br min To visualize cellecell adhesion
60 min. To visualize cellecell adhesion, immunofluorescence staining was chosen. The cultured ABT-888 were treated with the E-cadherin (CDH1) primary antibody (Takara Bio) overnight at 4 C and dyed with the secondary antibody conjugated with Alexa Fluor
555 (Life Technologies) for 30 min at room temperature. Both an-
tibodies were dissolved in 20-mM Tris buffer saline supplemented with 1% of BSA and 10-mM CaCl2 (Nacalai Tesque). The cell cyto-skeleton and nuclei were stained by Alexa Fluor 488 phalloidin (Life Technologies) and Hoechst 33342 (Life Technologies) for 20 min. All stained samples were imaged using a fluorescent microscope (EVOS FL Auto fluorescence microscope, Life Technologies).
The cellular morphologies (nuclear elongation factor, cytoplasm elongation factor, cytoplasm roundness, and nuclear area [AN] to cytoplasm area [AC] ratio [AN/AC]) were manually quantified by following the contour of each cell (n ¼ 32). The cytoplasm elon-gation factor, nuclear elongation factor, cytoplasm roundness, and AN/AC were calculated as the major axis/minor axis of the cyto-plasm, major axis/minor axis of the nucleus, 4(area)/(p[major axis]2) of the cytoplasm, and area of nuclear (AN)/area of the cytoplasm (AC), respectively. By definition, the cytoplasm round-ness is equal to 1 for a completely round cell.
2.4. Real-time polymerase chain reaction
MDA-MB-231 and MCF-7 cells were seeded at the density of 2.0 104 cells cm¡2 on the chamber slide and cultured under normoxic, or hypoxic condition for a period of 3 and 7 days. The total RNA was extracted from the cells cultured for three days using Fast Gene™ RNA Premium Kit (Nippon genetics). Then the RNA was subjected to reverse transcription using a Transcriptor Universal cDNA Master (Roche) following the manufacturer's instructions. The reaction solutions included 5 mL of KAPA SYBR® FAST qPCR Master Mix (KAPA BIOSYSTEMS), 200 nM of forward and reverse predesigned primers (Table S1: Supplementary data), and 1/10 of cDNA template in a 10 mL of volume. The resulting cDNA yield was then subjected to real-time polymerase chain reaction using LightCycler® 96 system (Roche). The results were analyzed with LightCycler® 96 software (Roche). The cDNA samples were analyzed for expression of CDH1, transforming growth factor beta (TGFB), hypoxia-inducible factor 1a (HIF1A), and vimentin relative to the glyceraldehyde-3-phosphatase dehydrogenase as an internal standard for sample normalization. In addition, the cDNA samples of day 3 were analyzed for expression of N-cadherin (CDH2), snail family zinc finger 2 (SNAI2), and zinc finger E-box binding ho-meobox 1 (ZEB1).
2.5. Cellular migration
MDA-MB-231 and MCF-7 cells were seeded on the substrates and cultured under normoxic or hypoxic conditions. Both cells were seeded at the density of 5 103 cells cm 2 for 3 days and 1.25 103 cells cm¡2 for 7 days, precultured for 24 h to attach the cells to the substrates. An environment of 37 C temperature, 5% CO2, and 95% relative humidity or 94% N2, 5% CO2, and 1% O2 was applied during the time-lapse measurement. The cells were stained with 0.1 mg mL¡1 of Hoechst 33342 for 20 min before time-lapse exper-iment. The blue fluorescent images of the tracers were obtained every 15 min for 15e18 h by EVOS FL Auto fluorescence microscope (Life Technologies). To track the cell movement, the position of 20 individual centroids [x(t), y(t)] of the cell nucleus were obtained by a Mosaic particle tracker for ImageJ developed by Helmuth et al. . The start points of each cell were defined at the same position. The evaluation of cellular displacement and direction were conducted to obtain mean squared displacement (MSD), as described in a previ-ous article  (1.4. Evaluation of cellular displacement and 1.5 Characterizing the cellular migration: Supplementary data).
All data presented are expressed as mean±standard deviations. Statistical analysis was performed using one-way analysis of
variance with TukeyeKramer' post hoc testing, and significance was considered at a probability of p < 0.05.
3. Results and discussion
3.1. PLLA and PCL electrospun nanofibers
The morphology of obtained random and aligned electrospun polymeric fibers is shown in Fig. S1 (Supplementary data). All electrospun fibers have uniform, bead-free, and smooth surface morphology with average fiber diameters of ~1.5 mm as revealed by field emission scanning electron microscope observation (FE-SEM) (1. Characterization: Supplementary data).
Fig. S2 (Supplementary data) exhibits the fast Fourier transform (FFT) patterns of the FE-SEM images, i.e., pixel intensity plotted as a function of azimuthal angle. The reciprocal value of FWHW is proportional to the degree of orientation of the fibers. Owing to the branched structure of electrospun A-PCL fiber (Fig. S1c), the A-PCL fiber shows lower degree of orientation compared with A-PLLA fi-ber. Fig. S3 (Supplementary data) shows the stressestrain curves of electrospun fibers evaluated by the uniaxial tensile test.