Cell culture

The MCF-7 cell line (ATCC) was cultured in RPMI 1640 medium (Gibco) containing FBS, L-glutamine, and penicillin/streptomycin, and maintained in an incubator set at 37 °C with 5% CO2. Non-tumorigenic epithelial cell line MCF-12 A (ATCC) was cultured in DMEM/F12 medium (Gibco) supplemented with Nu serum (Corning), ITS Premix (Corning), penicillin/streptomycin (Sigma) and L-glutamine (Sigma) and incubated at 37 °C with 5% CO2. Human umbilical vein endothelial cells (HUVEC, Lonza) were cultured in EGM-2 containing FBS, penicillin/streptomycin and L-glutamine at 5% CO2 and 37oC culture conditions. The cell culture medium was changed every two days for all cultures.

Isolation and characterization of cancer stem cells from MCF-7 cell line

The isolation of CSCs from confluent MCF-7 cells was performed using the Magnetic-Activated Cell Sorting (MACS) method with anti-CD24 and anti-CD44 (Miltenyi) magnetic beads, following the manufacturer’s instructions. After centrifugation at 300xg for 10 min, CSCs were resuspended in FACS buffer. They were then incubated for half an hour with anti-CD24-PE and anti-CD44 FITC antibodies. Subsequently, the characterization of CD44+/CD24−/low CSCs was performed using the BD AccuriTM C6 Flow cytometer (BD Bioscience).

Co-culture and generation of 3D breast cancer model

A medium was prepared to support mammosphere formation properties of CSCs, containing EGF (Sigma), bFGF (Sigma), B-27 (Sigma), L-glutamine (Sigma), and penicillin/streptomycin in RPMI 1640 medium (Gibco). When isolated CSCs were co-cultured with MCF-12 A cells, the culture media and cells prepared for both cell types were mixed in a 1:1 ratio. Methylcellulose (Methocel® MC, Fluka), Matrigel (Sigma), and Collagen-I (Advanced Biomatrix) were added to the created suspension to support the formation of a 3D structure, and the cell culture was maintained for three days. To promote vascularization within the organoid-like structure, HUVECs (Lonza) were introduced into the co-culture, extending the cell culture period by an additional three days. Medium changes were performed by removing half of the old medium and adding an equal amount of fresh medium to sustain the culture. The cell population of the 3D culture was assessed by flow cytometry. Cells were harvested using 3 mg/mL collagenase-I (Sigma) and trypsin, resuspended in FACS buffer solution, and then labeled with anti-human anti-CD44-FITC and anti-CD-24-PE antibodies for half an hour. CD44 + and CD24- cells were identified using the BD AccuriTM C6 Flow Cytometer (BD Bioscience).

Evaluation of the effect of OLE on cells within the 3D breast cancer model

Western Blot was performed to assess the effect of OLE on tumorigenic MCF-7 cells, CSCs and on non-tumorigenic MCF-12 A cells. After cells were harvested and tripsinized, total protein extraction was performed by RIPA lysis buffer (Serva) with protease inhibitor cocktail (Roche). Concentration of the extracted total protein was determined by BCA Assay Kit (Thermofisher). 30 µg of total protein per sample was loaded into the wells of the 10% separating gel. After the run, Trans-Blot Turbo (Bio-Rad) was used for membrane transfer (Advansta). Following washing and blocking with 5% non-fat milk powder (TBS-T), Caspase-3 (Abcam), Caspase-9 (Biorbyt), Bax (Biolegend), Bcl-2 (Biolegend), GAPDH (Biolegend), HRP-anti rabbit, and HRP-anti-mouse (Biolegend) antibodies were used. Each antibody binding was performed by the following steps: blocking, binding of the primary antibody, washing with TBS-T, binding of the secondary antibody, washing, and image acquisition using ECL (Advansta). Protein bands were visualized by ProteinSimple.

Synthesis of methacrylated alginate (mALG)

Briefly, 2.5% w/v alginate (low viscosity alginate, Sigma) was weighed and prepared. Then, it was dissolved with distilled water (100 mL) to form a homogeneous solution. The same volume of methacrylic anhydride (MA, purity ≥ 94%, Sigma) was slowly added into the completely dissolved alginate. The reaction vessel was then covered with foil. The reaction pH control was done with 5 N NaOH, and it was adjusted to 7. The reaction continued in this way for three days. pH was frequently checked throughout the reaction process. Then, mALG was precipitated with ethanol (500 mL) and taken as a solid. Then, precipitated mALG was resolved in DI water. Solution was taken in dialysis membrane. It was dialyzed by using DI water for 7 days. After the dialysis process, the solution was taken into a beaker, frozen and freeze-dried using TeknoSEM brand lyophilizer device for two days.

