KPL-1 is a human breast cancer cell line established from the malignant effusion of a breast cancer patient (30). This cell line was derived from a patient with recurrent breast cancer that appeared during postsurgical adjuvant chemoendocrine therapy including tamoxifen and medroxy-progesterone acetate. The KPL-1 cell line is estrogen receptor (ER)-positive, progesterone receptor (PgR)-negative, and human epidermal growth factor receptor-2 (HER2)-negative (luminal A subtype) (31). Xenograft athymic mice injected with KPL-1 cells often develop lymph node metastasis (30). KPL-1 cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Sigma, St. Louis, MO, USA) with 10% fetal bovine serum (FBS; Gibco-BRL, Grand Island, NY, USA) in 5% CO₂/95% humidified air at 37°C.

MA used for in vitro experiments was purchased from Sigma and dissolved to 5 mg/500 μl (32.6 mM) in ethanol and stored at −30°C. MA was diluted in DMEM containing 10% FBS to achieve final concentrations of 16.25, 32.5, 75, 150 and 300 μM.

KPL-1 cells were seeded at 2×10³ cells/well in 96-well plates in DMEM with 10% FBS. The cells were treated with ethanol (final concentration, 0.9%) or the indicated concentration of MA for up to 72 h. Cell proliferation was monitored by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) assay as previously described (32), and IC₅₀ value was calculated.

Animals were housed in groups of four or five in plastic cages with paper bedding (Paper Clean, SLC, Hamamatsu, Japan) in a specific pathogen-free room maintained at 22±2°C and 60±10% relative humidity with a 12-h light/dark cycle (lights on at 8:00 AM and lights off at 8:00 PM). The mice were randomly divided into two groups, the control diet group (n=15) and the MA diet group (n=10). Both experimental diets were modifications of the AIN-76 diet (33) and contained the same amount of nutrients but had different fatty acid compositions (Table I). The MA diet contained 5% SUNTGM33 (a kind gift from Suntory Wellness, Tokyo, Japan), which contains 48.0% MA (Table II). SUNTGM33 is a microbial oil obtained by fungal fermentation (34). Olive oil purchased from Nacalai Tesque (Kyoto, Japan) was used for the control diet. The fatty acid compositions of SUNTGM33 and olive oil are listed in Table II. Each experimental diet was formulated by Oriental Yeast (Tokyo, Japan). Four-week old female athymic BALB/c mice purchased from Charles River Japan (Kyoto, Japan) were fed either a control or MA diet starting at 4 weeks of age. In comparison to the control diet, the MA diet for 8 days starting at 4 weeks of age did not cause weight gain. Therefore, the MA diet group was switched to the control diet for 4 successive days to accelerate weight gain; thereafter, the MA diet group was treated with a cyclic feeding regime of 5 days on the MA diet followed by 2 days on the control diet throughout the experiment. The control group was treated with the control diet throughout the experiment. Schematic representation of the in vivo experiment is shown in Fig. 2. At 6 weeks of age, both groups were inoculated with 2.5×10⁵ KPL-1 cells in 100 μl DMEM supplemented with 10% FBS into the thoracic mammary fat pad (MA diet group was on the third day of MA diet). During the experiment, the dose of diet ingested, body weight and locally growing tumor volumes were measured once per week. Tumor volume was calculated by using the standard formula: width² × length × 0.5. Thirty-seven days after tumor cell inoculation (fifth day on MA diet), body weight was measured, and bromodeoxyuridine (BrdU; Invitrogen, Camarillo, CA, USA) was injected (50 mg/kg animal body weight) via the abdominal cavity, blood was sampled and then animals were sacrificed by exsanguination from aortic transection. At autopsy, all organs were examined macroscopically, and the primary tumors and local axillary lymph nodes were examined histologically; half of the primary tumor was snap frozen and used for fatty acid analysis. Tissues were fixed in 10% neutral buffered formalin, embedded in paraffin, and stained with hematoxylin and eosin (H&E). The study protocol and animal procedures were approved by the Animal Care and Use Committee of Kansai Medical University (permit no. 13-060). Throughout the experiments, animals were cared for in accordance with the Guidelines for Animal Experimentation of Kansai Medical University.

