The human ovarian cancer cell line, SKOV-3, was purchased from The Shanghai Stem Cell Institute, and was cultured in DMEM (Invitrogen; Thermo Fisher Scientific, Inc.) supplemented with 10% FBS (Gibco; Thermo Fisher Scientific, Inc.) and 1% penicillin-streptomycin mixture (Sigma-Aldrich; Merck KGaA) at 37°C in a humidified atmosphere with 5% CO₂.

CD44⁺CD117⁺ cells were then separated from SKOV-3 cells using the magnetic-activated cell sorting kit (Miltenyi Biotec GmbH) according to the manufacturer's instructions. The kit included CD44⁺ and CD117⁺ antibodies. CD44⁺ cells were sorted using a mouse anti-human CD44⁺ antibody combined with magnetic microbeads. After magnetic column depletion, among CD44⁺ SKOV3 cells, CD117⁺ cells were isolated using CD117⁺ antibody combined with magnetic microbeads. Moreover, among CD44⁻ cells, the CD117⁺ subset was isolated using CD117⁺ antibody combined with magnetic microbeads. The CD44⁺CD117⁺ stem-like subsets and CD44⁻CD117⁻ cells were harvested after magnetic column depletion and maintained in a DMEM/F12 medium containing 5 µg/ml insulin (Sigma-Aldrich; Merck KGaA), 10 ng/ml basic fibroblast growth factor (bFGF; Sigma-Aldrich; Merck KGaA), 20 ng/ml human recombinant epidermal growth factor (EGF; Sigma-Aldrich; Merck KGaA) and 0.5% FBS.

According to GenBank (NM_017617.4), a siRNA sequence of human Notch1 was designed. The sequence was as follows: siNotch1-1 forward (F), 5′-AGUGGACAUCAGUACUGUA-3′ and reverse (R), 3′-UACAGUACUGAUGUCCACU-5′; siNotch1-2 F, 5′-AUGCGGGCAAGUGCAUCAACA-3′ and R, 3′-UGUUGAUGCACUUGCCCGCAU-5′; and siNotch1-3 F, 5′-GCUACACAGGGAGCAUGUGUA-3′ and R, 3′-TACACATGCTCCCTGTGTAGC-5′. The negative control (siNC) was designed as follows: F, 5′-UUCUCCGAACGUGUCACGUTT-3′ and R, 3′-ACGUGACACGUUCGGAGAATT-5′. All primers were synthesized and recombined to the plasmids (pcDNA3.1) by Shanghai GenePharma Co., Ltd.. After cells were inoculated overnight in a 6-well plate with 60% confluency, 20 µl siNotch1 or siNC recombinant were transfected into cells using Lipofectamine^® 2000 (Invitrogen; Thermo Fisher Scientific, Inc.), according to the manufacturer's protocol at 37° and 5% CO₂ in an incubator and for the following experiments 48 h later as indicated.

In total, 16 male nude mice (age, 6–8 weeks; weight, 25–30 g) were purchased from the Experimental Animal Center of the Shandong University and raised at the Experimental Animal Center. Mice were housed at 25±1°C with 12-h light/dark cycles, 50±5% humidity and free access to food and water. The mice were randomly divided into four groups: i) CD44⁺CD117⁺ + PTX (Beijing Solarbio Science & Technology Co., Ltd.) group (n=4); ii) CD44⁺CD117⁺ + PTX + CCL20 (Beijing Solarbio Science & Technology Co., Ltd.) group (n=4); iii) CD44⁻CD117⁻ + PTX group (n=4); and iv) CD44⁻CD117⁻ + PTX + CCL20 group (n=4). For xenografts, CD44⁺CD117⁺ (2×10⁶) subsets or CD44⁻CD117⁻ subsets cells (2×10⁶) suspended in 0.5 ml PBS were subcutaneously injected into the left flank of each nude mouse to establish tumors. Then, PTX (10 mg/kg) and/or CCL20 (1 mg/kg) were injected intraperitoneally simultaneously on days 1, 3 and 7. Tumor size was monitored on the 3rd, 7th, 14th, 21st and 28th days (15). At the end of the experiment and under continuous anesthesia, the animals were euthanized by an overdose of pentobarbital (125 mg/kg). All experiments were approved by The Animal Care and Use Committee of the Shandong Provincial Hospital, Cheeloo College of Medicine, Shandong University.