Characterization of mALG

The unmodified and modified alginate were characterized by using Fourier Transform Infrared (FTIR) and Proton Nuclear Magnetic Resonance Spectroscopy (1H NMR) analyzes. A Jasco FT/IR-6700 spectroscopy was used for FTIR and Varian UNITY INOVA instrument was used 1H NMR for characterization studies.

Preparation of OLE loaded mALG microparticles

Microparticles production was carried out with the microfluidic device. The microfluidic device design was prepared using PDMS material as a droplet-based flow-oriented system. It was prepared from PDMS material, attached to the microscope glass. In the designed system, uniform sized microparticles were produced with the oil-water emulsion method. OLE-loaded mALG microbeads were obtained by using the following procedure. Briefly, 0.2% (w/v) of OLE was weighed and transferred into a glass tube. Then, 1 mL of photoinitiator solution in DPBS (TEA, 1.875% (w/v), VC 1.25% (w/v) and Eosin Y disodium salt (0.5mM)) was added. It was left for 1 h under constant stirring at 250 rpm at 50oC. Afterwards, 10% by weight of mALG was added and allowed to stir for an additional 1 h under same conditions. To prepare the OLE loaded mALG beads, the OLE loaded mALG solution was used as a dispersed phase, while 0.5% w/w Span80 in Mineral oil was used as a continuous phase to prepare a water-in-oil emulsion. Both phases (continuous phase and dispersed phase) were injected separately into the inlet microchannels of the one step emulsification microfluidic device using syringe pumps (Harvard Apparatus PHD 2000, Holliston, MA). The syringes were connected to the luer-stub inlets using polyethylene PE-5 tubing with an outer diameter of 1.32 mm and an inner diameter of 0.86 mm. The flow rate of the continuous phase was kept constant at 100 µL/min, while the flow rate of the dispersed phase was adjusted to obtain droplets of ~ 186 μm. Continuity in droplet formation was observed when the dispersed phase was 3.33 µL/min and the continuous phase was 1.66 µL/min. Then, microparticles were produced and collected in an eppendorf tube. Then, they were crosslinked ionically by using 1 M CaCl2 and were separated from the oil. Finally, the separated microparticles were cured by using visible light, a small LED light source (VALO Light Curing Device, Ultradent, USA) for 240s to make microparticles chemically crosslinked.

Characterization of OLE loaded mALG

The structural analysis of mALG and OLE-loaded mALG microparticles were carried out by using Fourier Transform Infrared (FTIR) analysis. FTIR spectra were recorded on Jasco FT/IR-4600 using ATR adapter in the wavelength range of 400–4000 cm−1. The dimensions and surface pore morphology of mALG and OLE-mALG microparticles were investigated by scanning electron microscopy (SEM) using ESEM-FEG (JEOL JSM 5600.). ImageJ software was used to evaluate the sizes of microparticles.

Drug release studies

Samples were prepared as microparticles, and total drug content was calculated as a function of sample weight and designed weight ratio. Following this, drug release behavior was tested by immersion in 20 mL of phosphate buffered saline (PBS) solution at 37 °C. At the time points, 3 mL of PBS was removed from the solution and the system was refilled with 3 mL of fresh PBS. The absorbance of UV light of the received PBS solution was measured at 250 nm which is a characteristic absorbance value for OLE with a UV-vis spectrophotometer (VALO Light Curing Device, Ultradent, USA). The concentration was then interpreted via an absorbance-concentration calibration curve ranging from 5 to 100 ppm. The calibration curve was calculated as y = 0.0055x + 0.006 and R2 = 0.9996. Here, Y represents the absorbance value of the solution at 250 nm while X represents the OLE concentration (ppm). With the help of the obtained calibration curve, cumulative drug release curves were drawn over time. Experiments were carried out in triplicate.