The fatty acid composition of the total phospholipid fraction of serum was determined. Total lipids were extracted by the method of Bligh and Dyer (35). The total phospholipid fraction was separated by thin-layer chromatography. For an internal standard, 1,2-diheptadecanoyl- sn-glycero-3-phosphocholine (Avanti Polar Lipids, Inc., Alabaster, AL, USA) was added. Total phospholipid fractions were transmethylated with HCl-methanol and then the fatty acid composition was analyzed by gas chromatography (GC-2014; Shimadzu Corporation, Kyoto, Japan) with a capillary column DB-225 (0.25 mm × 30 m × 0.25 μM; J&M Scientific, Folsom, CA, USA). The entire system was controlled with gas chromatography software (GCsolution; Shimadzu Corporation). The fatty acid composition of the total lipid fraction of KPL-1 tumors was determined. In brief, frozen tumor tissues were thawed, minced and homogenized three times in 8 ml chloroform-methanol (2:1) by a polytron homogenizer (Kinematica, Lucerne, Switzerland) for 10 sec. The fatty acid analysis of total lipids in the tumor was performed by the same method as previously mentioned (35,36).

Microvessel density of primary KPL-1 tumors was evaluated by anti-CD34 antiserum (Abcam, Cambridge, UK). The cell kinetics (cell proliferation and cell death) in primary KPL-1 tumors were evaluated. Cell proliferation was evaluated by anti-BrdU antibody (clone B44, 1:50; Becton-Dickinson, Franklin Lakes, NJ, USA) by using an LSAB staining kit (Dako, Glostrup, Denmark). Cell death was evaluated by anti-phospho-histone H2A.X (γ-H2AX) antibody (Ser139, 1:100; Cell Signaling, Danvers, MA, USA), an immunomarker of the DNA damage response. Each slide was scanned with a high-resolution digital slide scanner (NanoZoomer 2.0 Digital Pathology; Hamamatsu Photonics, Hamamatsu, Japan) to prepare digital images. The ndpi image files were opened in color mode with NDP.view software (Hamamatsu Photonics). The images were changed to jpeg files at ×40 magnification in five randomly selected areas within each tumor that were used to analyze immunohistochemical staining (37–39).

Cell blocks were prepared from KPL-1 cells cultured with or without 214.2 μM MA for 72 h (IC₅₀ value for 72 h). Cell blocks were prepared from cultured KPL-1 cells; cells were centrifuged at 1,000 rpm for 5 min, and cell pellets were fixed in 10% neutral buffer formalin and embedded in paraffin. Antisera used to detect CD34, VEGF, VEGFR1, VEGFR2 and E-cadherin were anti-CD34 antiserum (1:20; Abcam), anti-VEGF antiserum (1:50), anti-VEGFR1 antiserum (1:500) (both from Santa Cruz Biotechnology, Santa Cruz, CA, USA) and anti-VEGFR2 antiserum (1:400) (Cell Signaling Technology, Danvers, MA, USA), respectively, by using an LSAB staining kit (Dako), and anti-E-cadherin antibody (NCH38 ready to use) by using Histofine MAX-PO (both from Nichirei Biosciences, Tokyo, Japan) according to the manufacturer’s instructions. The reaction products were visualized with 3-3′-diaminobenzidine tetrahydrochloride. CD34 immunohistochemistry was applied for angiogenesis evaluation in KPL-1 tumor grown in athymic mice, and VEGF, VEGFR1, VEGFR2 and E-cadherin immunohistochemistry, respectively, were applied in cultured KPL-1 cells in cell pellets.

KPL-1 cells cultured with or without 214.2 μM MA for 72 h (IC₅₀ value for 72 h) were sampled. VEGF was quantified in the culture supernatant, and VEGFR1, VEGFR2 and E-cadherin were measured in the cell lysate. For cell lysate preparation, after washing cells with PBS (−), cell pellets were homogenized with RIPA buffer (Wako, Osaka, Japan), and the cell lysate was centrifuged at 14,000 rpm for 15 min at 4°C to obtain supernatant containing cell lysate. Protein concentrations were measured by the DC protein assay kit (Bio-Rad, Hercules, CA, USA). VEGF, VEGFR1, VEGFR2 and E-cadherin levels were measured by enzyme-linked immunoabsorbent assay (ELISA) with a human VEGF assay kit (IL-226; IBL, Fujioka, Japan), VEGFR1 human ELISA kit (ab119613), VEGFR2 human ELISA kit (ab100665) (both from Abcam), and E-cadherin human ELISA kit (R&D Systems, Minneapolis, MN, USA), respectively, according to the manufacturers’ protocols. Each protein standard or protein purified from each group was incubated overnight at 4°C against VEGF or E-cadherin; 90 min at 37°C against VEGFR1; or 150 min at room temperature against VEGFR2. After the final colors were developed with the addition of each color reagent, absorbance was measured at 450 nm. Cell sample concentration was calculated from a standard curve and corrected for protein concentration. The limits for detection for VEGF, VEGFR1, VEGFR2 and E-cadherin were 8, 156, 70 and 39 pg/ml, respectively.