In a 24-well plate, CD44⁺CD117⁺ or CD44⁻CD117⁻ cells were harvested, counted and seeded with a low attachment (Corning, Inc.) at a density of 1.0×10³ per well after isolation or treatment. Then, the cells were maintained in a DMEM/F12 medium (Sigma-Aldrich; Merck KGaA), without FBS, supplemented with 4 mg/ml heparin (Sigma-Aldrich; Merck KGaA), 20 ng/ml EGF (Sigma-Aldrich; Merck KGaA), 20 ng/ml bFGF (Sigma-Aldrich; Merck KGaA) and 2% B27 (Invitrogen; Thermo Fisher Scientific, Inc.). After 10–14 days, spheres were analyzed and images were captured with an inverted fluorescent microscope (Nikon Ni-U; Nikon Corporation; magnification, ×100).

The IC₅₀ of cells exposed to PTX was determined using a Cell Counting Kit (CCK)-8 assay (Dojindo Molecular Technologies, Inc.) according to the manufacturer's instructions. In brief, cells were inoculated at a concentration of 4×10³/100 µl per well and cultured at 37°C for 12 h in a 96-well plate. The cells were then maintained at 37°C with 200 µl fresh medium containing different concentrations of PTX (0–50 nM) for 48 h. Following this, the cells were incubated with 20 µl CCK-8 solution at 37°C for 24 h. Then, the optical density value of each well was detected using a microplate reader at 450 nm (Thermo Fisher Scientific, Inc.).

For ROS detection, 5×10⁵ cells were seeded overnight in a 6-well plate and then treated with PTX (5 nM) at 37°C for 48 h. Subsequently, 5 µM carboxy 2′-7′-dichlorofluorescein diacetate was added in Hanks' Balanced Salt Solution and incubated with cells at room temperature for 0.5 h. Fresh tissues were harvested, followed by a single cell suspension preparation. Tissue was digested with trypsin to obtain single cell suspension. Cells were then inoculated in a 6-well plate as aforementioned. The levels of intracellular peroxide were measured using a flow cytometer (BD FACSAria; BD Biosciences) at 485 nm excitation and 520 nm emission to analyze the ROS generation.

The apoptosis of cells was determined using flow cytometry and Hoechst assays. SKOV3 cells were seeded in a 6-well plate for flow cytometry and treated with PTX (5 nM) at 37°C for 48 h. The adherent and floating cells were harvested, washed with cold PBS and stained with FITC-conjugated Annexin V/PI (Beijing Solarbio Science & Technology Co., Ltd.), according to the manufacturer's instructions. The double-stained cells were then detected via Accuri C6 plus flow cytometry (BD Bioscience).

For Hoechst staining, 5×10⁵ cells were seeded overnight in a 6-well plate and treated with PTX (5 nM) at 37°C for 48 h. Cells were then stained with 0.5 µg/ml Hoechst 33342 dye in a 5% CO₂ incubator at 37°C for 25 min and were analyzed using a Nikon Ni-U fluorescence microscope (Nikon Corporation; magnification, ×100).

The tumors were removed and were fixed with 4% paraformaldehyde at 37°C for >24 h. Specimens were then dehydrated in isopropyl alcohol by xylene and embedded in paraffin. According to the manufacturer's instructions, specimens were cut into slices (thickness, 5 µm) and stained with H&E (Nanjing Jiancheng Bioengineering Institute) at 37°C for 24 h. For subsequent analysis, the slices were dehydrated and sealed with neutral resins. Images were captured with an inverted fluorescent microscope (Nikon Ni-U; Nikon Corporation; magnification, ×100). A total of 3 field of view were observed by microscopy.

The apoptosis of tumor tissue slices was assessed using the Clicl-iT Plus TUNEL assay (Thermo Fisher Scientific, Inc.). The tumors were removed and were fixed with 4% paraformaldehyde at 37°C for 24 h. According to the manufacturer's instructions, specimens were sectioned at 5 µm and stained with TUNEL reagent (Nanjing Jiancheng Bioengineering Institute) at 37°C for 1 h. After staining, the sections were sealed with a Prolong Gold antifade reagent with DAPI and images were captured with an inverted fluorescent microscope (Nikon Ni-U; Nikon Corporation; magnification, ×100). A total of 3 field of view were observed by microscopy.