The OLE loading of the microparticles was estimated according to the following method [16]. The particles were immersed into 20 mL of PBS and broke were broken the particles completely and incubated over 48 h. The absorbance of the supernatant solution was measured by using a UV-Vis spectrophotometer (the specific absorbance value of OLE). The OLE loading was determined by using the standard curve of OLE release and was calculated by the following equation (Eq. 1):

$$Drug \ Loading \ Efficiency \left(\%\right)=\frac \times100$$

(1)

Equation 1. Calculation of OLE loading efficiency.

Physical properties of mALG and OLE loaded mALG

To determine the swelling behavior of OLE loaded mALG microparticles, the prepared hydrogel was pipetted into silicone molds and was exposed to visible light for cross-linking. It was then frozen and lyophilized. The samples prepared for measurement were weighed dry and incubated in PBS solution at 37oC. It was then weighed at certain time intervals (1 h, 3 h, 5 h, 7 h and 24 h). Samples were prepared as 3 pieces, n = 3 and mALG-based samples without OLE were accepted as controls. The swelling ratio of the microparticles was calculated using the equation below (Eq. 2). Here, W0 represents the initial weight and Wt indicates the wet weight of the samples.

$$Swelling \ Ratio \left(\right)=\frac \times100$$

(2)

Equation 2. Calculation of the swelling ratio of the microparticles.

The degradation behavior of the mALG and OLE loaded mALG microparticles were evaluated by immersing them in completely dry and pre-weighed cylindrical samples in PBS solution containing 10 µg/mL of collagenase at 37oC in 24 h. After this process, the samples were removed from incubation and lyophilized. Weight reduction of samples was determined by calculating the dry mass of the sample before and after incubation. All specimens were measured as triplicate.

Effect of OLE-mALG on cell viability

Water Soluble Tetrazolium-1 (WST-1) assay (Roche) was conducted to determine the cytotoxic effect of OLE and to assess cell viability in the created 3D breast cancer model cells. For viability assessment, 10 µL of WST-1 was added to cells cultured in a 96-well plate with 100 µL of medium, in both groups treated and untreated (control) with OLE. After a 2-hour incubation period, the absorbance value (OD) for each sample was measured with a microplate reader at a wavelength of 450 nm.

Effect of OLE-mALG on tumorigenicity and apoptosis

RT-qPCR and Western Blot experiments were carried out to evaluate the anti-tumorigenic and apoptotic properties of CSCs following OLE-mALG treatment.“Initially, the organoid-like structure was dissociated into a single-cell suspension using collagenase (Sigma) and trypsin (Sigma). The cells were then harvested, and total protein extraction for Western Blot Analysis was conducted using RIPA lysis buffer (Serva) and a protease inhibitor cocktail (Roche). The following steps were performed as described in the previous section. The protein levels were assessed in both mALG-OLE treated and untreated (control) cell groups using anti-human anti-Slug (ST Johns Lab.), anti-Vimentin (Biolegend), anti E-cadherin (Biolegend), anti-GAPDH (Biolegend), and HRP-anti-mouse antibodies. Gene expression analysis (NANOG, OCT3/4, SOX2, SURVIVIN, CYCLIN D1, P21, GAPDH, Table 1) was performed by RT-qPCR (LightCycler® 480 II, Roche, Germany). Total cellular RNA was isolated using the “HibriGen RNA Isolation Kit” (Turkey). The isolated RNA concentration was measured at 260 nm using the NanoDrop ND-1000 spectrophotometer (Thermo Fisher Scientific Inc., USA) and the A260/A280 ratio was used to measure the RNA quality. Complementary deoxyribonucleic acid (cDNA) synthesis was performed from RNA samples using the High-Capacity cDNA Reverse Transcription Kit ProtoScript® II First Strand cDNA Synthesis Kit (NEB). Quantitative PCR condition: 10 min denaturation step at 95°C; A total of 45 cycles were set as a PCR step at 95°C for 10 seconds, at 60°C for 30 seconds, at 72°C for 1 second, and at 40°C for 30 seconds as a cooling step. Three biological replicates were run for each different experimental condition and three technical replicates were performed for each sample. Crossing point (Cp) or threshold cycle (Ct) value was calculated for target and reference genes using LightCycler® 480 II software.

Table 1 Primer sequences for detection of mRNA levelsStatistical analysis

SPSS Inc. was used for both parametric and non-parametric statistical analyses. t-tests and ANOVA analyses were conducted to compare different experimental groups. Data with a p value less than 0.05 were considered statistically significant. Unless otherwise stated, each sample was run three times.