Values are expressed as the means ± standard error of the mean (SEM). Body weight, tumor volume, tumor weight, fatty acid composition, CD34-positive area and the number of BrdU- and γ-H2AX- positive cells per 1 mm² among the groups were analyzed by t-test. The incidence of metastasis was analyzed with the χ² test.

The KPL-1 cells were treated with 5 concentrations (16.25–300 μM) of MA for up to 72 h. The MTT assay revealed that MA induced growth inhibition in a dose- and time-dependent manner (Fig. 3). The IC₅₀ value of MA against KPL-1 cells was 214.2 μM (65.7 μg/ml) for a 72-h treatment.

During the experiment, the daily dose of MA ingestion was 63–80 mg/mouse/day (mean, 74 mg/mouse/day). Body weight gain in the MA group was significantly smaller during the experiment but at the end of the experiment (eating the respective diet for 8 weeks), the difference in body weight between the MA and control diet group was not statistically significant (Fig. 4). No organs or tissues were macroscopically abnormal. Throughout the experiment, the MA diet group ingested less food than the control diet group (average, 92.1%).

Locally growing KPL-1 tumors in the MA diet group grew slower than in the control diet group. The final tumor volume was significantly smaller (Fig. 5A) and the final tumor weight was significantly lighter (Fig. 5B) in the MA diet group as compared to the control diet group (p<0.01 and p<0.05, respectively). Tumor volume two days before the termination of the experiment was 872±103 and 376±66 mm³, and the final tumor weight was 1,000±116 and 517±84 mg in the control and MA group, respectively. Regional (axillary) lymph node metastasis was found in both groups (Fig. 5C). Similarly, as compared with the control diet group, metastasis was significantly suppressed in the MA diet group (66.7%: 10/15 vs. 10%: 1/10).

Although the growth of KPL-1 tumors was suppressed in the MA diet group, the morphology was comparable in both groups (data not shown). The angiogenesis of KPL-1 tumors at the primary site (Fig. 6A) as well as the CD34-stained area was compatible between the control and MA diet groups (Fig. 6B).

To compare the cell kinetics of KPL-1 tumors (cell proliferation and cell death), the number of BrdU-positive cells and γ-H2AX-positive cells per 1,000 cells in primary tumors from the control and MA diet groups of mice was compared. The proliferation and apoptotic ratios are shown in Fig. 7A and B, respectively. The proliferation ratio in the control and MA diet groups was 11.8±0.7 and 8.7±0.9%, respectively (p<0.05), while the apoptotic ratio in the control and MA diet groups was 8.7±1.3 and 9.6±1.2% (not statistically significant).

Fatty acid composition of sera and KPL-1 tumors in the control and MA diet group is shown in Fig. 8A and B, respectively. In the sera, the concentration of MA was significantly higher in the MA diet group as compared to controls (237.2±17.6 vs. 23.8±1.1 μg/ml). In contrast, LA and AA were significantly lower in the MA diet group as compared with controls. In the KPL-1 tumor, the fatty acid profile was generally similar to that of the fatty acid composition in the serum. However, the serum OA composition was significantly higher in the control diet group, as MA was replaced by OA in this diet, while the OA concentration in KPL-1 tumors was comparable. The MA diet significantly decreased the n-6/n-3 ratio in the sera and the tumors (Fig. 9).

KPL-1 cells strongly expressed VEGF, VEGFR1 and VEGFR2 in the cytoplasm and E-cadherin at the cell surface. Although VEGF expression was unchanged, VEGFR1 and VEGFR2 expression tended to diminish and E-cadherin expression tended to increase in KPL-1 cells treated with the IC₅₀ dose of MA for 72 h (214.2 μM) (Fig. 10).

Levels of VEGF, VEGFR1, VEGFR2 and E-cadherin in KPL-1 cells cultured with or without the IC₅₀ dose of MA for 72 h were compared. Although the VEGF levels in KPL-1 tumor cells were similar (Fig. 11A), MA treatment significantly decreased VEGFR1 and VEGFR2 levels and tended to increase E-cadherin levels (Fig. 11B–D).