IHC was used to detect the expression level of CCR6 in xenograft mice. The tumors were removed and were fixed with 4% paraformaldehyde at 37°C for 48 h. To retrieve the antigen, 5-µm-thick slices from the center of paraffin-embedded tumor samples were collected, deparaffinized in xylene, rehydrated with gradient ethanol, immersed in a 10 mM citrate buffer (pH 6.0) and then heated in a microwave oven for 10 min. Subsequently, slices were incubated with 0.3% H₂O₂ at room temperature for 30 min to block the endogenous peroxidase activity and then with 1.5% normal goat serum at 37°C for 30 min (Takara Biotechnology Co., Ltd.) for blocking non-specific protein binding. Subsequently, the slices were immersed in the primary antibody in a humidity chamber at 4°C overnight. Slices were then incubated with a biotinylated goat anti-mouse IgG antibody (1:1,000; cat. no. AP124; Sigma-Aldrich; Merck KGaA) at 37°C for 30 min and visualized using the 3′3-diaminovbenzidine method. For further analysis, slices were then counterstained with hematoxylin for 30 min at 37°C and sealed with neutral resins. Images were captured with an inverted fluorescent microscope (Nikon Ni-U; Nikon Corporation; magnification, ×100).

TRIzol^® reagent (Takara Biotechnology Co., Ltd.) was used to isolate total RNA samples from cells and tissues according to the manufacturer's instructions. A RT-qPCR kit (Takara Biotechnology Co., Ltd.) was used for cDNA synthesis, according to the manufacturer's instructions. With cDNA as a template, RT-qPCR was performed using SYBR Green (Takara Biotechnology Co., Ltd.) in an ABI 6500 system (Applied Biosciences; Thermo Fisher Scientific, Inc.) under the following conditions: Initial denaturation at 95°C for 90 sec, followed by 40 cycles of 95°C for 15 sec, 58°C for 21 sec and 72°C for 24 sec. The primers of genes are presented in Table I. With GAPDH as the internal control, the relative gene expression was calculated using the 2^−ΔΔCq method (21).

The cells and tissue samples were lysed with a RIPA lysis buffer containing protease inhibitors (Sigma-Aldrich; Merck KGaA). The protein suspension was then collected and quantified using the BCA Protein Assay kit (Pierce; Thermo Fisher Scientific, Inc.). Then, 5 µg protein was boiled with a load buffer for 10 min, separated with 10% SDS-PAGE and transferred onto PVDF membranes (MilliporeSigma). Then, 5% skimmed milk was used to block the membrane at room temperature for 0.5 h, followed by primary antibody incubation at 4°C overnight. After washing with TBS −0.05% Tween-20 and polysorbate buffer, the membrane was incubated at room temperature with the secondary antibodies for 1 h. The secondary antibodies were horseradish peroxidase-labeled goat anti-rabbit IgG (1:5,000; cat. no. AP156P; Sigma-Aldrich; Merck KGaA) or goat anti-mouse IgG (1:5,000; cat. no. AP-308P; Sigma-Aldrich; Merck KGaA). The protein bands were visualized using an enhanced chemiluminescence kit (Beyotime Institute of Biotechnology) according to the manufacturer's protocol and quantified using the Gel-Pro-Analyzer 4.0 software (Media Cybernetics, Inc.). The primary antibodies were: CCR6 (1:1,000; cat. no. GTX71397; GeneTex, Inc.), ATP binding cassette subfamily G member 1 (ABCG1; cat. no. ab201776; 1:1,000; Abcam), Oct4 (1:1,000; cat. no. ab200834; Abcam), Notch1 (1:1,000; cat. no. 4380; Cell Signaling Technology, Inc.) and GAPDH (1:5,000; cat. no. SAB2103104; Sigma-Aldrich; Merck KGaA).

GraphPad Prism 7.0 (GraphPad Software, Inc.) was used for statistical analysis. All experiments were conducted in triplicate, and the data are presented as the mean ± SD. Comparisons between groups were conducted using paired Student's t-test, and multiple comparisons were made using a one-way ANOVA, followed by Tukey's test. P<0.05 was considered to indicate a statistically significant difference.

