDA-MB-453, and HCC-70 cells in a dose-dependent manner (effector cells:target cells = 10:1; P = 0.040, P = 0.040, and P = 0.040, respectively).