The results demonstrated that the expression levels of CD44 and CD117 in the total subclones was significantly increased (P<0.05; Fig. S1). Once cells were isolated, CCR6 and Oct4 expression was determined using RT-qPCR. It was found that both Oct4 and CCR6 expression levels were significantly increased in CD44⁺CD117⁺ cells compared with CD44⁻CD117⁻ cells (P<0.0001 and P<0.001; Fig. 1A). CD44⁺CD117⁺ cells also showed a higher sphere formation ability compared with CD44⁻CD117⁻ subgroup cells, as detected in the sphere formation assay (Fig. 1B). Moreover, CD44⁺CD117⁺ subgroup cells exhibited significant higher protein expression levels of Oct4, CCR6, Notch1 and ABCG1 than CD44⁻CD117⁻ cells (P<0.001; Fig. 1C and D).

Considering these findings, the PTX resistance of cells was determined using CCK-8 assay. The results indicated that the IC₅₀ of CD44⁺CD117⁺ subgroup cells for PTX at 24 h (10.22 nM) and 48 h (5.04 nM) was higher compared with that in CD44⁻CD117⁻ subgroups at 24 h (4.99 nM) and 48 h (2.51 nM), respectively (Fig. 1E). Moreover, the ROS level was lower in CD44⁺CD117 cells compared with CD44⁻CD117⁻ cells after treatment with PTX (Fig. 2A and B). Furthermore, a lower apoptotic rate was observed in CD44⁺CD117⁺ subgroup cells compared with that in CD44⁻CD117⁻ subgroup cells after treatment with PTX (P<0.01, Fig. 2C and D). The CCK-8 assay also indicated that CD44⁺CD117⁺ cells showed increased proliferation compared with CD44⁻CD117⁻ cells after treatment with PTX (P<0.05 and P<0.001; Fig. 2E). These findings suggest that increased stemness and PTX resistance were observed in CD44⁺CD117⁺ subgroup cells than in CD44⁻CD117⁻ cells, which could be attributed to the involvement of the Notch1 signaling pathway.

In order to investigate the underlying pathway involved in the regulation of stemness and PTX resistance, Notch1 was silenced in CD44⁺CD117⁺ cells that were treated with CCL20, a ligand of CCR6, followed by stemness and PTX resistance analyses. It was found that transfection of siNotch1 significantly inhibited the mRNA and protein expression levels of Notch1 (P<0.05, P<0.01 and P<0.001; Fig. S1).

The RT-qPCR analysis revealed that CCL20 treatment could significantly increase the expression level of Oct4, but silencing Notch1 could markedly reverse this upregulation in CD44⁺CD117⁺ cells (P<0.001; Fig. 3A). Moreover, via western blot analysis, it was identified that the protein expression levels of Oct4, ABCG1 and CCR6 in CD44⁺CD117⁺ cells were significantly be increased by CCL20, but that siNotch1 could markedly reverse the effect of CCL20 in elevating Oct4, ABCG1 and CCR6 expression in CD44⁺CD117⁺ cells (P<0.001 and P<0.001; Fig. 3B and C). Sphere formation assays also demonstrated that CCL20 could increase the sphere formation capacity of CD44⁺CD117⁺ cells, while siNotch1 could notably impair the promotive effect of CCL20 (Fig. 3D). Thus, it was suggested that the CCL20/CCR6 axis may increase the stemness of CD44⁺CD117⁺ cells via the Notch1 pathway.

Next, the PTX resistance of CD44⁺CD117⁺ cells after transfection with siNotch1 and CCL20 treatment was determined. The results demonstrated that CCL20 could markedly increase IC₅₀ of CD44⁺CD117⁺ cells (IC₅₀=17.51 nM, 8.41 nM) to PTX compared with the control group (IC₅₀=10.98 nM, 4.86 nM) at 24 and 48 h, while siNotch1 could significantly reverse the effect of CCL20 on PTX resistance of CD44⁺CD117⁺ at 24 h (IC₅₀=7.92 nM, 4.22 nM; Fig. 4A), respectively. After further analysis, it was observed that CCL20 could also markedly ameliorate the ROS level in CD44⁺CD117⁺ cells induced by PTX, while siNotch1 could attenuate the effect of CCL20 in alleviating ROS levels in CD44⁺CD117⁺ cells after treatment with PTX (Fig. 4B). Moreover, the apoptosis of CD44⁺CD117⁺ cells was significantly decreased by CCL20 after treatment with PTX (P<0.0001), but siNotch1 could partially reverse the effect of CCL20 in suppressing apoptosis of CD44⁺CD117⁺ cells induced by PTX (P<0.001; Fig. 4C-E). In addition, the CCK-8 assays identified that CCL20 could increase the proliferation of CD44⁺CD117⁺ cells, and the promotive effect of CCL20 after treatment with PTX could be impaired by siNotch1 (Fig. 4F). This evidence suggested that the CCL20/CCR6 axis nay increase the stemness and PTX resistance of CD44⁺CD117⁺ cells via the Notch1 signaling pathway.

To confirm the role of the CCL20/CCR6 axis in PTX resistance, SKOV3 CD44⁺CD117⁺ and CD44⁻CD117⁻ subgroup cells were subjected to the xenograft assay, followed by CCL20 and/or PTX treatment. The results demonstrated that CCL20 could significantly increase the tumor volume of CD44⁺CD117⁺ and CD44⁻CD117⁻ cells in nude mice compared with the respective PTX treatment alone groups (Fig. 5A and B). From the tumor growth curve analysis, it was found that tumors formed by CD44⁺CD117⁺ cells grew more quickly than those formed by CD44⁻CD117⁻ (P<0.001; Fig. 5B). The growth of all tumors could be significantly increased by treating with PTX, but the volume of tumor-derived from CD44⁺CD117⁺ cells was significantly higher compared with that from CD44⁻CD117⁻ cells (P<0.01; Fig. 5B). Moreover, the weight of tumors derived from CD44⁺CD117⁺ cells was significantly greater than that derived from CD44⁻CD117⁻ cells treated with PTX (P<0.0001; Fig. 5C). The tumor weight was significantly increased by CCL20 treatment in CD44⁺CD117⁺ and CD44⁻CD117⁻ cells treated with PTX. However, tumor weight in the CD44⁺CD117⁺ group was significantly higher than that in the CD44⁻CD117⁻ group (P<0.001; Fig. 5C).

H&E staining analysis tumors of the CD44⁺CD117⁺ groups revealed poorly differentiated morphology compared with the CD44⁻CD117⁻ groups, as characterized by the number of small, round cells with hyperchromatic nuclei and scanty cytoplasm, after treatment with PTX. CCL20 treatment could notably decrease tumor differentiation in the CD44⁺CD117⁺ group compared with in the CD44⁻CD117⁻ group (Fig. 5D). A higher expression level of CCR6 in the tumors was identified in the CD44⁺CD117⁺ group compared with the CD44⁻CD117⁻ group via the IHC analysis. Furthermore, CCL20 treatment could markedly increase the expression level of CCR6 in tumors of the CD44⁺CD117⁺ group compared with in the CD44⁻CD117⁻ group (Fig. 5E).

As observed in further investigations, the CD44⁺CD117⁺ group tissue presented with higher proliferation but lower apoptosis compared with the tumor tissue of the CD44⁻CD117⁻ group after treatment with PTX (Fig. 6A and B). In the CD44⁺CD117⁺ and CD44⁻CD117⁻ groups, CCL20 treatment could significantly accelerate proliferation but inhibit apoptosis after treatment with PTX. However, these changes in the CD44⁺CD117⁺ group were higher compared with those in the CD44⁻CD117⁻ group (Fig. 6A and B). Furthermore, ROS levels of tumors in the CD44⁺CD117⁺ group were markedly lower compared with those in the CD44⁻CD117⁻ group after treatment with PTX or CCL20 (Fig. 6C). The stemness and involved signaling pathway were also determined in tumor tissue to further confirm the underlying mechanism. The results indicated that the expression levels of Oct4, CCR6, ABCG1 and Notch1 in tumor of CD44⁺CD117⁺ and CD44⁻CD117⁻ groups could be significantly accelerated by CCL20 treatment, while such facilitation was increased in the CD44⁺CD117⁺ group compared with the CD44⁻CD117⁻ group (P<0.0001; Fig. 6D and E). Taken together, these findings suggested that CCL20/CCR6/Notch1 signaling could markedly facilitate the stemness and proliferation of the CD44⁺CD117⁺ subgroup cancer cells to promote ovarian cancer resistance to PTX.